混合动力车辆无线转矩测试系统设计及参数辨识  

Design and parameter identification of the wireless torque test system for hybrid vehicles

作　　者：李杰 周一健 耿冲 庞金录[3] LI Jie;ZHOU Yijian;GENG Chong;PANG Jinlu(School of Mechanical-electronic and Vehicle Engineering,Beijing University of Civil Engineering and Architecture,Beijing 100044,China;Beijing M-Stars Technology Co.,Ltd.,Beijing 101103,China;Beijing North Vehicle Group Corporation,Beijing 100072,China)

机构地区：[1]北京建筑大学机电与车辆工程学院,北京100044 [2]北京鼎昱晨星技术服务有限公司,北京101103 [3]北京北方车辆集团有限公司,北京100072

出　　处：《机械传动》2025年第4期93-99,共7页Journal of Mechanical Transmission

基　　金：国家自然科学基金项目(51675494);北京建筑大学金字塔人才培养工程(JDJQ20200308)。

摘　　要：【目的】在复杂路况下,车辆实际行驶产生的转矩在传统的台架测试中动态精度较差,难以模拟真实行驶状态;而且实车测试存在空间狭小、振动较大以及环境恶劣等问题。对此,创新性地设计了一种基于无线技术的嵌入式转矩测试系统。【方法】考虑车辆传动系统内部空间结构,基于遥测技术建立了转子、定子系统,并对试验主体进行了模块定制化设计,构造了一套动态性能强、抗干扰性好的测试系统;进行道路试验以获得实车转矩、转速数据;采用系统辨识法对试验数据进行迭代分析,获得6阶辨识值,建立转矩辨识模型。【结果】试验系统标定线性度为0.703%,线性误差小,精度较高;辨识模型均方根误差为1.6785;在急加速、缓加速、紧急制动3种动态工况下,模型最大误差分别为8.4569、1.78、5.4 N·m,平均相对误差在4%以内。上述结果验证了测试数据的可靠性,为转矩模型的精准预测提供了有效的数据基础。[Objective] In complex road conditions,the dynamic accuracy of the torque generated by the actual running of the vehicle is poor in the traditional bench test,it is difficult to simulate the real driving state,and the real car test has problems such as narrow space,large vibration and bad environment.To solve this problem,an innovative embedded torque test system based on the wireless technology was designed.[Methods] Considering the internal spatial structure of the vehicle transmission system,a rotor-stator system was established based on the telemetry technology.A modular customized design was implemented for the test subject,resulting in a test system with strong dynamic performance and excellent anti-interference capability.Road tests were conducted to obtain the real vehicle torque and speed data.System identification method was employed for the iterative analysis of the test data,yielding a 6th-order identification value and establishing a torque identification model.[Results] The calibration linearity of the test system reaches 0.703%,demonstrating small linear error and high accuracy.The root mean square error of the identification model is 1.678 5.Under three dynamic conditions—rapid acceleration,gradual acceleration,and emergency braking,the maximum model errors are 8.456 9 N·m,1.78 N·m,and 5.4 N·m respectively,with average relative errors within 4%.These results validate the reliability of the test data and provide an effective data foundation for the accurate torque model prediction.

关 键 词：混合动力 动态转矩 无线测试 系统辨识法 精准预测 

分 类 号：U467.21[机械工程—车辆工程]