两阶段CP-Copula的风电机组异常数据清洗算法

A two-Stage CP-Copula Algorithm for Clearing Abnormal Data of Wind Turbine

风电机组在实际运行过程由于运行环境影响及人为调控等因素影响,导致风功率曲线中存在大量异常运行数据,给风电机组的监测与控制带来严重干扰。提出一种基于变点分组(Change Point)和Copula理论组合的两阶段异常数据清洗算法。根据风电机组异常运行数据的分布特征和产生原因,将异常数据划分为堆积型异常数据和分散型异常数据;利用变点算法最大限度的清洗大部分堆积形异常数据和少量分散型异常数据,提高正常数据占比;结合Copula函数计算风速和功率的依赖关系,并依据依赖关系建立基于Copula的概率功率曲线,进一步清洗剩余分散型异常数据。通过内蒙古某风电场实际运行数据验证了算法的有效性,结果表明清洗效果好,可有效识别出三类异常数据,具有一定的工程实用价值。

In the actual operation of wind turbines,due to the influence of the operating environment and human control and other factors,there are a lot of abnormal operation data in the wind power curve,which brings serious interference to the monitoring and control of wind turbines.Therefore,a two-stage abnormal data cleaning algorithm based on the combination of change point grouping and Copula theory is proposed.First,according to the distribution characteristics and causes of abnormal operation data of wind turbines,the abnormal data were divided into accumulation abnormal data and scattered abnormal data.Secondly,the change point algorithm was used to remove most of the accumulated abnormal data and a small amount of scattered abnormal data to the greatest extent,and increase the proportion of normal data.Finally,the Copula function was combined to calculate the dependence of wind speed and power,and the probability power curve based on Copula was established according to the dependence to further clean the remaining scattered abnormal data.The effectiveness of the algorithm was verified by actual operating data of a wind farm in Inner Mongolia.The results show that the cleaning effect is good,and three types of abnormal data can be effectively identified,which can be applied in practical engineering.