基于能耗参数辨识和路况预测的剩余续驶里程数据驱动算法

Remaining Driving Range Data-Driven Algorithm Based on Energy Consumption Parameter Identification and Road Condition Prediction

提出了一种基于大数据监控平台的纯电动汽车剩余续驶里程(remaining driving range,RDR)估计算法。首先利用历史运行数据实现车辆纵向动力学参数辨识,并忽略车辆参数变化和其他附件能耗变化。然后采用k近邻法(k-nearest neighbor,kNN)结合实时监控数据完成道路工况预测,进行准确的能耗估算,最终实现RDR估计。与传统的RDR估计相比,提出的方法能将最大误差控制在3km以内,具有更高的估算精度。

This paper presents a remaining driving range(RDR)data-driven algorithm based on Bigdata platform.Firstly,the historical data is used to realize vehicle longitudinal dynamics parameter identification,without considering changes in vehicle parameters and other accessory energy consumption.Then based on the real-time data,the k-nearest neighbor(k NN)is used to predict the road conditions,and the accurate energy estimation is carried out to finally realize the RDR estimation.Compared with the traditional RDR estimation algorithms,the proposed method can control the maximum error within 3 km,which has a higher precision.