受约束的多峰值问题的遗传算法优化

On the GAS Optimization of Constrained Multi-Peak Value Problems

针对工程设计最优化中多峰值、多参数，且参数间结构性强的特点，提出一种遗传算法（GAs）的综合实现方案，整体上采用多子群体并行演化的策略。提出以双链多染结构编码表示，算法借鉴了模拟退火技术的思想，对约束条件的处理及控制参数建立了自适应机制。目标是要提高遗传算法应用于受约束的多峰值优化问题的收敛性和稳定性。

The optimization of engineering design is characterized by multi - peak value, multi - parameter and strict structure between parameters. A GAS comprehensive realization scheme is correspondingly advanced, which takes the tactics of multi subpopulation evolution,forms the structural code of the. double - chained multi - chromosome and, by learning from the annealing technology and establishes an adaptive system in algorithm for handling constrain factors and controlling parameters. In doing so,the paper is aimed at improving the convergence and stability of the GAS optimization when used in constrained multi -peak value problems.