Research On Distributed Assembly Shop Scheduling Problem Based On Intelligent Optimization Algorithms

With the rapid development of science and technology,China's manufacturing industry is to informatization,integration,intelligent direction.The introduction of computer information technology provides a practical basis for flexible and efficient production mode.Considering the change of market demand,the large-scale discrete manufacturing system based on multiple varieties and small batch production gradually replaces the traditional production mode,which also drives the development of distributed manufacturing.As an important research direction of distributed manufacturing in assembly shop,Distributed Assembly Shop Scheduling Problem can be better applied to assembly manufacturing systems with a high degree of complexity and improve the coordination and agility of assembly manufacturing systems.At the same time,as an extension of the traditional assembly shop scheduling problem,the Distributed Assembly Shop Scheduling Problem belongs to the NP-hard problem,which is difficult to solve,and needs to be solved by efficient intelligent computing and optimization technology.This paper focuses on two types of extension distributed assembly shop scheduling problem is studied,in order to makespan for scheduling goal,design improved hybrid estimation of distribution algorithm(IHEDA)and improve hybrid clonal selection algorithm(ICSA-SA/ICSA-TS),and whose effectiveness is verified by simulation and practical cases,finally develop a scheduling algorithm software.The main research contents are as follows:(1)The background significance,research status and related theories of distributed assembly shop scheduling are described,and the basic steps and applications of estimation of distribution algorithm and clonal selection algorithm are described.(2)In order to solve the Distributed Parallel Assembly Permutation Flowshop Scheduling Problem(DPAPFSP),the corresponding structural model and mathematical programming model were constructed through problem analysis,and then an improved mixed estimation of distribution algorithm(IHEDA)was proposed.The algorithm adopts single-layer permutation encoding and decoding strategy based on the rule of Earliest Finished Time(EFT),introduces the local search mechanism based on the critical path,and designs five neighborhood structures.In addition,the algorithm introduces a double sampling strategy based on Repeat Rate and conducts simulated annealing search for the current optimal solution.Finally,IHEDA is compared with other algorithms through simulation experiments to verify the effectiveness of the improved algorithm.(3)In order to solve the Distributed Assembly Unrelated Parallel Machine Scheduling Problem with the limitation of machine and parallel assembly machine(DAUPMSP?L?PA),the corresponding structural model and mathematical programming model were constructed through problem analysis,and then an improved hybrid clone selection algorithm was proposed.In this algorithm,single-layer permutation encoding and decoding strategy based on the rule of EFT is adopted,and roulette based clone selection operation and ladder mutation number based mutation operation are introduced in the immune operation.At the same time,a local search mechanism is introduced to improve the search efficiency of the algorithm.Finally,ICSA-SA and ICSA-TS are compared with other algorithms through simulation experiments to verify the effectiveness of the improved algorithm.(4)Taking a knitted garment production line of a textile manufacturing enterprise in China as an example,a practical application case was carried out to verify the improved algorithm.Meanwhile,a set of intelligent optimization software for distributed assembly shop scheduling was developed to provide theoretical and decision support for the practical application of distributed assembly shop scheduling.