New energy station abnormal data detection and diagnosis method and system fusing domain knowledge

The invention discloses a new energy station abnormal data detection and diagnosis method and system fusing domain knowledge. The method comprises the steps of obtaining new energy station data; constructing a feature project based on the physical characteristics of the new energy station data; an abnormal value recognition model is constructed based on the new energy station data and feature engineering training by using an isolation forest algorithm, and then abnormal value recognition is performed on the new energy station data based on the abnormal value recognition model to obtain abnormal data; and diagnosing the abnormal data based on a pre-constructed expert knowledge base to obtain an abnormal type to which the abnormal data belongs, and outputting a diagnosis result. The method has the advantages of rapid and accurate diagnosis and the like. 本发明公开了一种融合领域知识的新能源场站异常数据检测诊断方法及系统，方法包括步骤：获取新能源场站数据；基于新能源场站数据的物理特性构建特征工程；使用隔离森林算法并基于新能源场站数据和特征工程训练构建得到异常值识别模型，再基于异常值识别模型对新能源场站数据进行异常值识别，得到异常数据；基于预先构建的专家知识库对异常数据进