Objective: With rapid technological advancements and continuous restructuring of the global automotive industry, innovation has become a central factor in industrial upgrading and value chain repositioning. However, existing studies often focus on industrial growth and technological change separately, overlooking how innovation also affects export performance, production foundations, and regional integration into global value chains. Against this background, this paper investigates the role of innovation in China's automobile manufacturing industry, focusing on its dual effects on industrial evolution and spatial restructuring. It examines whether innovation functions merely as a growth driver or as a structural mechanism linking industrial upgrading and spatial reorganization. Methods: This study constructs a multi-source dataset combining the China Industrial Enterprise Database, firm registration data, the Chinese patent database, and the United Nations Comtrade database. The empirical strategy proceeds as follows. First, a descriptive analysis is conducted to characterize the long-term evolution of China's automobile industry since the reform and opening-up, focusing on changes in production scale, ownership structure, and spatial distribution. Second, the industry's evolution is analyzed through a global value chain framework, highlighting its transition from joint ventures to an innovation-driven structure led by new energy vehicles and intelligent connected technologies. Third, econometric models are used to evaluate how innovation capacity affects export scale, comparative advantages, regional production capacity, and the likelihood of regional entry into the automotive value chain. Innovation is classified into basic and applied innovation types to capture heterogeneous effects, with spatial interaction terms included to consider potential regional spillover effects. Results: The empirical findings suggest several key patterns: First, innovation capacity is positively associated with multiple dimensions of industrial upgrading, leading to stronger export performance and more stable comparative advantages in international markets. Second, innovation boosts regional production capacity, thereby increasing the likelihood of regional participation in the automotive value chain. This suggests that innovation impacts firm-level performance and integrates regions into global production networks. Third, at the spatial level, innovation helps facilitate the diffusion of automotive industry activities across regions, leading to a more dispersed yet increasingly interconnected industrial spatial structure. Fourth, significant heterogeneity exists between types of innovation. Basic innovation tends to generate longer-term and broader spatial spillover effects, whereas applied innovation is more closely associated with short-term improvements in regional performance. Fifth, innovation is closely associated with emerging sectors such as new energy vehicles, where technological advancement and market expansion reinforce each other. Conclusions: Overall, innovation should be understood not only as a driver of industrial growth but also as a structural mechanism connecting industrial upgrading and spatial restructuring. Innovation supports China's automobile industry's transition from extensive expansion to high-quality development and enhances its position in global value chains. The differentiated effects of basic and applied innovation suggest that innovation policy should balance long-term capability building with short-term application support. These findings provide empirical evidence for understanding the trajectory of China's automobile industry and implications for innovation-driven development and value chain optimization.
Objective: The unified national market is continuously expanding. Against this backdrop, fully activating the innovation spillover effects of national value chain (NVC) participation and accelerating the establishment of a "new development paradigm with domestic circulation being the mainstay and the two circulations reinforcing each other" have become crucial pathways for building an independent and controllable modern industrial system. Methods: This study utilized embedded multiregional input–output tables and provincial panel data to examine how NVC participation affects regional innovation performance across 30 Chinese provinces from 2007 to 2017. Fixed-effects models and robustness checks were employed to validate hypotheses. The threshold effect of intellectual property protection (IPP) was also examined. The core explanatory variable, NVC participation degree, was calculated using a production decomposition model based on input–output tables. This approach distinguishes between intraprovincial, national, and global value chains, enabling a nuanced assessment of NVC participation. Control variables, including economic development level, informatization level, openness degree, human capital, financial development, marketization, transportation infrastructure, and innovative industrial policies, were incorporated. The explained variable, research and development (R&D) capital stock, was measured using the perpetual inventory method, which accounts for the cumulative effect of R&D investments over time and provides an accurate reflection of innovation capabilities. Results: NVC participation significantly boosted regional innovation levels. The innovation-promoting effect was more pronounced in western regions and provinces with lower levels of participation in the global value chain. Moreover, an IPP threshold effect was observed: NVC participation yielded substantial innovation gains only when IPP reached a specific level. Stronger IPP reduced imitation risks, incentivized cross-regional knowledge sharing, and created stronger innovation incentives. When IPP was below the threshold, firms tended to prioritize capacity expansion over technological upgrades to protect their knowledge assets. Conclusions: Deepening NVC collaboration within a unified national market is pivotal for innovation-driven growth. Policy implications include reducing administrative barriers to factor mobility, establishing cross-regional industrial alliances to bridge innovation gaps between eastern and western regions of China, and strengthening IPP frameworks to encourage NVC-driven innovation. These findings provide actionable insights for fostering dual-cycle synergies and achieving high-quality technological autonomy in China.
Objective: As ideal hydraulic machines for high-head, low-flow hydraulic energy conversion, Pelton turbines are widely deployed in high-altitude mountainous regions. They play a crucial role in hydropower generation and water resource utilization. However, long-term operation with high-velocity sediment-laden water leads to severe erosion of key components such as the water distribution ring, injectors, and runner buckets. This erosion reduces turbine efficiency, shortens service life, increases maintenance costs, and poses safety risks. Thus, systematically investigating how sediment particle size affects the erosion characteristics of these components is theoretically and practically crucial, thereby providing technical support for optimal design, anti-erosion modification, and safe operation of the Pelton turbine. This study aims to examine the influence of sediment particle size on the erosion characteristics of the water distribution system. Methods: This study employs the SST k–ω turbulence model, coupled with the discrete phase model, to perform unsteady three-phase (water–air–sand) numerical simulations of a six-nozzle distribution system. Results: The results show the following: 1) The multi-nozzle distribution system exhibits a notable flow-diversion effect. Particle trajectories show that the downstream channel receives a substantially reduced number of particles; consequently, erosion damage is primarily concentrated in the upstream channel and the first three injectors. 2) Secondary flow phenomena, particularly Dean vortices, arise from the combined effects of inertia and curvature within the water distribution ring. Erosion distribution correlates strongly with these flow structures, with damage concentrated near elbow sections, bifurcated pipe junctions, and internal vortex regions within the injectors. This indicates that erosion in the water distribution system is predominantly governed by secondary flow dynamics. 3) Particle properties have a remarkable impact on injector erosion. Fine sediment particles, due to their strong flow-following characteristics, mainly erode the needle surface. Moreover, residual Dean vortices are reactivated in the contraction section of the injector, intensifying particle accumulation in the vortex region and increasing the collision rate with the needle surface, thereby exacerbating needle erosion. 4) Large sediment particles exhibit strong inertial effects and poor flow-following characteristics, causing their trajectories to deviate considerably from streamlines and mainly erode the nozzle shell. 5) In a mixed particle-size group, a notable synergistic erosion effect is observed: fine particles primarily damage the needle surface, whereas large particles are more likely to impact the nozzle shell, further intensifying erosion in that area. Conclusions: The multi-nozzle water distribution system demonstrates a pronounced flow diversion effect, substantially reducing the number of particles in the downstream channel. Consequently, erosion is primarily concentrated in the upstream channel and the first three injectors. Secondary flow and Dean vortices are prominent in the water distribution ring, with erosion predominantly occurring near elbows, bifurcated pipes, and internal vortex regions of the injectors, driven by secondary flow dynamics. Sediment particle size considerably affects the distribution and intensity of injector erosion. Fine particles, with strong flow-following abilities, erode the needle surface, whereas residual Dean vortices in the contraction section exacerbate this erosion. By contrast, large particles, with strong inertia and poor flow following abilities, deviate from streamlines and erode the nozzle shell. Mixed particle sizes exhibit a synergistic erosion effect, intensifying overall injector damage.
