摘要
针对电梯群控系统的特点,设计了以候梯时间、乘梯时间、能耗为群控目标的多目标函数,为找到一个最优解实现多目标整体最优,引入了基于遗传算法的文化算法,实现了在此模型上的应用。种群空间利用遗传算法进行选择,交叉,变异,知识空间由状况知识构成,两空间独立演化,通过接受函数和影响函数进行沟通,在一定状态下知识空间引导种群空间进化,达到增强搜索能力加快搜索速度的目地。仿真表明,文化算法优于单纯的遗传算法解决电梯群控多目标寻优问题。
According to the features of elevator group control system,the multi-object function was composed by average waiting time,average riding time and energy consumption.To find the optimal solution,cultural algorithm was introduced to this mathematical model.The population space based on genetic algorithm included select,cross and variation.The knowledge space was comprised of Situation Knowledge.The two spaces evolved independently and connect by accepting function and influence function.The knowledge space guided the direction of the population space under a certain condition to prove the search ability and the speed of genetic algorithm.The results of simulation show the advantages of cultural algorithms.It is a better way to solve the multi-object function problem in research of elevator group control system.
出处
《计算机仿真》
CSCD
北大核心
2013年第7期300-303,387,共5页
Computer Simulation
基金
国家自然科学基金资助项目(60874037)
关键词
电梯群控
文化算法
遗传算法
多目标优化
Elevator group control system
Cultural algorithms
Genetic algorithm
Multi-object function
