Dynamic Compensation Pricing Scheme for Demand Resources Based on Deep Reinforcement Learning

As demand response is playing a more and more important role in power system operation, the pricing of those kinds of service becomes critical to the economic operation. In this paper, a pricing method based on deep reinforcement learning is proposed to maximize the cumulative gain of the load aggregator. The responsive behavior of demand resources is affected by the price generated by the load aggregator which dynamically adjusts its compensation price in an optimal way to maximize its return in a longer time horizon. Experimental simulation is conducted to illustrate the effectiveness of the proposed pricing model.