Cloud-based task distribution system infrastructure for group of mobile robots

Abstract

One of the possible approaches to the construction of control systems for groups of mobile robots using distributed cloud technologies is considered, for which a scheme of access to information resources and a mechanism for distributing resources of a cloud computing system with linear decomposition are developed: The solution to the problem is divided into a series of smaller, simpler, subtasks in a hierarchical tree based on the linear distribution method. The specifics of the workspace model are shown, the goals of the functioning of robots are formalized, and the optimal energy evolutionary algorithm for solving the problem of distributing tasks in the team is proposed taking into account the initial and current levels of battery power, the energy consumption of each robot and the energy needed to perform individual tasks. Parameters for evaluating the effectiveness of the obtained solutions are determined, and genetic algorithms are synthesized, for which a coding form of the solution in the form of a chromosome is proposed and specific fitness functions are compiled. An algorithm has been developed for calculating the fitness function, implemented taking into account the specifics of its work in the cloud. Experimental results were obtained when checking the operability of the algorithms on the available onboard computing means of mobile robots, and the effectiveness of using distributed computing resources of a group of robots was estimated when implementing cloud services on their basis.