There are three independent control degrees of freedom in dual active bridge DC-DC converters (DAB) with triple phase shift (TPS) control. The current stress in a DAB can be reduced by a reasonable combination of changes in the phase shift degrees of freedom. However, soft switching analyses show that the system efficiency is not necessarily optimized when the DAB current stress is optimized. This paper presents a performance comprehensive optimization method that considers both the current stress and the efficiency based on current stress optimization control. Firstly, a DAB loss model for TPS control is developed based on the switching combination laws. A commutation loop analysis is used to separately predict the on-state losses and switching losses of the semiconductor devices and the copper losses and iron losses of the magnetic component for any combination of switches. Then, an efficiency optimization weight is introduced to construct the optimal objective function balancing the current stress and efficiency. A two-dimensional traversal algorithm is then used to optimize the DAB working point. The DAB performance is compared to measurements when the current stress or efficiency is optimized separately. Proper selection of the weight increased the prototype efficiency when transmitting 6 kW by 0.83% with the current stress only increased by 0.5 A. Thus, this method effectively improves the system efficiency without substantially increasing the current stress.

As the power level of regional energy routers increases, the router topologies and control strategies are becoming more complicated. Simulations based on the complete circuit model are very inefficient. This article presents a high-efficiency equivalent transient model method for regional energy routers that is much more efficient than previous models. The model decouples the various stages to divide the large complex circuit network into several small sub-networks with the Thevenin equivalent method applied to each module and module combination to reduce the number of nodes in each sub-network. This significantly reduces the order of the nodal admittance matrix which improves the simulation efficiency. This method is applied to a regional energy router to model steady operation, load switching or load failure conditions. This method can also be applied to regional energy router simulations when connected to the gird. Simulations of a 10 kV/1 MW four-port regional energy router show the high computational speed and accuracy of this method.

Boost converters are widely used for photovoltaic power generation, energy storage and electric vehicles. However, since the right half plane of the Boost converter is zero in the frequency domain, traditional PI control algorithms for the converter have serious dynamic response limitations which can result in large output voltage fluctuations leading to over or under voltage faults during fast transitions with large load changes. Predictive control can improve the dynamic response and avoid control parameter tuning with additive system constraints, but predictive control can lead to output voltage errors at steady state. A load-current sensorless sliding-mode-predictive control algorithm was developed here to reduce the output voltage fluctuations during sudden load transitions while also maintaining the proper steady-state characteristics. The outer loop uses a sliding surface to generate the inductor current reference while the inner loop with a deadbeat predictive control regulates the inductor current. The load current is estimated by a sliding mode observer. This control algorithm reduces the output voltage fluctuations during load transitions and the transition times. The algorithm also limits the output voltage errors at steady state. In addition, the algorithm does not require a load current sensor. Tests with a Boost converter confirm the effectiveness of this control strategy.

The synchronous acquisition and analysis of information about arc shape and current variations during the polarity commutation process of variable polarity tungsten inert gas (VP-TIG) arc welding are helpful to understand the arc dynamics mechanism. However, the polarity commutation process is relatively short, which leads to higher accuracy requirements for synchronous acquisition of the current information and arc image, as well as the acquisition frequency and image resolution of the image acquisition equipment. In this paper, the synchronous input and output functions of a data acquisition board were related to the synchronous trigger of a high-speed camera to realize synchronization acquisition of the welding current and arc image during polarity commutation with the synchronization error controlled to within 2 μs. In addition, use of the difference between the trigger voltage pulse and variable polarity periods enabled approximately continuous observation of the arc images with a lower acquisition frequency and higher image resolution, which reduces the camera speed requirement. The observations of the welding arc changes with the current showed that the arc shape variation lags behind the current change during the polarity commutation process and arc extinction does not occur at or around the time when the current value passes zero. In addition, the arc with the lowest gray level appears in the current recovery stage after the polarity commutation process. The results indicate directions for further improving the arc burning stability of VP-TIG arc welding.

