To solve the difficulty in online life prediction of silicon carbide metal-oxide-semiconductor field-effect transistor (SiC MOSFET) under practical working conditions, a digital implementation method for SiC MOSFET module life prediction based on particle swarm optimization-back propagation (PSO-BP) neural network was proposed. First, the saturation voltage drop of SiC MOSFET was extracted by a saturation voltage drop platform as the temperature-sensitive electric parameter, and a junction temperature prediction scheme based on experimental data was established. Second, a life prediction scheme based on PSO-BP neural network was established by using a power cycling accelerated aging experimental platform to extract the aging characteristic data. Third, the junction temperature prediction scheme and life prediction scheme were transplanted to field programmable gate array to realize the digitization of SiC MOSFET life prediction. Finally, a circuit was designed to verify the proposed method. Experimental results show that the error between the digital junction temperature and real junction temperature was 4.73 ℃, and the percentage of error between the predicted life times and real life times was 4.1%, which proves that the proposed life prediction method is realized digitally and can accurately predict the life times of SiC MOSFET module.

For a hybrid cascade H-bridge inverter with a DC-side voltage ratio of 1:1:2, if the hybrid carrier disp-osition modulation strategy is adopted, the problem of output power imbalance in the low-voltage unit will occur al-though there is no current backflow phenomenon and the harmonic performance of output voltage is good. To solve this problem, the power imbalance is analyzed at first. Then, an improved hybrid modulation strategy is proposed, under which the high-voltage unit performs step wave modulation and the low-voltage unit performs PWM, and two low-voltage units adopt different processing methods for the modulation wave. The good harmonic performance of output voltage is kept, the number of triangular carriers is reduced, and the control process is simplified, with a frequency doubling effect. Third, this strategy is optimized, and the switching signal of low-voltage unit is logically calculated, which can solve the power imbalance problem of low-voltage unit in two carrier cycles. Finally, the feasibility was proved by simulation and experimental results.

Aimed at the problems of a traditional LLC resonant converter in wide voltage applications such as a wide switching frequency range and a poor voltage regulation performance, a voltage doubling two-phase parallel resonant converter is proposed. There is a parallel double half-bridge LLC structure on the primary side of this converter, and a bidirectional switch is introduced into the full-bridge rectifier network on its secondary side to form a reconfigurable voltage doubling rectifier network. Fixed frequency control is adopted during operation. The lower half-bridge on the primary side changes the input voltage of the resonant tank by changing the duty cycle, while the upper half-bridge works with a fixed duty cycle. The rectifier network on the secondary side realizes full-bridge and voltage doubling hybrid rectification under the bidirectional switch, which can achieve 4 times of voltage gain. At the same time, this converter has a good soft switching performance, its voltage gain is independent of magnetizing inductance and load, and a larger magnetizing inductance can be selected to reduce the switch-off loss and conduction loss. Finally, the feasibility of the proposed converter was verified by simulation and experimental results.

Aimed at the problem that the accuracy of photovoltaic array fault diagnosis based on support vector machine (SVM) is not high and it is easily affected by the kernel function and penalty factor parameters, a photovoltaic array fault diagnosis method based on SVM optimized by the seagull optimization algorithm (SOA) is proposed. The SOA is introduced to optimize the parameters of the SVM model, and an SOA-SVM fault diagnosis model based on the optimal parameters is established. MATLAB software is used to build a photovoltaic array simulation model, and the characteristic parameters under different fault types are extracted and further inputted into the SOA-SVM model for fault diagnosis. Experimental results show that the fault diagnosis accuracy of the SVM model optimized by SOA is significantly improved. Compared with the ABC-SVM and PSO-SVM models, the SOA-SVM model converges faster in the optimization process and has a higher fault diagnosis accuracy.

As the penetration rate of renewable energy resources in a new power system continues to rise while the proportion of traditional thermal power units continues to decline, the new power system faces severe frequency control problems. Distributed battery energy storage systems (BESSs) provide an effective way to solve these problems. On this basis, a robust load frequency control (LFC) method for distributed BESSs based on sparse communication network is proposed. To suppress the uncertainties related to system operation, a two-tier model predictive control (MPC) is designed to improve the response characteristics of BESSs, thus improving the performance of LFC. To minimize the area control error, the proposed method can satisfy various operating physical constraints of the system. In addition, the influence of communication delay on the performance of frequency modulation participated by BESSs is also considered, and a fuzzy coordination control device is designed to coordinate BESSs and the traditional generator, so that the mis-operation of the traditional generator under the condition of long delay can be avoided. Finally, simulation results show that the response capability and frequency modulation effect of distributed BESSs are better than the traditional methods under parameters such as different values of capacity, rated power, charge and discharge coefficient, state-of-charge and time constant.

