[Objective] Anthropogenic heat is a crucial feature of urbanization and human activities. Large amounts of anthropogenic heat emissions can affect the regional land-atmosphere process and exacerbate the local climate change in cities. In winter, the anthropogenic heat emissions caused by central heating increase significantly in the North China, thereby enhancing the urban heat effect. The urban heat effect may have a substantial impact on winter snowfall. Therefore, studying the impact of anthropogenic heat on urban winter precipitation can help improve our understanding of the impact of urbanization and human activities on urban precipitation.[Methods] Here, the weather research and forecasting (WRF) model was used to analyze the impact of anthropogenic heat on urban winter precipitation by simulating a winter precipitation event in Beijing. To study the impact mechanism of anthropogenic heat on the urban water and thermal environment in winter and the possible threshold, sensitivity experiments were conducted using the default anthropogenic heat built in the WRF, which is an ideal experiment to explore the possible threshold of anthropogenic heat magnitude. By setting built-in anthropogenic heat values of different intensities, the impact mechanism and intensity threshold of different anthropogenic heat intensities on urban water and thermal environment in winter were compared and analyzed. Based on the conclusion of the sensitivity experiment, anthropogenic heat with high spatiotemporal resolution calculated using the large-scale urban consumption of energy (LUCY) model in Beijing was used as the input data for WRF. The WRF-LUCY coupling model was applied to investigate the impact of anthropogenic heat on winter precipitation in Beijing. By setting up simulated scenarios with and without anthropogenic heat, the impact of anthropogenic heat on winter precipitation was compared and analyzed. Furthermore, the results of the WRF-LUCY coupling model were compared with the existing effects of anthropogenic heat on summer precipitation, and the differences between summer and winter precipitation are summarized.[Results] The following results were obtained:1) Anthropogenic heat primarily affects the phase state of mixed precipitation and has little effect on single-phase precipitation. With the increase in anthropogenic heat, snowfall and rainfall gradually decrease and increase, respectively. 2) No evident threshold exists for the impact of anthropogenic heat increase on air temperature; however, a threshold for the impact height, with a maximum impact height of about 1.3 km, is noted. 3) The increase in anthropogenic heat intensifies the vertical wind shear and water vapor convergence, thereby increasing the water vapor content in the boundary layer and providing favorable developmental conditions for the convective system. 4) For summer precipitation, the effect of anthropogenic heat demonstrates large scope and magnitude, and the impact mechanism is relatively complex; in contrast, the impact on winter precipitation is primarily concentrated in the urban interior, mainly affecting the precipitation phase.[Conclusions] A threshold for the impact height of anthropogenic heat is found, while no evident threshold for the anthropogenic heat magnitude is observed. Compared with the complex mechanism of summer precipitation, the impact of anthropogenic heat on winter precipitation is more concentrated. This study analyzes the impact of anthropogenic heat on urban winter precipitation, which has a crucial scientific application for exploring the urban land-atmosphere process and improving the urban environment.

[Objective] Water discharge prediction in canals under complex conditions is a fundamental problem with prominent practical significance in improving farmland irrigation water efficiency, conserving water resources, and reducing involved costs. The state-of-art solution of prediction is establishing nonlinear partial differential equations with numerical calculation methods, with time cost being exponential to the fineness of the spatiotemporal division. Moreover, the current time step calculation depends on the result of the last time step, i.e., the calculation cannot be parallelized, which results in a tradeoff between accuracy and efficiency. In actual irrigation areas, the control of gate openings in canals primarily relies on human experience, which has an extremely long feedback process. Therefore, it is challenging to employ human experience and numerical calculation methods when multiple gate changes are required. The rapid development of artificial intelligence-related technologies has yielded more opportunities for modernizing conventional industries. In this study, the input and output were definite for the water discharge prediction task, which corresponds to the "regression" problem-one of the two types of fundamental problems that neural networks are good at solving. This study presents new insights to leverage the neural network to solve the water discharge prediction problem end-to-end. The neural network only needs to be trained once, and further, multiple results can be obtained with high efficiency during testing. Therefore, the proposed approach overcomes the shortcomings of the conventional methods, which involve extremely high time costs.[Methods] Based on the Internet-of-Water theory of "real-time perception, water-information interconnection, process tracking, and intelligent processing", this study introduced a novel approach for water discharge prediction. First, we investigated the sequence features of the upstream and downstream canal water discharge gate control and introduced the static features of the gates and canal. Second, we proposed a novel predicting method for canal discharge based on a long short-term memory (LSTM) neural network, in which the gating mechanism allows better modeling and prediction of problems with sequential information. Feature discretization and normalization were applied to the static features to improve the generalization ability of the model to predict unseen data. Layer normalization was performed on the output of the LSTM network to adjust the distribution of the output to the unsaturated region of the activation function, making the neural network more sensitive to the input and output, as well as accelerating its convergence.[Results] The following comparative experimental results were obtained:1) The proposed model can complete the prediction task with an accuracy rate exceeding 97% in every canal segment, which is significantly better than all baselines, indicating the effectiveness of using the hidden sequence features inside the canal and the gating mechanism of the LSTM neural network. 2) Under normal circumstances, introducing static features as part of the model's input improves the prediction performance. 3) The proposed model demonstrates good robustness. It successfully learns and shows good prediction performance without too much data fed into it. Hence, it is extremely useful in situations of data shortage and when requiring model migration to other canals. 4) Compared to the conventional numerical calculation method, the proposed model demonstrates 308 times higher prediction efficiency, reducing the prediction time from 950 h to about 3 h on 100,000 pieces of data.[Conclusions] This study verifies the feasibility of artificial intelligence-based methods in improving the conventional canal discharge prediction problem, achieves a win-win situation between accuracy and efficiency through a reasonably designed deep learning model, and provides a new idea for applying artificial intelligence-based methods in solving hydraulic problems.

