برنامه ريزي مقاوم بلادرنگ مسير كوادكوپتر در برابر اغتشاش خارجي در محيط ناشناخته

In recent years, the use of quadcopters has grown significantly across a wide range of applications, including environmental monitoring, search an‎d rescue operations, an‎d aerial transportation. Nevertheless, ensuring safe an‎d optimal motion in complex environments, particularly in the presence of external disturbances such as wind, remains a critical challenge in path planning an‎d flight control. This thesis addresses three fundamental problems related to safe trajectory planning for quadcopters an‎d proposes an innovative solution for each. First, a novel method is introduced to modify paths generated by classical path-planning algorithms, such as the Jump Point Search (JPS) algorithm, in cluttered environments. The objective is to enable effective re-routing when encountering obstacles while maintaining proximity to the originally computed optimal path. To this end, a virtual force field is applied to the path waypoints, reshaping the trajectory in response to obstacle proximity while preserving its near-optimal characteristics. The performance of the proposed method is eva‎luated for various obstacle geometries an‎d compared with an alternative path modification approach. Second, external disturbances can substantially degrade quadcopter performance an‎d, if neglected, may lead to mission failure. To address this issue, a new trajectory re-planning framework under wind disturbances is proposed. The uncertainty introduced into the quadcopter dynamics due to wind is modeled as a safe ellipsoid using Hamilton–Jacobi reachability analysis. To adapt the trajectory to these conditions, a virtual field proportional to both the disturbance intensity an‎d the proximity of the safe ellipsoid to surrounding obstacles is applied. This field modifies the trajectory while preserving key optimality properties. A comparative study, supported simulation results, demonstrates the effectiveness of the proposed method in terms of computational cost. Finally, this thesis extends the proposed framework to cooperative trajectory re-planning for a group of quadcopters operating under wind disturbances. To prevent inter-vehicle collisions, at each instant within the re-planning horizon, a temporary hyperplane is constructed between the equivalent safe ellipsoids of two quadcopters. This hyperplane acts as a separating boundary, facilitating the computation of mutual virtual forces an‎d enabling independent yet coordinated trajectory deformation for each vehicle. The effect of wind uncertainty is reflected in the size of the corresponding safe ellipsoids, directly influencing the formation of the separating hyperplane. The effectiveness of the proposed multi-agent strategy is eva‎luated through a comparative study supported simulations, highlighting its computational cost.

5

158142