基于预测控制的无人驾驶车辆爆胎转向控制
|
摘要
针对无人驾驶车辆爆胎后的转向控制问题,考虑实时性和控制性能的要求,提出了连续时域自适应预测控制方法.爆胎使得滚动阻抗系数和侧偏刚度等轮胎参数在短时间内产生剧烈的变化,从而导致转向控制失灵,进而引起无人驾驶车辆偏离道路甚至侧翻.为此,对无人驾驶车辆标称动力学模型进行反馈线性化,结合泰勒展开预测无人驾驶车辆的运动趋势.在此基础上,将爆胎引起的参数变化转化为不确定,利用模糊系统万能逼近原理,通过设计自适应模糊观测器进行在线观测.并同时考虑控制输入的饱和约束,利用连续预测控制方法设计解析控制律,以满足系统控制的实时性要求.最后,与传统非线性预测控制以及PID控制进行了仿真对比.从仿真结果可以看出,当车辆发生爆胎后,轮胎滚动阻抗系数瞬时增大了29倍、侧偏刚度瞬时降低了72%,如不施加额外的转向控制作用,无人驾驶车辆将在1 s内偏离原车道约5.5 m.而施加本文所提方法后,系统的实时性和控制性能均优于给定传统算法.一方面,与非线性预测控制和PID控制相比,所提方法计算速度提高了约150倍,计算时间缩短约50%;另一方面,在给定的控制输入饱和约束范围内,所提方法仍能够控制无人驾驶车辆在爆胎后只发生微小偏移,偏移量仅为传统算法的2.5%左右.
To control the steering of unmanned ground vehicles with tire blowout,we designed a continuous-time adaptive model predictive control method that meets the requirements of real-time and control performance. In a very short period of time,tire blowouts cause drastic changes in the tire parameters,including the rolling resistance coefficient and cornering stiffness,which lead to steering control failure and deviation of the unmanned ground vehicle from the lane of travel or even a rollover. In this paper,we linearize a nominal dynamic model of an unmanned ground vehicle using feedback linearization,and predict the motion trend of the vehicle by Taylor expansion. On this basis,the parameter changes caused by the tire blowout are transformed into uncertainties,so we designed an adaptive fuzzy observer to perform on-line observation based on the universal approximation theorem of fuzzy systems.Using control input saturation constraints,we designed an analytic control law using continuous model predictive control to satisfy the real-time requirement of the control system. Finally,we performed simulations to compare the performance of the proposed method with those of traditional nonlinear model predictive control and PID control. The simulation results showed that,after a tire blowout,the tire rolling resistance coefficient experiences an instantaneous 29-fold increase,and the cornering stiffness experiences an instantaneous decrease of 72%. Without additional steering control,the unmanned ground vehicle would deviate from the original lane by about 5.5 meters in one second. We validated that the system using the proposed method exhibited better real-time and control performance than those using the traditional algorithms. We found the proposed method to improve the computing speed 150-fold and to cut the computing time 50%,respectively,compared to the nonlinear model predictive control and PID algorithms.In addition,the proposed method can guarantee that the vehicle experiences only a slight deviation from its lane of travel after a tire blowout within the given control input saturation constraints,with the deviation being only about 2.5% of that of the traditional algorithm.
To control the steering of unmanned ground vehicles with tire blowout,we designed a continuous-time adaptive model predictive control method that meets the requirements of real-time and control performance. In a very short period of time,tire blowouts cause drastic changes in the tire parameters,including the rolling resistance coefficient and cornering stiffness,which lead to steering control failure and deviation of the unmanned ground vehicle from the lane of travel or even a rollover. In this paper,we linearize a nominal dynamic model of an unmanned ground vehicle using feedback linearization,and predict the motion trend of the vehicle by Taylor expansion. On this basis,the parameter changes caused by the tire blowout are transformed into uncertainties,so we designed an adaptive fuzzy observer to perform on-line observation based on the universal approximation theorem of fuzzy systems.Using control input saturation constraints,we designed an analytic control law using continuous model predictive control to satisfy the real-time requirement of the control system. Finally,we performed simulations to compare the performance of the proposed method with those of traditional nonlinear model predictive control and PID control. The simulation results showed that,after a tire blowout,the tire rolling resistance coefficient experiences an instantaneous 29-fold increase,and the cornering stiffness experiences an instantaneous decrease of 72%. Without additional steering control,the unmanned ground vehicle would deviate from the original lane by about 5.5 meters in one second. We validated that the system using the proposed method exhibited better real-time and control performance than those using the traditional algorithms. We found the proposed method to improve the computing speed 150-fold and to cut the computing time 50%,respectively,compared to the nonlinear model predictive control and PID algorithms.In addition,the proposed method can guarantee that the vehicle experiences only a slight deviation from its lane of travel after a tire blowout within the given control input saturation constraints,with the deviation being only about 2.5% of that of the traditional algorithm.