[Objective] The parameterization schemes for the surface solar radiation effect in the climate model primarily include the plane-parallel radiative transfer scheme, two-dimensional (2-D) solar radiation effect scheme, and three-dimensional (3-D) solar radiation effect scheme. Because the plane-parallel radiative transfer scheme assumes a flat topography, it cannot ascertain the impact of terrain on radiation, which results in systematic biases in simulating the land surface processes of the climate model. The 2-D scheme only considers the impact of the terrain slope and aspect on the solar incident angle. However, it overestimates the surface radiation during early morning and late afternoon, leading to systematic biases in simulating precipitation over mountainous areas. In contrast, the 3-D scheme is based on the solid physical foundation of the mountain radiation theory, which considers the radiation impact of 3-D structures along the sub-grid terrain on the surface radiation. The weather research and forecasting (WRF) model incorporating the 2-D scheme in simulating land surface processes also encounters problems related to the precipitation overestimation and cold bias over the Tibetan Plateau.[Methods] This study coupled the WRF model with the 3-D scheme to accurately represent surface radiation processes over complex terrains such as the Tibetan Plateau, thus improving the model performance in simulating surface energy balance and precipitation. Based on the mountain radiation theory, the WRF model was incorporated with the following three types of incoming solar radiation:the solar radiation flux, diffuse radiation flux, and solar radiation flux reflected by the surrounding terrains. The influence of local (small-scale) topography on the land-atmosphere process was considered from the perspective of thermodynamics to improve the simulation ability of a regional climate model for precipitation over the plateau. Accordingly, a high-resolution WRF simulation study was conducted in the Yadong Valley, located in the central Himalayas.[Results] The results showed that:(1) The improved scheme reduced the precipitation overestimation in the Yadong Valley. The simulated surface solar radiation and precipitation were closer to the measured data. The precipitation simulated using the default scheme and that using the improved scheme were 15.7 and 12.1 mm, respectively. In addition, the correlation coefficient between the precipitation simulated using the improved scheme and the measured data was 0.46. Moreover, the improved 3-D scheme reduced the overestimation of simulated solar radiation during sunrise and sunset and increased the surface solar radiation at noon, while the correlation coefficient between the solar radiation simulated using the improved scheme and the measured data was 0.95. (2) The improved 3-D scheme better reflected the impact of complex terrains on the surface radiation distribution. During the daytime, the mountain slope was heated by a large amount of incident radiation (0.09- 0.20℃). The generated updrafts and subsequent formation of uphill winds (0.02-0.08 m/s) on the slope carried the water vapor in the valley to the upper slope. The water vapor rose and formed precipitation over the slope, which resulted in a decrease in precipitation in the valley (-4.54- -3.34 mm) and an increase in precipitation on the upper slope (0.59-2.82 mm).[Conclusions] This research can explain the mechanism behind the influence of terrain on precipitation in mountainous areas, making it a reference for future research in precipitation simulation.

