混动电四驱车辆纵向车速估计  

Longitudinal Velocity Estimation for Four Wheel Drive Hybrid Electric Vehicle

作　　者：李欢 王金航 刘东升 陈立华 崔光日 徐浩 Li Huan;Wang Jinhang;Liu Dongsheng;Chen Lihua;Cui Guangri;Xu Hao(Automotive Research&Development center,Guangzhou Automobile Group Co.Ltd,Guangzhou 511434,China)

机构地区：[1]广州汽车集团股份有限公司汽车工程研究院,广东广州511434

出　　处：《湖北汽车工业学院学报》2025年第1期33-39,共7页Journal of Hubei University Of Automotive Technology

基　　金：广东省科技计划项目——广东省汽车清洁动力与能源应用技术重点实验室(2024B1212020006)。

摘　　要：混合动力汽车的实时纵向车速是车辆稳定性控制和动力系统控制的重要状态参数。针对四驱车辆各车轮驱动打滑特性导致实时纵向车速直接测量可靠性不足的问题,提出基于轮速传感器、加速度传感器和车辆模型的融合在线车速估计算法。为减少高阶非线性模型带来的标定难度和鲁棒性挑战,建立了混动电四驱车辆系统动力学和车辆运动学的二阶线性估计模型,基于伦伯格观测器设计了自适应调整观测增益的四驱纵向车速估计算法。通过实车验证,在冰面、雪面、城市道路等不同附着路面和不同车速下,该算法估计精度达到均方根误差3 km·h^(-1)、相对误差6%以内。Longitudinal velocity is a key parameter for both vehicle stability control and powertrain control.To address the issue of insufficient reliability in directly measuring real-time longitudinal velocity due to the wheel slip characteristics of four-wheel-drive(4WD) vehicles,an algorithm to estimate velocity online featured by the fusion of wheel speed sensors,acceleration sensors,and vehicle models was proposed.To reduce the calibration difficulty and robustness challenge of high-order nonlinear models,a second-order linear estimation model for 4WD hybrid electric vehicle systems dynamics and vehicle kinematics was built.Based on the Luenberger observer,a 4WD longitudinal velocity estimation algorithm with self-adaptive adjustment of observation gain was designed.Validated through real vehicles,under different road surfaces adhered by various materials,such as ice,snow,and pitch,and different vehicle velocity levels,this algorithm can achieve an estimation accuracy with a root mean square error within 3 km·h~(-1) and a relative error within 6%.

关 键 词：混动电四驱系统 纵向车速估计 模型线性化 伦伯格观测器 增益自适应 

分 类 号：U461.1[机械工程—车辆工程] TP273[交通运输工程—载运工具运用工程]