On the navigation of mobile robots in narrow passages: a general framework based on sliding mode theory

Abstract

This paper establishes a general framework for the passing of mobile robots between closely spaced obstacles and through narrow passages. Under such circumstances, the artificial potential function method often leads to lock-up situations (local minima) and oscillatory motion. Sliding mode theory is utilized to examine the gradient of an artificial potential field in the vicinity of the equi-distance line between the passage constraints. State-dependent and time-dependent smoothing methods are introduced that enable tracking of the gradient lines and decrease the influence of the obstacles in this region, ensuring safe passing. The methodology is illustrated in several numerical case studies in known and unknown environments.