基于K-means算法和积灰损耗系数的中国西北地区光伏电站清洗策略建模分析

MODELING AND ANALYSIS OF CLEANING STRATEGY FOR PV POWER STATIONS IN NORTHWEST CHINA BASED ON K-MEANS ALGORITHM AND DUST ACCUMULATED LOSS COEFFICIENT

针对当前中国西北地区光伏电站缺乏科学合理的清洗策略来清除光伏组件表面积灰的问题,提出基于K-means算法和积灰损耗系数的光伏电站清洗策略模型。利用K-means算法对光伏电站的发电日志进行聚类得到晴天数据样本,利用该样本数据拟合光伏电站发电能力与积灰持续时间曲线,求解积灰损耗系数,联立光伏电站清洗成本建立清洗策略模型;以中国西北地区某光伏电站为例进行分析,将得到的清洗经济阈值与该光估电站长期的运行经验模型计算结果对比后,发现该模型准确性较高。以期本研究为中国西北地区光伏电站清洗策略的研究提供新思路。

In response to the lack of scientific and reasonable cleaning strategies for removing surface dust from PV modules in the Northwest China,this paper proposes a PV power station cleaning strategy model based on the K-means algorithm and dust accumulation loss coefficient.Cluster the power generation logs of PV power station using the K-means algorithm to obtain sunny data samples.Use this sample data to fit the curve of PV module power generation capacity and dust accumulation duration,solve the dust accumulation loss coefficient,and establish a cleaning strategy model in conjunction with the cleaning cost of PV power station;Taking a PV power station in Northwest China as an example for analysis,after comparing the obtained cleaning economic threshold with the long-term operating experience model calculation results of the PV power station,it is found that the accuracy of the model is relatively high.This study aims to provide new ideas for the research of cleaning strategies for PV power stations in Northwest China.