A Genetic Algorithm enhanced with Fuzzy-Logic for multi-objective Unmanned Aircraft Vehicle path planning missions

The fast growth of computational-intelligence and computer vision-based approaches in the field of robotics have initiated the development of novel unmanned aircraft vehicles (UAVs) with additional capabilities for mission-critical applications, such as fire detection or delivery services. Path planning is a crucial part of the automated operation of a UAV to find the optimal path from a starting point to a target-goal point, satisfying in parallel specific criteria and constraints. To this end, traditional path planning algorithms, such as graph-based, heuristics or metaheuristics, have been adopted to address this problem. However, the complex operational environment combined with the demand for energy efficiency have emerged the need for dynamic and multi-objective path planning meeting the special characteristics of a UAV flight. To address the above challenges, in this study a novel Genetic Algorithm (GA) enhanced with Fuzzy Logic (GAF) is developed for solving the UAV multi-objective path planning problem. The proposed algorithm aims to find an optimal path with respect to contradicted objective terms, such as distance, energy efficiency and path’s curvature. In the proposed developments, the novelty of GAF rises via a two-step procedure: (i) the evaluation stage of the energy efficiency of the paths; and (ii) the evaluation stage of the GA for the development of an energy efficient, smooth path of shortest possible traveled distance. The efficacy of the approach is illustrated via a comparative experimental evaluation.