基于风电机组控制原理的风功率数据识别与清洗方法

Identification and Cleaning of Wind Power Data Methods Based on Control Principle of Wind Turbine Generator System

针对变速变桨距控制风电机组在风功率曲线拟合中所面临的非正常运行数据识别与清洗难题,提出了一种基于风电机组控制原理的风功率数据识别与清洗方法。从控制原理出发,首先通过有功功率和发电机转速对风功率散点进行预处理,然后分析功率时序分布,识别出限功率点,最后结合四分位法对叶尖速比散点和风功率散点进行清洗,在完成功率散点筛选的同时,识别出机组在不同风速下的运行状态。结果表明:本文方法对不同形态特征的风功率散点都具有很好的清洗效果,而且算法执行效率高。

In view of difficulties encountered in abnormal data recognition and cleaning on wind power curve fitting for the variable-speed variable-pitch wind turbine,an effective algorithm based on control principle was proposed.Starting from wind turbine control principle,a data pretreatment of wind power scatters was adopted by active power and generator speed signals,and then the power limiting data was recognized by analyzing the time series distribution of power data.Finally,wind speed tip speed ratio scatters and wind power scatters were cleaned by the quartile method,while screening for power scatters was completed,the operation status of the unit under different wind speeds was identified.Results show that the algorithm developed has a good cleaning effect on wind power scatters with different morphological characteristics,which is of high execution efficiency.