An Improved Population Migration Algorithm Introducing Local Search Mechanism Of Leaping Frog Algorithm And The Crossover Operator

As a new kind of evolutionary algorithms, the intelligent group has become a newresearch characteristic in the optimization field. Having completed the theoretical andapplied research proved that the swarm intelligence algorithms are an effective newmethod to solve most of the global optimization problem. In particular, the group ofpotential parallelism and the characteristics of distributed intelligence to deal with a greatdeal of data in the form of data, provides a fundamental guarantee.Population Migration algorithm (PMA) is proposed a kind of global optimizationsimulation principle of population Migration intelligent algorithm by a famous scholar ofour country Zhou Yonghua in2003. Population migration algorithm is a simulatedprinciple of population migration global optimization intelligent algorithm. Compared withthe other traditional optimization algorithm, the population migration algorithm has fastconvergence rate, the superior quality to solve the problem and the advantages of goodrobustness in multi-dimension function optimization, dynamic objective function. Becausethe population migration algorithm has the slow searching speed, easily falls into the localoptimal solution, wastes a long time of the contract and has a low accuracy shortcomings.So the study of algorithm is a necessary task.On the basis of general population migration algorithm, the author of this paper made afurther improvement and promotion, at the same time obtained satisfactory results. Themain work of the paper can be summarized the following several aspects:First, because the population migration algorithm dealing with the complex functions, iseasy to fall into local optimal and has low convergence precision of faults, the authorintroduces the local search mechanism of the leapfrog algorithm and the crossover operatorto improve the migration of the population migration algorithm strategy, effectively avoidsthe premature problem of the population migration algorithm. So we can improve searchrate of complex problems of the population migration algorithm.Secondly, by using the theory of stochastic process, demonstrates that the improvedpopulation migration algorithm is a feasibility and effectiveness of the algorithm. At thesame time, Experiments conducted on solving complex function optimization and simulation. Results show that the success rate of obtaining accurate solution is very high,which can reflect the improved algorithm has strong adaptability, stability robustness andglobal search ability.