基于改进BP神经网络的新能源汽车销量预测

SalesForecast Of New Energy Vehicles Based On Improved BP Neural Network

针对新能源汽车销量数据包含非随机性波动序列和随机性波动序列,采用标准BP神经网络预测精度不高的问题,提出一种基于改进BP神经网络的新能源汽车销量模型。首先,采用SARIMA模型对新能源汽车销量数据中非随机性波动序列进行预测,采用BP神经网络模型对新能源汽车销量数据中随机性波动序列进行预测;然后,将预测结果相乘,得到新能源汽车销量最终预测结果;最后,通过仿真实验对提出方法进行了验证。结果表明,本研究基于改进BP神经网络可准确预测新能源汽车销量的趋势,平均预测准确率为90.82%,可用于实际新能源汽车月度销量预测。

Aiming at the problem that the sales data of new energy vehicles include non random fluctuation series and random fluctuation series,and the prediction accuracy of standard BP neural network can not meet the actual demand,a new energy vehicle sales model based on Improved BP neural network is proposed by combining SARIMA model and BP neural network model.Firstly,SARIMA model is used to predict the non random fluctuation Series in the sales data of new energy vehicles,and BP neural network model is used to predict the random fluctuation Series in the sales data of new energy vehicles.Then,the prediction results are multiplied to obtain the final prediction results of new energy vehicle sales.Finally,the proposed method is verified by simulation experiments.The results show that the improved BP neural network can accurately predict the sales trend of new energy vehicles,and the average prediction accuracy is 90.82%,which can be used to predict the actual monthly sales of new energy vehicles.