Application of fuzzy Q-learning for electricity market modeling by considering renewable power penetration

By increasing renewable resource penetration, the need for developing fast and reliable market modeling approaches in the presence of these resources has gained greater attention. In this paper, fuzzy Q-learning approach is proposed for hour-ahead electricity market modeling in presence of renewable resources. The proposed approach is implemented on IEEE 30-bus test system. The effectiveness of the proposed approach is evaluated and compared with Q-learning approach for both normal and stressful cases. Simulation results indicate that the proposed approach is able to model electricity market for a range of continuous multidimensional renewable power penetration in considerably less iterations compared with Q-learning approach. Moreover, the probability of finding Nash equilibrium is becoming higher by using fuzzy Q-learning approach, while the other indices such as average social welfare, average of locational marginal prices (LMPs), and average standard of deviation of LMPs do not change considerably.