Owing to its advantages of good controllability, high efficiency and reliability, and adaptability to the connection of renewable energy resources, the direct current (DC) distribution system has become the development direction of next-generation power distribution systems. The characteristics of energy resources and loads, disturbance resources, and operation mechanism of DC distribution networks are different from those of alternating current (AC) distribution networks, which lead to new features of power quality issues. Aimed at the widespread applications of DC distribution systems, their typical topologies with the corresponding grounding modes are introduced. With the combination of operation characteristics, the phenomena of various possible power quality issues in DC distribution systems are analyzed, including voltage band, ripples, imbalance, flicker, voltage dip (short-time interruption) and voltage swell. In addition, the main technical indices that can be used to describe the power quality of DC distribution systems are proposed. The results of an example demonstrate that these indices can well quantitate the related power quality phenomena. Finally, the research directions for power quality issues in DC distribution systems in the future are also discussed. The research aims to analyze the power quality issues in DC distribution systems in depth, providing support for the comprehensive power quality measurement and analysis of DC distribution systems.

The output characteristics of an LCC-S-type magnetically coupled resonant wireless power transfer (WPT) system are quite different under different parameter configurations. Aimed at the characteristic that the load power on launch vehicles varies in a wide range, a steady-state efficiency model of the WPT system between the rocket and ground was established, and a parameter design method for a resonant network with the average power-efficiency product as an optimization objective was proposed for the first time, so as to improve the system's overall output characteristics in the full power range. On this basis, a time-domain model of inverter output current considering high-order harmonics was built to achieve zero voltage switching of the inverter. In view of the influence of resonance parameter disturbance on indexes such as the switching losses of inverter and the voltage gain, a disturbance design method for resonance network was also given. Finally, the rationality of the model and the superiority of the parameter configuration method were verified by simulation results.

The conventional power supply methods for rotary mechanisms have problems such as complex structures, monotonous functions and security risks. Aimed at these problems, a radial-flux rotational wireless power transfer (R-WPT) system with a rotation speed detection function was proposed. First, the relationship between the mutual inductance of a radial-flux coupling coil and rotation angle was deduced, and a circuit model of the R-WPT system was built. Second, an output voltage processing circuit was designed to convert the waveform, which was further used for motor control. Finally, an R-WPT experimental platform driven by a stepping motor was constructed and verified at different target speeds. Results show that the output voltage waveform well conformed to the expectation of deduced mutual inductance. With the voltage signal processing and the closed-loop speed control method based on the stepping motor, the starting process of the R-WPT device with speed regulation can be performed at higher target speeds, indicating the feasibility of the system.

Aimed at the shortcomings of large size, susceptibility to damage and high cost associated with electrolytic capacitors in permanent magnet synchronous motor (PMSM) drive systems, an active damping control strategy for PMSM without electrolytic capacitors is proposed by incorporating an RD damping circuit. First, the relationship between the selection of the film capacitor on the DC-bus side and the grid-side input power is analyzed, and the dynamic equations for the DC-bus voltage and capacitance are derived. Then, an active damping power decoupling method based on d-axis repetitive control is put forward to enhance the grid-side current power factor. Meanwhile, a control strategy based on q-axis power transformation is introduced to improve the motor's dynamic negative-impedance characteristics. The RD damping circuit is utilized to simplify control commands and absorb the overshoot voltage. Finally, simulation and experimental results indicate that with the use of small-capacity film capacitors, the active damping control effectively reduces the current distortion and suppresses the torque ripple.

Aimed at the problem that the control system for a primary permanent magnet linear motor (PPMLM) is susceptible to uncertain factors such as load disturbance and parameter change, an action-dependent heuristic dynamic programming control algorithm is used to improve the control performance and anti-disturbance capability of the system while reducing the thrust force fluctuation. First, the mathematical model of the direct thrust force control system for PPMLM is constructed. Second, the speed tracking problem is transformed into an optimal control problem, which is further transformed into a problem of solving the Hamilton-Jacobi-Bellman equation by minimizing the performance index function and the speed tracking error. The numerical solution of the Bellman equation is obtained by using a strategy iteration algorithm based on the Bellman criterion. Finally, an accurate tracking of speed is achieved. To solve the problem that the BP neural network is easy to fall into local optimum, an adaptive learning rate is designed. In addition, in view of the problem of approximate error existing in the BP neural network, a robust compensation term is added to the control law to improve the accuracy of speed tracking. Afterwards, the stability of the closed-loop system is analyzed. Simulation results show that the proposed control strategy can improve the speed tracking performance of the system, reduce the thrust force fluctuation and restrain load disturbance and other nonlinear factors.

