节能赛车全路段滑行车速预测  

Energy Saving Race Car Speed Prediction for All Road Sections

作　　者：吴展骞 吴闯 赵治国[1] Wu Zhanqian;Wu Chuang;Zhao Zhiguo(Tongji University,Shanghai 201804)

机构地区：[1]同济大学,上海201804

出　　处：《汽车技术》2021年第12期29-34,共6页Automobile Technology

基　　金：国家级大学生创新创业训练计划项目(G202019002)。

摘　　要：为在节能车比赛中预测加速点位置并提醒车手,利用CarSim软件建立原型车模型开展仿真,将仿真数据用于反向传播(BP)神经网络训练,并通过实车数据进行验证,建立了原型车全工况BP神经网络降速预测模型,实现了不同工况下原型车降速规律和加速点位置的预测,并结合车载设备通过声光提醒辅助车手操作。实车试验结果表明,该模型平均提前5.15 s预测加速点位置,辅助实现了单次平均4.07%的燃油消耗量下降。In order to predict the location of the acceleration point and remind the driver in the energy-saving car competition,the CarSim is used to establish a prototype car model to carry out the simulation,and the simulation data is used for Back Propagation(BP)neural network training,and the model is verified by vehicle data.The deceleration prediction model based on BP neural network of this energy-saving car realizes the prediction of the deceleration and the position of the acceleration point under different working conditions.This model is combined with the on-board equipment to remind the driver by means of sound and light.Vehicle test results show that the model predicts the location of the acceleration point 5.15 s in advance,and assists in achieving a single-time average 4.07%reduction in fuel consumption.

关 键 词：节能车 降速模型 反向传播神经网络 

分 类 号：U467[机械工程—车辆工程]