A functional extension of the sufficient condition to design a family of sensor-based deadlock-free path-planning algorithms

In this paper, we propose a functional extension of the previous sufficient condition to design a larger family of sensor-based path-planning algorithms operating in an uncertain world. Almost all previous sensor-based path-planning algorithms keep their deadlock-free characteristics by the asymptotical approach of an automaton toward the goal based on the Euclidean distance. This paper shows a generalization of the sufficient condition by replacing the Euclidean distance with different types of distances. By using an arbitrary distance function, we can easily design a lot of sensor-based path-planning algorithms. As an example, we design a learning sensor-based path-planning algorithm fitting adaptively the distance to world shape. The algorithm makes a deadlock-free path shorter than those generated by the previous algorithms without any learning