改进的模拟退火遗传算法在函数优化中的应用

针对模拟退火遗传算法中可能出现的早熟收敛和后期进化较慢问题,提出了多规则选择算子,同时对交叉和变异算子进行了改进,引入了小生境技术解决早熟收敛问题。在此基础上针对函数优化问题设计了改进的模拟退火遗传算法。仿真实验表明,改进的算法在函数优化中,特别是在对多变量函数寻优中,收敛速度和收敛精度都有一定提高。

Application of Improved Simulated Annealing Genetic Algorithm in Function Optimization

Having as concern the problems of early-maturing constringency and relatively-slow evolution in later stage which may arise in simulated annealing genetic algorithm,multi-method selecting operator is formulated.At the same time crossover and mutation operator is improved and niche is introduced to solve the premature problem.On that basis,aimed at functional optimization the improved functional simulated annealing genetic algorithm is designed.Simulated experiment shows a certain increase in both speed and accuracy of constringency in the improved algorithm which is in functional optimization,especially in those of multi-variables.