Maximizing Swarm Coverage: Hunting for Members of a Moving Population

We explore search methods for finding and removing members of a large population of mobile particles in different environments. We begin by simulating the movement of individual swarm members biased by a map image with four search patterns and find that simulation duration affects the results. We then treat particle motion as a Markov process in an environment bounded with walls that retain a portion of the swarm and in an environment that attracts the particles into a normal distribution at the center of the space. We test six search patterns with several parameter variations. We show that a two-step greedy algorithm has the best performance in all cases but second best performance varies with the parameters of the swarm and the search.