Objective: The intake pool of a deep-tunnel pump station is highly prone to air-entraining vortices due to its unique structural design, which can significantly impair pump performance. However, the mechanisms behind the initiation, development, and interaction of vortices in this specific intake pool configuration are not well understood. The lattice Boltzmann method combined with large-eddy simulation (LBM-LES) has been demonstrated in other fluid dynamics areas to match the accuracy of traditional LES methods while providing benefits in boundary handling and parallel processing. To examine vortex characteristics in the intake pool of a deep-tunnel pump station, this study uses the LBM-LES approach along with particle image velocimetry (PIV) flow-field measurements and numerical simulations on a model of the pump station. By comparing experimental data with simulation results, the study verifies the reliability and precision of the LBM-LES method. The combined experimental and numerical findings are then used to analyze vortex distribution and flow-field features of the intake pool under various operating conditions. Methods: A 10∶1 scale model of the deep-tunnel pump station intake pool was built in this study. To analyze the development of air-entraining vortices, three operating conditions— "below the critical Reynolds number Re," "at the critical Re," and "above the critical Re" —were set for the experiments and simulations, all using the same water depth. PIV flow-field measurements were then performed under these conditions, and vortex evolution of the scaled model was simulated with the LBM–LES method. The numerical results from LBM-LES were validated by comparing the average velocities and velocity components along measurement lines in both the simulations and experiments. Finally, the experimental and numerical data were used to examine vortex characteristics across different sections of the intake pool, including vortex distribution, intensity, and scale. By comparing vortex features under various operating conditions, the evolution of vortices near the critical condition was determined. Results: The numerical and experimental results showed the following: 1) The LBM-LES model aligned well with the experimental data. The velocity-component errors at key measurement points were within 5%, and the predicted number, distribution, and vorticity strength of the vortex structures matched the experimental observations. 2) The flow above the right outlet pipe was affected by multiple vortex systems. As Re increased, the vortex system near the right wall tended to merge, the vortex structures became more stable, and the vorticity gradually intensified. The flow above the left outlet pipe was dominated by a single vortex. With increasing Re, this vortex shifted from the left wall to a position directly above the pipe near the rear wall, and its vorticity progressively increased. Meanwhile, numerous vortices with opposite rotational directions formed around the main vortex system and extended toward the left and right walls of the intake-pool expansion section. 3) Vortices tended to form above outlet pipes near the rear wall. As Re increased, the negative vorticity above the pipe in Section 1 increased, while the vortex structure above the pipe in Section 2 moved closer to the free surface, and its vorticity gradually increased. Conclusions: The LBM–LES method was verified as a reliable and accurate approach for simulating vortex evolution in the intake pool of a deep-tunnel pump station, providing a new mesoscopic tool for similar intake-flow studies. Meanwhile, the investigated vortex characteristics will deepen the understanding of the vortex formation mechanism in this special type of intake pool, offering valuable engineering guidance for optimizing its structural design.
Objective: The unsteady flow within a centrifugal pump, particularly the pressure distribution on rotating impeller blades, is fundamental to its performance and stability. Direct, high-fidelity measurement of this pressure field remains a significant technical challenge due to the complexity of transmitting data from a high-speed rotating frame. In response, this study develops and validates a novel wireless measurement system designed to acquire dynamic pressure data directly from rotating pump blades. The objective is to use this system to conduct a detailed experimental investigation of the spatiotemporal characteristics of blade surface pressure under various operating conditions. This approach provides crucial data for validating computational fluid dynamics (CFD) models, optimizing hydraulic design, and understanding the root causes of pump vibration and noise. Methods: The study was conducted on a centrifugal pump with a design flow rate (Qd) of 18 m3/h, a rotational speed of 1 450 rpm, and an eight-blade impeller. A wireless measurement system was custom-designed, integrating four miniature pressure sensors (0–200 kPa range, 0.5% accuracy) flush-mounted on the blade surface to minimize flow disturbance. A specially machined main shaft with internal grooves routed the sensor wires to a compact shaft-mounted module containing signal amplifiers, a multi-channel data acquisition (DAQ) card, and a Wi-Fi transmitter. The DAQ system synchronously sampled data from all four channels at 1 024 Hz. Measurement points were located on the pressure side at 30% (PA) and 70% (PB) of the chord length, with corresponding locations on the suction side (SA and SB). Experiments were conducted over a range of flow rates, and nonstationary signal processing techniques, including time-domain analysis and Fast Fourier Transform, were applied. Results: The experiments provided detailed insights into time-averaged and unsteady pressure characteristics. Time-averaged results confirmed that on both the pressure and suction sides, the pressure at the trailing edge was higher than that at the leading edge. The pressure difference between the pressure and suction sides increased linearly with flow rate. Notably, the leading edge exhibited greater sensitivity to flow rate variations compared with the trailing edge, indicating that improvements in inlet flow conditions more significantly impact the loading at the blade front. The dynamic pressure signals showed strong periodicity driven by rotor–stator interaction (RSI) with the volute tongue. Frequency-domain analysis revealed that the dominant pulsation frequency at all locations was the shaft rotation frequency (Fi≈24.17 Hz), followed by the blade passing frequency (Fb≈193.36 Hz) and the third harmonic of the shaft rotation (≈72.5 Hz). The amplitude of the dominant frequency varied non-monotonically with flow rate, decreasing initially, then increasing, and finally decreasing again. Furthermore, the dimensionless peak-to-peak value (97% confidence level) exhibited a monotonic trend with increasing flow rate. Conclusions: A stable and reliable wireless measurement system for acquiring dynamic surface pressure on a rotating centrifugal pump impeller was successfully developed and validated. The system enables multi-point, synchronized, high-fidelity data acquisition, overcoming the limitations of conventional methods. The results offer direct quantitative insights into the effects of RSI and flow rate on blade loading and pressure pulsation, providing a valuable experimental database for CFD validation and a key foundation for the design of high-performance centrifugal pumps.
