The Improment Of Aco With PSO Based On Excellence Mechanism

As a subdiscipline of bionics algorithm, Ant Colony Optimization(ACO)initiates a new way to resolve the combinatorial optimization. In the years,withthe effort of the research staff,the algorithm has been more mature. But with theharsher system condition, former ideal conditions do not fit the system any more.And it makes the pure ACO cannot resolve the complex problem. So how tomake the algorithm suit the developed technology has become the main subjectto us.First,the principle,parameter definition and current research of ACO aregiven. Compared with other bionics algorithms, ACO is more objective and haswide application. Meanwhile, it has the disadvantages such as time-consumingsolving processes and weak accuracy. So if integrated with other algorithms tomake up for the shorts, ACO could increase the ability to resolvethe combinatorial optimization greatly.Then, in order to resolve the algorithm weak point, Partical SwarmOptimization(PSO) is brought in. With the excellent randomicity, PSO could beused as the former part of ACO, and then lead the whole system-resolve morequickly. Compared to original ACO, new algorithm enhances its randomicityand global optimum, and avoid falling into local optimum of the system.Finally, mixed with PSO and ACO deeply is appeared. It uses excellencemechanism to make up for the time-consuming solving processes because of theexcellent positive feedback ability. With the outstanding effort to system, the newalgorithm could accelerate the optimizing course, and finds out the best answerwithout damaging the accuracy. At the same time, because of the independenceof the every single ant, excellence mechanism could leads the deep positivefeedback to affect PSO system more or less, thus the PSO part is more quicklythan ever before. New algorithm has great power in optimizing speed and make sure the final answer is global optimum rather than local optimum, providing anew theory to the combinatorial optimization. At last,the article giving aparameter estimation method of eil51,Traveling Salesman Problem(TSP) inACO/ACO with excellence mechanism/PSO-ACO/ACEPSO, proving theeffectiveness of the algorithm proposed through the simulation in MATLAB...