Intelligent demand side management in PV–battery microgrids via an optimized adaptive neural network controller

Demand Side Management (DSM) is an effective method to optimize energy use in the commercial and residential sector in order to reduce peak demand, electricity cost and load factor. The conventional DSM strategies are based on rule-based systems, which have limited adaptability under dynamic operating conditions. To overcome this limitation, this paper proposes an Adaptive Artificial Neural Network (A2N2) based energy management strategy optimized by improved Doll maker Optimization (IDMO) for PV-battery microgrids. The proposed A2N2 controller receives the time varying system parameters such as PV generation, load demand and battery State of Charge (SoC) to determine optimal load shifting and energy management decisions. The Doll maker optimization (DMO) is improved using the chaotic equation, which increases the convergence speed and provided better accuracy. The proposed A2N2-IDMO has a lower operational cost of 2291.36 USD, a lower peak power demand of 88.48 kW and a lower standard deviation of load demand of 0.73 kW, which indicates better load scheduling performance. The simulation results in MATLAB confirm that the proposed DSM strategy is effective for optimal utilization of energy, cost reduction, and adaptive control for PV–battery microgrid applications.