Objective: Airborne wind energy (AWE) has attracted increasing attention because it can access stronger and more consistent winds at higher altitudes, offering high power density with reduced material consumption. Among various AWE technologies, the parafoil-based kite power generation system is a promising option because of its lightweight structure, operational flexibility, and suitability for pumping-cycle electricity generation. However, its practical application is constrained by the difficulty of maintaining stable cyclic operation under realistic turbulent wind conditions. Existing studies have mainly focused on steady inflow conditions or limited turbulence scenarios, and the combined effects of turbulence intensity and control parameter settings on system stability remain insufficiently understood. Therefore, this study investigated the stability of a parafoil-based kite power generation system under turbulent wind fields and explored effective control parameter adjustment strategies to improve operational stability. Methods: A coupled numerical framework was established for a pumping-cycle parafoil-based kite power generation system, incorporating parafoil aerodynamics, tether dynamics, atmospheric modeling, and dual closed-loop proportional–integral–derivative (PID) control. Parafoil aerodynamics were described using a segmented aerodynamic analysis method, the tether was modeled by the lumped-mass method, and the atmospheric model accounted for variations in wind speed and air density with altitude. Three-dimensional turbulent wind fields generated by TurbSim were imposed as inflow conditions to represent realistic atmospheric disturbances. Different turbulence intensities and target-attractor elevation angles were examined to assess their coupled effects on system stability. For each operating condition, multiple independent turbulent wind fields were simulated, yielding a total simulation time of 20 000 s. Under noncrash conditions, more than 80 complete power-generation cycles were achieved, ensuring that the conclusions were supported by sufficient statistical samples. System stability was evaluated using two indicators: the average operating duration and the average crash probability per power-generation cycle. Results: For all target-attractor elevation-angle settings, the average operating duration decreased progressively with increasing turbulence intensity, whereas the average crash probability per power-generation cycle increased continuously, indicating that strong turbulence significantly weakens system stability. Moreover, when the turbulence intensity reached a certain range, both the indicators exhibited accelerated deterioration, suggesting threshold-like behavior. Statistical analysis revealed that the transition stage between the power-generation phase and the recovery phase was the most vulnerable part of the pumping cycle. During this stage, the parafoil was more sensitive to aerodynamic disturbances, and lateral and vertical wind fluctuations could drive local angle-of-attack excursions beyond the normal operating range, leading to rapid aerodynamic degradation, trajectory deviation, and ultimately a crash. When fluctuations in lateral and vertical wind speeds within a 1 s timescale exceeded approximately 2.5 m/s, the system faced a pronounced crash risk. At turbulence intensities of approximately 15%–18%, such fluctuations became considerably more frequent, and the crash probability increased sharply. Under the same turbulence intensity, a higher target-attractor elevation angle generally yielded a longer operating duration and a lower crash probability, and this stabilizing effect became more pronounced under high-turbulence conditions. Conclusions: The findings indicated that the stability degradation of the parafoil-based kite power generation system in turbulent wind fields is not caused by a single factor but results from the coupled interaction among parafoil aerodynamic characteristics, transient flight states, turbulent disturbances, and the limited compensation capability of a fixed-parameter PID controller. Under low to moderate turbulence, the system maintained stable pumping-cycle operation, whereas under strong turbulence, the stability deteriorated rapidly in a threshold-like manner. Appropriately increasing the target-attractor elevation angle can enlarge the flight-envelope safety margin and improve the system's tolerance to disturbance, delaying stability degradation and enhancing operational robustness. These findings offer practical guidance for control parameter selection and stable operation of parafoil-based kite power generation systems in complex wind environments.
Objective: The air spring with an auxiliary airbag (ASAA) is a novel dual-chamber throttling tube air spring featuring a lightweight design, with low cost and flexible installation layout. Different from conventional air springs with constant-volume auxiliary chambers, the ASAA employs an elastic airbag whose volume varies with internal pressure, enabling a wide adjustable range of stiffness and damping. Because of this structural feature, the ASAA exhibits strong adaptability across diverse operating conditions and shows great potential in intelligent air suspension systems. However, under large-amplitude excitation, the coupled nonlinear effects of airbag expansion, the complex viscoelastic behavior of the rubber bladder, and turbulent airflow within the throttling tube lead to highly nonlinear dynamic characteristics. These strong nonlinearities drastically limit the prediction accuracy of traditional physical models. To address this problem, a physics-data hybrid modeling approach is proposed to enhance the prediction accuracy of dynamic stiffness and lag angle under large-amplitude conditions. Methods: A simplified physical model of the ASAA was developed based on the thermodynamic theory, flow resistance analysis of the throttling tube, and viscoelastic modeling of the rubber bladder. For the compressed air part, governing equations were derived under adiabatic assumptions. The nonlinear airbag volume variation term, which is difficult to linearize, was simplified to reduce model complexity, whereas the nonlinear airflow term in the throttling tube was linearized using a first-order Fourier series expansion. The rubber bladder was modeled by combining a Coulomb friction model, which captured amplitude-dependent hysteresis, and a fractional-order Kelvin-Voigt model, which reproduced frequency-dependent viscoelastic effects. Static and dynamic experiments were conducted to identify the key structural parameters. Dynamic experiments were further carried out at excitation frequencies ranging from 0 to 6 Hz and amplitudes ranging from 5 to 40 mm to obtain the experimental data on dynamic stiffness and lag angle. A deep neural network (DNN) was subsequently constructed to learn the residual errors between the experimental measurements and the physical model calculations. Excitation frequency and amplitude were used as input features, and the corresponding errors in dynamic stiffness and lag angle were taken as output targets. The DNN had two hidden layers with rectified linear unit activation and dropout regularization and is trained using the Adam optimization algorithm on normalized datasets to ensure convergence stability and generalization capability. Results: Validation results demonstrated that the proposed physics-data hybrid model considerably improved the prediction performance, particularly under large-amplitude excitation. In the test cases with excitation amplitudes of 35 mm and 40 mm, which were not included in the training dataset, the hybrid model demonstrated strong generalization ability. For these test cases, the maximum relative errors of dynamic stiffness prediction reduced to approximately 3.22% and 3.98%, respectively, representing an improvement of approximately 6% compared with the conventional physical model. For lag angle prediction, the maximum relative errors were 17.21% and 15.83%, respectively, corresponding to an error reduction of approximately 35% compared with the conventional physical model. Although lag angle prediction remains more sensitive because of its small baseline value and the strong nonlinear stiffness-damping correlation, the hybrid model effectively captures the overall variation trend and achieves substantially improved accuracy. Conclusions: By integrating physics-based modeling with machine learning-based residual compensation, the proposed hybrid approach effectively overcomes the limitations of simplified physical models in describing strong nonlinear behavior. The method substantially enhances the prediction accuracy of dynamic stiffness and lag angle under highly nonlinear, large-amplitude conditions, providing theoretical support and practical guidance for the design and dynamic modeling of intelligent air suspension systems.
Significance: Autonomous driving maps are essential for the safe operation of intelligent vehicles, especially on Chinese urban roads, where road structures are complex and traffic conditions change frequently. For a long time, map production has relied on specialized surveying and extensive manual annotation, resulting in high costs, low efficiency, and long update cycles, making large-scale coverage and continuous updating difficult to sustain. Under these conditions, end-to-end autonomous driving map generation has gradually become an important direction in this field. Its fundamental goal is to infer structured map outputs more directly from sensor observations, thereby reducing dependence on complex intermediate procedures and intensive manual intervention. With the continued development of deep learning, the two core tasks of autonomous driving map generation—lane network generation and lane topology prediction—have both shifted toward end-to-end approaches. Concurrently, industrial end-to-end systems have been developed for city-scale map production and updating, while large models, vision-language models, and agent-based systems have further expanded the research space in this area. Nevertheless, several challenges remain unresolved, including generalization in complex scenes, dynamic change detection, interpretability, and consistency in multi-vehicle collaboration. Therefore, it is important to systematically review existing research to provide clearer guidance for the future development of this field. Progress: For lane network generation, the technical route has evolved from early convolutional neural network (CNN)-based methods to Transformer-based and hybrid architectures. CNN-based methods were widely adopted at an early stage due to their mature operators, stable training, and low deployment cost. They are effective in modeling local geometry; however, their ability to preserve long-range structural consistency becomes more limited in complex scenes. Transformer-based methods later emerged as a major direction because query-based decoding is well suited for structured instance prediction and global relation modeling. Subsequent studies further expanded this line of work to include geometric constraints, map element representation, prior-guided prediction, and temporal consistency in online mapping. Hybrid architectures combine convolutional feature extraction, Transformer-based reasoning, graph modules, and temporal memory, enabling local geometric precision, topological consistency, and engineering feasibility to be addressed within a unified framework. A similar shift can be observed in lane topology prediction. Early methods mainly relied on local connection inference, whereas later studies increasingly treated topology as a structured prediction problem involving order, connectivity, and global consistency. Transformer-based methods strengthened long-range dependency modeling and gradually incorporated geometry, order, connectivity, and generation into a more unified framework. In contrast, graph neural network-based methods explicitly represent node relationships, edge constraints, and multi-scale connectivity patterns through graph structure. Hybrid methods further combine segmentation, sequence modeling, graph reasoning, and redundant supervision to improve robustness in complex road scenes. The main challenge is no longer limited to recovering local geometric shapes, but increasingly lies in maintaining structural consistency when map elements, topological relations, temporal information, and prior knowledge are considered together. Beyond these academic methods, industrial practice has also developed end-to-end map generation systems for large-scale urban deployment, aiming to improve automation, reduce production costs, and shorten update latency. Related studies also cover online high-definition map construction and pseudo-label learning under weak or missing annotations. Meanwhile, large models and vision-language models have gradually entered this line of research. Their potential in map generation has attracted increasing attention; however, their practical use is still constrained by data quality, structural priors, engineering controllability, and the requirements of real production environments. Conclusions and Prospects: Overall, end-to-end autonomous driving map generation is reshaping the conventional map construction paradigm and has shown clear advantages in process simplification, timeliness, and structured prediction. With the introduction of large models, vision-language models, and agent-based systems, map generation may gradually move beyond geometric construction toward a stage that also involves semantic understanding and human–machine interaction. Further progress in this direction will depend not only on model capability but also on the establishment of stable data pipelines, quality control mechanisms, and closed-loop engineering workflows, thereby supporting the large-scale and stable deployment of autonomous driving systems.
