Enhancing frequency management in renewable-powered smart grids with metaheuristic optimization of cascaded controller

Abstract

The growing integration of Renewable Energy Sources (RES) needs advanced approaches to ensure power grid stability and reliability. This study improves frequency regulation by using metaheuristic optimization to fine-tune a cascaded Proportional Integral Derivative with Derivative Filter and Proportional Integral (PIDn-PI) controller, addressing issues such as load variability, communication delays, incorporation of RES, Electric Vehicles (EVs), Redox Flow Batteries (RFBs) and thermal power. Simulations are conducted across various load scenarios featuring fixed, variable and realistic disturbances. Sine Cosine Algorithm (SCA), Grey Wolf Optimization (GWO), Particle Swarm Optimization-Grey Wolf Optimization (PSO-GWO), Salp Swarm Algorithm (SSA), Tunicate Swarm Algorithm (TSA), Particle Swarm Optimization (PSO) and Chaos Game Optimization (CGO) are used for this purpose. SCA consistently performs best and ensures the fastest stabilization with minimal frequency fluctuations, overshoot and undershoot. GWO, TSA, and PSO-GWO work well but take longer to reach steady-state. PSO balances performance and computational time, but performs worse than SCA, whereas SSA and CGO have the worst control performance. The performance of SCA: PIDn-PI demonstrates that when there is 1 % load change in both areas, the frequency response in zone 1 and zone 2 settles in 1.64 s and 1.13 s, respectively, whereas it settles in 1.22 s and 1.60 s in area 1 and area 2, respectively, when the load changes by 5 %. Moreover, the robustness of SCA: PIDn-PI is tested under various RES combinations and penetration levels, Denial-of-Service (DoS) cyberattacks, and sensitivity analysis is carried out.