Sandholm Algorithm with K-means Clustering Approach for Multi-robot Task Allocation

Abstract

Multi-robot systems are becoming more and more significant in industrial, commercial and scientific applications. The current attempts made by the researchers concentrate only on minimizing the distance between the robots and the targets, and not much importance is given to the balancing of workloads among robots. Auction based mechanism are popularly used to allocate tasks to multiple robots. This paper attempts to develop mechanisms to address the above two issues with objective of minimizing the distance travel by ‘m’ robots and balancing the work load of ‘N’ targets between ‘m’ robots equally. The proposed approach has three stages, stage I bundles the ‘N’ targets into ‘n’ clusters of targets using commonly adopted K-means clustering technique with the objective of minimizing the distance between the ‘n’ targets and its cluster centroids, this gives the legal bundles and also reduces the search space. Stage II calculates the biding distance based of the shortest path from the current robot position to bundle or bundle combinations. In stage III bundles are allocated to the each robot using Sandholm algorithm. The performance of the proposed method is tested with small and large size bench mark problem instances.