[Objective] Shallow soil in the Loess plateau has undergone significant physical changes as a result of the extensive grain for green project, which aims to reduce soil erosion and surface runoff. These changes have altered the hillslope-valley hydrological responses, which are vital processes influencing slope stability. Many studies shows a potential correlation between vegetation restoration and shallow landslides on the Loess plateau. By increasing water infiltration and the vertical variability of soil physical properties, vegetation restoration both induces landslides and increases the volume of landslides. Moreover, shallow landslides have a negative effect on the basin's efforts to restore its ecological balance by sharply increasing the runoff rate and sediment yield rate on the slope surface. However, the impact mechanism of vegetation restoration type and degree on shallow landslides in the Loess plateau is undiscovered, especially how the vertical heterogeneity of shallow soil caused by vegetation restoration affects the occurrence of landslides and the depth of the sliding surface. Therefore, the only way for soil and water conservation and ecological restoration in the new period is to explore the influence mechanisms of vegetation on shallow landslides in the Loess plateau and reduce shallow landslides as much as possible while retaining soil and water.[Methods] To elucidate the impact of vegetation-restoration-induced vertical heterogeneity of soil-saturated hydraulic conductivity on shallow landslides in the Loess plateau, a three-dimensional finite element mesh is established in the Loess plateau's Shejiagou catchment, and simulation accuracy is improved through model calibration and verification of infiltration, runoff, and soil water movement. Forty scenarios are run based on the saturated soil hydraulic conductivity measured under different vegetation restoration conditions, combined with the rainstorm and continuous rainfall processes. The Integrated Hydrological Model is used to simulate the processes of precipitation-infiltration and runoff production, and the soil moisture variables at any time and in any position are coupled to the infinite slope stability model to calculate the slope stability.[Results] The simulation results showed that:(1) As infiltrated water accumulated in a certain soil layer determined by the heterogeneity of soil-saturated hydraulic conductivity, slopes covered by grasslands and shrubs were more unstable than slopes of bare soil, and the risk of landslide under a single rainfall storm increased with the vegetation recovery period. (2) The high soil-saturated hydraulic conductivities in all soil layers which led to quick vertical and lateral drainage, lowered the risks of shallow landslide on forest-covered slopes. (3) The high antecedent water content caused by continuous rainstorm infiltration significantly increased the risks of slope instability in terms of landslide depths and volumes, especially on forest-covered slopes with deeper infiltration paths. (4) The greater the vertical heterogeneity of saturated hydraulic conductivity between layers, the more conducive to water accumulation between layers, and the occurrence of slope instability.[Conclusions] These results reveals the impacts of vertical heterogeneity of soil-saturated hydraulic conductivity caused by vegetation plant restoration on slope stability, which can serve as theoretical guidance for future vegetation restoration as well as soil and water conservation on the Loess plateau.

[Objective] A self-paced treadmill (SPT) is key human-robot interactive equipment for virtual reality, which can enable a user to walk at the intended speed by using a re-positioning technology. Realizing multimode interactions of SPTs is crucial for enriching their applications. However, existing studies only realizes a few interaction modes. To realize multimode interactions in self-paced treadmills, a novel multilayer control framework is proposed in this paper.[Methods] In this study, the control system is divided into two layers:the recognition layer and the control layer. First, a novel hybrid spatial-temporal graph convolutional neural network is proposed to realize user-independent human locomotion mode recognition based on plantar pressure insoles in the recognition layer. The proposed network separates the pressure and acceleration signals and dealt with them individually. It is also utilized to extract the natural spatial topology between pressure nodes. Long short-term memory layers are used to individually extract temporal-dependent features of pressure and acceleration signals and to fuse multimodal features for final recognition. A multilayer perceptron is utilized to map the fusion features to the locomotion modes. By extracting the natural spatial-temporal features of multimodal data during human locomotion, a high generalization capability of the recognition results can be expected. Second, control strategies for different locomotion modes are designed in the control layer according to the stability condition of different human locomotion modes. Meanwhile, a walking speed feedforward control strategy is proposed to re-position the user and ensure natural gaits for the walking mode. Variable gain control strategies are adopted to manipulate the acceleration for the running and back walking modes. A buffer control strategy is proposed to improve the stability during jump landing for the jumping mode. Then, a finite state machine is used to automatically switch the control strategies. The states are transited based on the recognition results.[Results] 1) The proposed locomotion mode recognition method was evaluated on a dataset that comprises eight subjects with five locomotion modes through the leave-one-subject-out cross validation. Then, it was compared with the convolutional neural network (CNN) and domain-adversarial neural network (DANN). Experimental results indicated that the mean and standard deviation classification accuracies of the CNN, DANN, and HSTGCN are (90.26±8.54)%, (97.71±3.60)%, and (97.37±1.40)%, respectively. These results validated that the proposed method can achieve high generalization capability without any dependency on the data of target subjects. Hence, the burden of repeated data collection and network training was reduced. 2) Based on the recognition results, experiments on the multi-locomotion mode human-robot interaction were conducted using a finite state machine. Experimental results indicated that a user can freely change the locomotion modes on the treadmill, and the balance was not significantly affected by the treadmill acceleration.[Conclusions] The proposed framework can automatically combine the recognition results with the treadmill control and can realize the control of multi-locomotion mode human-robot interactions. Further, experimental results validate that the proposed multilayer control strategy can achieve a stable and smooth multi-locomotion mode human-robot interaction, ensure natural gaits and posture stability of the user, and meet the requirements of multi-locomotion mode human-robot interactions for self-paced treadmills.

