Abstract: Under the goals of carbon peaking and carbon neutrality（CPCN）, how to invoke and stimulate flexibility of resources and form reasonable energy structure becomes an urgent problem to be solved during the construction of the new-type power system. In order to fully reveal the influence of flexible resources and incentive mechanisms, we came up with an equilibrium analysis method of electricity market combining bi-level model and deep reinforcement learning. Firstly, we sorted out the flexible resources towards the goals of CPCN such as new energy, pumped storage and demand response, and from the dimensions of flexible resources integration, optimizing dispatching and reasonable settlement in spot market, multiple kinds of auxiliary service products and joint optimization with the main market, long-term investment leading, etc., to classify and conclude the market mechanisms. Secondly, the bi-level equilibrium model was constructed by the decision-making models of various kinds of resources, and joint clearing model of electricity and flexible ramping products, and the deep reinforcement learning was adopted for iterative solution. The model and algorithm together constituted a general equilibrium analysis method suitable for multiple flexible resources to participate in different market mechanisms. Finally, by constructing examples with a high-proportion new energy, we simulated market equilibrium results in a variety of scenarios, including different flexible resource combinations, various kinds of incentive mechanisms and different power structures, and further discussed the function of flexible resources and incentive mechanisms. It can provide references for the construction of new-type power system and the design of market mechanism for the goals of CPCN.