An active path-planning algorithm based on pattern-classification

Path-planning is a key technology for mobile robots and its quality has close contact with the whole mobile robot´s performance. The shortcoming of the traditional solution based on potential field is that all information is reduced to a resultant force, so that it loses some valuable information of obstacles. Furthermore, it potentially gets into the local minima. In this paper, a new robot´s path-planning algorithm by classifying patterns is presented. This algorithm is easy to implement, and it improves the adaptability and reliability of path-planning