Intelligent Residential Demand Response: Achieving Resilient Voltage Management with Consumer Preference

This paper aims to develop and optimize an efficient residential demand response (DR) strategy, which ensures a reliable voltage supply to customers, reduces network losses, and minimizes load switching. To maintain comfort levels, this strategy considers compensation costs for voltage management and prioritizes consumer preferences. The network voltage management involves the utilization of the modified binary teaching learning-based optimization algorithm (BTLBO) to determine the optimal combination of household appliances for switching. Additionally, the solar and wind power model is trained using the Levenberg Marquardt Artificial Neural Network (LM-ANN) technique. The proposed method is evaluated on a modified IEEE 33-bus distribution system with significant overload and high integration of renewable energy sources. Two worst-case scenarios are examined to improve network voltage magnitude during DR operations. Simulation results demonstrate that the proposed approach, incorporating direct load control, effectively improves network voltage and addresses the challenges associated with high renewable energy integration.