A Three-Dimensional UCAV Path Planning Approach Based on Immune Plasma Algorithm

AbstractPath planning of an unmanned aerial vehicle (UAV) or its variant supported with weapon systems, also called unmanned combat aerial vehicle (UCAV), is one of the most crucial steps for the autonomous flight and requires compelling decisions being made by considering the existence of the enemy defence, obstacles, and battery or fuel capacity of the vehicle optimally before starting the flight. Immune plasma algorithm (IP algorithm or IPA), inspired by the convalescent plasma treatment which is a medical method being popularized again with the rise of new coronavirus or COVID-19, has been introduced recently as a new optimization technique and the promising performance of the mentioned algorithm validated through a set of numerical and engineering problems. In this study, the IP algorithm was specialized as a three-dimensional UCAV path planner for the first time. Its performance was investigated particularly by using three different battlefield environments and assigning various constants to the population size. Moreover, the effect of two IPA-specific control parameters that are responsible for determining the number of donors and number of receivers on the qualities of the calculated paths were analyzed in detail. The results obtained by the IPA were compared with the results of other three-dimensional path planners guiding optimization algorithms such as simulated annealing (SA), gray wolf optimizer (GWO) and symbiotic organism search (SOS). Experimental studies showed that the main idea lying behind the usage of qualified solutions as donors and transferring information directly from them to the receivers representing the poor solutions of the problem gives a positive contribution to the exploitation-dominant operations of IPA and fully complies with the challenging requirements of the path planning problem. Especially in the battlefields for which an optimal or near optimal path contains complex maneuvers, the advantages from the subtly balanced operational steps of the IPA become more apparent by setting the number of donors and receivers appropriately and IPA is found to be more stable and successful compared to the other tested algorithms.