Research On Robust Project Scheduling Problem Based On Scattered Buffer

Most traditional project scheduling studies are conducted in a deterministic environment,which means,parameters such as activity duration and resource requirements are unique.However,the actual project environment is full of uncertainties,interference and risks,and the progress plan based on the determination of parameters may deviate greatly from expectations and even become unfeasible when implemented.Robust project scheduling,as an effective method to solve project scheduling problems in uncertain environment,aims to produce a scheduling plan with high stability and certain ability to resist uncertainties.In this paper,first of all,the robustness of the existing project scheduling problem was summarized and concluded.Then based on scattered buffer basic model,the robustness of project scheduling using robust deviation cost measure the robustness of the plan.In order to describe uncertain parameters,after briefly introducing the uncertainty theory,uncertain variables are used to represent uncertain construction period,and a method of uncertain simulation is adopted to calculate various uncertain variables involved in the process of solving robust plan,such as the actual starting time of activity,the deviation cost of activity reality and plan.Scattered buffer is based on the shortest time limit for a project baseline scheduling plan,before the activity start time inserts.In order to solve the resource conflict in the process of distributed buffer insertion,an improved heuristic algorithm for constructing resource flow network is proposed,and then an improved simulated annealing algorithm is designed to optimize the dispersed buffer size.In the experimenrt,examples from PSPLIB standard database were selected to obtain the actual project data by controlling different levels of uncertainty,project deadline and marginal cost,and to test whether the improved resource flow network algorithm and the improved simulated annealing algorithm could effectively formulate a robust plan based on decentralized buffer.Experiments show that the improved scattered resources flow network and simulated annealing buffer algorithm can make better robustness plan.The actual project management is a multi-objective decision problem.This article build double objective robust scheduling model to minimize project period and robustness cost,and use the improved multi-objective evolutionary algorithm NSGA-II.According to the dominant relationship and the crowd distance,it is possible to sequence the possible solutions,improve the population update strategy,and propose a local search operator to improve the optimization of the algorithm.In the experiment,Hypervolume index was used to test whether the improved algorithm could improve the quality of pareto optimal solution,and other indicators were designed to test the quality of the multi-objective evolutionary algorithm to obtain the optimal solution.Experiments show that the proposed algorithm can obtain a better pareto optimal solution.