Розробка IPSO-TVAC для адаптивного керування мережеутворюючими інверторами в мережах з низьким SCR з апаратними обмеженнями

The object of the study is the REGFM-B1-based battery energy storage system (BESS) grid-forming inverter control system operating in weak grids with a short-circuit ratio (SCR) below 2.0. The work focuses on optimizing the virtual synchronous machine (VSM) controller parameters using an improved particle swarm optimization with time-varying acceleration coefficients (IPSO-TVACs). The primary challenge addressed is the failure of conventional tuning methods to converge under three concurrent hardware constraints, including current saturation at 1.2 pu, measurement latency of 10 ms, and ADC quantization noise of 0.01 pu, which form a non-convex search landscape. The proposed IPSO-TVAC is benchmarked against standard PSO (Std-PSO) and gradient-based algorithms, which often converge to physically infeasible solutions under the specified hardware restrictions. The findings reveal that IPSO-TVAC greatly outperforms the standard approaches, with integral of time-weighted absolute error (ITAE) decreased by 16.1%, convergence standard deviation below 1 × 10⁻4, and active power ripple lowered from 0.03 pu to below 0.005 pu. These gains suggest that IPSO-TVAC is highly effective in robust transient performance across all investigated constraint combinations. The method’s major benefit lies in its fractional-order inertia decay and derivative-penalized cost function, which enable simultaneous management of current-saturation non-convexity and ADC noise sensitivity within a single optimization cycle. The findings imply that IPSO-TVAC is especially advantageous for utility-scale battery storage in distant microgrids, island-grid and offshore wind farms, where consistent frequency stability under inverter overcurrent restrictions and ADC noise during grid transitions is critical