Bidirectional converter with spike suppressed switching and optimized neuro restrictive steady-state controller for hybrid AC/DC microgrid

Abstract

A bidirectional converter (BC) is the core of a hybrid AC/DC microgrid that connects the AC and DC subgrid. However, existing hybrid AC-DC converter technologies continue to have major drawbacks and obstacles including high switching loss and stability issues. Hence a novel, Bidirectional Converter with Spike Suppressed Switching and Optimized Neuro Restrictive Steady-State Controller has been proposed to reduce switching loss and improve the steady-state performance of the microgrid. The existing converter-controlling methods fail to account for parasitic effects that impact multi-component interactions, specifically magnetic coupling during power conversion.  To overcome this issue a novel Spike Suppressed Switching Circuit is proposed, which uses a Transient Voltage Suppresser Diode (TVSD) and Field Decoupler to reduce switching loss during converter operation. The Synergic TVSD effectively absorbs voltage spikes and reduces the extreme voltage damage to converter components. The Dualistic Field Decoupler, which consists of the Twin Ultra-Fast Recovering Diode (TUFRD) and the Mid-Snubber unit, reduces magnetic coupling between inductors and transformers, thus lowering eddy current production. Moreover, the wind-up effect accumulates errors in integral terms, which leads to an overshoot or a protracted settling period when the system returns to the operational range. Therefore, an Optimized Neuro Restrictive Integration Regulator is proposed, in which the Saturation Attending Freezing Algorithm with an optimized RBFN (Radial Basis Function Network) is used to reduce the wind-up effect and enhance the steady-state performance of the Hybrid AC/DC microgrid. The results obtained from the proposed model have a low switching loss, high efficiency, and low error.