[Objective] Dissimilar metal welds (DMWs) between martensitic heat-resistant steels and nickel-based alloys filled with nickel-based filler metals are widely used in advanced ultra-supercritical power plants. A large composition gradient is present at the interface between weld metals and martensitic heat-resistant steels due to the obvious difference in chemical composition between these two materials, resulting in an abrupt change in microstructure and mechanical properties across the interface. For DMWs exposed to creep conditions, interfacial failure is a commonly seen failure mode that reduces the creep life to less than half of the expected life. To eliminate the interface with a large composition gradient, a graded transition joint (GTJ) between the martensitic heat-resistant steel (named COST E) and 617B nickel-based alloys is fabricated using a dual-wire tungsten inert gas (TIG) welding technique in the present study.[Methods] The key part of the GTJ is a functionally graded material (FGM) in the middle, of which the chemical composition varied gradually from martensitic heat-resistant steels to 617B nickel-based alloys over a distance of 14 mm or less. During fabrication, the feeding rates of the two wires are varied in a controlled manner to obtain the desired dilution rates. After FGM fabrication is completed, the two ends of the FGM are joined by similar welds with corresponding base metals, thus fabricating a GTJ between COST E steels and 617B nickel-based alloys. Optical microscopy, scanning electron microscopy, energy dispersive spectrometry, electron probe microanalysis (EPMA), electron back-scattered diffraction, and a microhardness tester are used to investigate the chemical compositions, and the microstructure and hardness of the GTJ in as-weld condition are characterized. The dynamic kinetics module of Thermo-Calc, DICTRA, is used to investigate why mixed austenite (A)+martensite (M) formed based on the Scheil solidification equation.[Results] The results showed that the chemical composition gradient was greatly reduced compared with conventional DMWs, as expected, and the microstructure from the steel side to the nickel-based side varied from quenched martensite, mixed A+M, and finally, full γ nickel-based microstructure. A hardness peak as high as 500 HV0.2 was found in the quenched martensite region, and hardness decreased sharply to lower than 200 HV0.2 once entering the martensite and austenite dual-phase region, followed by a gradual increase to 250 HV0.2 or less in the full γ nickel-based microstructure region. The dynamic kinetics module revealed that Ni, Cr, and Mo tended to segregate into interdendritic regions during solidification.[Conclusions] Ni, Cr, and Mo segregation are due to the reason that the equilibrium partition coefficients of the three elements are less than 1, meaning that these elements are higher in concentration in the liquid than in the solid at the liquid/solid interface. Therefore, the concentrations of these elements in the newly formed solid are higher than those in the previously formed solid. The segregation of these elements in interdendritic regions lowers the martensite transformation starting temperature Ms below the ambient temperature, and thus, the austenite in interdendritic regions is stable even at room temperature, while the austenite in the dendrite core region transforms into martensite in the following cooling process, thus forming the A+M dual-phase region.

[Objective] With the increasing of the disabled elderly population, the demand for in-bed rehabilitation robots increases. However, the clinical utilization rate of in-bed rehabilitation robots remains low because biomechanical studies on lying posture rehabilitation training are few. The function of an in-bed rehabilitation robot is relatively simple. However, its rehabilitation efficiency should be improved. Therefore, this study aimed to evaluate the joint motion and muscle activation with different movements of lying posture and provide a theoretical basis for designing the motor function of lower-extremity rehabilitation robots.[Methods] We designed a measurement experiment of three typical in-bed rehabilitation training movements, including cycling and straight leg raising in supine and lateral decubitus positions. Furthermore, different variables of velocity and amplitude/distance were set for each movement. Ten healthy subjects performed three movements during the experiment. Kinematics data were collected using a Vicon motion capture system and electromyography data were collected using a Noraxon electromyography acquisition device. A musculoskeletal model for the simulation of supine motion was developed using the software OpenSim. This model included 23 degrees of freedom and 92 muscles of the trunk and lower limbs, which could simulate a larger range of hip and knee flexion than the usual models. Further, a weld constraint was added between the trunk and the ground in the musculoskeletal model to compensate for human-ground contact force. Kinematics data were then imported into the OpenSim model for model scaling, inverse kinematics, and static optimization calculation steps. Then, joint angle and muscle activation were obtained. Electromyography data were compared to the simulation data to verify the musculoskeletal model's reliability.[Results] The OpenSim model was confirmed to be reliable and accurate for simulation. Cycling in supine position showed a higher range of motion (ROM) in the knee and ankle. However, the overall muscle activation was lower than that of the other two movements. Additionally, the greater the movement's ROM during cycling, the higher the muscle activation. Concurrently, the subjects' translation of the center of mass relative to the ground became larger, which should be avoided during patients' in-bed rehabilitation. Straight leg raising in supine position improved hip flexion ROM and activated related muscle groups, such as iliopsoas, sartorius, and rectus femoris. Straight leg raising in lateral decubitus position improved hip abduction ROM and activated related muscle groups such as the gluteus maximus, gluteus medius, gluteus minimus, piriformis, and tensor fascia lata. Muscle activation became higher when subjects lifted their leg faster. However, the subjects' displacement of the center of mass relative to the ground became larger when they lifted their leg faster. Overall, when the angle of the leg lift increased, the mean value of muscle activation decreased and subjects' displacement of the center of mass relative to the ground increased.[Conclusions] Three typical in-bed rehabilitation movements have different benefits to the joints and muscles. Various movement combinations in supine and lateral decubitus positions can improve the rehabilitation effect in clinical training. The rehabilitation robot should provide more sagittal and coronal rehabilitation training functions.

