Multi-Agent Working Together Based on the Improved k-Means Algorithm

Aiming at the multi-agent problem in collaborative multi-objective, it is easy to fall into the optimal local solution. This paper proposes a quantitative analysis method based on k-means and ant colony algorithm (Improved K-means & ACO) to study the robot ball picking path planning problem. First, the clustering algorithm divides the sample into several clusters so that each cluster is covered by only one robot. Subsequently, in addition, second-order clustering is used to distribute the task amount in the cluster evenly, and the ant colony algorithm is used for global optimization, which improves the efficiency of the optimal time solution, and the simulation results show that the algorithm can complete the task in a short time and solve the multi-traveling salesman problem in the ordinary environment, which is of great significance for the research and development of multi-agent collaboration and path planning.