Research On Trajectory Tracking Control For Intelligent Vehicle

Intelligent vehicle(IV), i.e. the intellectualization of vehicles, is the main direction of development of vehicle technologies. IV technology is the combined result of vehicle technology, control theory, information technology and artificial intelligence. IV technology will greatly promote the safety and automation of vehicles and improve the handle method. The development of IV does not only impact the vehicle engineering discipline. It also poses great challenges for the control discipline. Nowadays, IV technology has attracted researchers from various fields and disciplines.Environment recognition, path planning and driving control are three main aspects of IV technology. One important issue in driving control is the highprecision control of longitudinal and lateral motion, which plays an important role in autonomous driving, vehicle platoons control and trajectory tracking. This paper studies two maneuvers: the longitudinal velocity and yaw rate tracking control in longitudinal and lateral motion and the trajectory tracking control in steering maneuver. We hope this research can be one of the feasible methods for speed stabilization and yaw stability control for vehicles steering with high speed and the trajectory tracking control for lane change and obstacle avoidance.First of all, this paper analyzes the the modeling issue in longitudinal and lateral coupling control and present a 4-Do F dynamics model which illustrates the longitudinal motion, lateral motion, yaw motion and roll motion of the vehicle. Since the order of the model is too high, it is difficult to design a controller. So, we determine factors which have main effects on lpongitudinal and lateral coupling control through simulation. According to the simulation results, the roll motion is neglected, and a 3-Do F model is promoted. Further more, two simlified 3-Do F models are pesented for the ease of controller design. These models are compared with the initial one under different maneuves via MATLAB/Simulink so as to determine the model used for the controller design. After that, time varying parameters are selected and the model is converted into a polytopic quasi LPV(Linear Parameter Varible) model.For the longitudinal velocity and yaw rate tracking control problem, the steering angle in 3-Do F dynamic model is regarded as a known disturbance, and the yaw rate and the reciprocal of longitudinal velocity are selected as time-varying parameters in this paper. So, this problem is converted to the controller design problem of a quasi LPV controller with disturbance rejection. To reduce the conservativeness of the controller and satisfy the requirements of practical applications, the ranges of the parameters are analyzed based on the driving condition and the LPV system is converted into a convex polytope with limited vertices by using convex deposition. The controller design and the solving process are greatly simplified by designing the H∞ controller of each vertices and combining these controllers linearly to derive the control varibles for arbitrary timevarying parameter. For the H∞ controller design of each vertex, the proposed method introduces an integral term for longitudinal velocity error and yaw-rate error. This method can place the closed-loop poles in a specific disk, minimizes the H∞ perefomance and satisfy the control value constraints. After that, this paper discussed a torque vectoring method which can decompose the left and right moments into the moments of four wheels. To evaluate the performance of the proposed controller, this paper simulated various situations in the high precision vehicle dynamic simulation software ve DYNA. The simulation results show that the proposed LPV robust controller has good performance and robustness in tracking desired longitudinal velocity and yaw rate.To solve the trajectory tracking control problem in intelligent vehicle, this paper first models the trajectory of vehicles with lateral displacement of vehicle centroid and driving direction in inertial coordinate system. Further more, the dynamic equations of vehicle trajectory are presented based on the following state varibles: the lateral velocity and yaw rate of vehicle centroid in vehicle body coordinate system and the lateral displacement of vehicle centroid and driving direction in inertial coordinate system. Based on the dynamic equations, this paper proposed a receding horizon optimization method for steering angle. Combined with the vehicle longitudinal velocity and yaw rate tracking controller, the desired trajectory can be tracked. The method presented in this paper is a hierarchical control structure. The upper controller plans the steering angle of vehicles and the lower one is used to track the desired longitudinal velocity and yaw rate. This method has a simple structure, which has the same architecture as the steering and driving distributed control system in real vehicles. The simple structure makes it easier to implement and apply. The method is simulated in ve DYNA by detecting the tracking performance of double-lane-change maneuver with different velocity. Simulation results show effectiveness of the proposed method.