Method for detecting wind speed-power abnormal data of wind generating set

Abstract

The invention discloses a method for detecting wind speed-power abnormal data of a wind generating set. The method comprises the steps of collecting the operation state data of the set, and performing first anomaly detection on the abnormal data according to the set operation state corresponding to the wind speed-power data; performing second anomaly detection on the data passing the state detection according to the constructed normal data main band range; constructing a DBSCAN model, and performing third anomaly detection on the data passing the state detection and the power detection; performing independent anomaly detection on the data above a rated wind speed; and outputting normal data passing the detection. According to the technical scheme, the physical mechanism and data morphological characteristics are formed by integrating the abnormal data to construct the technical scheme, preliminary detection of the wind speed-power abnormal data is achieved, further detection is conducted through the DBSCAN model, final detection of the abnormal data is achieved in combination with a credibility mechanism and detection of the data above the rated wind speed, and therefore, the detection of the wind speed-power abnormal data of the wind generating set has higher accuracy.

Description

technical field [0001] The invention relates to the technical field of wind power generation, in particular to a method for detecting abnormal wind speed-power data of a wind power generating set. Background technique [0002] As an important information management method in wind farms, SCADA data acquisition and supervisory control system (SCADA) is of great significance to the control optimization and performance analysis of wind turbines. Therefore, cleaning the SCADA data and improving the quality of the SCADA data is of great significance for improving the operation analysis and economic analysis of wind power generating units. [0003] Some data show that the existing SCADA unit operation data anomaly detection methods can be mainly divided into the following categories: 1. Anomaly detection technologies based on mathematical statistics theory, such as variance cleaning, quartile cleaning, etc., such technologies are mainly based on abnormal data The feature of being ...

AI Technical Summary

Problems solved by technology

[0006] The present invention mainly solves the technical problems of the original technical scheme that a large number of data points are mistakenly or omitted to be deleted, the robustness is poor, and the distribution of scattered points is not smooth after data cleaning, and provides a method for wind speed-power abnormal data detection of wind power generating sets The method of combining the physical mechanism of abnormal data formation and data morphological characteristics to construct a technical plan, using wind turbine status detection and power detection to realize the preliminary detection of wind speed-power abnormal data, using the DBSCAN model for further detection of abnormal data, combined with the reliability Mechanism and data detection above the rated wind speed realize the final detection of abnormal data, making the detection of wind speed-power abnormal data of wind turbines more accurate

Images