Power quality enhancement based disturbance rejection controller in microgrid system using walrus optimization algorithm and multi-functional recurrent fuzzy neural network

• A novel FOADRC control strategy is proposed for enhancing power quality in microgrid systems. • The Walrus Optimization Algorithm (WOA) is applied to fine-tune the controller parameters effectively. • Multi-Functional Recurrent Fuzzy Neural Network (MFRFNN) predicts control parameters for adaptive regulation under varying conditions. • The method significantly reduces total harmonic distortion (THDv) and improves voltage profile across multiple scenarios: sag, swell, imbalance, and fault conditions. • Four comparative investigations are presented of control parameters, various optimizations, other controllers, and robustness assessment under noise and dynamic conditions aimed at enhancing the efficiency and resilience of the UPQC scheme. The incorporation of renewable energy systems into the power utility substantially affects the electrical system’s power quality (PQ). Despite their intermittent nature, renewables are viewed as unreliable. To evaluate their impact on voltage superiority and harmonic assessment, the deployment of a Unified Power Quality Conditioner (UPQC) has emerged as a highly effective alternative. This research concentrates on employing UPQC to tackle PQ challenges linked with renewables, encompassing voltage fluctuations, harmonics, and fault conditions. The UPQC’s control is executed through a Fractional Order-Active Disturbance Rejection Controller (FOADRC) that employs the Innovative Walrus Optimization Algorithm (WOA) to ascertain optimal gain values for the control method across numerous PQ scenarios, while the Multi-Functional Recurrent Fuzzy Neural Network (MFRFNN) predicts control parameters for adaptive system regulation. The efficacy of the presented approach is confirmed using the MATLAB/Simulink environment under diverse scenarios. Four comparative investigations are presented of control parameters, various optimizations, other controllers, and robustness assessment under noise and dynamic conditions aimed at improving the efficiency and resilience of the UPQC typology. The results illustrate that the introduced control strategy significantly enhances voltage quality and harmonic mitigation in the presence of grid voltage distortions. Specifically, the THD v was minimized from 12.25 % to 1.93 % during sag mode, from 8.95 % to 1.70 % during swell mode, from 8.48 % to 1.22 % during imbalance mode, and from 22.90 % to 1.95 % during fault condition. The WOA-based FOADRC tuning reduced computation time to 116.244 s and an objective function value of 2.371 in the voltage sag scenario. In addition, WOA-MFRFNN maintained > 96 % accuracy at a 5 dB signal-to-noise ratio (SNR), surpassing DWT-ANN (87.6 %). Furthermore, WOA-MFRFNN achieved 63 % faster than DWT-ANN in the case of detected load and renewable energy fluctuations.