Cognitive UAS Path-Planning for Large Spatial Search

Abstract

Search and Rescue/Destroy missions are some of the most high-risk situations in modern engineering. Every mission nearly always presents a life or death scenario for one or more individuals, with the penalty for failure often being human lives. Modern Search and Rescue/Destroy missions implement the use of autonomous systems in the form of giving an unmanned autonomous aerial system(s) the task of searching a given area in the attempt of discovering one or more objects of interest. Though this ingenuity has already benefited the line of work, these unmanned systems are still not being used to their full potential. Some means of planning how to search the area must be developed, with the most basic means of accomplishing this task being creating a predefined path which is guaranteed to cover all known areas. To increase the rate of success and decrease necessary search time, a pseudo-random search method, known as meta-heuristics, is used to develop a new path planning algorithm to search the field in an intelligent manner. This work develops a means of turning meta-heuristic optimization into a cognitive navigation with autonomous path-planning algorithm that is decoupled from apriori information, with minimal requirements for initiation. To account for the higher performance requirements of such a method, novel guidance methods were developed to meet said demands. Simulations suggest that the proposed search method performs better on average than the current accepted basis.