用鱼群算法求解通风系统风机定位优化问题.pdfVIP

维普资讯 ComputerEngineeringandApplications计算机工程与应用 2008，44(1) 243 用鱼群算法求解通风系统风机定位优化问题 黄光球，姚玉霞，陆秋琴 HUANGGuang-qiu，YAOYu—xia，LU Qiu—qin 西安建筑科技大学 管理学院，西安 710055 SchoolofManagement。Xi’an UniversityofArchitecture ＆ Technology，Xi’an710055，China E—mail：huangnan93@sohu．con HUANG Guang—qiu．YAO Yu-xia，LU Qiu-qin．Locatingoptimizationoffansinventilation system based onfish—swam algorithm．ComputerEngineeringandApplications，2007，44(1)：243-248． Abstract：In orderto solve the optimization problem oflocation offans in a ventilation system ofunderground mine，a large— scale nonlinearmixed integerprogramming modelisestablished for this problem．In the optimization model，the solving space is discretized based on actualconsideration ofspace constraintsofvariablesand precision ofsolutions．Thepolicyofdiscretization is to discrete each realvariable into finite values and quantize the continuousintervalofeach variable into finite levels．Because the dimension ofthe modelisSO enormousthattraditionaloptimization techniquescannotfind itsoptimalsolution effectively，an improved fish—swarm algorithm isused to solve the problem．In the improved algorithm，in orderto delete constraints and con— pressgreatly the variables ofthe mode1．allvariablesare grouped to reduce the numberofvariables．An evolutionary algorithm based on theneighborcompetitive evolution operatorisapplied SOthatthebasicpropertiesofglobaland localsearch aremixed togetherand abalance between efficiencyand effectivenessisrealized．Themulti—agent—based serf-leaningand self-adaptivevari— ation operator，theorthogonalcrossoveroperator，theneighbor competitive evolution operatorand SO on areused to improve all behaviorsofartificialfish—swarm algorithm．An application resultshowsthatthe speedandreliabilityofthealgorithm hasa signifi— cantimprovementintheoptimization algorithm ，andthesolutionofthe