Management of Behavior of a Swarm of Robots Applicable to the Tasks of Monitoring a Some Territory

Abstract

The work is devoted to the design of bio-inspired methods and algorithms for controlling the collective behavior of robots based on adaptive behavior models of multi-agent biological systems of the region: detection of harmful substances; search and rescue operations in areas of natural and technological disasters, as well as combat zones, patrolling the borders of a certain area. The fundamental idea of swarm control is “swarm intelligence”. The control of a homogeneous swarm of robots is based on the principle of force relaxation. The search for a solution is carried out in an affine space whose elements are n-dimensional points (positions). Each robot ri calculates the objective function f (Xi(t)) - the value of the desired substance at the point Xi(t). The fundamental problem that was solved in this paper is related to the development of the structure of an affine position space that allows one to display and search for interpretations of solutions with integer parameter values. Coordinate values must be discrete and independent of each other. In contrast to the canonical particle swarm method, in order to reduce the weight of affine bonds, by moving the pi particle to a new position in the affine solution space, a directed mutation operator has been developed, the essence of which is to change the integer values of the genes in the chromosome. New chromosome structures have been developed for representing solutions. The temporal complexity of the algorithm lies in the range O (n2)–O (n3), where n is the number of robots.