An Empirical Comparison of Search Approaches for Moving Agents

Abstract

This study empirically compares existing search approaches used for path planning of moving agents, namely, incremental and real-time search approaches. The comparisons are performed in both stationary and moving target search problems separately. In each problem domain, well-known representatives of both approaches are evaluated in partially observable environments where the agent senses a limited area based on its sensor range. In addition to the available algorithms, we propose two algorithms to be used in each problem. The simulations conducted on random grid and maze structures show that the algorithms behave differently and have advantages over each other especially as the sensor range varies. Therefore, the proposed study enables the agent to determine the most appropriate algorithm depending on its priorities and sensor range. © 2017 Elsevier B.V., All rights reserved.