[Objective] In the past decades, the lifestyle of the Chinese population has changed because of rapid economic development. The incidence and prevalence of diabetes are increasing in China. Diabetes and its complications can cause substantial personal and social burdens. Recent experimental studies have shown that cold exposure may enhance insulin sensitivity and prevent the development of obesity-related insulin resistance and hyperglycemia. However, there is insufficient evidence of the correlation between cold exposure and diabetes in the real environment of daily life. In contrast, although people spend most of their life indoors, existing evidence usually focuses on the relationship between the outdoor thermal environment and diabetes without considering the effect of the indoor thermal environment. This study preliminarily examines the association between the real indoor thermal environment during the winter and the risk of diabetes.[Methods] The study was analyzed from provincial and individual perspectives. At the provincial level, the analysis included 17 provinces. The diabetes prevalence data were obtained from the cross-sectional study on diabetes prevalence in China through Thyroid Disorders, Iodine Status, and Diabetes Epidemiological Survey (TIDE). The level of cold exposure was calculated based on the prevalence of central heating, according to the statistical yearbook data of the country and each region. Spearman correlation analyses were conducted to reveal the correlation between cold exposure and diabetes. The data were obtained from the China Health and Retirement Longitudinal Study (CHARLS) database at the individual level. Subjects' age, gender, body mass index (BMI), smoking, drinking, socioeconomic status, and whether they experienced indoor cold exposure in the winter were considered. Subjects were divided into two categories based on whether they experienced indoor cold exposure in the winter from 2011 to 2015. Chi-square and logistic regression analyses were applied to explore the correlation between indoor cold exposure and diabetes at the individual level.[Results] At the provincial level, the results indicated a significant negative correlation between indoor cold exposure in the winter and diabetes prevalence (Spearman's correlation coefficient=0.505, P=0.039). At the individual level, 1) 8 403 subjects were eligible for analysis, and 1 138 (13.5%) had diabetes. 2) A total of 7 879 subjects had experienced indoor cold exposure in the winter. A significant difference was discovered in the prevalence rate between the group with and without cold exposure (P<0.001). 3) After age, gender, BMI, and smoking factors were adjusted, the group with indoor cold exposure in the winter was significantly associated with a 23% reduction in the risk of diabetes compared with the group without indoor cold exposure (OR=0.77, 95% CI:0.61~0.97).[Conclusions] At the provincial and individual levels, indoor cold exposure in the winter is significantly associated with diabetes. This study provides evidence for the correlation between the indoor thermal environment and diabetes prevalence on a long-time scale. This supports the conclusions of laboratory studies on cold exposure and diabetes and guides indoor thermal environment design.

[Objective] Deep learning methods have mastered sophisticated structural design skills via feature extraction and data-driven learning, thus significantly advancing intelligent building structural design. The performance of data-driven intelligent structural design typically depends on the quantity and distribution of training data. However, only a few studies have investigated the relationship between the learned data features and the intelligent design, thus limiting its performance. This study aims to study the influence of training data distribution and quantity on structural design and subsequently improve intelligent design quality.[Methods] Two typical datasets are generated by collecting the shear wall architectural-structural design drawing data from two distinct regions (Beijing and Shandong) with different design habits. Based on the Shandong dataset, the data distribution characteristics are studied, and the correlation between the design conditions, the structural planer, the vertical shape, and the equivalent ratio of the shear wall to the architectural plan area are investigated using linear regression analysis. Subsequently, a generative adversarial network-based intelligent design method for building structures is adopted to extract and learn high-dimensional data features. According to the quantity and the distribution characteristics of data, this study has proposed a data augmentation method and a two-stage training method, wherein all data is hybridized for training in the first stage, and the data grouped by different design conditions is used in the second stage, thereby improving the design quality of intelligent design models. In addition, the Beijing dataset, which is substantially different from the training data, is used to evaluate and validate the study results using various training methods and training data quantities. Finally, to validate the results and illustrate the performance of the intelligent structural design, typical cases with three different design conditions are used, namely, 7 degree seismic intensity and 27 m structural height, 7 degree seismic intensity and 54 m structural height, and 8 degree seismic intensity and 39 m structural height.[Results] The analysis results demonstrated the following. 1) The regression analysis based on low-dimensional data features could not guide the design generation properly for complicated structures, such as shear walls. In contrast, high-dimensional feature learning based on deep learning might effectively capture the potential design laws and optimize the design generation. 2) With the improvement of the quantity of training data and the training strategy, the quality of the intelligent design structure increased by an average above 20%; however, the quality of the intelligent design was compromised when the properties of the test and training data were considerably different (with a shear wall ratio difference of over 50%). 3) Moreover, the analysis results were evaluated using relevant case studies. Regarding the plane's design and the mechanical performance of the overall structure, the intelligent design and the engineer's design had a high degree of resemblance, and the maximum interstory drift ratio varied by up to 8% on average.[Conclusions] Consequently, by assessing the data distribution, design conditions, and quantitative properties, the generative adversarial network-based intelligent structural design may provide high-quality designs with suitable training datasets. Furthermore, this study provides a reference for future research on intelligent structural design based on deep learning and the influence of data features.

[Objective] In the context of carbon peak and carbon neutralization, hydrogen utilization becomes a promising measure to solve the energy shortage and reduce total greenhouse gas emissions. Commonly produced during many industrial processes, byproduct hydrogen acts as a hydrogen source that is widely available, cheaply produced, and sufficiently clean, thereby having a large potential market. However, the lack of large-scale storage, corresponding logistics supply chains, and untapped markets hinder the further use of byproduct hydrogen.[Methods] Given the low cost of byproduct hydrogen and the need for large-scale hydrogen storage, this paper proposes a business model in which salt caverns purchase byproduct hydrogen from chemical plants. The decision-making process of chemical plants and salt caverns is modeled and studied as a mixed-integer nonlinear optimization problem. During the planning stage, the proposed model optimizes transportation routes, modes, and hydrogen processing capacity, and during the operation, it optimizes hydrogen processing volume based on electricity price fluctuations to improve the profit of the upstream supply chain. The constraints of the optimization problem in the proposed model include the dynamic process of hydrogen transportation between salt caverns and chemical plants, the fluctuation in market demand with changes in hydrogen pricing, and the state of charge of salt caverns. The objective is to maximize the benefits of salt caverns and chemical plants. Given the characteristics of the optimization problem, this paper combines genetic algorithms and a commercial solver of linear programming to obtain the optimal solution. Finally, an envisioned case is used to study the economic benefits brought about by the optimization of supply chain decision-making and sensitivity analysis.[Results] (1) Different scenarios in the supply chain for hydrogen transportation achieved a net income with room for profit, making the proposed business model viable. (2) The optimization model proposed in this article optimized transportation routes, transportation modes, and hydrogen processing unit capacity during the planning phase. During the operational phase, it optimized the hydrogen processing volume based on electricity price fluctuations, thereby increasing the upstream supply chain benefits of byproduct hydrogen. (3) Sensitivity analysis showed the benefits of joint transportation under changing costs, and there existed an optimal pipeline capacity for a given market demand, beyond which increasing pipeline capacity would not further increase profit. (4) Varying the production scale of hydrogen by chemical plants, transportation distance, and cost showed that small and medium-scale chemical plants were more likely to engage in joint transportation, while large-scale chemical plants tended to transport independently. Increasing transport costs encouraged joint transportation to reduce costs. (5) Modifying the linear demand function parameters for the market showed that increasing demand and reducing price sensitivity increased the profit of the upstream supply chain. Improving hydrogen transportation technology to lower costs also increased profit.[Conclusions] The business model proposed in this paper provides a new source of income for chemical plants and salt caverns, improves resource utilization by reducing industrial exhaust emissions, realizes the rational use of natural resources, and provides a new way to accelerate the energy transition.

[Objective] In the context of increasingly stringent pollutant emission standards, more and more circulating fluidized bed (CFB) boilers are pursuing low-NO_x combustion and high-efficiency desulfurization in the furnace. How to decouple the contradiction between the two is one of the research hotspots. Previous studies found that when the separator efficiency is suitable, a reasonable adjustment of the limestone particle size can significantly improve the desulfurization efficiency in the furnace while effectively inhibiting the negative effect of limestone on NO_x emission. However, this phenomenon is only at the experimental stage and lacks theoretical analysis.[Methods] In this paper, a commercial 550 MWe ultra-supercritical CFB boiler was calculated with the help of the established comprehensive one-dimensional two-phase hybrid mathematical model for CFB combustion to obtain the rules and the effect of limestone particle size on 烧NO_x emission desulfurization in a furnace. The comprehensive mathematical model for CFB combustion consists of three parts, namely, material balance (including ash, limestone, raw coal, and coke), gas balance, and energy balance. Moreover, this model specifically considers the local gas/solid fluidization state and gas/heat transfer conditions in different regions of the CFB combustor. Some two- or three-dimensional problems, such as bubble breakage over dense bed surface, secondary air injection, core annular flow structure, and particle clusters in freeboard, are also considered in the modeling. For nitrogen-containing catalytic reactions on the limestone surface, CaO has significant catalytic activity for many nitrogen-containing reactions, which is one of the important reasons for the increase of NO_x emission caused by limestone desulfurization in the CFB furnace. This paper considers four types of nitrogen-containing catalytic reactions on the surface of CaO particles, including oxidation of NH₃, hydrolysis of HCN, oxidation of CO, and catalytic reduction of NO by CO. The integral CFB model is validated against the field test data obtained from the commercial CFB boiler. Favorable comparisons were conducted between the calculated and measured results, involving the median diameter of each ash sample and the concentration of the main flue gas components.[Results] The calculation results showed the following:(1) Due to the significant catalytic effect of CaO on each nitrogen-containing reaction, increasing the calcium-sulfur mole ratio could improve the desulfurization efficiency in the furnace but could also lead to an increase in NO_x emission. (2) In addition to the negative effects on NO_x emission, increasing the calcium-sulfur mole ratio also had different degrees of negative effects on boiler efficiency and flue gas dedusting, among others. (3) With the reduction of limestone particle size, the total CaO effective reaction area in the furnace and near the coal feeder was reduced, which inhibited the overall rate of relevant catalytic reactions and was conducive to alleviating the negative effect of limestone on the reduction of NO_x emission.[Conclusions] By calculating and analyzing the relationship between the limestone desulfurization in the furnace and NO_x emission, a new idea is provided to decouple the contradiction between desulfurization and low-NO_x combustion. That is, under the same calcium-sulfur mole ratio, appropriately reducing the limestone particle size or even using ultrafine limestone particles matched with the high-efficiency separator can effectively improve the in-furnace desulfurization efficiency and alleviate the negative effects of limestone on the reduction of NO_x emission.

