基于agent的连续双向拍卖人工金融市场交易机制涌现分析-系统工程专业论文.docxVIP

II II Abstract Agent-based artificial financial market research is a focus of the economic research. In this paper, based on systematic review of the literature of artificial financial market, it was found that: most of the artificial financial market model were based on entirely rational expected utility theory and ignored the bounded rationality of the traders; at the same time, the field lacked a unified simulation system theory. According to the principle of the Agent-based computational economy called ACE, with the use of agents’ autonomy and reactivity characteristics, psychological factors have been put into agents’ behaviors to establish a continuous double auction artificial financial market simulation platform to study the emergence of trading mechanism. First, using object-oriented modeling method, an open and unified architecture of agent-based simulation platform for artificial financial market was proposed. Then, the prospect theory was applied to the model of the behavioral strategies of the traders and an agent-based continuous double auction artificial financial market model was built according to the ACE principle. Further, using AnyLogic modeling software tool, an agent-based continuous double auction artificial financial markets simulation platform was developed and implemented. Finally, simulation experiments of the continuous double auction artificial financial market were carried out at the simulation platform. The experimental results show that: The agent-based continuous double auction artificial financial market model is reasonable and effective; Price limits curb the liquidity of financial markets, price limits are not a one-to-one correspondence to the volatility, the relaxation of price limits can increase the mobility of financial markets, but does not significantly affect the volatility of financial markets; When the tick size is reduced, the quoted spread decreases, the mobility of financial markets increases, and there does not exist