基于数据驱动的光伏电站理论功率计算方法

Calculation Method of AGC Theoretical Power of Photovoltaic Power Station Based on Data Driving

随着光伏电站的大规模并网,自动发电控制(Automatic generation control,AGC)作为必备的控制功能在各光伏电站广泛应用。在光伏电站AGC控制过程中,当前AGC可控理论功率的计算是亟待解决的问题。本文通过基于相关性的特征选择算法计算出特征相关性高的逆变器作为样板逆变器,然后分别使用线性回归和随机森林建立预测模型,最后以样板逆变器作为输入通过预测模型计算后的输出与实际自然发电的功率作比较,得到误差最小的计算方法,最终得到一种光伏电站理论功率计算的方法。

With the large-scale grid connection of photovoltaic power plants,Automatic Generation Control(AGC)is widely used as a necessary control function in various photovoltaic power plants.In the AGC process of photovoltaic power plants,the calculation of the controllable theoretical power of AGC is an urgent problem to be solved.In this paper,the Feature selection algorithm based on correlation is used to calculate the inverter with high feature correlation as the sample inverter,and then linear regression and Random forest are used to establish the prediction model respectively.Then,the output calculated by the prediction model is compared with the power of actual natural power generation with the sample inverter as the input,and the calculation method with the smallest error is obtained.Finally,a method for calculating the theoretical power of photovoltaic power station is obtained.