Flyable path planning for a multi-UAV system with Genetic Algorithms and Bezier curves

Abstract

Unmanned Aerial Vehicles (UAVs) are used in numerous military and civil application areas, and they have gained prominence in the research community. A UAV has to operate in a complex environment and checks the control points in the mission area by satisfying different constraints of the assigned task. Therefore, the path planning problem is one of the important areas in UAV researches. If the number of control points increases in the Unmanned Aerial System (UAS), finding a feasible solution in this large search space takes up a great deal of time. Nowadays low-cost UAVs are available, and this enables the use of multi-UAV systems to perform different tasks more efficiently and quickly. This usage increases the complexity of effective path planning and task allocation problem. This paper presents a flyable path planning for multi-UAV systems by using a Genetic Algorithm (GA) in a known environment at a constant altitude. A feasible path is firstly calculated by GAs, and then this path is smoothed by using Bezier curves. Experimental results indicate that the proposed approach produces effective and feasible paths for each UAV in a multi-UAV system. System is implemented in Java with a GUI for presenting results. The paper also draws future works that can be done on this topic.