Model Predictive Trajectory Tracking for a Ground Vehicle in a Heterogeneous Rendezvous with a Fixed-Wing Aircraft

Abstract

Abstract The maneuver of landing a fixed-wing unmanned aerial vehicle (UAV) autonomously on a moving ground platform requires precise spatial synchronization of both agents. Depending on the desired control strategy for the maneuver, the unmanned ground vehicle (UGV) must be capable to track the UAV’s trajectory robustly with respect to the ground plane even in the presence of disturbances such as wind gusts. In this paper, a linear model predictive trajectory tracking controller for a UGV based on a kinematic bicycle model is presented, assuming that the UAV aims at following a straight flight path with a given velocity. The vehicle model is discretized and linearized in each sampling step, resulting in a quadratic optimization problem which yields the optimal steering angle and motor current demand of the UGV. In the optimization problem, actuator constraints as well as hardware-related dead-times are taken into account. By constraining the yaw rate of the UGV, sideslip of the UGV is prevented, preserving the consistency of the kinematic model. Main requirements for the controller are the ability to allow sufficiently precise trajectory tracking with a longitudinal and lateral deviation of less than 0.5 m, i.e., within the dimensions of the landing platform in the given hardware setup, and real-time capability. Hardware-in-the-loop simulations and experimental results with a model-scale ground vehicle are presented that indicate the validity of the proposed control scheme.