INTELLIGENT STRATEGY PLANNING AND CONTROL OF A GROUP OF MOBILE ROBOTS UNDER CONDITIONS OF INCOMPLETE INFORMATION

Different problems of strategy planning and control a mobile robot group under complex dynamicconditions with incomplete information about the external environment are considered. Approachesto solving effective work scheduling conditions inconstant active groupcomposition, searching for source nonstationary concentration field, supervisory ofdiscrete-event systems presented. An original mathematical model formulated in terms work-shiftscheduling problem-oriented modification evolutionary algorithms specializedset heuristics its efficient solution developed problem top-level groupwork. Searching monitoring field is carried out usinga decentralized multi-agent that combines elements bionic gradient approaches,as well as method generating artificial potential fields. The considered has lowcomputational complexity, high variability respect to types fields surveyed, easily scalableto any available number robots. latter special importance, particularwhen considering parallel independent multiple sources. It proposedto use means logical inference, namely automatic theorem proving calculus positivelyconstructed formulas, solve various supervised theory discrete-eventsystems used at different levels robotic hierarchical system. Features calculusallows dynamic control, processing controllingevents based on environmental data real time process inference efficiently. Theapproach positively constructed formulas allows studying properties automata-baseddiscrete-event systems, synthesize finite automata construction realizationof monolithic modular supervisors. A general scheme combining approachesfor controlling robots scales within single hierarchicalcontrol system proposed.