Optimal Scheduling of Hierarchical Energy Systems with Controllable Load Demand Response

A hierarchical energy system management framework based on the demand response is developed to address the impact of controllable loads on the power system. Stimulated electric vehicle clusters (EVs) and temperature-controlled load clusters (TCLs) can quickly respond to the scheduling strategies of load aggregators to reduce the impact on the grid caused by the large number of flexible loads connected to the grid. First, a hybrid model of convolutional neural network and long- and short-term memory network is used to predict each part of the load, and the load aggregator dispatches controllable flexible loads to maximize the fit of the predicted load profile. The load aggregator performs Peer to Peer(P2P) power trading with the power operator according to the current scheduling strategy and applies distributed optimization to solve the maximum benefit for both parties. For the remaining energy demand after local energy trading, a multi-objective optimization model for system operating cost, carbon emission, and wind energy spillover is considered, and the Pareto frontier of this model is solved using NSGA-II with centralized optimization and verified by arithmetic cases in IEEE30 node system. The simulation results show that under the proposed energy management strategy, a balance between the supply and demand of electric vehicles and temperature-controlled loads is achieved while bringing good economic and environmental benefits to the power system.