The electricity industry consists of multiple parties on generation, transmission, distribution, trading and market supervision of electricity. Models for a single electricity firm are only suitable for studying the operational decisions of the particular firm. In order to study the bidding behavior of all the players in the electricity market, this paper proposes a hybrid simulation model (HSM) that combines agent-based simulation (ABS) and system dynamics simulation (SDS). With the proposed hybrid model, some input variables required for one simulation model can be obtained by the output of another simulation model. In order to improve the competitiveness of agents that participate in bidding, this paper incorporates the Reinforce Learning (RL) algorithm into the agents. Each agent can obtain information and adapt to the environment through the continuous interaction with the environment. With the hybrid simulation model, the dynamics of the entire market remain stable, the market clearing prices converge, and the market share is relatively uniform. (C) 2019 Elsevier Ltd. All rights reserved.