Reinforcement Learning-Based Smart Bidding Strategy in Electricity Market for Effective Distributed Consumer Participation

Distribution operators can change consumer usage patterns through a variety of demand-side management strategies, thanks to smart grid technology. This entails the utility and the customers working together to modify customer loads in a way that benefits the stakeholders in various sectors. The main challenge posed by system operators nowadays is creating simulations of the system market systems that effectively choose generation resources under changing load patterns. An agent-based bidding technique for markets is presented in this article and implemented for a five-bus grid. The real power load model is created to enable connections to load buses from various load profile types, including commercial, industrial, and residential. A comprehensive DC OPF model serves as the basis for the Locational Marginal Prices (LMPs) and using Reinforcement Learning, GenCos proactively modify their supply offers over time. Two scenarios are compared relating the features of the day-ahead market: one in which the load is connected as a single, homogenous entity, and the other in which it is segmented into various client kinds.