A reinforcement connectionist approach to robot path finding in non-maze-like environments
This paper presents a reinforcement connectionist system which finds and learns the suitable situation-action rules so as to generate feasible paths for a point robot in a 2D environment with circular obstacles. The basic reinforcement algorithm is extended with a strategy for discovering stable solution paths. Equipped with this strategy and a powerful codification scheme, the path-finder can: (i) learn quickly, (ii) deal with continuous-valued inputs and ouputs, (iii) exhibit good noise-tolerance and generalisation capabilities, (iv) cope with dynamic environments, and (v) solve an instance of the path finding problem with strong performance demands.
Bibliographic Reference: Article: Machine Learning, Vol. 8 (1992) pp. 363-395
Record Number: 199210756 / Last updated on: 1994-12-02
Original language: en
Available languages: en