In the stock market, it is unrealistic to completely exclude noise traders. However, because noise traders do not have clear investment strategies and do not respond to the rational behavior required by the efficient market hypothesis, it is difficult to conduct economic modeling research on them. The thesis integrates the psychological factors of the behavioral decision-making process into the agent behavioral decision-making model, the investor agent model is built based on prospect theory, and noise traders and long-term investors are distinguished according to the goals of the agent. By changing the agent’s investment objectives and its confidence level, the impact of noise traders’ behavior on the operation of the stock market is explored. It is explored that frequent trading of noise traders and excessive self-confidence of traders will produce higher stock market trading volume, which in turn affects price fluctuations and market returns.