【摘要】 给出了一种进行局部快速微调的遗传算法———在变异中,将适应值高的个体和适应值低的个体分别进行诱导和随机动态区域变异;在交叉操作中,划分为搜索阶段和微调阶段,分别采用随机线性交叉和部分确定性诱导交叉.应用该算法对全局最优解邻域进行搜索,能在较短的时间内找到高精度的数值解.对10个典型测试函数优化问题的实验表明,该方法具有快速、稳定和易于实现的优点.如果合理调整控制参数,其全局搜索的收敛速度和解的质量明显优于传统GA.更多还原

【Abstract】 A genetic algorithm with fast local adjustment is presented.In the mutation,high fitness individuals in population are executed inducing mutation and others are executed stochastic-dynamic-range mutation;in the crossover,the algorithm is divided into searching phase and adjusting phase.And in two phases,the stochastic linear combination crossover and partial certainty inducing crossover are adopted respectively.The high accurate numerical solution can be found in the short time when the FLAGA adopted in an adjacent domain of the global optimization solution.By the experiments of ten typical test functions’ optimization,it’s proved that the FLAGA is fast and stable and easy to realize.In the global search,the FLAGA’s convergence rate and solution quality exceed GA’s obviously when it’s control parameters are adjusted rightly.更多还原