Objective: Aiming at the problems of low convex quality in single-neighborhood search, insufficient adaptability to dynamic obstacles, and poor controllability in multi-convex region switching for autonomous vehicle trajectory planning, this paper proposes a dynamic trajectory planning method based on Variable Differential Neighborhood Search (VDNS). The core objectives are to improve the coverage and quality of drivable convex space, enhance the smoothness of trajectory under neighborhood switching, and ensure the stability and convergence of multi-convex shape transition, so as to provide a safe, efficient and robust trajectory planning solution for autonomous driving in dynamic urban traffic scenarios. Methods: Firstly, a differential neighborhood model integrating vehicle kinematics and real-time environmental perception is established. By embedding the Maximum Volume Inscribed Ellipse method into the differential neighborhood search framework, the maximum differential neighborhood at the current moment is generated, which provides a strict lower bound of safe convex polyhedron space and balances the quality and efficiency of neighborhood generation. Secondly, a comprehensive evaluation function considering longitudinal driving distance, lateral deviation and safety margin is constructed, and an adaptive weight adjustment mechanism based on scene risk and motion urgency is introduced to realize the dynamic optimization of the next moment differential neighborhood. Then, the logarithmic barrier function is adopted to transform the trajectory smoothness optimization with inequality constraints into an unconstrained quadratic programming problem, and the Newton iteration method is used to solve it, which suppresses trajectory oscillation and meets vehicle dynamics constraints. Finally, the average dwell-time method is applied to establish the switching stability index of variable differential neighborhood search. Each differential neighborhood is regarded as an independent subsystem, and the sufficient conditions for exponential convergence of the switched system are derived to guarantee the controllability of multi-convex shape transition. Results: The joint simulation based on PreScan, Simulink and CarSim shows that: 1) Compared with the single-neighborhood search (NS) algorithm and iterative regional inflation (IRIS) algorithm, the proposed VDNS algorithm improves the coverage of drivable convex space by 20% in dynamic obstacle environment and 7% in static obstacle environment, with significantly enhanced heuristic search performance. 2) In two consecutive obstacle avoidance maneuvers, the convergence time of neighborhood switching stability is controlled within 0.6 s and 0.4 s, and the maximum velocity overshoot is only 1.00% and 2.94%, showing strong switching stability. 3) The trajectory generated by VDNS is continuous and smooth without oscillation, which overcomes the shortage of poor smoothness in traditional IRIS algorithm. 4) With the increase of obstacle number, the single-step time complexity of VDNS is lower than that of IRIS, and the computational efficiency is higher in complex environments. 5) The adaptive weight mechanism achieves better balance among safety, efficiency and smoothness, which is superior to the fixed weight strategy in trajectory quality and obstacle avoidance performance. Conclusions: The VDNS-based dynamic trajectory planning method effectively improves the convex quality of neighborhood search and the controllability of multi-convex region switching. It not only expands the drivable convex space and enhances the dynamic adaptability to obstacles, but also ensures the smoothness of trajectory and the stability of neighborhood switching. This method can be applied to urban dynamic traffic scenarios with static and dynamic obstacles, and provides a new technical approach for real-time, safe and reliable trajectory planning of autonomous vehicles.
Objective: Advanced equipment is the cornerstone of industrial modernization. New-energy vehicle manufacturing requires large servo presses to improve the quality and efficiency of automotive body stamping. However, the power fluctuation in such presses significantly increases grid capacity requirements. This study investigates the characteristics of power fluctuation in large servo presses with six-link mechanisms. Methods: Based on the principle of the six-link transmission mechanism, the kinematic model of the transmission system of the press was established, and the motion relationships between the bars, slider, and crank were obtained. Specifically, the expressions for the load torque, inertia torque, load power, and inertia power transmitted to the main motor shaft were derived, considering the stamping load and system inertia. Finally, an analytical model for the output power of the main motor was obtained. Taking the 2 500 t servo press produced by Jinan No. 2 Machine Tool Factory as an example, the fluctuation law of the output power of the main motor of the press during the stamping cycle and the cause of the peak power were analyzed by combining the analytical model and experimental data. Results: The established power analytical model can effectively capture the power fluctuation characteristics of a large servo press with a six-bar linkage. The main factors affecting the inertia power are the inertia and acceleration of the main transmission system of the press. Conversely, the main factors affecting the load power are the load applied to the slider and its speed. During the slider movement from the upper dead point to the starting point of the workpiece drawing, the inertia power initially increased and then decreased; the load power increased rapidly when the drawing pad started to operate. In the deep drawing stage of the workpiece, the inertia power fluctuated around zero, but the load power increased. During the upward movement of the slider from the bottom dead point, the inertia power initially decreased and then increased; the load power gradually decreased. The largest power peak was observed when the drawing pad started to operate, and the slider just crossed the bottom dead point; this is attributed to the speed fluctuations at both phases. When the drawing pad started to operate, speed fluctuation was caused by the rapid load increase, leading to a large positive power peak; when the slider just crossed the bottom dead point, speed fluctuation was caused by the change in the load direction, resulting in a large negative power peak. Conclusions: To reduce the impact of power fluctuations in servo presses on the grid side, it is necessary to reduce the load impact when the drawing pad starts to operate, and the slider just crosses the bottom dead point. When the hydraulic drawing pad starts to operate, the speed difference when it contacts the slider can be reduced by optimizing the motion control algorithm of the drawing pad. The load impact when the slider crosses the lower dead point can be suppressed by reducing the link clearance of the transmission system.
