A hybrid interactive simulation method for studying emission trading behaviors

Understanding human participants´ trading behaviors is important for market designers. When analyzing the sensitivity of a control variable, it is of crucial importance to keep all other variables fixed, as well as human behaviors. In this paper, a hybrid interactive simulation methodology is introduced and utilized to model emission trading behaviors of generation companies. By extracting information from human-subjected experiments under several scenarios, a stochastic model for multi-agent is constructed and verified. Furthermore, influences of different emission reduction level on emission prices are studied based on multi-agent simulation. Combining the complementary features between experimental and agent-based computational methods, the hybrid interactive simulation can help studying human behaviors involved research problems.