基于MATLAB的局域网多机并行遗传算法求解不确定优化问题  

Solution of optimization under uncertainty by parallel genetic algorithm based on MATLAB under local area network

作　　者：蒋峥[1] 刘斌[2] 方康玲[2] 

机构地区：[1]湖北省武汉科技大学信息科学与工程学院自动化系,430081 [2]武汉科技大学

出　　处：《微计算机信息》2006年第05X期215-216,222,共3页Control & Automation

基　　金：国家科技攻关计划;编号2005BA102C;"基于FPGA技术的个人定位导航与监控系统"

摘　　要：本文提出一种新的不确定非线性优化问题的确定化形式,采用局域网多机并行遗传算法求解不确定优化问题,应用常用的计算软件MATLAB来实现该算法。通过在主机上开辟共享缓冲区,在其中设立数据文件,主机和从机都用MATLAB语言中读写文件的命令来交换数据,以实现局域网中的并行计算环境。该算法编程简便,充分利用操作系统提供的数据共享传输功能,性价比较高。The deterministic framework of the nonlinear programming under uncertainty is presented in this paper. The parallel genetic algorithm under the local area network is used to solve the framework. The algorithm is implemented with common software MATLAB. The shared buffer area, in which the data file is used, is created in the master computer, and the master computer and the slave computers can exchange information by using the file operation commands. The algorithm is implemented simple, effectively utilizing the operating system＇s performance of data sharing and transforming.

关 键 词：局域网 并行计算 不确定优化 遗传算法 

分 类 号：TP311.1[自动化与计算机技术—计算机软件与理论]