A new artificial life algorithm to solve time-varying optimization problem

In the application of optimization problems, there are a lot of time-varying problems whose parameters are shifting with time. The standard evolutionary algorithms can´t trace the leap of system peaks because of the loss of genotype diversity. In the model of artificial life, the system evolution is displayed through the interactions of individuals, and the loss of diversity can be avoided. A new evolutionary algorithm is designed by this principle. The species competition rule is applied to this algorithm to solve time-varying problem. The conclusion is proved by the experiment result.