Aspiration-Point-Based Multi-Objective Path Planning Method for an Unmanned Aerial Vehicle

Abstract

This research presents a new multi-objective path planning method for an unmanned aerial vehicle (UAV) using evolutionary computation. The proposed method searches for a desirable Pareto-optimal solution using an "aspiration point" and an "ideal point." The aspiration point refers to the preference information for a decision maker (DM), and the ideal point represents a virtual solution that optimizes all objective functions simultaneously. All of the solutions generated in using evolutionary computation evolve toward the aspiration region, which is determined by the aspiration point. If a solution that is closer to the ideal point than the aspiration point is generated in the search process, the aspiration point is moved to the position of the solution point. This process is repeated until specific termination conditions are satisfied. Some results of the benchmark test problems show that the proposed method can efficiently generate the Pareto-optimal solution for the DM and a high probability compared to the existing method called the "weighted-sum method." The usefulness of the proposed method is also shown by applying it to a multi-objective path planning problem that assumes an aerial photo-shoot mission using a UAV.