Online ecological observation data anomaly detection method and system

The invention discloses an online ecological observation data anomaly detection method and system, and the method comprises the following detection processes: carrying out the learning of a data drift detection method from historical ecological observation data, and building a historical data anomaly detection model needed by data drift detection and a data drift segmentation list; on the basis of the historical data anomaly detection model, performing fine adjustment to obtain an online anomaly detection model of the current drift segment, and performing online anomaly detection by the online anomaly detection model; when the training data is insufficient, similar data are matched from the drift segment list of the historical ecological observation data, and the training amount of the fine-tuning data is enhanced. According to the method, drift detection is carried out on the online observation data by utilizing the historical data learning model and the historical data enhancement training sample, and the o