PSO-Tuned Predefined-Time Sliding-Mode Control of a Grid-Forming Inverter

This paper introduces a novel predefined-time sliding-mode control (PSMC)-based cascaded approach. Two PSMCs are designed separately to regulate the capacitor voltage and inverter current of an islanded grid-forming (GFM) inverter with an L C filter. Particle Swarm Optimization (PSO) is employed to optimally tune the PSMC parameters, thereby ensuring optimal controller performance. In the proposed design, a new sliding surface and control law are formulated based on the concept of predefined-time stability (PTS), ensuring that the tracking errors converge to the origin within a predefined time. A variable exponent is introduced in both the sliding surface and the control law, guaranteeing PTS while adaptively adjusting the control effort depending on whether the states are far from or near the equilibrium, thereby ensuring fast convergence, robust performance, and avoiding additional chattering. Furthermore, chattering is mitigated by incorporating the integral of the switching term within the control law. Finally, simulation results confirm the effectiveness of the proposed method compared to a commonly used PI controller from the recent literature, particularly under system disturbances such as load variations and reference voltage changes.