一种改进的遗传算法:Family GA  

Family GA - An improved genetic algorithms

作　　者：李建华[1] 王孙安[1] 杜海峰[1] 

机构地区：[1]西安交通大学机械工程学院,陕西西安710049

出　　处：《控制与决策》2004年第9期999-1003,共5页Control and Decision

基　　金：国家973项目(G1998030405);陕西省自然科学基金资助项目(2001X17).

摘　　要：分析了影响遗传算法性能的因素,在遗传算法(GA)的基础上设计了一种新的家族遗传算法(FGA).该算法改造了选择和变异算子,其目的在于提高收敛速度、避免早熟.同时,该算法提出在优良解附近构造最优家族,在此微型空间中进行精确搜索,确保了算法收敛速度和解的精度.最后给出4个典型函数的模拟例子,由对比实验结果可以看出,FGA提高了收敛速度及解的精度,说明该算法具有应用的潜力.Based on the analysis of any factors that affect the performance of genetic algorithm, a novel genetic algorithm (Family GA) is presented. In this algorithm, election operator and mutation operator are redesigned in order to increase the convergent speed and avoid premature phenomenon. And the optimum solution family around quality individuals is constructed at the same time. Continuous search is done in this micro-space. So the convergent speed of the algorithm and accuracy of the solution can be affirmed. Four typical function tests are given. The result indicates that there is much improvement in the speed of convergence and the accuracy of the optimal solution.

关 键 词：遗传算法 种群规模 个体空间 早熟现象 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]