Application of improved k-medoids algorithm in charging station planning for mobile robot
Qingdan Yuan, Jun Liu · 2017
In order to solve the problem of battery replacement when the battery is consumed to a certain extent by mobile robot working in outdoor environment, the clustering analysis is introduced to solve the problem of charging station planning for mobile robot. In this paper, combined with motion energy model, the criterion that similarity between two nodes is measured according to the distance in the traditional k-medoids is adjusted. Aiming at the limitation of random selection of the initial clustering centers, an improved k-medoids clustering algorithm is formed, finally obtaining a reasonable number and location of the charging stations, achieving a good balance between less motion energy consumption of mobile robot returning the charging station and the economic benefits of charging station construction. Experiment shows that application of improved K-medoids algorithm in charging station planning for mobile robot is effective and practical.