To solve the problem of identification and control of potential harmful transactions in electricity market

a method for identifying potential harmful behaviors based on the cloud model and the fuzzy Petri net is proposed to realize real-time warning of potential hazards in the transaction process. The cloud model and the fuzzy Petri net are applied to pattern recognition

which effectively solves the problem that the potential hazard behavior is difficult to quantify

and overcomes the defects that the general recognition method is not easy to understand. Firstly

a general framework and procedure for identifying potential harmful behaviors is designed. Secondly

two methods of building feature cloud based on reverse cloud generator and expert description are described. Meanwhile

a fuzzy Petri net structure and recognition algorithm for potential harmful behaviors is designed. Finally

an example is made and analyzed by using the real transaction data of monthly auction market in Jiangxi Province

which shows the results given by the proposed method are highly consistent with the results obtained by artificial analysis and actual data.