基于GRA-SSA-BP神经网络的电力负荷预测方法  

Power Load Forecasting Method Based on GRA-SSA-BP Neural Network

作　　者：金丽丽[1] JIN Lili(Liuzhou Railway Vocational Technology College,Liuzhou 545616,China)

机构地区：[1]柳州铁道职业技术学院,广西柳州545616

出　　处：《红水河》2022年第3期92-96,共5页Hongshui River

基　　金：广西高校中青年教师科研基础能力提升项目(2021KY1395)。

摘　　要：为了解决现有的电力负荷预测方法存在准确率低、效率不高和精度不足等问题,笔者提出基于GRASSA-BP神经网络的电力负荷预测方法。首先利用灰色关联度分析确定电力负荷与温度、压力、湿度和压强之间的关联程度,再利用麻雀搜索算法优化BP神经网络中的权重值和偏置值,解决BP神经网络预测准确度不高的问题。实验结果表明,基于GRA-SSA-BP神经网络的电力负荷预测结果与实测值基本吻合,预测效果理想,决定系数达到0.945,为电力负荷预测提供一种有效的方法。In order to solve the problems of low accuracy,low efficiency and insufficient accuracy of existing power load forecasting methods,a power load forecasting method based on GRA-SSA-BP neural network is proposed.Firstly,the grey correlation analysis is used to determine the correlation degree between power load and temperature,pressure,humidity and pressure.Then,the weight value and bias value of BP neural network are optimized by sparrow search algorithm to solve the problem of low prediction accuracy of BP neural network.The experimental results show that the power load forecasting results based on GRA-SSA-BP neural network are basically consistent with the measured values,and the coefficient of determination reaches 0.945.The prediction effect is ideal,which provides an effective method for power load forecasting.

关 键 词：电力负荷预测 灰色关联度分析 麻雀搜索算法 BP神经网络 

分 类 号：TM715[电气工程—电力系统及自动化]