基于随机森林的短期电量负荷精准预测方法

Accurate Short-Term Power Load Forecasting Method Based on Random Forest

为解决电能供给量增加,短期电量负荷情况难以预测,无法制定准确的电能分配策略的问题,本文采用基于随机森林的短期电量负荷精准预测方法,深入分析短期电量负荷预测影响因素(气象、时间、电价与随机干扰因素),选取适当的模型输入变量,包括历史电量负荷数据、温度数据与日类型等内容,使用随机森林算法构建短期电量负荷预测模型,并重复确定相似日的选取规则;采用粒子群优化算法寻找预测模型参数最佳值,将样本集输入至模型中,解决短期电量负荷预测误差较大的问题。得出结论:当输入变量数量达到一定值后,短期电量负荷预测时延稳定在0.55s左右,短期电量负荷预测误差接近为0,由此分析得出短期电量负荷精准预测方法应用性能较佳。

In order to solve the problem of increasing electricity supply,difficulty of predicting short-term electricity load,and incapablity of generating accurate electricity distribution strategy,this paper uses the short-term electricity load accurate forecasting method based on random forest,deeply analyses short-term electricity load forecasting influencing factors(weather,time,electricity price and random interference factors),and selects appropriate model input variables,including historical electricity load data,temperature data and day type,etc.The short-term electricity load forecasting model is constructed by combining the stochastic forest algorithm,and the selected rules of similar days are determined repeatedly.The particle swarm optimization algorithm is used to find the best value of the parameters of the forecasting model,and the sample set is input into the model to solve the problem of large short-term electricity load forecasting error.It is concluded that when the number of input variables reaches a certain value,the short-term electricity load forecasting delay is stable at about 0.55s,and the short-term electricity load forecasting error is almost zero.Through the result analysis,it is concluded that the short-term electricity load accurate forecasting method has better application performance.