Optimization of bidding strategy in the day-ahead market by consideration of seasonality trend of the market spot price

Due to the liberalization of the electricity market, evaluation of competitor behaviors, as an uncertainty factor, is a critical information for a Generation Company (GenCo) to maximize its profit by optimizing bidding strategies. In this paper, a new bidding strategy model has been presented based on the Genetic Algorithm and a refined Monte Carlo simulation model. This process is done through the similarity function and consideration of the seasonality trend as the main characteristic of the electricity spot price. The main contributions of this paper include: (a): Consideration of the similarity value for all days in historical dates in the database, (b): Consideration of the seasonality trend of market clearing price by applying K-Means algorithm for clustering historical data based on demand, (c): Application of the proposed model for each cluster's data, (d): Performance evaluation of the fitness function of each generated strategy by a simulation model based on historical data. The proposed model has been tested for the 10 subsets of Iran's electricity market 2016. The obtained results show that the proposed model is statistically efficient, and the prediction accuracy of MCP by the proposed model can be improved by more than 25% and 11% compared with a simple simulation model and the hybrid of simulation and Q-learning model. © 2020 Elsevier Ltd