With the scale expansion of a subsea observation network, the stability of its high-power power supply system has attracted attention. First, the impedance models of key parts in the subsea power supply system are established. Considering the characteristics of high power electronic penetration rate, multi-bus cascading and adjacent bus interactive coupling of the subsea DC power supply system, the stability and influencing factors of the system are explored by using the step-by-step analysis method. The analysis result shows that the integral parameter of the controller is the dominant parameter that leads to the instability of the Buck converter, and the proportional parameter of the controller is the dominant parameter that results in the instability of the junction box subsystem. Both an increase in the impedance parameter of the optoelectronic composite cable and a decrease in the inductance parameter are beneficial to improving the system stability. The simulation results based on the PLECS simulation software verify the stability analysis results.

With the large-scale network entry of electric vehicles (EVs), their disorder charging further increases the load peak-valley gap, which has a negative impact on the stable operation of power system. A two-stage optimization scheduling strategy which takes into account the EV load and the energy storage system of batteries is proposed. First, an orderly charging scheduling model for EV is established, which aims at minimizing the absolute peak-valley gap between user charging cost and load. The improved particle swarm optimization algorithm is used to solve this model to avoid peak charging. Second, an optimal scheduling model of peak-shaving and valley-filling for the energy storage system is established with an objective of minimizing the variance of load and the combined cost of energy storage life, which is solved by the improved Harris Hawks optimization algorithm to reduce the peak-valley gap of load. In addition, the optimization results are evaluated and analyzed based on the evaluation index of peak-shaving and valley-filling. Finally, a simulation experiment is carried out with the measured load power of one power network as an example. Results show that under the proposed two-stage optimization scheduling strategy, the peak load decreases by about 147 kW, the valley load increases by about 223 kW, and the peak-valley gap deceases by 46.73%, indicating that this strategy can effectively improve the load curve, alleviate the pressure on power supply during the peak load period and ensure the safe and stable operation of power grid.

Compared with the traditional silicon(Si) devices, the gallium nitride(GaN) devices have lower parasitic parameters, a faster switching speed and a smaller on-resistance, which will easily lead to the phenomenon of continuous oscillation during their switching-on process and further result in the circuit instability. Therefore, it is necessary to suppress this phenomenon in practical circuits. Under this background, a negative conductance model of a bridge circuit under the conventional driving scheme is established at first, and the oscillation stability of the circuit is analyzed. Then, by adding optimization to the conventional driving scheme, the corresponding negative conductance model is established. The optimization schemes of series damping represented by changing the resistance and adding ferrite beads and those of parallel low impedance represented by adding RC snubber are selected, respectively. With this model, the influence of adding the driving optimization schemes on the oscillation stability of the circuit can be identified, and the changes in the stability before and after the addition were verified by experimental results, providing a reference for the driving circuit to select its appropriate driving optimization scheme.

With the continuous advancement of medical technology, implantable medical devices (IMD) are increasingly applied in clinical practice. Since the traditional battery-powered method will bring additional tissue damage and surgical costs to patients, the use of wireless power transfer (WPT) technology to power IMD will become a trend in the future. However, how to design a high-efficiency IMD-WPT system in a limited space is very challenging. To this end, the performance characteristics of five WPT technologies suitable for IMD are compared. Then, the magnetic resonance WPT technology is taken as an example to introduce the key issues in the design of a magnetic resonance IMD-WPT system. Finally, the application status of part of the magnetic resonance WPT technologies in some typical IMD is combined, and the research direction of IMD-WPT technology in the future is discussed.

Aimed at the problems of current unobservable area and zero drift error in the traditional space vector pulse width modulation with single-sensor phase current reconstruction method, an error self-correction complementary non-zero vector pulse width modulation method is proposed. Through the analysis of the DC bus sample principle, the minimum sample time is defined, the complementary non-zero vector is used to replace the zero voltage vector, and the current sampling window is extended, thus eliminating the sector boundary unobservable area. At the same time, the generation mechanism of error amplification effect is revealed, and the zero drift is detected and self-corrected by means of double-sampling complementary non-zero vector, which realizes the compensation for current zero drift reconstruction. Experimental results show that the reconstruction error of the proposed method was lower than 1.26%, and the phase current THD was lower than 6.15%.