A Type‐3 Fuzzy‐Based Model Predictive Control Approach for Management of Constant Energy

ABSTRACT Despite direct current (DC) microgrids' benefits over alternating current microgrids, these systems confront several difficulties. The instability brought on by constant power loads (CPLs) is one of the main challenges. Because of their rising negative impedance characteristics, CPLs reduce system performance. This study looks at a DC microgrid that has a photovoltaic (PV) battery grid connected to a CPL. To tackle this challenge, a disturbance observer and a continuous‐time model predictive control based on type‐3 fuzzy logic systems (T3‐FLSs) are applied to the bidirectional DC‐DC converter. Even in the face of load and PV generation fluctuations, this method aims to regulate the nonlinearity of the CPL and preserve system stability across a broad operating range. A new learning scheme based on square root cubature Kalman filter is developed for online deep learning of identified T3‐FLS model. Finally, the proposed strategy is compared with the conventional model predictive control method, which demonstrates the superiority of the suggested method in terms of flexibility, reduced response time, reduced fluctuations, and improved stability. Tests were performed using a variety of circumstances, including variations in PV power and CPL load, to assess the system's performance. All these experiments and simulation results were performed in the MATLAB software. Simulation results show that the proposed network has very good performance in terms of dynamic response, precise tracking, and stability over a wide range. The voltage fluctuations are improved by about 20%. The suggested algorithm significantly impacts the energy efficiency of the microgrid by voltage stabilizing, enhancing microgrid adaptability to changing conditions, disturbances, increasing renewable integration, and reducing operational costs.