Uncertain Scheduling Research On Limited Resource

The uncertain scheduling problem with limited resources(USPLR)is a special schedule problem.In reality,some accidental,inevitable,etc.uncertain information always exist in the scheduling problem and disturb the performing results.Especially,when scheduling resources are limited,these uncertain information could affect the scheduling effectiveness and results at a large extent.In this case,studying this kind of scheduling problem,this paper could provide supports for the scheduling in reality to face various uncertain information and save the sources in need.This paper studies the case that making a schedule for the self-optimization agile Earth observation satellite(AEOS)which considers nearly real time cloud.With the gradually wide application and study of self-optimization AEOS,the study of this paper has the relevant research meaning and background.At the same time,due to the limited storage on the satellite and the uncertain cloud information,this study belongs to USPLR.At first,this paper constructs the triangle mathematic model which could help to capture the nearly real time cloud information.Then this paper gets the objective-constraint mathematic model by including related constraints.Next,this paper designs the algorithm to solve the model based on the limited resources and nearly real time cloud and other constraints.In the following,this paper adopts commonly used approaches of greedy search algorithm(GSA),dynamic programming algorithm(DPA)and local search algorithm(LSA)in AEOS to solve the present problem in this paper.Besides,this paper optimizes the GSA according to the characteristics of the studied problem when presenting the iterative stochastic greedy search algorithm(ISGSA)to solve the problem.At last,this paper designs related simulation experiments.The results of the experiments show that ISGSA could better solve the studied problem.The experiments devised in this paper have the parameters that observation period T is set as 3600s and the biggest customer satisfaction time is 90s.Besides,it compares the task profits,average image quality of observing,the satellite storage usage rate and the average task completion time.In a period that its task number is 100,ISGSA can improve observation profits by 5.72%,16.63%and 21.39%than GSA,LSA,DPA respectively.In this case,ISGSA could be better suitable to solve the studied problem in this paper.