摘要
随着国家对光伏产业的大力推进与扶持,考虑到光伏发电不同于传统发电方式的独特性,以及国家补贴政策的长期性,出现了很多以虚假记录发电量骗取光伏补贴的行为,称光伏窃电。针对现有的分布式光伏窃电手段,提出了1种考虑天气类型指数基于神经网络算法的防光伏窃电智能检测办法。通过将天气类型划分为特殊日与常规日,以历史气象数据为基础对待识别日进行分时发电量预估,结合实时发电数据计算出各类型窃电系数,以此对光伏窃电程度进行有效判定。仿真结果表明,该方法对光伏窃电行为识别的有效性。
With the country's strong promotion and support of the photovoltaic industry,considering that the unique characteristics of photovoltaic power generation is different from the traditional power generation,as well as the long-term national subsidy policy,there have been a lot of false record power to deceive PV subsidies,named as PV stealing.This paper proposes an intelligent detection method for PV stealing based on neural network algorithm with considering weather type index.By dividing the weather type into a special day and a regular day,the time-ofday generation is forecasted based on the historical weather data,and various types of stealing coefficients are calculated combined with real-time power generation data,which can effectively determine the degree of PV stealing.The simulation results show that the method is effective to identify the behavior of PV stealing.
出处
《分布式能源》
2017年第4期13-19,共7页
Distributed Energy
关键词
光伏窃电
天气类型
窃电系数
常规日
特殊日
PV stealing
weather type
stealing coefficient
regular day
special day