Objective: Type Ⅳ high-pressure hydrogen storage vessels have become a key development direction for the global low-carbon transition due to their high hydrogen storage density and light weight. However, the complex stress states and continuously varying layup angles in the dome region present significant design challenges for composite material layups. Traditional grid theory, which primarily focuses on cylindrical body stress analysis, makes it difficult to ensure dome region strength and a safe burst mode. To address this issue, this study proposes an improved grid theory that accounts for dome stresses and establishes fundamental layup arrangement rules through combined finite element analysis. Methods: This study integrates theoretical derivation with finite element simulation. Based on traditional grid theory, spiral-direction and hoop-direction correction coefficients were introduced to ensure dome region strength and to regulate the burst mode. The minimum winding angle of the cylindrical body is calculated using the geodesic winding principle. A reaming winding strategy is employed to prevent fiber accumulation at the polar opening and to enhance the dome transition region strength. A total of 18 layup schemes are designed, which include three experimental groups with different angle combinations and one control group based on traditional grid theory. A 1/36-axisymmetric finite element model of the type Ⅳ hydrogen storage vessel is established in Abaqus, utilizing the WoundSim plugin. Periodic boundary conditions, fixed-end constraints, and internal pressure loads are applied. The Endcap factor in the plugin accurately simulates fiber turning points in the dome region and fiber accumulation at the polar opening, thereby constructing a high-precision finite element model of the composite layup. The peak stress, stress distribution, and failure mode of the composite layup are analyzed under an internal pressure load equal to the minimum guaranteed burst pressure. Results: Simulation results from the 18 design schemes indicate that in the control group, designed using traditional grid theory, the peak stresses in both the dome and cylindrical body regions are similar and significantly exceed the material’s ultimate strength, confirming the necessity of revising traditional grid theory. Among all designs, four qualified schemes are identified, revealing a critical spiral-direction correction coefficient of 1.667 and a critical hoop-direction correction coefficient of 1.153. Comparative analysis of the schemes reveals the following: 1) Increasing the proportion of high-angle spiral layers reduces fiber accumulation and slippage at the polar opening while improving fiber stress distribution uniformity in the dome’s non-polar opening region, thereby reducing design redundancy. 2) The arrangement sequence of layup angles influences the location and severity of stress concentration. 3) Placing the hoop layer adjacent to the liner reduces hoop stress in the cylindrical body’s middle section but increases stress levels in the winding initiation zone. 4) The ratio of the spiral-direction correction coefficient to the hoop-direction correction coefficient critically regulates the vessel’s burst mode. Conclusions: By introducing hoop- and spiral-direction correction coefficients, an enhanced design of the composite vessel dome region can be achieved. The layup angle configuration and arrangement rules derived from these finite element results provide a theoretical basis and engineering guidance for the rapid design and verification of fiber layup schemes for type Ⅳ hydrogen storage vessels.
Objective: Coating thickness uniformity is a key indicator of spray quality, and its improvement depends on process parameter optimization. Therefore, a simulation method combining computational efficiency with physical interpretability is required to predict coating thickness distribution under different process parameters. Existing methods primarily include physics- and model-based simulations. The former offers high physical interpretability but suffers from high computational cost and long solution times, while the latter is efficient but relies heavily on fitted experimental data and often lacks robustness when spray conditions or workpiece geometry change. To address these limitations, this study proposes a particle system-based simulation method for robot spray coating thickness prediction and trajectory optimization. Methods: A particle system-based simulation framework for robot spray coating thickness is established in Unity. By retaining essential particle attributes, such as mass, velocity, and acceleration, and simplifying complex multiphysics processes such as atomization, airflow coupling, droplet breakup, and splash, the spray process is efficiently modeled. The workpiece surface is discretized into grid cells, and coating thickness is obtained by accumulating particle impacts on each grid element. For particle emission modeling, the droplet size is characterized by the volume median diameter, and the particle emission rate is determined based on the paint flow rate. A probability density approximation method for the initial particle emission velocity direction is established based on rejection sampling, enabling the simulation to be theoretically applicable to arbitrary distribution models. Furthermore, an elliptical double-beta distribution is adopted to describe the initial particle emission velocity direction. For spray gun motion modeling, trajectory interpolation is employed, with trapezoidal velocity interpolation used for position and spherical linear interpolation used for orientation to ensure smooth and continuous motion. In addition, a transfer efficiency parameter is introduced to represent the effective deposition ratio after the combined effects of diffusion, evaporation, and splash, such that the dissipative effects in the particle flight and deposition processes can be considered in a simplified manner. Coating thickness is then calculated based on the particle volume and material properties. Furthermore, a joint simulation and optimization framework is constructed by integrating Unity and Python through ML-Agents, enabling interaction between the simulation environment and external optimization algorithms. The performance of the proposed method is evaluated in terms of computational efficiency and prediction accuracy under typical spraying conditions. Results: The experimental results are as follows: 1) The computational efficiency of the proposed framework is mainly affected by the particle emission rate and grid density. Under appropriate parameter settings, the computation time of a single simulation ranges from several seconds to tens of seconds, which is significantly shorter than the tens of hours required by traditional physics-based simulation methods; 2) In planar specimen validation, the relative error between simulated and measured coating thickness is within 12%, indicating that the proposed method achieves acceptable prediction accuracy for practical applications. Conclusions: By combining simplified physical modeling with an efficient particle system framework, this study efficiently predicts coating thickness distribution under different process parameters while balancing computational efficiency and physical interpretability. The proposed method maintains an acceptable prediction accuracy and can provide effective support for spray process parameter optimization.
Objective: Aircraft operations at high-altitude airports and during cruise phases are subjected to low-pressure environments, which significantly alter the physical and optical properties of smoke aerosols generated via cargo combustion, presenting severe challenges for conventional detection systems. Existing research has predominantly focused on macroscopic combustion parameters, while the evolutionary patterns of key microscopic parameters, such as the Sauter mean diameter (SMD) and particle number concentration, remain underexplored. This study aims to systematically elucidate the mechanisms underlying the evolution of multiwavelength smoke aerosol characteristics under low-pressure conditions, thereby providing theoretical support for enhancing the reliability of aviation-smoke detection. Methods: This study established an integrated experimental platform based on a full-scale, dynamic-pressure and temperature-controlled chamber, capable of simulating a pressure range of 10–101 kPa. A triple-wavelength laser detection system—comprising three independent laser–detector pairs at 0.450 μm (blue), 0.532 μm (green), and 1.064 μm (infrared)—was designed and implemented. The optical path for each wavelength was independently calibrated, and the actual optical path lengths were precisely determined. The SMD and particle number concentration were retrieved using the multiwavelength extinction method and the Beer–Lambert law, combined with Mie scattering theory. Four representative fuels—beech wood, corrugated paper, n-heptane, and polyurethane—were selected to simulate typical cargo-compartment fire scenarios under smoldering and flaming conditions. Experiments were systematically conducted at three pressure levels: 90, 70, and 50 kPa. For each pressure condition, optical data were recorded at 1 s intervals across 1 500 repetitions and each set of experiments was repeated three times to ensure reproducibility. With a broad spectral span from blue to infrared wavelengths, the system provided enhanced sensitivity to particle size variations within the typical smoke aerosol range of 0.100–1.000 μm. The introduction of the third wavelength (green) served as an independent constraint, effectively reducing the common inversion multiplicity problem encountered in single-or dual-wavelength systems. Results: This study systematically revealed, to the best of our knowledge, for the first time, the differential responses of smoke parameters to pressure variations across combustion modes. 1) Under smoldering conditions, the SMD slightly increased with decreasing pressure (beech: 0.359→0.376 μm; paper: 0.292→0.318 μm), with variations being only < 0.020 μm, showing remarkable size stability. In contrast, flaming aerosols showed significant SMD reduction; n-heptane aerosols exhibited substantial variations exceeding 0.200 μm, while polyurethane aerosols varied < 0.050 μm, indicating higher pressure sensitivity for pure hydrocarbon fuels. 2) Regarding the particle concentration: smoldering smoke displayed a nonmonotonic trend (initial increase followed by decrease), inversely correlated with the optical power; the n-heptane flaming concentration continuously increased with decreasing pressure, whereas the polyurethane concentration decreased due to oxygen-limitation–induced pyrolysis suppression. 3) Method validation confirmed that dispersion values were < 10% for the triple-wavelength system, considerably enhancing the reliability of the particle size and concentration measurements under low-pressure conditions. Conclusions: The evolution of smoke aerosol size and concentration in low-pressure environments is strongly governed by the combustion mode and fuel characteristics. Smoldering smoke exhibits notable size stability, whereas flaming smoke demonstrates significant pressure sensitivity, with fuel volatility and chemical structure being key influencing factors. The triple-wavelength extinction method, through multiwavelength collaborative constraints, effectively addresses the technical challenges of aerosol characterization at low pressures, providing crucial methodological support and a theoretical foundation for optimizing next-generation aviation smoke detection systems.
