Complete Coverage Path Planning Based on BINND and MK-means++
Siyue Liang, Xin Li, Ying Zhang · 2024
Aiming at the problems of low information utilization efficiency and poor quality in the search field, a complete coverage path planning method is proposed, which introduces the concept of biological incentives to improve backtracking search and dynamically allocates areas using the modified K-means++ algorithm (MK-means++). A dynamic function is designed to carry the probability information in the environment, in order to solve the problems of low path planning efficiency and low environmental information utilization caused by the lack of probability prior knowledge. In order to improve the problems of high cost of escape paths and low value of temporary target points in traditional methods, the method Biologically Inspired Neural Network Decision (BINND), combining the idea of biological incentives with the backtracking search method, is designed. In order to solve the problem of unreachable areas due to the distribution of obstacle areas and starting points in the same allocation area, the area allocation mechanism, based on MK-means++ with obstacle information, is designed. The simulation results show that the path planning method has a better modeling effect and can improve the reachability of collaborative traversal process, which verifies the effectiveness, progressiveness, robustness and real time performance of the method.