Silicon carbide (SiC) devices possess advantages such as high voltage resistance, low losses and high thermal conductivity, making them of significant importance for the development of the electric vehicle industry. A design for a SiC MOSFET power module utilizing large-chip packaging was proposed, and experiments were conducted to analyze the module's electrical performance. Simulations were set up to compare the module temperature under two conditions, i.e., electrical characteristics only and a combination of electrical characteristics and temperature feedback. Simulation results indicate that under identical operating conditions, the SiC MOSFET power module designed with large-chip packaging exhibited stronger conduction current capability, smaller temperature variations and improved electrical performance.

With the improvement of the integration degree of power modules, the optimization of their heat transfer structures has become a focus in the development. The topology optimization(TO) can maximize the cooling performance by transforming the morphology and structure of heat sinks, thus receiving extensive attention. However, in the TO process, the temperature distribution of modules and heat sinks needs to be calculated in each iteration step, consuming a large amount of computing resource and calculation time. To accelerate the TO process of traditional heat sinks, a fast iterative method combining neural network (NN) synchronous learning and the traditional solid isotropic material with penalization (SIMP)-based TO methods is put forward. First, an NN prediction model based on the encoder-decoder structure is constructed, which can iteratively evolve the shape of heat sinks to achieve a fast prediction of optimized structures. Second, the NN model is integrated into the TO process of the heat sink based on the SIMP method, and the NN is trained synchronously using the intermediate morphology obtained in the iteration process. Finally, aimed at the single-chip and dual-chip modules, the results obtained by the new method and traditional iterative methods are compared to validate the accuracy and rapidity of the proposed NN synchronous leaning method.

Power semiconductor devices are the core of electric energy conversion and electric drive based on the power electronics technology, which have broad application prospects in new energy generation, transportation, aerospace and other fields. However, the problems such as degradation, failure and reliability caused by heat generation have become bottlenecks that limit their further development, and it is urgent to explore effective thermal management methods to improve their reliability and service life. In this paper, based on the introduction of thermal management methods for power modules, the research progress in active thermal management methods is reviewed in detail, and these methods are divided into device-level, system-level and multi-parameter comprehensive methods according to the difference in control parameters. In addition, various methods are analyzed and compared. Finally, the development trend and prospect of technologies for power devices which are related to junction temperature are put forward, providing a reference for the subsequent research and applications.

As a novel and extensively applied switching device, silicon carbide metal-oxide-semiconductor field-ef fect transistor(SiC MOSFET) offers a faster switching speed and lower device loss in practical applications, thereby en-hancing the converter efficiency and delivering a superior performance. Aimed at the driving characteristics of SiC MOS-FET, the influence of parasitic parameters on its performance was analyzed. To investigate the relationship between the gate-source voltage and turn-on time of SiC MOSFET, a two-pulse experimental platform was also established. However, there are certain drawbacks with the existing domestic SiC MOSFET. Based on the experimental platform and other power products, the changes in conduction time, driving loss and negative voltage amplitude after replacing the imported SiC MOSFET with domestic devices were analyzed.

Aimed at the problem that DC-DC converters with coreless transformers must operate at high frequencies, a DC-DC converter with a coreless transformer based on free decay oscillation is proposed. First, a circuit model in series-series topology with no excitation sources is established, the evolution of the eigenvalue is discussed, and the free decay oscillation behavior of the system is analyzed. Second, the DC-DC converter with a coreless transformer based on free decay oscillation is designed. The switching frequency of this converter can be significantly lower than the operating frequency of the coreless transformer, and the output power from the converter and the operating frequency of the coreless transformer can be described by the eigenvalue of the free decay oscillation system. Finally, an experimental prototype was constructed, which was under constant duty control. The switching frequency of the converter was reduced to one half(103 kHz) and one third (69 kHz) of the operating frequency of the coreless transformer (206 kHz), respectively. The converter efficiency achieved 91.2%, and the transformer efficiency was always higher than 96%.

