Consensus achievement method for a robotic swarm about the most frequently feature of an environment

Abstract

Decentralized multi-agent robotic systems have many advantages over centralized ones. Such systems make it possible to distribute computational and communication operations between all its elements, and are also more resistant to the loss of individual elements of a swarm, but they complicate the implementation of complex high-level tasks. An example of such a problem is the selection of one of the possible alternatives, in which the swarm must choose the most favorable solution from a list of possible alternatives. This paper proposes an algorithm for reaching a consensus for robotic swarms deployed in scenarios in which they can choose one of the two most common signs in the external environment. The proposed algorithm is based on the calculation of the measured features of the environment, as well as the distribution of this data between robots. The algorithm was tested using the ARGoS simulator.