考虑介入惩罚因子的智能车辆人机协同控制架构  

Human-machine Cooperative Control Framework for Intelligent Vehicle Considering Intervention Punishment

作　　者：冯吉伟 殷国栋[1] 梁晋豪 庄伟超 彭湃 卢彦博 采国顺 徐利伟 FENG Jiwei;YIN Guodong;LIANG Jinhao;ZHUANG Weichao;PENG Pai;LU Yanbo;CAI Guoshun;XU Liwei(School of Mechanical Engineering,Southeast University,Nanjing 211189)

机构地区：[1]东南大学机械工程学院,南京211189

出　　处：《机械工程学报》2024年第14期238-251,共14页Journal of Mechanical Engineering

基　　金：国家自然科学基金(52025121,52394263);国家重点研发计划(2023YFD2000301);汽车安全与节能国家重点实验室(KFZ2201);江苏省应用数学科学研究中心(BK20233002);江苏省科技成果转化(BA2021023)资助项目。

摘　　要：人机交互中的不确定性易引起驾驶人与智能辅助驾驶系统的冲突,进而削弱驾驶性能。为此,提出一种基于动态介入惩罚的智能车辆人机协同控制架构,以提升车辆的驾驶性能与侧向稳定性。针对人机共享控制系统的不确定性,建立考虑驾驶人时变预瞄行为与轮胎非线性特性的车辆模型;引入影响共享驾驶权限分配的人机介入惩罚因子,并根据驾驶人手力矩和实际预瞄点侧向偏差建立模糊规则,减少由个性化驾驶行为导致的驾驶人与辅助驾驶行为之间的冲突;提出基于极点配置的线性参数时变(Linear parameter varing,LPV)控制策略以提升控制系统鲁棒性。最后,采用Matlab/Carsim联合仿真和基于NI-LabVIEW-RT系统的硬件在环测试(Hardwave-in-loop,HIL)进行试验验证,结果表明,所提出的人机协同控制架构能够在保证车辆操纵性能的同时有效缓解人机冲突。The uncertainties of human-machine interaction would cause conflicts between the driver and the intelligent assisted driving system,and thus deteriorating the vehicle driving performance.To enhance the drivability and lateral stability of the vehicle,an intelligent human-machine cooperative control framework,which considers the dynamic intervention penalty,is proposed.First,to well address the uncertainties in the human-machine shared driving system,the time-varying driver preview behavior and the tire nonlinear characteristics are considered in the vehicle system modelling;Second,to attenuate the conflicts between drivers and assistance steering actions due to the personalized driving behaviors,the human-machine intervention penalty factor is introduced into the driving authority allocation,and the fuzzy rule is established based on the dynamic driver torque and lateral deviation of actual preview point.Third,a linear parameter varying(LPV)controller based on the system poles placement is developed to improve the control system robustness.Finally,to verify the feasibility and effectiveness of the proposed control strategy,the Matlab/Carsim joint simulation and the hardware-in-loop(HIL)test based on NI-LabVIEW-RT system are conducted.The results show that the proposed human-machine cooperative control framework can effectively mitigate the human-machine conflict while guaranteeing the vehicle handling performance.

关 键 词：车辆稳定控制 LPV鲁棒控制 人机协同控制 车辆系统动力学 

分 类 号：U461[机械工程—车辆工程]