The large grid inductance in weak grid may cause a grid-connected converter to be unstable. Therefore, an impedance model of grid-connected converter is built at first, and the influence of grid inductance on the stability of grid-connected converter is analyzed according to the impedance ratio criterion. Then, aimed at the problem of low adaptability of the grid-connected converter to inductive grid impedance, a virtual impedance control strategy based on band-pass filter is proposed, and the influence of virtual resistance value on the adaptability of grid-connected converter to weak grid is studied. Furthermore, a selection principle for the virtual resistance value is also given. Finally, a system simulation model is built, and simulation results verify the correctness of theoretical analysis and the effectiveness of the proposed control strategy.

A single-phase transformerless inverter for photovoltaic power generation is proposed, which has a common ground between its input and output and thus eliminates the common-mode current in the photovoltaic system. Meanwhile, its voltage gain is higher than that of the traditional single-phase quasi-Z-source inverter. In addition, the use of thin film capacitor makes the inverter more reliable and prolongs its service life. The working principle of this inverter is also analyzed in detail. Finally, a 140 W prototype was built, and results verified the correctness of the proposed inverter topology and the effectiveness of the control strategy in islanding and grid-connected operation modes.

The magnetic leakage from an indoor air-core reactor leads to problems such as serious nearby electromagnetic pollution and heat generation by metal equipment. To solve these problems, based on the ANSYS finite element simulation platform, the indoor air-core reactor and its surrounding ancillary facilities and building structure of a 500 kV substation are modeled in three dimensions, and the building roof is equipped with a shielding plate. Through the optimization analysis of shielding materials, thickness and gap width, a scheme is formulated to improve the electromagnetic environment surrounding the indoor reactor. Simulation results show that the magnetic induction intensity of the steel structure can be effectively reduced by adding a shielding plate which is overlapped by 3 mm thick aluminum plates above the steel beam. The temperature change of the reactor during its operation meets the insulation heat resistance requirement for the air-core reactor. The edge section temperature of the shielding device is higher, with a peak of 53.26 °C. The average temperature of surrounding buildings and facilities increases by about 5-10 °C, which meets the requirements of safe operation.

The three-port converter studied in this paper includes three ports, i.e., power supply, energy storage and load, in which the load can directly obtain electric energy from the power supply or the energy storage port. First, a mathematical model of Euler-Lagrange form is established, a passivity-based controller is designed, and the simulation of passivity-based control (PBC) is carried out. The PBC strategy has advantages of strong global stability and strong robustness to system parameter deviation and external disturbance in the nonlinear system. Based on the PBC theory, the passivity of the three-port converter system is demonstrated. Through the PBC and single-phase-shift control of dual-active-bridge (DAB), the energy transfer between the three ports is realized. Finally, simulation results in Matlab/Simulink are given, which proves that the load in the device can directly obtain electric energy from the energy storage port and run safely, stably and reliably.

A soft-switching DC-DC converter with low current ripple and high gain is proposed, which can be applied to new energy generation systems. Based on the conventional interleaved Boost converter, the proposed converter can achieve high gain by introducing a coupled inductor, diodes and a capacitor Boost unit. The coupled inductor transmits energy during the entire switching cycle, thus improving the utilization rate of magnetic core. The input Boost stage works in an interleaved mode, and the current ripple of the two-phase inductor can cancel each other, so as to obtain a lower input current ripple. Due to the existence of leakage inductance of the coupled inductor, the reverse recovery problem of rectifier diodes are alleviated. Meanwhile, an active clamp circuit is adopted to absorb the leakage inductance energy, thereby achieving the zero-voltage soft-switching of all switches, restraining the turn-off voltage spike of switches, and improving the converter's conversion efficiency. The working principle, circuit characteristics and soft-switching realization method of the converter are analyzed in detail. Finally, a 200 W experimental prototype was built to verify the theoretical analysis.