To fully consume the distributed photovoltaic (PV) while reducing the impact of electric vehicle charging on the distribution network, PV-storage-charging integrated systems are rapidly developing. However, the secondary harmonic current of the DC bus in a PV-storage-charging integrated system will cause problems such as oscillations of PV output power and an increase in the current stress on the switching tube. To ensure the efficiency and power density of the system while suppressing the secondary harmonic current, the performance of different power decoupling circuits is evaluated from several dimensions at first. Then, based on the selected circuit topology, a soft-switching modulation method and the corresponding closed-loop control strategy are proposed, which can effectively improve the harmonic current suppression capability while realizing the full-range zero-voltage switching (ZVS). Finally, the effectiveness of the power decoupling circuit and the proposed control strategy was verified by simulation and experimental results.

Aimed at the phenomenon of capacity recovery and the problem of accurate prediction during the cyclic usage of lithium-ion batteries, an approach integrating feature analysis with an innovative variational pyramid split attention long short-term memory (V-PSA-LSTM) network model is proposed. This method performs the feature analysis of lithium-ion batteries, and Pearson correlation coefficients are employed to screen out the health features that are closely associated with the battery’s state of health (SOH), thereby revealing the key characteristics such as dynamic capacity restoration. Then, a novel V-PSA-LSTM network model is put forward, in which a periodic variation activation mechanism is incorporated into the pyramid split attention (PSA) algorithm to optimize the long short-term memory (LSTM) network for the first time. This design aims at enhancing the LSTM model’s capability to precisely capture the evolution patterns of SOH in lithium-ion batteries, particularly the fluctuating changes in capacity during the cyclic process. It is verified that the V-PSA-LSTM model significantly reduces prediction errors during the SOH prediction for lithium-ion batteries, e.g., the root mean square error (RMSE) and mean absolute error (MAE) are 0.002 2 and 0.001 8, respectively, indicating the exceptional performance in solving the problems of capacity recovery and accurate prediction.

To improve the accuracy of remaining useful life (RUL) prediction for lithium-ion batteries, a method combining complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), random forest (RF), improved hazelnut tree search (IHTS) algorithm and improved gated recurrent unit (GRU) is proposed. First, CEEMDAN is used for modal decomposition and noise reduction on the original capacity data. Second, RF is adopted to calculate the weight of each modal component relative to the original capacity data. Third, chaos initialization optimization is employed to improve a hazelnut tree search algorithm to obtain the IHTS algorithm, which is further used to optimize and improve GRU to establish an IHTS-GRU model. Finally, the IHTS-GRU prediction results of each modal component and the weight of each modal component are weighted and averaged, thus realizing the lithium-ion battery RUL prediction. The NASA data is used for verification, and it is found that the prediction accuracy of the CEEMDAN-RF-IHTS-GRU model is improved by 3.2% and 2.1% on average compared with those of the CEEMDAN-GRU and CEEMDAN-RF-GRU models, respectively, indicating significant increases in the prediction accuracy. Therefore, the proposed method improves the accuracy of RUL prediction for lithium-ion batteries.

The state of health (SOH) estimation method for lithium batteries is crucial for battery management systems used in electric vehicles. At present, most of the SOH estimation methods often require the charging and discharging data during a long time, resulting in a poor real-time performance. To address this issue, an SOH estimation method based on random charging segments for feature extraction analysis is proposed. First, the complete constant current (CC) charging process is divided into several fixed-length data segments according to voltage intervals. Then, health indicators (HIs) are extracted for different voltage intervals. Finally, based on the characteristics of HIs, a back propagation neural network (BPNN) optimized by the particle swarm optimization (PSO) algorithm is trained, which can extract features from any CC charging segment to complete SOH estimation. NASA and Oxford datasets verify that this method has high flexibility and accuracy.

An efficient battery thermal management system is crucial for the performance and safety of batteries. A liquid cooling-air cooling coupling thermal management system for lithium-ion batteries was proposed, and the effects of coolant flow rate and radial thermal conductivity of batteries on the system’s cooling performance were analyzed. In addition, the comprehensive effect of the liquid cooling-air cooling composite system on the heat dissipation characteristics of batteries was also investigated. Results show that increasing the coolant flow rate can improve the cooling performance while leading to an increase in pressure drop, resulting in higher system power consumption and a lower economic efficiency. The more pronounced the anisotropic characteristics of battery thermal conductivity, the worse the uniformity of temperature distribution. By increasing the radial thermal conductivity of batteries, the maximum temperature of batteries was effectively reduced, and the temperature uniformity was also improved. Compared with a single liquid cooling system, the liquid cooling-air cooling composite system significantly optimized the temperature distribution and heat dissipation performance of the battery module, e.g., the maximum temperature was reduced by 3.20 K, and the maximum temperature difference was reduced by 3.36 K. Moreover, adopting an intermittent air cooling operation strategy can effectively reduce the system energy consumption while maintaining the operation of batteries within the optimal temperature range.

Under an unbalanced power grid, the grid-connected inverter system may experience an imbalance in grid-connected current or load voltage. To achieve a grid-connected current balance or a load voltage balance in the grid-connected inverter system under linear and symmetrical load, a balance control strategy based on a negative-sequence compensation algorithm is proposed. On the basis of adding a compensation impedance device between the local load and the point of common coupling (PCC), the positive- and negative-sequence components of current on the inverter side are controlled by the reference value of output voltage from the inverter, which is generated by superimposing the positive-sequence loop output with the negative-sequence compensation loop output with different balance control objectives, so as to achieve the balance control of these objectives. The reference values of the positive- and negative-sequence current loops on the inverter side are obtained using the grid-side current loop or the local load identification algorithm. Finally, the feasibility and the effectiveness of the proposed strategy were verified by simulation and experimental results.

In a weak grid, the phase-locked loop (PLL) of a grid-connected converter will dynamically interact with the current loop and further lead to a small-signal (or harmonic) instability problem. To solve this problem, a small-signal modeling method is used to establish an equivalent impedance model of the grid-connected converter. On this basis, the impedance model of the converter is combined with a system stability judgment method, and the instability mechanism of the grid-connected converter under the weak grid condition is analyzed by using the pole variation trajectory and the Bode plot. For a multiple-input multiple-output system, an active damping control strategy based on minimum phase digital filters is proposed while considering multi-loop coupling, so as to improve the system stability margin. At the same time, a specific implementation scheme of the control strategy, a parameter design method and the impact of the control strategy on the multi-loop coupling part are given. In addition, to study the effectiveness of the proposed control strategy under different operating conditions, the system stability margin under the proposed control strategy in the cases of different grid strengths, PLL bandwidths and current loop bandwidths is analyzed, and it is compared with that under the traditional control. Finally, the correctness of the proposed theory was verified by the result of a hardware-in-the-loop experiment.

With the advancement of dual-carbon goals, many of the existing hydropower stations plan to integrate nearby wind power within a certain range, form hydro-wind renewable clean energy bases and use the existing hydropower transmission lines for power delivery. As the wind power capacity increases gradually, a power system may encounter sub-/super-synchronous oscillation issues. To ensure the safe and stable operation of the system, a typical operation scenario for a system with hydro-wind renewable energy delivered via LCC-HVDC is constructed. The key factors causing sub-/super-synchronous oscillations are examined using the impedance method, and a virtual admittance-based suppression strategy for sub-/super-synchronous oscillations is put forward. This method reduces the interaction between LCC system and the hydro-wind renewable energy generation system by lowering the equivalent impedance of LCC system at the oscillation frequency, thus suppressing oscillations. In addition, the effectiveness of the proposed control strategy is verified through MATLAB/Simulink simulations.

The flexible DC distribution network has characteristics such as a fast rising speed and a large amplitude of fault current. Therefore, the fast and reliable identification of DC line faults is a key technology that the DC distribution networks urgently need to break through. Aimed at the problems of poor selectivity of single-end protection and a slow speed of multi-end protection in the existing flexible DC power grids, a fault protection method for flexible DC power grid based on the Canberra distance is proposed. First, according to the difference in polarity of current differential between the two terminals under internal and external faults in the flexible DC power grid, the Canberra distance is introduced to calculate the polarity similarity of fault current differential sampled at both ends of the line, so as to perform diagnosis of the fault area. Second, based on the fault characteristics of fault and non-fault lines at the time of fault occurrence, a ratio is formulated using the numerical difference in current differential between the two poles, thus selecting the fault pole. Finally, different fault scenarios are set to verify the applicability of the proposed method under different working conditions. Simulation results show that the proposed protection scheme can achieve correct fault location and pole selection under the conditions of different fault locations, different values of current-limiting reactance and different transition resistances.

The admittance measurement of grid-connected inverters plays a crucial role in the stability analysis of renewable energy grid-connection. On this basis, a high-precision measurement method for the frequency coupling admittance of a grid connected inverter is proposed. First, the admittance measurement process based on disturbance injection is taken as a measurement process of deterministic components. Then, the signal with disturbance injection is considered as a broadband signal, and a method for measuring deterministic components of the broadband signal based on polynomial fitting is put forward. Finally, a frequency coupling admittance measurement framework for the inverter based on deterministic components is established. Simulation and experimental results show that compared with the existing methods, the proposed method can achieve high-precision measurement of broadband signals and accurately measure the frequency coupling admittance characteristics in various grid frequency scenarios, indicating its good applicability for the measurement of deterministic components.