능동 SLAM을 위한 계층적 커버리지 경로 계획 기법

초록

In this study, we propose a hierarchical coverage path planning method for active SLAM. Active SLAM is a technique that plans robot paths to revisit areas with high uncertainty in SLAM, thereby improving SLAM accuracy. However, visiting only areas with high uncertainty can lead to divergence in SLAM due to a continuous increase in covariance. The proposed method suggests an enhanced coverage path planning under the constraint of not exceeding acceptable uncertainty. To achieve this, the ant colony optimization method and the genetic algorithm were approached in a hierarchical structure. In simulations, performance comparisons and analysis were conducted between the proposed method and conventional approaches. The proposed method demonstrated results with shorter path lengths while simultaneously not exceeding acceptable uncertainty, compared to conventional methods. In the future, we plan to demonstrate that the covariance of SLAM is reduced and the map is accurately constructed by the proposed method in real time.

키워드