Industrial control anomaly detection method for complex uncertain unbalanced data set

The invention provides an industrial control anomaly detection method for a complex, uncertain and unbalanced data set, and the method comprises the following steps: carrying out the targeted training of an anomaly detection model through employing a real, complex, uncertain and unbalanced data set in an industrial control system, and obtaining a basic model # imgabs0 #; the AC-GAN model is trained to reach Nash equilibrium; using the synthesized training data set to train an anomaly detection model; and performing weighted combination on the two results according to a weight strategy to obtain final classification judgment. According to the method, challenges of complex, uncertain and unbalanced data sets are effectively dealt with in the field of industrial control anomaly detection, so that the model is more suitable for data features in a real industrial environment, and the accuracy and robustness of anomaly detection are improved. 本发明提供了一种面向复杂不确定不平衡数据集的工控异常检测方法，包括以下步骤：使用工业控制系统中真实复杂不确定且不平衡数据集，对异常检测模型进行有针