A synchronous reluctance motor (SynRM) exhibits significant magnetic saturation effect, which will lead to rotor position observation errors when using traditional high-frequency voltage injection methods. To address this issue, a rotor stationary method is employed to detect the flux linkage model of SynRM, and a nonlinear model that accounts for magnetic saturation effect is established. Based on this model, a high-frequency current response model for SynRM is constructed, and its stability is analyzed and optimized. Under heavy-load conditions, the magnetic saturation effect of SynRM intensifies. By introducing injection tilt angle signal compensation in a high-frequency injection reference frame, the instability issue of the rotor position observer is resolved, enhancing the control stability. Under light-load conditions where the saliency of the motor is weak, a q-axis minimum current limitation strategy is designed to improve the motor’s saliency. Experimental results demonstrate that the proposed control method achieves stable operation of SynRM under all the operating conditions at zero and low speeds, exhibiting anexcellent control performance.

To enhance the parameter robustness of the current decoupling controller of a high-speed permanent magnet synchronous motor (HSPMSM), first, a current decoupling controller with high parameter robustness is proposed. Then, a parameter optimization design method is put forward. Based on the zero-pole cancellation principle, the proposed method directly designs the current decoupling controller in the discrete-domain, which can effectively overcome the delay effect of the digital control system under low carrier ratio conditions, thus greatly improving the parameter robustness of the controller. Finally, experimental results based on the HSPMSM verified the effectiveness of the proposed method.

Deadbeat predictive current control (DPCC) has attracted wide attention of scholars in recent years owing to its advantage of fast dynamic response, and a lot of research results have been achieved. To improve the low-speed steady-state performance of DPCC for permanent magnet synchronous motors (PMSM), a modified dead-time compensation method is proposed. First, the voltage error caused by the dead-time and the nonlinearity of an inverter is analyzed. Second, the d- and q-axis reference currents are utilized to replace the filtered currents, and three-phase current reconstruction is carried out considering the one-step delay in DPCC. Finally, the influence of current polarity on voltage error calculation is analyzed, and a modified sine function is adopted instead of the conventional switching function to calculate the compensation voltage. Three DPCC methods were compared experimentally, i.e., no dead-time compensation, conventional dead-time compensation and the proposed dead-time compensation algorithm. Results show that the proposed algorithm can effectively eliminate the current distortion at the zero-crossing point and further reduce the current harmonics.

To solve the problem of poor performance in traditional full-order flux observers (FOFO) applied under low-speed scenarios, first, the composition of an FOFO is taken as the starting point, and stator resistance estimation criterion is designed to address the problem of changes in stator resistance during operation. A feedback gain matrix is designed under the condition of ensuring the stable operation of the system, which is further combined with resistance estimation and rotor electrical angular velocity estimation. Then, a model predictive torque control (MPTC) system for the induction motoris is studied, its value function is optimized, and the delay compensation for the control system is performed. Finally, the improved FOFO is applied to the MPTC system, so as to realize speed sensorless control. The enhancement to the low-speed performance of the induction motor by the improved FOFO is verified by simulations. Results show that the MPTC system with the improved FOFO increases the control performance of motor at low speeds, improves the resistance of motor to load disturbances, and enhances the robustness to stator resistance.

The modular multilevel converter (MMC) is a core component in the voltage sourced converter based high-voltage direct-current transmission, and its safe and reliable operation is crucial for ensuring the security of power supply. Insulated gate bipolar transistor (IGBT) devices in converter valves will produce significant electromagnetic interference (EMI) due to the high du/dt and di/dt, which can easily affect the normal operation of low-voltage control circuits and even lead to system failures. The gate driver circuit is directly connected with the highvoltage and large-current IGBT, and it supplies switch control signals to the system. Therefore, it is the most susceptible unit in low-voltage control circuits. The electromagnetic radiation from IGBT devices and the EMI issues within the gate drive voltage are investigated. First, it is clarified that the distance between the driver circuit and the power module is much less than the wavelength of electromagnetic waves generated by IGBT switching, thus focusing on the near-field radiation issues. With the consideration of the complex layout and wiring on the gate driver circuit and the presence of localized copper plating, it is revealed for the first time that the electromagnetic radiation sources on the gate driver circuit are composed of electromagnetic sources generated by the switching of IGBT devices and the electromagnetic scattering sources caused by copper plating on the gate driver circuit, which exhibit characteristics of multi-source interaction. Then, an actual converter valve sub-module was taken as a research object, and experiments were conducted under various switching conditions to fully and accurately determine the spatial magnetic field distribution characteristics of the converter valve sub-module in terms of location, frequency and intensity. In addition, the electromagnetic scattering characteristics on the gate driver circuit and the induced voltage at the driver signal port under the combined effects of electromagnetic scattering and radiation were also verified.

