Dataset of Wind Power Abnormal Data Cleaning and Quality Control Results for a Wind Farm Based on Combined Model

Effective data cleaning can improve data quality and make new energy power prediction, power generation prediction, and other results more accurate. Collect wind speed and power data from a certain wind farm, and based on exploring the distribution characteristics of the data and the causes of abnormal data, construct a combination model of "DBSCAN quartile" and "variance change rate quartile" to clean the abnormal data.

At the same time, combine the DBSCAN algorithm with the variance change rate criterion to construct a combination model of "DBSCAN variance change rate" and "variance change rate DBSCAN", and carry out quality control research on cleaning multiple combination models. The results indicate that the combination model can effectively identify abnormal data. Among them, the DBSCAN quartile combination model has the highest correlation coefficient between wind speed and power after cleaning, the lowest root mean square error relative to the modeling curve, and the best cleaning effect.