Layered stochastic approach for residential demand response based on real-time pricing and incentive mechanism

摘要

This paper proposes a layered stochastic optimization approach for residential demand response (DR) under real-time pricing (RTP) and an incentive-based mechanism, which contains three steps. In the first layer, an independent system operator (ISO) announces day-ahead RTP to a residential load aggregator (RLA). The RLA predicts individual household loads (step 1) and aggregates the loads to minimize electrical cost (step 2). In the second layer, the RLA announces incentives to homes, and home energy management systems (EMS) control the loads to maximize the reward in real-time (step 3). In Step 1, probability based individual load prediction models are developed. In Step 2, a stochastic optimization model is developed to aggregate controllable loads of residential consumers. In Step 3, an incentive-based mechanism is proposed, based on which, a real-time load control model for individual homes is developed to benefit the RLA and homeowners. A highly efficient real-time control algorithm for home EMS is developed. The case studies show that, with 10% controllable energy integration, the peak demand is reduced by 17.5% and the energy cost of the controllable loads is reduced by 28%. The proposed mechanism can effectively aggregate many individual residential controllable loads to participate in an electricity market.