To achieve the monitoring of line-end and internal coil ground insulation of the stator winding for an inverter-fed machine, an insulation monitoring method based on common-mode circuit is proposed. First, based on the common-mode equivalent model of the motor, the impedance characteristics of the common-mode circuit of the motor system at high-frequency in MHz and medium-frequency in kHz were analyzed. Then, based on the common-mode impedance characteristics, the dominant mode and path of common-mode current were analyzed. Theoretically, the relationship among the oscillation frequency of medium- and high-frequency common-mode current, the terminal ground capacitance and the internal ground capacitance of the stator winding was derived. The variational mode decomposition algorithm was used to separate the two modes of common-mode current, and the insulation degradation statuses of the line-end and internal coil of machine stator winding were monitored through high-frequency common-mode current and medium-frequency common-mode current, respectively. Experimental results show that the proposed method can sensitively reflect the slight changes in the insulation at the line-end and internal coil of machine stator winding.

To enhance the diagnostic accuracy of circuit breaker faults, an improved particle swarm optimization (IPSO) algorithm is used to optimize the relevant parameters of support vector machine (SVM). Then, a novel IPSO-SVM circuit breaker fault diagnosis model is established, in which the IPSO algorithm enhances the traditional particle swarm optimization (PSO) algorithm by incorporating chaotic mapping, dynamically adjusting the inertia coefficient and updating the optimal particle strategy. Therefore, the convergence speed and accuracy of the algorithm are improved. Furthermore, the IPSO-SVM fault diagnosis model is constructed based on several typical high-voltage circuit breaker fault types and the actual operation data of one specific circuit breaker. Finally, a comparison with other circuit breaker fault diagnosis models using the same training dataset is performed. Experimental results show that the IPSO-SVM model outperformed other models in terms of diagnostic accuracy and reduced the need for manual intervention, achieving effective fault identification of fault types and ensuring the reliable operation of the circuit breaker’s opening and closing coils.

DC-DC power supplies are simultaneously affected by long-term low-dose-rate radiation and temperature in a space environment. At present, there is no space radiation-temperature coupling effect acceleration model, and how to evaluate the life of DC-DC power supplies under the coupling effect is a challenge. A life evaluation method based on a generalized coupled acceleration model and a response surface acceleration model is proposed. First, the particle swarm optimization algorithm is used to select the optimal acceleration model and estimate its parameters. Then, the final coupling effect acceleration model is established through model averaging. Finally, a certain type of DC-DC power supply is taken as the object, the coupling relationship analysis of acceleration performance degradation data obtained under different stress combinations is carried out, and the proposed method is used to evaluate the life of DC-DC power supply under the coupling effect, thus verifying the effectiveness of the novel method.

Aimed at the problem of system voltage imbalance caused by single-phase grounding high-resistance faults in inverter lines in an islanding operation mode of microgrids, a zero-axis high-resistance fault detection and fault-tolerant control method based on a residual generator is proposed to reduce the degree of system voltage imbalance. First, a state space model for the dq0 axis of a four-bridge-arm inverter is established, and the impact of high-resistance faults on output voltage is analyzed. Second, based on the theory of left and right coprime decomposition and Euler parameterization, a residual generator is designed to obtain residuals, and a Butterworth bandpass filter is used to generate zero-axis residual components for fault detection. Third, from the perspective of zero-axis disturbance cancellation, a zero-axis fault-tolerant controller is designed using the model matching theory, which is further solved using a linear matrix inequality method. Finally, high-resistance fault detection and fault-tolerant control experiments were conducted using a PXI semi-physical simulation platform to verify the feasibility of the proposed method.

Due to the trade-off relationship between breakdown voltage (BV) and specific on-resistance R_on,sp, to minimize the decrease in breakdown voltage while reducing on-resistance, an improved U-trench metal-oxide-semiconductor field-effect transistor (UMOSFET) structure with PN-Bottom is proposed. First, on the basis of traditional 100 V U-trench metal-oxide-semiconductor (UMOS) devices, deep trench etching technology is adopted to introduce a PN-Bottom wrapping region at the bottom of the trench, which wraps the P-type region inside the N-type region. Then, the new structural parameters and fabrication process of the device are introduced. Next, using the TCAD simulation tool, a detailed discussion was conducted on the doping concentration, width, and depth of the N-Bottom and P-Bottom encapsulation regions, and their effects on the breakdown voltage and specific on-resistance of the device were studied. Finally, the simulation results show that under the same conditions, an optimized design with BV of 110 V and R_on,sp of 1.15 mΩ·cm² was obtained.