Objective: The complex working environment of oil and gas pipelines, characterized by widely distributed risk sources, often hinders employees' ability to identify risks, which is a key cause of workplace safety incidents. Supplementary safety signs act as important information carriers that complement the content of main safety signs, effectively nudging employees toward better risk identification behavior. However, their effectiveness is significantly influenced by core attributes such as color, shape, and format, but systematic research on this issue is currently lacking in the oil and gas pipeline industry. This study aims to identify how different attributes of supplementary safety signs influence employees' visual attention and nudge risk identification behavior, providing a theoretical framework for optimizing safety sign design and enhancing hazard recognition efficiency in oil and gas pipeline enterprises. Methods: Based on a major oil and gas pipeline network enterprise in China, this study selected three representative operational scenarios: anti-corrosion, pressure testing, and maintenance and repair of pipelines and storage tanks. Using 38 risk identification points as sign content, 18 groups of supplementary safety signs were designed following national standards. An orthogonal design was used to combine different attributes of color (red, blue, black) × shape (triangle, circle, rectangle) × format (text, icon). The study recruited 180 participants with at least 2 years of work experience in oil and gas pipelines, who were randomly divided into 18 groups to participate in the orthogonal experiments. A Tobii Pro Glasses 2 wearable eye tracker was used to collect four eye-tracking metrics: time to first fixation, first fixation duration, total fixation duration, and fixation count. Furthermore, two behavioral metrics—response time and accuracy of risk identification—were recorded. Multivariate analysis of variance (MANOVA), Least Significant Difference post hoc tests, and interaction plots were employed to systematically evaluate the main and interaction effects of color, shape, format, and their combinations on visual attention characteristics and behavioral nudge effects. Results: Color and shape were the primary drivers of visual attention orientation and behavioral accuracy. Specifically, red and triangular designs significantly shortened the time to first fixation (by 411 ms and 213 ms, respectively) and improved behavioral accuracy by 5.79% and 2.20%, respectively. Notably, color (Partial η2 = 83.6%) and shape (Partial η2 = 55.2%) exhibited the strongest main effects on time to first fixation. The format was the main driver of sustained visual attention. Specifically, text signs drew more engagement than icon signs, increasing first fixation by 363 ms, total fixation by 350 ms, and fixation count by 1.93. The findings also demonstrated that color and format were critical factors influencing risk identification response time. Red signs outperformed blue and black variations, shortening the response times by 564 ms and 540 ms, respectively. Similarly, icon signs shortened the response time by 878 ms relative to text signs. Color (Partial η2 = 90.3%) and format (Partial η2 = 96.3%) exhibited the strongest main effects on response time. Significant synergistic effects were observed among attributes: the red–triangle combination yielded the highest behavioral accuracy, outperforming blue and black combinations by 7.93% and 8.92%, respectively. Furthermore, the red–icon combination resulted in the fastest response time, reducing it by approximately 792 ms compared with blue–icon and black–icon combinations. Finally, shape and format interacted significantly regarding total fixation duration; triangles and circles attracted longer fixation durations in text format but shorter durations when paired with icons. Conclusions: The color, shape, and format differentially influence employees' visual attention orientation, sustained attention, and risk identification nudge effects among employees in oil and gas pipeline enterprises. Red and triangular attributes rapidly capture attention and enhance behavioral accuracy, text facilitates sustained attention, and icons accelerate behavioral responses significantly. Synergistic designs further amplify these nudge effects. This study recommends that the oil and gas pipeline industry treat supplementary safety signs as independent management units using optimal color–shape–format combinations. The findings offer theoretical and empirical support for optimizing safety signage in high-risk, large-scale industrial scenarios. Future research should validate the generalizability of these findings in real-world operational settings.
Objective: Liquid fuel leakage on sloped porous media, such as loess, is a common hazard in oil storage, transportation, and the chemical industry in the loess-covered areas of China. On inclined terrain, the gravity effect changes the seepage and the heat and mass transfer of liquid fuel inside the porous medium, which accelerates fire spread and expands the thermal influence range. Most existing studies on liquid fuel fire spread over porous media adopt quartz sand as the research medium, while the pore structure, permeability, and adsorption characteristics of natural loess are obviously different from those of quartz sand; therefore, the existing conclusions cannot be directly applied to loess fire scenarios. In addition, the coupling mechanism between base inclination angle and ignition position on fire spread behavior has not been systematically clarified. It is therefore of practical significance to explore the fire spread rules of kerosene-infiltrated loess under different inclination conditions, which can provide theoretical support for fire prevention, risk assessment, and emergency disposal in loess areas. Methods: In this work, a self-designed experimental platform with an adjustable inclination angle was adopted to carry out a series of fire spread experiments. Dry loess with a particle size range of 0.116–3.675 mm was paved evenly in the experimental tank to form a uniform porous medium bed. Kerosene was injected slowly into the loess bed in multiple small doses until the medium reached full saturation. Seven working conditions with equivalent inclination angles of −9°, −6°, −3°, 0°, 3°, 6°, and 9° were set by combining different base angles and ignition positions. Anhydrous ethanol was used as the ignition source in the designated ignition area to initiate combustion, and the total duration of fire spread was set to 20 min. A high-definition camera with a frame rate of 50 frames per second was arranged 100 cm away from the experimental platform to record the evolution of flame morphology and the spread process. Meanwhile, 12 K-type armored thermocouples with a probe diameter of 0.5 mm were arranged in a 6 × 2 array to synchronously measure the temperature distribution on the surface and inside the loess bed along the axial direction. All flame characteristics, spread velocity, and temperature data were collected and analyzed quantitatively. Results: The experimental results revealed that the coupling effect of inclination angle and ignition position exerted a prominent influence on flame structure and propagation. Under the same inclination angle, the flame height and propagation distance of lower-end ignition were greater than those of upper-end ignition. For the upward fire spread on positive inclination bases, the height of the yellow flame zone increased with the rise of inclination angle; for the downward fire spread on negative inclination bases, the height of the yellow flame zone gradually decreased as the inclination angle increased. The flame spread at a constant velocity under all working conditions, and the spread velocity rose monotonically with the increase of inclination angle. The velocity of upward spread was obviously higher than that of downward spread under the same absolute inclination value. When the flame front arrived at the measuring points, an obvious layered heat transfer feature was observed: the surface heating rate of loess was far higher than the internal heating rate. Compared with the horizontal base, the inclined base presented a lower steady combustion temperature, a shorter time to reach thermal stability, and a larger internal temperature gradient inside the loess bed. Conclusions: The base inclination angle changes the fuel seepage path and the intensity of heat feedback through gravity, and further regulates flame morphology, spread velocity, and the internal temperature field of the loess bed. The layered heat transfer characteristics of the loess bed are determined by differences in heat transfer mechanisms between the surface and the internal porous structure. The results clarify the mechanisms by which inclination angle and ignition position influence fire spread in kerosene-wetted loess. These findings can provide guidance for the design of fire isolation zones, the development of emergency response plans, and the optimization of firefighting strategies in sloped loess terrains. They also lay a foundation for further study on the evolution of fire hazards associated with liquid fuel leakage on loess substrates.