To improve the wire and arc additive manufacturing (WAAM) numerical simulation efficiency of aluminum alloy cylindrical walls, a segmented cambered body heat source model was developed with its mathematical derivation clarified based on segmentation on circular path of the Goldak body heat source with secondary development on ABAQUS. The moving heat source, the material models and the boundary conditions were verified in a finite element model with the numerical results agreeing well with experimental data. The accuracy and efficiency of the WAAM numerical simulations were evaluated using three segmented computing strategies. Comparisons of the results using the three segmented computing strategies with those using the moving heat source show that the thermal cycles are similar with peak temperatures having less than 8% errors, the residual stress distributions are consistent and the results are not sensitive to the number of segments in the segmented sources. The use of one, four or eight segmented heat sources reduces the total computational time by 98.24%, 77.51% and 65.96%.

Space manipulators need to pass tests in zero-g simulation systems before being used in orbit. The unloading ratio of zero-g simulation systems was evaluated here using a simulation manipulator with force and torque sensors. A dynamic model of a seven-joint manipulator was then developed for operation in a zero-g environment. The joint torques were calculated and used as the evaluation index. A continuous friction model was used to identify the friction parameters for the system with the friction characteristics then modified so that the predicted torques matched the experimental data. The results show that the theoretical model can accurately predict the torques and the space manipulator model parameters without sensors when operating in a zero-g environment to guarantee the safety of the experiments.

The dynamics characteristic of parallel mechanisms are very important. This study analyzed the dynamics of a 2UPU/SP parallel mechanism in a hybrid robot for aerospace composite machining. A kinematics model was developed using the virtual work principle for the three-degree-of-freedom parallel mechanism including the plane constraints of the two UPU kinematic branches. The maximum acceleration of the moving platform when one driving force is a unit force and the other driving forces are less than or equal to the unit force was used to evaluate the parallel mechanism. This performance index was used to separately evaluate the effects of linear acceleration and angular acceleration of the 2UPU/SP parallel mechanism compared with a traditional 2UPS/UP parallel mechanism. The results show that the present 2UPU/SP mechanism has better acceleration characteristics.

This paper presents lattice Boltzmann method (LBM) analyses of three-dimensional, particle-fluid two-phase flows based on the MPI+OpenMP mixed programming model where the cluster nodes used the MPI parallelism with further fine-grained nodes using the OpenMP parallel calculation. The algorithm is optimized according to the characteristics of particle-fluid flows. The algorithm was then used to analyze a particle deposition problem to test the acceleration capabilities of the parallel algorithm. The results show that the algorithm has good speedup and scalability and that its computational complexity is insensitive to the number of particles, which is good for large-scale two-phase flow analyses.

The safe operation and management of metro systems are related to the metro station design and the passenger flow rates into and out of stations and between stations. This study examines the passenger flow rates in and out of each station at various times on various days by filtering and standardizing the automatic fare collection (AFC) system data of the Shenzhen metro system. The data was analyzed using a principle component analysis (PCA) and a Gaussian mixture model (GMM) to cluster passenger movement data on weekdays and weekends to simplify analysis of the passenger flow patterns. The results were then used to identify time periods and types of metro stations having greater risks as a passenger flow risk analysis method for metro stations. This big data analysis can identify risk conditions for large passenger volumes to avoid mass incidents like stampedes so as to protect passenger safety.

3-D radiation field data is used for dose evaluations and design optimization. This paper presents a mesh generation method that uses variable weights based on gradient changes in a radiation field to improve the calculational speed and 3-D visualizations of the radiation field. The 3-D mesh has different densities that are arranged based on the radiation field gradients to balance the computational accuracy and efficiency. The point kernel integral method was used to verify the method. The results show that this non-uniform mesh generation method with variable weights effectively shortens the computing time for the point kernel integral method results. This method also provides better visualization of the radiation field. Thus, this study reduces the calculational time for the point kernel integral problem and gives better results for visualization and dose assessments of 3-D radiation fields in nuclear facilities.

Inaccurate subbasin definitions will affect soil and water assessment tool (SWAT) model results in plain irrigation districts. The SWAT model watershed delineation method was modified by preprocessing of digital elevation model (DEM) and joint use of automatic and pre-defined methods. This improved method was then applied to the Ninth Drainage Ditch Watershed of the Hetao Irrigation District in Inner Mongolia, China. The results show that the improved method generates specific subbasins corresponding with specific levels of the drainage ditches. This method increases the subbasin boundary agreement rate by 61% compared with the original automatic delineation method, improves the river agreement rate by 4% and improves the total watershed coverage rate by 4%. The hydrological simulation results of the improved SWAT model agree well with the measured data; thus, this method is more practical for the special management requirements of plain irrigation districts.