Autonomous Demand Response Controller Based on Load Voltage Sensitivities and Artificial Intelligence

Increasingly higher renewable penetrations and distributed generation (DG) in the existing power system may require more flexible resources. The classical method of adding more conventional, mainly fossil-fueled, generation to improve system's flexibility is no longer desired in the recent push for more renewable energy sources (RES) integration. This paper introduces a new source of system-level flexibility, through demand response (DR) using a load voltage sensitivity-based control approach. The proposed control approach autonomously manages each DR resource based on local measurements of its voltage and its voltage sensitivity due to the changes in load power. Using the voltage sensitivity as an input helps ensure fair contributions of all DR resources distributed throughout the distribution system without the need for communication. A robust back-propagation function approximation has been used for deriving the controller parameters. Simulation results verify the effectiveness of the proposed autonomous control approach in improving the system voltage profile and its robustness against varying operational settings.