The Research Of Cooperative Co-evolutionary Algorithm And Its Application

Cooperative co-evolutionary algorithm (CCA) is a heuristic search algorithm. Although it has not been developed for a long time, it has been successfully applied in many combinatorial optimization problem including many that are NP-hard. CCA is a kind of evolutionary algorithms, but it differs from traditional evolutionary algorithms in a point that in the cooperative co-evolutionary algorithm, there is competitive relation among the individuals of the population and at the same time there is cooperative relation among the individuals of different sub-population. At present, many scholars have done many research works in the aspect of cooperative co-evolutionary algorithm.In this paper, first, we analyze the research works that scholars are going along with in the aspect of cooperative co-evolutionary algorithm. Second, to overcome the blindness of selecting representative when calculate the fitness of individual, we put forward the cooperative co-evolutionary algorithm based on shared-memory. In this algorithm, we import shared-memory that play two important roles: first it can realize communication among the sub-populations; second it can preserve the excellent combination gained from the algorithm implementing process and in some degree it can improve the performance of the cooperative co-evolutionary algorithm. Meanwhile we have studied and discussed the parallelization of the cooperative co-evolutionary algorithm based on shared-memory. Finally, we apply the cooperative co-evolutionary algorithm to practical problem, e.g., the problem of the kinetics parameters estimating and process simulating inautocatalytic oxidation of cyclopean and the problem of data clustering.