A physics-informed deep transfer reinforcement learning (PIDTRL) strategy is proposed for enabling advanced control and modulation of a dc solid state transformer (DCSST) in a dc microgrid. The strategy involves three stages: (i) Centralized training of deep reinforcement learning agents to balance power and reduce current stress in the DCSST; (ii) effective knowledge transfer from a Source-simulation system to a Target-experimental system using minimal experimental data; and (iii) deployment of multiple agents for online control in the DCSST. The proposed method adaptively determines optimal modulation variables (duty cycles and phase shifts) in stochastic and uncertain environments without requiring accurate model information. Experimental results validate the effectiveness of the proposed PIDTRL algorithm.

This paper focuses on the reduction of the radial force harmonics of a consequent-pole permanent magnet vernier machine by current harmonic injection. The current harmonic injection has been applied based on dq-axis frame. In the proposed method, a sixth harmonic of d-axis current is injected to reduce a sixth harmonic of the radial force with zero order mode. The effectiveness of the current harmonic injection is verified by finite element analysis. The simulation results show that sixth harmonic of radial force is reduced by 45.7%, resulting in the reduction of a vibration at the respective harmonic by 42%. To demonstrate the controllability of the current regulator for the proposed method, a circuit simulation has been carried out with a proportional, integral, and resonance (PIR) controllers. The simulation results show that the PIR controller can realize superior performance of the current harmonic regulation compared with the conventional PI controller.

Cascaded H-bridge (CHB) static compensator (Stat-Com) enhances stability of the electricity power grid, which is becoming ever increasingly vulnerable as a consequence of replacing fossil fuel-based synchronous generators by intermitten renewable sources. A CHB StatCom needs large dc capacitors in the submodules to operate, which is a major drawback that hinders a widespread use of this technology. In order to overcome this drawback, the low-capacitance StatCom (LC-StatCom) technology was developed. Other than reducing the capacitors size, improvement in power quality and reduced losses were listed as further advantages of the LC-StatCom. On the other hand, control challenges and reduced stable operating range are the main challenges of this technology. This study aims to provide a better understanding of the prospects and challenges of the LC-StatCom technology as compared to the conventional$C$HB StatComs.

This paper presents a novel zero-sequence voltage injection method for discontinuous modulation (DM) in star-connected cascaded H-bridge (CHB) static compensators (Stat-Coms). Conventional DM methods in CHB StatComs suffer from control interactions during unbalanced grid conditions. This paper proposes a DM that eliminates this control interaction by inserting a clamping transition within a third of the fundamental period. The transition instant is decided by comparing a duty cycle, which is calculated to mitigate the fundamental-frequency in the zero-sequence voltage for discontinuous operation, with a triangular carrier signal with a frequency of three times the fundamental. The clamping intervals under the proposed DM are affected by the phase-shift of this carrier. Specifically, the paper analyses the effects on the harmonic content of the proposed zero-sequence voltage, switching loss, twice-fundamental-frequency capacitor voltage ripple, and current total harmonic distortion, for different phase-shift angles of the DM carrier signal. Besides, the feasibility of the proposed DM in dealing with a grid voltage sag is also discussed.