基于局部搜索机制MOGA的多目标ADGHP建模及优化  

Modeling and Optimization for Multi-objective ADGHP Based on Local Search MOGA

作　　者：张玉州[1,2] 陈文莉[2] 李锐[3] 江克勤[2,4] 李建华[1] 

机构地区：[1]中国科学技术大学计算机科学与技术学院,合肥230027 [2]安庆师范学院计算机与信息学院,安庆246133 [3]北京航空航天大学电子信息工程学院,北京100191 [4]华中科技大学控制科学与工程系,武汉430074

出　　处：《系统仿真学报》2013年第5期962-968,共7页Journal of System Simulation

基　　金：国家重点基础研究发展计划(973)项目(2010CB731805);安徽省高校省级自然科学研究项目(KJ2011B078);中国科学技术大学高性能计算与应用重点实验室开放课题基金(NHPCC-KF-1101)

摘　　要：依据航班的进离港过程,提出了一种航班优先系数计算策略,使得延误损失在进离港航班之间的分配合理化;在此基础上建立了一种进离港地面等待问题(ADGHP)多目标优化模型,以实现延误损失和续航航班延误时间的多目标优化。针对问题模型的复杂性以及现有多目标遗传算法(MOGA)的不足,提出了一种引入局部搜索机制的多种群遗传算法对问题求解,并改进优秀个体迁移策略,实现多目标的协同优化。最后,以国内某机场进离港航班为算例,使用所提算法进行计算,并与其它典型算法的求解结果对比,实验结果表明了所提模型与算法的有效性。A strategy of computing flight priority coefficient was brought forward according to arrival departure process, in order to assign delay cost between arrival flights and departure flights reasonably. Moreover, a multi-objective optimization model was proposed for Arrival-Departure Ground-Holding Problem （ADGHP）, so as to achieve multi-objective optimization of minimizing delay cost and delay time of flights with continuation of the journeys. Due to the complexity of the model and inefficiency of existing Multi- Objective Genetic Algorithm （MOGA）, a multi-population multi-objective genetic algorithm was proposed for the model, which included with local search strategy, and improved elitist selection and individual migration strategy to realize multi-objective co-optimization. Finally the simulation was implemented with the data of arrival-departure flights from an airport in China. The experimental results show the validity of the proposed algorithm and model, comoarin~ with the results of other representative algorithms.

关 键 词：进离港地面等待问题 改进多目标遗传算法 多目标优化 优先系数 

分 类 号：V355[航空宇航科学与技术—人机与环境工程] TP18[自动化与计算机技术—控制理论与控制工程]