Optimization-Based Control Design for Challenging Maneuvers in Nonholonomic Vehicles

Abstract

The thesis explores a diverse range of nonholonomic vehicles and addresses the challenging maneuvers pertaining to their control. The concerned nonholonomic systems span from hybrid Unmanned Aerial Vehicles (UAVs), fixed-wing UAVs, multicopters, drifting mobile robots, to in-wheel motor vehicles. Each of these systems presents unique challenges due to its dynamical constraints, making them compelling test beds for advanced control techniques. Starting from the modeling aspect, we obtain motions for these vehicles executing the challenging maneuvers using optimization-based techniques, namely optimal control and Nonlinear Model Predictive Control (NMPC). The thesis proposes contributions across various applications. Each solution is tailored to the specific dynamics and constraints of each system, pushing the boundaries of what can be achieved with optimal control and NMPC in these areas. In the thesis, we first provide an extensive review of hierarchical control architecture and analyze the critical transition phase between modes of operation of hybrid UAVs. Next, a deep-stall maneuver ending and concatenating with a perching maneuver is proposed, demonstrating a novel approach to recover a fixed-wing UAV. After that, we further design a semi-globally asymptotically stable NMPC for a hovering problem of a multicopter, which addresses stability challenges while ensuring performance. Moving on, we develop a stabilizing NMPC scheme with terminal ingredients tailored to the complexities of maintaining stability for a drifting mobile robot. In the end, we tackle the problem of a double lane change maneuver, a benchmark problem for four in-wheel-motor vehicles, with an optimal control framework that ensures both safety and performance. Comprehensive simulation results demonstrate the effectiveness of the optimization-based techniques employed and also provide insights into their potential implementation in realworld scenarios.