The Study On Lot-sizing And Scheduling Model And Its Algorithms Of An Automotive Interior Parts Production

It is an important aspect for making fine production andenhancing the competitiveness of enterprises that modern advancedmethods and techniques are used to optimize production operations andresource allocation and production planning and scheduling. And Withthe change of market environment and the needs of the customer productpersonalization, more variety and small batch production pattern hasgradually become the main mode of production of the manufacturing.Therefore, research about the optimization algorithms for flexible flowshop lot-sizing and scheduling problems has important theoretical valueand practical significance to improve the development of advancedmanufacturing.In this paper, the concept of flexible flowshop lot-sizing andscheduling problem and its features are summarized. And thedevelopment process and research of lot-sizing and scheduling problemare introduced. Besides, models of single-stage multi-product and multi–stage multi-product lot-sizing problem have been briefly summarized.Different algorithms for solving the lot-sizing and scheduling problemsare introduced. And then the basic idea of lagrangian relaxation, appliedresearch and its solving framework have been described briefly.Firstly, processing technique and scheduling features of car dashboard production flow shop are summarized. In order to solve thepractical problems of the company and change the situation that the plansof production planning and scheduling are calculated and adjustedmanually, a mixed integer programming model for lot-sizing andscheduling of flexible flow shop is established, which considerssequence-dependent setup times, due date and limited productioncapacity constraints. The optimization algorithm based on lagrangianrelaxation contains heuristic algorithms based on network analysis and theidea of dynamic programming for the relaxation of the sub-problems.And then designing a Lagrangian relaxation heuristic algorithm toconstruct feasible solutions and improve the upper bound value.On the basis of the analysis of the basic methods, an improvedsubgradient algorithm with the introduction of the concept of fuzzy isdesigned to update the lagrangian multipliers, while introducing Bundlemethod for comparative analysis; setting16different size of the problemand160example experiments. And the C++language is used to compeeralgorithms for processing simulation tests. The results prove thefeasibility and effectiveness of the algorithms based on lagrangianrelaxation.