Robotic action planning with the application of explanation-based learning

Domain-dependent knowledge and searching engine are very effective for enhancing synthesis speed and capability of a robotic action planning system. However, it is rather difficult to acquire the domain-dependent knowledge and searching engine, especially to recognize, acquire and compile them automatically. In this paper, a new learning based approach is developed for robotic action planning by applying explanation-based learning, which is best at acquiring domain-dependent searching engines. An example study shows that this method is feasible and effective