XGBoost model-based electricity market collusion behavior identification method

The invention discloses an electricity market collusion behavior identification method based on an XGBoost model. The method comprises the following steps: firstly, according to electricity market collusion behavior characteristics and modes, establishing generator set collusion behavior identification parameters, and then constructing a data set; thirdly, the XGBoost classification model is trained through the data set, and the trained XGBoost classification model is obtained; and finally, classifying the to-be-identified data so as to complete electricity market collusion behavior identification. The electricity market collusion behavior identification method is beneficial to avoiding the influence of subjective factors in electricity market collusion behavior identification, improves the real-time performance, the identification accuracy and the operability, and is continuously updated according to the development of the electricity market so as to improve the identification effect. 本发明公开了一种基于XGBoost模型的电力市