Intelligent vehicle perpendicular parking path planning and tracking control

This paper presents a novel automatic perpendicular parking algorithm, taking into account the initial heading angle of the intelligent vehicle. The feasible parking starting point is located based on the vehicle’s minimum turning radius, the relative pose between the vehicle and the parking spot as well as the coordinates of right vertex of the parking space. The type of parking (i.e. single-step or multi-step) is determined on the basis of the collision avoidance constraint. Considering the constraint of the vehicle steering actuator, the quartic spline curve is used to smooth the single-step parking path and the last path into the parking space of multi-step parking, so as to improve the path characteristics. Then, a desired heading angle function of the vehicle is introduced, and the lateral displacement deviation in path tracking tends to zero when the actual heading angle approaches the desired heading angle. Subsequently, a nonlinear integral sliding mode controller (NISMC) with the hyperbolic tangent function based on dynamic boundary layer as the switching function is proposed to track the desired heading angle, so as to complete the path-tracking process. Co-simulations and Hardware-in-the-Loop (HIL) tests are carried out to verify the effectiveness of the designed algorithm, simulation and test results show that the desired parking path can be feasibly planned even when the vehicle has an initial heading angle; meanwhile, the tracking performance based on the proposed NISMC controller not only has small tracking error, smooth control output, but also strong response rate and robustness.