Combined Map-Based and Case-Based Path Planning for Mobile Robot Navigation

This paper presents an integrated approach to mobile robot path planning by combining grid-map based and case-based methods. The map-based path planner is used to suggest new innovative solutions for a particular path planning problem. A probabilistic path transform method is introduced for grid-map based path planning which proposes several alternative solutions for a single path planning problem. The casebase is used to store the paths and evaluate their traversability. Using the casebase allows the robot to beneet from its former experiences while planning a new route. As the dynamic properties of the environment may change the exploration as well as the evaluation of the paths will allow the system to self-organize by forming a set of paths that will be safest to follow.