• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2023, Vol. 59 ›› Issue (14): 190-201.doi: 10.3901/JME.2023.14.190

• 运载工程 • 上一篇    下一篇

扫码分享

轮毂电机独立驱动电动汽车线性时变模型预测主动安全控制

魏洪乾1,2, 赵文强1,2, 艾强1,2, 张幽彤1,2, 王洪荣3, 赖晨光4, 邹喜红4   

  1. 1. 北京理工大学机械与车辆学院 北京 100081;
    2. 清洁车辆北京市重点实验室 北京 100081;
    3. 中国汽车工程研究院股份有限公司 重庆 401122;
    4. 重庆理工大学车辆工程学院 重庆 400054
  • 收稿日期:2022-01-25 修回日期:2023-01-07 出版日期:2023-07-20 发布日期:2023-08-16
  • 通讯作者: 张幽彤(通信作者),男,1965年出生,博士,教授,博士研究生导师。主要研究方向为混合动力汽车的优化设计和控制、汽车内部异构网络的轻量化防护、太阳能汽车设计优化、电机设计与控制等。E-mail:youtong@bit.edu.cn
  • 作者简介:魏洪乾,男,1992年出生,助理研究员。主要研究方向为网控汽车的主动安全控制、汽车总线网络的入侵检测和混合动力汽车的能量管理策略。E-mail:bit_hongqian@126.com
  • 基金资助:
    国家重点研发计划(2021YFB3101500)、2020年工业互联网创新发展工程-智能网联汽车安全检测平台(TC200H01S)、国家自然科学基金(52202461)和中国博士后科学基金(2022TQ0032,2022M710380)资助项目。

Line-time-varying Model Predictive Active Safety Control of In-wheel Motor Driven Electric Vehicles

WEI Hongqian1,2, ZHAO Wenqiang1,2, AI Qiang1,2, ZHANG Youtong1,2, WANG Hongrong3, LAI Chenguang4, ZOU Xihong4   

  1. 1. School of Mechanical Engineering, Beijing Institute of Technology, Beijing 100081;
    2. Key Laboratory of Low Emission Vehicle in Beijing, Beijing 100081;
    3. China Automotive Engineering Research Institute Co., Ltd., Chongqing 401122;
    4. School of Vehicle Engineering, Chongqing University of Technology, Chongqing 400054
  • Received:2022-01-25 Revised:2023-01-07 Online:2023-07-20 Published:2023-08-16

摘要: 为解决轮毂电机独立驱动电动汽车的主动安全控制和参数估计问题,提出自适应质心侧偏角观测器和基于线性时变理论的模型预测横摆稳定控制策略。首先,建立指数型的滑模观测器识别质心侧偏角,并利用闭环侧向力积分算法消除观测器的稳态误差。然后,引入线性时变理论(Line-time varying,LTV)构建轮胎的准线性模型,并基于此设计横摆稳定系统的模型预测控制器(Model predictive controller,MPC)。为减少控制器的计算负担并提高系统的瞬态响应,采用Laguerre网络拟合全时域的预测控制序列。通过求解具有不等式约束的二次规划函数,获取所需要的外部横摆力矩,并将其分配至4个轮毂电机。数值仿真和实车试验结果表明,提出的自适应滑模观测器可以精确地观测质心侧偏角,具有很强的鲁棒性;提出的横摆稳定控制器可以解决轮胎的非线性饱和问题,进一步提高车辆的主动安全性。

关键词: 电动汽车, 轮毂电机, 汽车安全, 模型预测控制, 自适应滑模观测器

Abstract: To address the active safety and parameter estimation of the in-wheel motor driven electric vehicles(IMDEV), an adaptive sideslip angle observer and a line-time-varying(LTV) model predictive controller(MPC) for the yaw stability are proposed. Firstly, an exponential slide mode observer is established and the closed-loop lateral force integration algorithm is utilized to eliminate the static error. Then, a quasi linear tire model is adopted to design the model predictive yaw stability controller. In addition, Laguerre networks are utilized to approximate the full-horizon control sequence. The external yaw moment is obtained by solving the quadratic programming function with inequality constraints, and thereby it is allocated into four in-wheel motors. Results of the numerical simulation and experimental test manifest that the proposed adaptive slide observer can accurately estimate the sideslip angle with high robustness, the proposed yaw stability controller can address the nonlinear tire saturation and enhance the active safety of IMDEV.

Key words: electric vehicles, in-wheel motor, vehicle safety, model predictive control, adaptive sideslip observer

中图分类号: