摘要
本文根据(N+M)容错系统优化模型高度非线性和复杂性的特点,采用具有高效、并行、全局性的遗传算法来实现此类复杂模型的优化求解.针对改进模型的特点,从编码、适应值、复制算子、交叉和变异算子以及控制参数的选取等多方面进行了遗传算法设计;用MATLAB语言实现了遗传算法,求得的结果验证了算法的正确性和有效性;算法性能分析证明了所设计的改进的交叉算子和随遗传代数成正态变化的变异概率等对改善遗传算法的搜索效率具有明显效果.
Genetic algorithm,which is of high efficiency,can perform parallel and global searches,is employed for solving the optimization cost model of (N+M) fault-tolerant system of hydropower station.This model features ligh nonlinearity and complication.Genetic algorithm for the model is designed with encoding,fitness,reproduction operator,crossover operator,mutation operator and control parameters taken into consideration,and is implemented by use of Matlab.Correctness and effectiveness of the algorithm are verified by calculation results,while the measures taken in the algorithm,for instance,the improved crossover operator and mutation probability,varying with Gaussian distribution,are proved by the performance analysis to be a good help in increasing the search efficiency.
出处
《内蒙古工业大学学报（自然科学版）》
2004年第3期185-191,共7页
Journal of Inner Mongolia University of Technology：Natural Science Edition
基金
教育部科学技术研究重点项目(00151)
教育部<高等学校骨干教师资助计划>项目
关键词
水电站
容错系统
遗传算法
费用模型
hydropower station
fault-tolerant system
genetic algorithm
cost model
