Research On Obstacle Avoidance Path Planning And Tracking Control Strategy Considering Roll Stability

With the rapid development of advanced sensors,chip arithmetic and artificial intelligence and other high technologies,automatic driving technology is facing unprecedented opportunities and challenges.The research on emergency collision avoidance function of automatic driving vehicles has an important role in promoting the construction of intelligent transportation system,the reduction of traffic accident rate,and the improvement of vehicle driving safety.However,the existing active collision avoidance system still depends on the driver’s own steering maneuvering behavior when the vehicle is running at high speed,and thus cannot fundamentally ensure the safety of vehicle collision avoidance.At the same time,because the steering subsystem and suspension subsystem of the vehicle are coupled with each other,the emergency steering maneuvering of the vehicle is likely to lead to the vehicle rollover.Moreover,under this working conditions,the nonlinear characteristics of vehicle dynamics are significantly enhanced,which brings great difficulties to the steering control and roll control of the vehicle.Aiming at the above problems,this paper carries out the research on the collision avoidance path planning strategy and path tracking control strategy considering the stability of vehicle roll for the emergency steering collision avoidance working condition of self-driving vehicles.The specific research work carried out is as follows:(1)A vehicle dynamics model with coupling of lateral and vertical directions is established to provide a basis for the analysis of the vehicle dynamics problem as well as the verification of the control algorithm.A magic tire model is established for accurately reflects the nonlinear characteristics of the system.A four-wheel time-domain road excitation mathematical model is constructed,and the accuracy of the constructed vehicle model is verified by Carsim,a mature commercial software for vehicle dynamics,under the doubl line change maneuver.(2)The research on emergency collision avoidance path planning strategies for self-driving vehicles is carried out.Considering the limitations of the original Sigmoid function,the MittagLeffle function is introduced,and the generalized Sigmoid function is proposed.The important properties of the function and the influence of the new parameters on the change of the shape of the function curve are analyzed.Then it is applied to emergency collision avoidance path planning,the collision avoidance path planning strategy based on the generalized Sigmoid function is established,the conditions of road geometry constraints and vehicle stability constraints are analyzed,and the multi-objective optimization method of collision avoidance path parameters is proposed.A path safety cost function is designed using Gaussian discrete convolution,and a vehicle lateral stability cost function is designed using an iterative learning algorithm.In order to improve the real-time performance of the path planning,a neural network training dataset is generated from the optimization results,and an online planning strategy for collision avoidance paths is established.(3)The research on path tracking control strategy based on fractional order calculus theory is carried out.A focus preview driver model is proposed,where a preview deviation estimation method based on the fractional-order G-L definition and the fractional-order accumulation algorithm is designed.Then a steering control model based on the focus preview deviation is established.The simulation test under the large curvature steering collision avoidance condition shows that the path tracking effect can be effectively improved by adjusting the control parameters,and the stability of the vehicle can be guaranteed.The hardware-in-the-loop experimental study of the emergency collision avoidance path tracking control is carried out,and the real-time performance of the fractional-order controller on the real hardware is analyzed.(4)The study on the roll control strategy based on magnetorheological semi-active suspension was carried out.The body roll mechanism during the vehicle steering control process is analyzed.A mathematical model for rollover prediction based on the driver’s model is derived,and the effects of the driver’s preview and time lag factors are analyzed.A fractional-order PID controller with the quarter-body velocity in vertical direction as the feedback quantity is designed,and a comprehensive evaluation function that integrates the tracking error,roll angle,body acceleration,suspension dynamic deflection,and wheel dynamic deformation is built.The quantum particle swarm optimization algorithm is adopted to rectify the system parameters in time domain.The control performance of passive suspension,integer order suspension and fractional order suspension is analyzed from the time domain.(5)A whole vehicle skyhook damping control strategy based on modal decoupling is proposed,and the critical damping control method of each order mode is designed under the principal coordinates.The existence conditions of fractional-order critical damping and the calculation formula are derived.Then the whole-vehicle fractional-order skyhook damping control strategy is established.A four-wheel time-domain impact road excitation model is constructed,and the simulation results show that the whole-vehicle skyhook damping control strategy improves in the body vibration response,the body lateral inclination response,and the pitch angle response.The fractional-order skyhook damping control is improved further on the basis of integer-order controllers.