OPTIMIZATION BASED ON COMBINING MODELS OF ADAPTIVE BEHAVIOR OF A SWARM OF AGENTS

A bionic search architecture has been developed to solve the problem of placing VLSI elementsbased on hybridization algorithms a bee colony and swarm chromosomes,which allows you get out "local holes" increases convergence placement algorithm. The initial iterations are implemented by algorithm provide broad overview ofthe area, final chromosome algorithm,which ensures exact localization extremum found Agents representedas population chromosomes, which genotypes for solving problem.The paper describes modified paradigm which, in contrast thecanonical method, provides possibility searching solutions affine space positionswith integer values parameters. In method optimization aswarm agents chromosomes. is thegenotype object. essence procedure successive changeof states object (chromosome) directed mutation operator thesearch optimal state. An affine-relaxation model (ARM) chromosomes isproposed - this graph whose vertices correspond arcs toaffine bonds between them. transition new state carried using arelaxation procedure. work, acts as means changing thesolution, change genes chromosome. Thepurpose reduce weight bond Themechanisms described. structure algorithmis proposed. For each base chromosome, probabilistic choice set chromosomeslocated vicinity implemented. It possible improve qualityof adjusting control time complexityof fixed size number generations O(n). Ingeneral, dependence running hybrid O(n2) – O(n3).