An obstacle-avoidance technique for autonomous underwater vehicles based on BK-products of fuzzy relation

摘要

This paper proposes a new heuristic search technique for obstacle avoidance of autonomous underwater vehicles (AUVs) that are equipped with a looking-ahead obstacle-avoidance sonar. The fuzzy relation between the sonar sections and the properties of a real-time environment is used as a core concept. The Lukasiewicz fuzzy implication operator is used in Bandler and Kohout's Triangle Subproduct to calculate the relationship between the fuzzy relation and its transposed relation. This product relation the reveals the characteristics and interrelationships of the sonar sections. A direction of the sonar section that has good characteristics is selected as the successive heading for obstacle avoidance of AUVs. The simulation results clearly demonstrate that the heuristic search technique, which uses the BK-product of a fuzzy relation with a sonar partitioned into seven sections, enables AUVs to navigate safely through the obstacle to the goal with the optimal path.