Mission-Oriented Trajectory Optimization for Search-and-Rescue Multirotor UAVs in Cluttered and GPS-Denied Environments

摘要

View Video Presentation: https://doi.org/10.2514/6.2022-3999.vid Planning Search-and-rescue (SAR) missions for UAVs in cluttered and GPS-impaired environments remains a challenging topic in both robotics and aerospace related research areas. Small multirotors, especially quadrotors, are deployed by many in indoor or confined spaces due to their structural simplicity, ability to hover and take off and land vertically, and good maneuverability. However, most of small multirotors suffer from less-than-30-min flight time on one charge imposing constraints on missions including searching for survivors and payload delivery. In this paper, we introduce a mission-oriented trajectory generation approach in order to improve the quantitative performance of multirotors by reducing single-flight energy and time consumption. In our methodology, each flight mission is initially planned using computationally cheap path planning algorithms to generate a set of waypoints as an initial guess of the trajectory. Next, a multi-phase optimal control problem is formulated with the mission-specific objective function. Apart from constraints of initial and final conditions, this multi-phase optimal control problem is also subject to dynamics constraints from both quadrotor equations of motion and rotor aerodynamics models as well as path constraints. The optimal control problem is then solved in the frame of multi-disciplinary design optimization (MDO) by setting the energy consumption, times, control inputs, and states as design variables, and it has the potential to be converted into a simultaneous design and control process when including quadrotor design parameters as new design variables. After that, simulations are performed to validate the effectiveness of proposed methodology and to compare its results with those from peer methods. Three mission scenarios are set up to validate the proposed methodology. The simulation results indicate that the proposed method can decrease the flight time by up to 6.69% or mechanical energy consumption by 3.00%, and it outperforms two other trajectory generation methods for specific mission performance.