基于PSO-LSSVM的新能源汽车充电负荷预测方法研究

Research on Charging Load Forecasting Method of New Energy VehiclesBased on PSO-LSSVM

“双碳”背景下,新能源汽车充电负荷的持续增长对电网安全稳定运行的影响愈发显著。针对现有的以用户侧为出发点的预测方法存在较多的不确定性、实用性不足的情况,从电网侧的角度,提出了将粒子群优化(PSO)算法与最小二乘支持向量机(LSSVM)应用于充电负荷预测。通过分析影响充电负荷因素的相关程度系数,选择合适的特征量作为输入量,并运用PSO算法优化LSSVM模型参数,建立了基于PSO-LSSVM的充电负荷预测模型。以国网上海市电力公司市北供电公司东升路充电站作为案例,验证了所提出的方法具有较好的数据拟合效果,预测的数据准确度较高。这为新能源汽车充电负荷的预测提供了一种新思路。

Under the background of"dual carbon",the continuous and stable growth trend of new energy vehicle charging load has an increasingly significant impact on power grid operation.In view of the many uncertain factors and insufficient practicability of the existing forecasting methods based on the user side,particle swarm optimization(PSO)and least square support vector machine(LSSVM)are proposed to forecast the charging load from the perspective of the grid side.By analyzing the correlation coefficient of the factors affecting the charging load,the appropriate characteristic variables are selected as the input value.PSO algorithm is used to optimize the model parameters of LSSVM,and the charging load prediction model of PSO-LSSVM algorithm is established.In this paper,the charging Dongsheng Road station state grid shibei power supply company,SMEPC,is taken as an example to verify that the proposed method has good data fitting effect and high accuracy of the predicted data.This research provides a new way to predict the charging load of new energy vehicles.