Aimed at the problem that the prediction results of state-of-health (SOH) of lithium-ion batteries are inaccurate due to the failure to consider the capacity regeneration phenomenon, a nonlinear relationship method based on multi-time scale set empirical mode decomposition fusion time series mapping input and SOH is proposed. The battery capacity information considering capacity regeneration is decomposed into high-frequency main trend signals and low-frequency margin signals by an ensemble empirical mode decomposition (EEMD) algorithm, and the estimation of lithium-ion battery SOH is completed by the prediction results of the two types of signals. Results show that the proposed method can capture the phenomenon of capacity regeneration to improve the estimation accuracy and achieve the fitting of battery capacity degradation, and it is less affected by different prediction starting points. Both the predicted maximum RMSE (root mean square error) and MAE (mean absolute error) are not more than 1.5%, and the maximum MAPE (mean absolute percentage error) is not more than 2%.

Through a comparison of the economic advantages of different power decomposition methods for energy storage system capacity configuration and hybrid energy storage system (HESS) over one single energy storage system, a method based on improved moving average filtering and ensemble empirical mode decomposition (EEMD) is proposed to mitigate wind power fluctuations in view of the optimal capacity configuration of HESS. First, EEMD was used to decompose the HESS reference power which was derived by improved moving average filtering, so as to obtain several single-component signals, i.e., intrinsic mode functions (IMFs). Then, the time-frequency spectrum of each IMF was processed by Hilbert transform to identify the suitable gap frequency, and the HESS reference power was obtained by reconstructing the high- and low-frequency components. Meanwhile, the charge-discharge efficiency of the energy storage system and its state-of-charge were fully taken into account, and the rated power and capacities were configured. Finally, based on the life cycle cost theory, an economic cost calculation model for the energy storage system was established to compare the cost among different configuration schemes, thereby obtaining the optimal one with the highest economic efficiency.

Under the “Dual-carbon” target, aimed at the problems of excessive investment in microgrid energy storage systems and the arrangement of retired batteries of new energy vehicles, a dual-layer optimal configuration model of a multi-microgrid shared energy storage system considering the stepwise utilization of retired batteries was proposed. First, the multi-microgrid shared energy storage system based on retired batteries was modeled. Second, a dual-layer capacity configuration model was adopted with the investment and operation and maintenance cost of microgrids as its upper-layer objective function and the daily operation cost of energy storage as its lower-layer objective function. The particle swarm optimization algorithm and Cplex linear solver were combined to solve the capacity configuration and equipment running state of shared energy storage. Finally, the result of an example indicates that the proposed model can effectively reduce costs and improve the utilization rate of resources, and the corresponding effect is remarkable.

Lithium-ion batteries are usually taken as energy storage components to realize the storage and utilization of electrical energy, and an accurate estimation of their state-of-health (SOH) can provide a theoretical basis for their actual applications. However, SOH is difficult to measure directly. To estimate the SOH of lithium-ion batteries accurately and conveniently, a model for estimating the SOH of lithium-ion batteries is proposed based on residual connection and convolutional long short-term memory (Conv-LSTM) networks from the perspective of data driven. This method is based on residual connection, Conv-LSTM and convolutional neural networks, and it uses the voltage, current and capacity during the charging process of a lithium-ion battery to realize the SOH estimation in its service cycle. The experimental results on multiple battery test datasets show that compared with those of the existing models with better estimation effects, the estimation result of the SOH estimation model based on residual connection and Conv-LSTM was more accurate, and the maximum percentage error and mean absolute percentage error of SOH estimation on the test dataset were less than 3.3% and 1.7%, respectively.

The cell uniformity and connection mode of solid oxide fuel cell (SOFC) are important factors affecting the overall performance of the stack. The uniformity of flat-chip cells in the SOFC stack is evaluated by methods such as linear sweep voltammetry and electrochemical impedance spectroscopy. Then, the comprehensive effect of cell uniformity and connection mode on the stack's output performance and the underlying electrochemical mechanism are investigated by combining with the method of distribution of relaxation times. Results indicate that the cell uniformity and connection mode jointly affect the stack performance, i.e., the cell uniformity affects the stack lifetime, while the connection mode affects the stack's output power. In a series mode, a low cell uniformity will cause the poor-performance cell to over-discharge, which may accelerate the deterioration of the stack performance. Compared with the series mode, the parallel mode will cause a higher power loss, which mainly comes from ohm resistance, oxygen exchange and charge transfer processes but affects the stack lifetime to a small degree. In general, the peak power of the series stack is high, but the lifetime is greatly affected when the cell uniformity is low. The power loss of the parallel stack is obvious, but the influence on the stack lifetime is weak.