[Significance] Marine environmental pollution has become a globally concerning environmental issue, which has attracted the great attention of governments and scientists in coastal countries. The release of large quantities of pollutants into the oceans has inevitable lasting effects and harm to marine life and can even bring about irreversible changes leading to the degradation or deterioration of the ecosystem. However, marine organisms under pollution stress do not always await their doom. Considerable evidence suggests that marine organisms could respond to pollution stress at the individual, population, and community levels, leading to molecular-level to ecosystem-scale effects and eventually reaching either adaptation/coexistence or degradation/death, with dynamic processes of interactions and changeable evolutionary outcomes between the two.[Progress] In general, organisms can adapt to complex environments in three mechanisms, namely, behavioral change, phenotypic alteration, and genetic evolution. The initial response to environmental stress from pollution is generally behavioral (such as escape or migration to favorable habitats or vigilance). However, the continuous degradation of the environment makes migration more difficult, and some sedentary organisms (e.g., corals) can hardly move once immobilized, forcing the organisms to immediately react in situ. Altering the phenotype appears to be the best solution. This rapid nongenetic mechanism is based on the ability of individual genotypes to produce different phenotypes in response to the environment, which allows marine organisms to adapt and survive in the polluted marine environment. However, such phenotypic adjustments can be highly variable in the life cycle within a generation or even across generations. Phenotypic plasticity within a generation may play an important role in survival through rapid environmental change. In marine pollution, organisms may alter their macro or micromorphology and sex specificity or adjust the transcriptome characteristics and metabolic pathway expression of different organs, which play a detoxification-oriented role, within a generation for self-preservation. Although plasticity is usually shown under environmental conditions experienced by the parental generation, these conditions can also affect their offspring from one to multiple generations. That is, populations can adjust their phenotypes within several generations, which is called transgenerational phenotypic plasticity. Environmental pollution persists across generations for most species, and these organisms may evolve through transgenerational phenotypic plasticity to fight against environmental pollution. This may help in understanding the effects of pollution on marine organisms. Accordingly, supposing that plasticity works for the adaptation of organisms, its primary role is likely to buffer the cost of evolutionary mismatches and facilitate genetic evolution by "gaining time". Consequently, the phenotypes could be more adaptive to the current conditions, and these genetically based adaptations would continue to evolve until completion.[Conclusions and Prospects] This review has done comprehensive literature research, collected various scenarios, and summarized the laws of the short-term response and long-term adaptation of marine organisms under pollution stress based on the viewpoint of evolutionary ecology and relevant principles in pollution ecology. From the perspective of behavior, adaptation, and evolution, this review preliminarily discussed the significance of marine pollution evolutionary ecology and attempted to summarize and explore the interactions and connections between pollution and organisms in the marine environment, providing new insights for the protection of the marine environment and the development of marine sciences.

