The Research Of Improved Immune Genetic Algorithm For Function Optimization

Immune Algorithm (IA) is a realization of natural immune system in evolutionary computation, simulating the knowledge processing system of immune system about learning, memory and forgetting, so that there is a higher intelligence and robustness in decomposing, handling and solving distributed complex problems. Immune algorithm has been successful in many fields, such as virus removal and invasive monitoring, function optimization, TSP problem, data analysis and mining, fault diagnosis. However, IA has the disadvantage of slow convergence.Genetic Algorithm (GA) is a highly parallel, random and adaptive searching probabilistic method based on biosphere natural selection and genetic mechanism. But the theory and method of genetic algorithm is not mature. Some insufficiencies of algorithm also remains to be further improved and perfected.In view of the above-mentioned facts and function optimization as a common method for evaluation of algorithm performance, immune genetic algorithm (IGA) is presented and improved to solve function optimization problems. Propose a new approach:assessment of expected reproduction rate, calculate concentration of the vector distance, immune memory, approximation method of the crossover and mutative probability of mutation operator, analyze three algorithms' operational mechanism, research the affinity, concentration, immune memory, crossover and mutation to, and using improved algorithm to solve the function optimization problem in MATLAB. Compared with genetic algorithm and immune algorithm, simulation tests select some typical low and high dimensional complex functions to indicate that the improved immune algorithm can effectively avoid the "premature" problem, improve the convergence speed and obtain the optimal solution.Finally, through concluding the research work, the advantages and the weaknesses have been clear, which provide certain reference for further research of IGA.