引文
[1]Wang F,Chen H,Guo H,et al.Constrained H∞control for road vehicles after a tire blow-out[J].Mechatronics,2015,30:371-382.
[2]Choi E H.Tire-Related Factors in the Pre-crash Phase[R].Washington,DC,USA:National Highway Traffic Safety Administration,2012.
[3]Tandy D F,Ault B N,Colborn J,et.al.Objective measurement of vehicle steering and handling performance when a tire loses its air[J].SAE International Journal of Passenger Cars-Mechanical Systems,2013,6(2):741-769.
[4]Qian J,Kim D S,Lee D W.On-vehicle triboelectric nanogenerator enabled self-powered sensor for tire pressure monitoring[J].Nano Energy,2018,49:126-136.
[5]Mirzaeinejad H.Robust predictive control of wheel slip in antilock braking systems based on radial basis function neural network[J].Applied Soft Computing,2018,70:318-329
[6]Zheng B,Anwar S.Yaw stability control of a steer-bywire equipped vehicle via active front wheel steering[J].Mechatronics,2009,19(6):799-804.
[7]Wang F,Chen H,Cao D.Nonlinear coordinated motion control of road vehicles after a tire blowout[J].IEEETransactions on Control Systems Technology,2016,24(3):956-970.
[8]Wang F,Guo H,Chen H.Nonlinear model predictive control of blowout tire vehicles and its map-based implementation[C]//Chinese Control Conference.Hangzhou,China,2015:4043-4048.
[9]Ding B,Pan H.Output feedback robust MPC with one free control move for the linear polytopic uncertain system with bounded disturbance[J].Automatica,2014,50(11):2929-2935.
[10]Zheng Y,Li S,Tan R.Distributed model predictive control for on-connected microgrid power management[J].IEEE Transactions on Control Systems Technology,2017,26(3):1028-1039.
[11]Lu P.Optimal predictive control of continuous nonlinear systems[J].International Journal of Control,1995,62(3):633-649.
[12]Panchal B,Kolhe J P,Talole S E.Robust predictive control of a class of nonlinear systems[J].Journal of Guidance Control&Dynamics,2014,37(5):1437-1445.
[13]Barnard R.Continuous-time implementation of optimalaim controls[J].IEEE Transactions on Automatic Control,1976,21(3):432-434.
[14]Lee H,Tomizuka M.Robust adaptive control using a universal approximator for SISO nonlinear systems[J].IEEE Transactions on Fuzzy Systems,2000,8(1):95-106.
[2]Choi E H.Tire-Related Factors in the Pre-crash Phase[R].Washington,DC,USA:National Highway Traffic Safety Administration,2012.
[3]Tandy D F,Ault B N,Colborn J,et.al.Objective measurement of vehicle steering and handling performance when a tire loses its air[J].SAE International Journal of Passenger Cars-Mechanical Systems,2013,6(2):741-769.
[4]Qian J,Kim D S,Lee D W.On-vehicle triboelectric nanogenerator enabled self-powered sensor for tire pressure monitoring[J].Nano Energy,2018,49:126-136.
[5]Mirzaeinejad H.Robust predictive control of wheel slip in antilock braking systems based on radial basis function neural network[J].Applied Soft Computing,2018,70:318-329
[6]Zheng B,Anwar S.Yaw stability control of a steer-bywire equipped vehicle via active front wheel steering[J].Mechatronics,2009,19(6):799-804.
[7]Wang F,Chen H,Cao D.Nonlinear coordinated motion control of road vehicles after a tire blowout[J].IEEETransactions on Control Systems Technology,2016,24(3):956-970.
[8]Wang F,Guo H,Chen H.Nonlinear model predictive control of blowout tire vehicles and its map-based implementation[C]//Chinese Control Conference.Hangzhou,China,2015:4043-4048.
[9]Ding B,Pan H.Output feedback robust MPC with one free control move for the linear polytopic uncertain system with bounded disturbance[J].Automatica,2014,50(11):2929-2935.
[10]Zheng Y,Li S,Tan R.Distributed model predictive control for on-connected microgrid power management[J].IEEE Transactions on Control Systems Technology,2017,26(3):1028-1039.
[11]Lu P.Optimal predictive control of continuous nonlinear systems[J].International Journal of Control,1995,62(3):633-649.
[12]Panchal B,Kolhe J P,Talole S E.Robust predictive control of a class of nonlinear systems[J].Journal of Guidance Control&Dynamics,2014,37(5):1437-1445.
[13]Barnard R.Continuous-time implementation of optimalaim controls[J].IEEE Transactions on Automatic Control,1976,21(3):432-434.
[14]Lee H,Tomizuka M.Robust adaptive control using a universal approximator for SISO nonlinear systems[J].IEEE Transactions on Fuzzy Systems,2000,8(1):95-106.