基于数据驱动的港口电动集卡续航能力预测模型研究  

作　　者：黄梓畅 张小锐 于泳[2] 韩德胜 孙会路 雷路 

机构地区：[1]北京经纬恒润科技股份有限公司,北京100015 [2]唐山港口实业集团有限公司,河北唐山063611

出　　处：《科技创新与应用》2025年第12期84-88,共5页Technology Innovation and Application

摘　　要：针对港口电动集卡计算续航里程会产生较大误差的问题,该文研究发现主要是因为乘用车循环工况不适用于港口电动集卡的特殊工况。该文采用基于对抗式生成网络的方法得到代表性循环工况,并利用基于动力学的能耗模型实现更准确的续航里程估计。该方法通过划分短行程,统计短行程特征,对特征降维、聚类等操作将数据划分为不同的短行程子集。用子集训练对抗式生成网络得到港口代表性循环工况,并根据动力学分析建立整车能耗模型,实现对续航里程的估计。通过与实际港口电动集卡续航里程的对比验证,该方法所预测的续航里程误差率仅为4.6%,充分展现该方法在预测港口电动集卡续航里程方面的高度精确性。To address the problem of big errors in calculating the range of electric container trucks in ports,this paper finds that it is mainly because the driving cycle of passenger cars are not applicable to the special working conditions of electric container trucks in ports.In this paper,a generative adversarial network-based approach is used to obtain representative driving cycle,and a dynamics-based energy consumption model is used to achieve a more accurate range estimation.The method divides the data into different subsets of short trips by dividing the short trips,counting the short trip features,and performing operations such as dimensionality reduction and clustering on the features.The subsets are used to train adversarial generative networks to obtain representative driving cycle in ports,and the whole vehicle energy consumption model is established based on kinetic analysis to achieve the estimation of range.Through the comparison and validation with the range of actual port electric container trucks,the error rate of the range predicted by this method is only 4.6%,which fully demonstrates the high accuracy of this method in predicting the range of port electric container trucks.

关 键 词：循环工况 对抗式生成网络 能耗模型 续航里程 电动集卡 

分 类 号：U469.62[机械工程—车辆工程]