A new Immune Genetic Algorithm (IGA) modeling was completed using Markov chain. The convergence rate of IGA to absorbed-state was deduced using norm and the analysis of transition probability matrix. According to the design and the performance of IGA, the detailed quantitative expressions of convergence rate to absorbed-state which include immune parameters in IGA was presented. Then the discussion was carried out about the effect of the parameters on the convergence rate. It was found that several parameters such as the population size, the population distribution, the string length etc. would all affect the optimization. The conclusions demonstrate that why IGA can maintain the diversity very well so that the optimization is very quick. This paper can also be helpful for the further study on the convergence rate of Immune Genetic Algorithm.
展开▼
