Based on the previous literature research and interviews, this study extracts fourteen key variables from the current Internet advertising traffic transactions to construct agents-based model of Internet advertising traffic with the help of computer simulation software Netlogo and focuses on the effects of six factors which include income, cost, net income, growth threshold, the proportion of fraudsters and the initial number of demanders of traffic transactions on the market equilibrium through six simulation experiments. It is found that the market environment factors, such as income, cost, net income and the proportion of fraudsters, have impact on the market equilibrium more significantly than the factors about demanders. High profit temptation and fraud tolerance are equal in the impact on market equilibrium. In addition, the model also provides a way to explore the critical state of the market (between full withdrawal and market saturation), which also provide some enlightenment on the governance of Internet advertising traffic fraud.