OPTIMIZATION OF SWARM ROBOTICS ALGORITHMS

Context. Among the variety of tasks solved by robotics, one can single out a number those for solution which small dimensions work are desirable and sometimes necessary. To solve such problems, micro-robots with needed, mass allows them to move freely in tight passages, difficult weather conditions, remain unnoticed. At same time, microrobot also impose some indirect restrictions; therefore, it is better use groups microrobots these problems. The efficiency using depends on chosen control strategy stochastic search algorithms optimizing group (swarm) microrobots.
 Objective. purpose this consider swarm (methods) belonging class metaheuristics. includes, particular, ant colony algorithm, possibilities were investigated traveling salesman problem, often arises when developing an algorithm behavior Method. first stage study, main parameters identified that determine flow characterize state at any time algorithm: input, control, disturbance parameters, output parameters. After identifying was developed, advantage lies scalability, as well guaranteed convergence, makes possible obtain optimal regardless dimension graph. second stage, code implemented Matlab language. Computer experiments carried influence output, convergence algorithm. Attention paid indicators direction method given time. In computational experiment, ants placed nodes network, amount pheromone, graph varied, iterations find shortest path, execution determined. final test modeling performance out.
 Results. Research has been application solving problem graphs random arrangement vertices; constant vertices change ants, different values coefficient Q; pheromone evaporation p; vertices. results showed methods good routes much faster than clear-cut combinatorial optimization methods. dependence found route established example networks iterations.
 Conclusions. studies make give recommendations microrobots.