Integrated Control For Vehicle Active Steering And Braking Based On Road Safety Boundary

Automobiles can be one of important symbols of freedom and modern civilization.Vehicle active safety relates to the life safety for both car users and road users.Various active safety control systems have been developed to improve vehicle safety,aiming to ensure vehicle stability while fully tracing driver intention.However,the traditional vehicle dynamics control systems normally only focus on "Vehicle" itself,and it is not easy to accurately determine whether the driver's operation will lead to accidents due to the lack of external environment information.In fact,a large amount of traffic accidents are caused by the driver factors.Recently,the road environment perception technology are becoming practical,providing a basis for the identification of improper driver operation.How to improve vehicle performance,especially the active safety,by using the road environment information has become an urgent problem to be solved,and the vehicle nonlinear dynamics is one of the key and difficult issues in the study.This dissertation proposes a vehicle dynamics integrated control strategy considering “Vehicle stability limit” and “Road safety boundary(RSB)” based on the assumptions of known information of vehicle states and environment.The research is implemented from "vehicle" to "road".To be specific,the vehicle stability limit is studied firstly,and then the RSB is studied to find the limited vehicle states under the road constraints in the view of vehicle dynamics.The RSB is defined as the boundary beyond which a collision will eventually occur no matter what future manoeuvres the driver will apply.When the driver's operation will lead to an accident,the active control is applied to ensure safety in road environment.The proposed integrated controller tracks driver's intention and ensure vehicle stability by optimally regulating active steering and active braking.When the driver intention exceeds the RSB,the control target is limited to the RSB to ensure vehicle safety.The objectives of this dissertation are described as below.Firstly,an observer for the longitudinal/lateral forces and cornering stiffness of four tyres are designed in order to improve adaptability of the designed controller and RSB solver.A Dual Extended Kalman Filter(DEKF)is designed to simultaneously estimate the tyre lateral forces and cornering stiffness based on nonlinear tyre model,in order to maintain satisfactory performance with respect to different tyre properties.The simulation results is implemented by using a full vehicle model in Carsim.Secondly,the integrated control strategy of active steering and braking is proposed based on MPC.A hierarchical structure is adopted.The upper layer of vehicle motion control is designed to calculate the desired front lateral force and additional yaw moment for tracing the nominal yaw rate and sideslip angle by using the MPC,which adopts a linear time varying 2-dof vehicle model with the front axle lateral force and additional yaw moment as inputs.In the lower execution layer,by using the nonlinear tyre model a tyre force optimization allocation algorithm calculates the optimal front wheel steer angle and braking forces of each wheel to trace the desired front axle lateral force and additional yaw moment obtained by the upper layer.Hence,the coupling and nonlinearity of the tyre forces,which is the most difficult part to deal with in the integrated control,is moved to the underlying layer,leading to a reduced influence of nonlinear factors on the upper controller.This is helpful to improve the control performance and to reduce controller design difficult.Meanwhile,the reference model of the MPC and tyre force allocation algorithm can be updated with the estimated tyre forces and cornering stiffness,which enhances the control adaptability.Simulation results show that the control performance maintain satisfactory with respect to different tyre properties and tyre-road adhesion conditions.Then,a RSB solver is proposed based on vehicle dynamics.Currently research involved with the collision risk prediction mostly based on kinematics without considering the extemely manoeuvre capabilities of a vehicle,while this dissertation proposes a RSB solver based on vehicle dynamics.By predicting the vehicle trajectory with a 2-dof vehicle model with the front axle lateral force as input,the RSB solver is formulated as a Linear Programming(LP)problem to obtain the critical vehicle state(yaw rate).Since the reference model uses front axle lateral force as input,the nonlinearity of tyre property is extracted from the optimization problem.Moreover,the vehicle handling capabilities is also fully considered by applying linear constraints on control input and vehicle states.Finally,"Human-Vehicle-Road" closed-loop simulations are implemented for the proposed algorithm by using Carsim vehicle model and driver model.The results show that,in the single lane change test,the integrated controller with only vehicle stability limits can ensure vehicle stability,but can not guarantee vehicle with respect to different drivers.The integrated controller with the RSB protection can intervene timely under improper driver operation to help driver pass the test successfully to avoid collisions.Moreover,the proposed integrated controller based on RSB is able to adapt to the variations of tyre-road adhesion coefficient and tyre property as well.