[Objective] Overlay targets with a small range and requiring extremely high measurement accuracy are used in the lithography system's alignment, exposure, and measurement processes. They are inevitably affected by many errors when performing specialized functions, e.g., line width and design difference, line edge roughness, and target edge effect. This paper conducts a simulation study on the micro diffraction-based overlay target (μDBO) in the condition of wafer stage overlay (WSO), which is commonly used in lithography, based on the design of overlay alignment targets in the field of measurement. Litho-target design simulation software is known to be more advanced than lithographic imaging simulation software, but its license is usually rarer and more expensive. We propose a novel method for simulating lithography target metrology results using lithography imaging simulation software. The method predicts the detected light intensity distribution in the experiment qualitatively and can calculate target performance in metrology.[Methods] This paper compares the similarities between metrology process imaging in the optical method and lithography imaging. In the metrology process imaging simulation, the target, lens, and imaging sensor modules have analogs in the lithography simulation, which are the mask, lens, and aerial image modules. Thus, metrology target patterns can be represented as mask patterns, and the output of an aerial image from lithography simulation software can be analyzed to obtain metrology information. We use WSO-μDBO targets, which have two different critical dimensions and suffer from a heavy marginal effect in experiments. The self-made DrM software in Matlab code with the rigorous couple wave analysis (RCWA) algorithm and the commercial program HyperLith with the finite-difference time-domain algorithm are used as tools to restore the phenomenon in the experiment. Both programs import the illumination patterns of the experimental instruments as well as the structures of WSO-μDBO targets. The Abbe imaging principle in DrM and the Hopkins approximation in HyperLith are used in the simulations. The pupil imaging simulation is applied to analyze the marginal effect on the original targets. In this simulation study, four key performance indicators (KPI) of WSO-μDBO performance are selected as evaluation indices.[Results] The simulation of WSO-μDBO targets showed that:(1) The special phenomenon of bright and dark lines in the experiment can be reproduced in simulation using both tools and algorithms. (2) In the simulation, this method can provide calculated KPI of μDBO targets in simulation. (3) The aerial image generated by this method was closed between both tools but with a relative error of 1% to 5%, resulting in a discrepancy of around 1 nm between the simulated overlay values. (4) This work presented a novel design of WSO-μDBO targets (Edge, Windmill, and Hybrid) with a relative improvement in simulated stack sensitivity(SS) of 202.6% over the original design. As a result, the simulated metrology error of the target was reduced by 50.9% to 1.3%.[Conclusions] The feasibility of our simulation method is verified by comparing simulated and experimental WSO-μDBO images, which can serve as a new simulation research platform for studying metrology targets. This method is also simple and easy to implement, and it has the potential to reduce the software cost investment in marking design during the production process.

[Objective] The treatment of subjective tinnitus is a global problem in hearing clinics. Currently, the technical defects and bottlenecks of tinnitus detection and treatment in hearing clinics include the lack of portable tinnitus detection and treatment equipment that integrates personalized and accurate detection with targeted rehabilitation, lack of precise treatment plans using pleasant music in the background to eliminate patients' discomfort and rejection to secondary hearing damage caused by clinical white noise masking therapy, and lack of effective treatment options for high-frequency tinnitus. This study pioneers the development of a personalized tinnitus detection and treatment instrument.[Methods] In this study, the STM32 embedded single-chip microcomputer is used as the core module to run the software, execute device control and algorithms, control and operate different functional modules, and provide audio decoding conversion module, power conversion module, earphones, and charging interfaces. The audio-decoding chip is designed to realize high-quality stereo multimedia digital signal processing. Staticrandom access memory (SRAM) provides the external memory for the microcontroller unit (MCU). The flash chip stores audio, text, and data. The SD card stores audio files and fonts required by tinnitus detection and treatment equipment, user information, and other large-capacity information. The sound output of this detection and treatment device is transmitted to the patient through the transducer (earphone), and the patient enables human-computer interactions with the device through buttons and screen, all of which form the operation adjustment of tinnitus detection and treatment. Batteries provide power to the whole device. The software adopts object-oriented programming and realizes functions by designing top-level, middle-level, and bottom-level functions. Top-level functions refer to the graphical user interface (GUI) part that targets user programs and functions related to user control. Middle-level functions include each specific page, functions inside the page, and the function of each button operation. Bottom-level functions target the hardware to control and configure peripheral register functions such as memory, screen, audio decoding, keys, SD card, LED, and others.[Results] The proposed tinnitus detection and treatment instrument enabled personalized and precise detection of tinnitus and targeted rehabilitation treatment. Its features included tinnitus ear selection, tinnitus detection, tinnitus treatment, data storage of prescriptions, equipment calibration, and information saving and scanning. It pioneered several comprehensive notching and masking personalized sound treatment solutions and resolved technical defects and bottlenecks of current clinical tinnitus detection and treatment equipment. The treatment results of the first four participants who finished the 8-week treatment using this therapeutic instrument showed a therapeutic effectiveness close to 100%. The tinnitus handicap inventory (THI) score and tinnitus loudness of patients with tinnitus showed a significant downward trend. Significant therapeutic effectiveness was observed among patients with high-frequency tinnitus. Obvious changes to the mismatch negativity occurred in patients with significantly lower THI scores and lower tinnitus loudness.[Conclusions] The proposed portable instrument, not only provides a portable treatment method integrating tinnitus detection and treatment for clinics, its personalized and accurate detection and targeted rehabilitation treatment functions, but also have laid the application foundation for the quantitative and objective diagnoses and treatment standards and prescription establishment.