Competition-Oriented Demand Response Strategic Bidding Model for Retailers Considering Backup Scheme

Abstract

With the fierce competition of electricity market, demand response (DR) amount is also traded in day-ahead market. From the perspective of retailers, just considering its inside customers’ characteristic is not enough, the competitiveness of DR bidding also matters, because it depends on the qualification of participating in DR market. Thus, this paper constructs a complicated DR strategic bidding model. Firstly, based on managed residential customers’ DR feature, optimize bidding considering competitors’ risk preference with deep reinforcement learning approach and guarantee the probability of winning DR bid as much as possible. Secondly, in the actual quotation process, the inaccuracy DR declaration amount or retailers’ personal bidding preference, aggressive or moderate style, leads to DR vacancy punishment or overage waste, so that produce the loss of income. Based on previous bidding model, design backup schemes for different types of retailers in advance to reduce loss. Then utilize real case to verify the effectiveness of proposed DR bidding models.