Significance: The control rod drive mechanism (CRDM) actuates the control rods to perform critical operational functions such as holding rod positions, incremental upward (step-up), downward (step-down) movements, and rapid drops (scram). Control rod assemblies typically include neutron-absorbing materials. By the precise adjustment of the axial positions of these assemblies within the reactor core, a nuclear reactor's reactivity can be effectively regulated. Therefore, the positions of the control rods serve as key indicators of the reactor's reactivity state and constitute one of the primary parameters requiring continuous real-time monitoring during operation. Rod position detectors are critical components that directly measure the actual positions of control rods, providing essential signals to the reactor control and protection system. Consequently, the accuracy and reliability of these detectors directly influence the safety performance of reactor operation. Progress: Inductive rod position detectors are the most widely used type in pressurized water reactors (PWRs) and small modular reactors (SMRs). Their configuration typically includes a primary coil and multiple secondary coil groups. As the control rod's measuring rod advances in discrete mechanical steps, the magnetic coupling between the primary and secondary coils varies, modulating the output voltage signals from the secondary coils. These detectors are specifically designed to interface with step-driven CRDMs. By encoding the voltage outputs from individual secondary coil groups, the detector produces discrete rod position "step" signals, each representing an integer multiple of the control rod's fundamental mechanical step distance. Accordingly, inductive rod position detectors provide only discrete positional information and cannot deliver continuous analog data. For reactors requiring precise rod position control, inductive detectors are inadequate due to their limited measurement accuracy. However, capacitance sensing technology has the potential to achieve continuous recognition of control rod positions, thus serving as a benchmark for precise regulation of nuclear reactor reactivity. Conclusions and Prospects: The nonuniformity of the sensitivity field is the key factor affecting the accuracy of capacitive rod position detectors. Measurement errors in detectors designed for SMRs are proportional to the product of the sensitivity field's nonuniformity and the insertion depth of the measuring rod. The nonuniformity of the sensitivity field depends on the radial displacement and the deflection angle of the measuring rod. Thus, to ensure the accuracy of capacitive control rod position detectors, the sensitivity fields must be properly designed and regulated. The sensitivity mechanism and rod position recognition model of the detectors are described herein, and the characteristics of the sensitivity fields are summarized. In addition, the design strategies for the sensitivity field are discussed. Future research directions for this technology are also proposed.
Objective: The control rod hydraulic drive system (CRHDS) is an innovative internal control rod drive technology developed by Tsinghua University for the low-temperature nuclear heating reactor NHR200-Ⅱ, in which the drive pump serves as the critical hydraulic power equipment. Its operational reliability is directly linked to reactor safety. However, the bypass circuit of the drive pump suffers from significant vibration and noise induced by the throttling of the control valve. Consequently, developing a high-efficiency multi-stage bypass energy dissipation and vibration reduction component is essential. This study aims to reduce pipeline vibration and enhance the operational reliability and safety of the CRHDS bypass loop through an optimization approach, thereby providing robust design criteria and theoretical guidance for bypass energy dissipation equipment. Methods: This study employed an integrated approach combining fluid-structure interaction (FSI) experiments, machine learning, multi-objective evolutionary optimization, and computational fluid dynamics (CFD) simulations. First, an FSI experimental loop simulating the CRHDS bypass was constructed, comprising a water tank, a centrifugal pump, adjustable flow resistance components, and pipe supports. Three series-connected ball valves were used to simulate a multi-stage flow resistance component, with system parameters recorded at a sampling frequency of 4 000 Hz using four fast-response pressure sensors, a three-axis pipe accelerometer, and an ultrasonic flow meter. A 125-group full-factorial FSI experiment was conducted using characteristic valve closing angles of 20°, 30°, 40°, 50°, and 60°. Second, a physics-informed neural network (PINN) surrogate model, evaluated via 5-fold cross-validation, was developed to predict the total pressure drop and synthetic vibration acceleration using the three-stage resistance coefficients. To improve prediction accuracy, a physical constraint stipulating that the total pressure drop increases monotonically with the sum of the resistance coefficients was integrated into the network's loss function. Third, the PINN surrogate was coupled with the non-dominated sorting genetic algorithm Ⅱ (NSGA-Ⅱ) to perform multi-objective optimization, targeting minimized synthetic vibration acceleration under rated pressure drop constraints. Based on the resulting Pareto optimal front, an optimized design was selected to fabricate a physical three-stage orifice component for experimental validation. Finally, CFD simulations were carried out to analyze the internal flow fields of the optimized component under high-temperature operating conditions. Results: The experimental, optimization, and simulation results indicated the following: 1) The loop vibration acceleration exhibited a non-monotonic trend, initially increasing and subsequently decreasing with increasing valve closing angles, with the final-stage valve angle exerting a dominant influence; 2) The prediction accuracy of the PINN model for the total pressure drop and synthetic acceleration improved by 27.9% and 29.4%, respectively, compared with the conventional radial basis function model; 3) The optimized multistage orifice component achieved an 84.9% reduction in synthetic vibration acceleration compared with the initial FSI tests while successfully satisfying the loop pressure-drop requirements; 4) Under the 230 ℃ high-temperature condition, the minimum pressure in the flow field was 3.58 MPa, which was considerably above the saturation vapor pressure of water (2.7 MPa), thereby ensuring an ample cavitation margin. Conclusions: By implementing an integrated framework of FSI experiments and multi-objective optimization, the optimal flow resistance configuration of multi-stage throttling components can be systematically determined to achieve vibration reduction and energy dissipation, thereby mitigating pipeline vibration induced by single-stage throttling. The optimized multi-stage orifice component achieved an 84.9% reduction in loop vibration acceleration under rated driving pressure and was verified to possess a sufficient cavitation margin under high-temperature reactor conditions.
Objective: With the rapid development of high-power electronic devices, microchannel thermal management systems have attracted widespread attention due to their high efficiency and compact structure. They are also widely used in digital polymerase chain reaction chips, fuel cells, and pharmaceutical microreactors. Owing to the multiphase flow in microchannels, such systems exhibit outstanding heat transfer performance. In particular, liquid–liquid slug flows achieve Nusselt numbers that are 400% higher than those of single-phase flows with better flow stability. However, manufacturing roughness inevitably and substantially changes the actual complex flow patterns and heat transfer characteristics. Most previous studies adopted ideal smooth-wall assumptions, thereby ignoring the effects of actual surface roughness on liquid–liquid slug flows and the coupled thermal behaviors of such flows in real-world applications. This study aims to systematically explore how random wall roughness affects slug flow dynamics and heat transfer in small-diameter co-flow microchannels, clarify the influence of key roughness parameters (including relative height and correlation length), and further reveal the underlying mechanisms to support effective engineering design and optimization of microchannel-based thermal management devices. Methods: A two-dimensional axisymmetric numerical model was built in Fluent using the fixed reference frame and the volume-of-fluid interface-capturing method. The continuous surface force model was used to simulate interfacial tension, and the pressure implicit with splitting of operator (PISO) algorithm was adopted for pressure–velocity coupling. Random wall roughness was generated via a Gaussian distribution approach to mimic real machining errors. Grid independence and model reliability were verified by comparing simulation results with published experimental data on flow velocity, liquid film thickness, and Nusselt number. Toluene and water were used as working fluids with equal superficial velocities of 0.125 m/s, and a constant heat flux of 50 kW/m2 was applied to the channel wall. A fast Fourier transform was adopted to analyze the slug generation frequency. The flow velocity field, slug movement characteristics, temperature distribution, and heat transfer performance under different roughness parameters were systematically studied. Results: Rough walls reduce the effective flow area and increase the internal flow velocity significantly. The maximum slug generation frequency in rough channels was 16.3% higher than that in smooth channels. The developed wall area of rough surfaces was up to 30% larger than that of smooth walls, greatly enhancing the heat transfer interface. The wall temperature and average fluid temperature were notably higher in rough microchannels. Roughness height has a more prominent effect on flow and heat transfer than correlation length, whose influence saturates when the roughness distribution reaches a certain density. Slugs in rough channels travel at least 6% farther than those in smooth channels over the period of 6.0–26.5 ms, and the heat transfer coefficient increases by up to 12% as roughness height increases. Conclusions: Random wall roughness accelerates slug formation and enhances heat transfer by narrowing the flow area and disturbing the near-wall flow field. Roughness height plays a dominant role in regulating flow and thermal performance, while correlation length has a limited and saturable effect. These findings provide clear theoretical support and practical guidance for the design and optimization of microchannel thermal management devices and can guide the formulation of reasonable machining tolerance standards to balance manufacturing cost and thermal performance. Furthermore, the findings can help in properly controlling surface roughness to achieve better thermal efficiency in real industrial applications, laying the foundation for further research on roughness-optimized microchannel structures.
Objective: Flexible interconnection technology can realize asynchronous closed-loop operation among multiple alternating current distribution feeders. It can support dedicated power flow transfer, balance feeder loading, reduce network losses, and enable optimal allocation of controllable resources. The thyristor-controlled hybrid transformer (TCHT) can be used to realize flexible interconnection in distribution networks, offering good economic performance and high reliability. In device operation and control, the power-transfer command—whether obtained from an optimal power flow algorithm or specified manually based on engineering experience—must be converted into voltage compensation tap positions that are executable by the TCHT. However, in practical engineering, the accurate determination of the equivalent circuit parameters of lines, transformers, and other components is challenging. Existing studies have shown that a feedback mechanism can be introduced for a single TCHT, whereby the compensation voltage that minimizes power transfer error can be determined through a shortest-path search method. However, this method is difficult to extend to coordinate control when multiple TCHTs are coupled through the network topology. Methods: This study proposed a coordinated operation and control method based on multiagent reinforcement learning (RL). First, the coordinated control of multiple TCHTs was modeled as a Markov decision process (MDP). The system state space was defined as the local power regulation error of each TCHT. The action space was defined as the neighboring points of the current three-phase voltage compensation tap position of the TCHT, which avoids the convergence complexities of the training and control processes caused by an excessively large action space. The reward function was defined as the maximum regulation error. Consequently, the original problem was transformed into an equivalent MDP. Second, an online solution method based on multiagent RL was developed. A parameterized policy function was adopted to handle the continuous state space. Generalized advantage estimation was applied to reduce the approximation variance of the policy gradient, after which policy parameters were updated via the proximal policy optimization algorithm. Based on this, a coordinated policy training algorithm for multiple TCHTs was developed. Guided by the global reward function, this method enabled coordination among different TCHT devices. Results: Numerical analyses were performed on typical public network topologies. A flexible simulation platform was constructed for the interconnection distribution network based on Python and pandapower, with three homogeneous TCHT devices installed in the process. The results showed that the policy iteration process of the proposed method was stable, with only small fluctuations. The reward value increased significantly from iterations 0 to 50. From iterations 50 to 350, the value converged gradually to the optimum. From iterations 350 to 400, it remained basically stable around the optimum. Additionally, each decision step required only one forward pass of the neural network. On an Intel i7-13700 processor, each computation required only 10 ms on average, meeting real-time requirements. Compared with the independent shortest-path search method, the proposed method reduced the average regulation steps and error by 36.6% and 58.9%, respectively. Conclusions: These results show that although existing methods cannot coordinate multiple TCHTs within a common distribution network in the absence of power grid parameters, our algorithm significantly improves the accuracy and efficiency of coordinated control. Thus, this study fills the methodological gap in the collaborative control problem of multiple TCHTs.
Objective: Large-animal single-photon emission computed tomography (SPECT) systems are crucial for preclinical cardiovascular research. An accurate system matrix (or system response matrix) is essential for high-quality iterative image reconstruction. However, directly measuring the system matrix on large-animal SPECT systems is often prohibitively time-consuming due to the extensive field of view and high spatial resolution needs. To address the lengthy measurement process for the system matrix in large-animal SPECT, this work proposes and validates a calculation method based on two-dimensional (2D) Gaussian fitting. This method leverages the inherent continuity of the projection probability density function (PPDF) in the image domain. Instead of measuring the system response for each voxel, the proposed approach demonstrates that an accurate system matrix can be built by acquiring only sparse point-source measurement data. Using 2D Gaussian fitting, the method effectively models the system's spatial response and connects sparse data points to synthesize the full matrix. Methods: The complete dataset of point-source projections for 37 210 voxels was collected on the large-animal SPECT system, with a total acquisition time of 124 h. Down-sampled subsets at ratios of 1/4 and 1/9 were created from the full dataset to simulate accelerated protocols and to thoroughly evaluate the feasibility and robustness of the proposed method. The accuracy of the fitted PPDFs was quantitatively assessed against the fully measured ground truth using two metrics: the relative root mean square error (RRMSE) and the structural similarity index measure (SSIM). Additionally, to evaluate the reconstruction performance of the system matrix derived from the fitted PPDFs both qualitatively and quantitatively, 3.5 and 4 mm hot-rod phantom images were reconstructed. The performance of the proposed method was compared with several traditional approaches, including fully sampled direct measurement, barycentric Lagrange interpolation, and cubic spline interpolation. Results: Quantitative analyses showed exceptional fidelity in the estimated system matrices. The matrix computed under the 1/4 sparse sampling condition achieved an RRMSE of 0.023 ± 0.068 and an SSIM of 0.997 ± 0.003. Even with the more aggressive 1/9 sparse sampling, the method produced an RRMSE of 0.035 ± 0.095 and an SSIM of 0.996 ± 0.006. The system matrix generated with this method successfully resolved 3.5-mm hot rods. Under the 1/4 sparse sampling, the reconstructed images of 3.5 and 4 mm rods had RRMSE values of 0.795 and 0.654 against the ground truth, and SSIM values of 0.981 and 0.988. For the same sampling, the reconstructed images showed RRMSE values of 1.042 and 0.797, with SSIM values of 0.971 and 0.981. These imaging results were visually and quantitatively superior to those obtained with other methods. Importantly, the computational time remained efficient, with calculations taking only 1.9 hours and 1.8 hours for the 1/4 and 1/9 sparse datasets, respectively. Conclusions: The proposed 2D Gaussian fitting approach effectively overcomes the traditional limitations in generating system matrices for large-animal SPECT systems. It significantly reduces measurement time and computational costs without sacrificing tomographic image quality. This method presents a practical and efficient solution for acquiring precise system matrices in large-animal SPECT imaging.
