Research And Application Of Collaborative Scheduling Optimization In Metal Structural Part Manufacture Workshop

Metal structural parts are primary units in heavy industrial machinery products.The main manufacture processes consist of cutting and nesting,machining,and component welding et al.,which are completed in cutting workshop,machining workshop,component welding workshop respectively.Metal structure parts perform small batch and many varieties,the coexistence of part manufacture and assembly manufacture in the production process.Then the produced intermediate parts are of various specifications.The absence of any intermediate parts will lead to the retention of all matched parts,result in chaos on the production site and delay the completion of products.Therefore,it is necessary to carry out scientific and reasonable collaborative production scheduling,so as to reduce wip inventory,shorten the total completion time of products and improve production efficiency.In view of the above characteristics in the heavy manufacture process of metal structural parts,the existing standardized or universal scheduling methods are not fully applicable this question.At present,there is no mature scheduling model and algorithm to solve the complex scheduling problems in steel structure manufacture.Otherwise,the existing shop scheduling researches usually consider a single workshop merely.The studies always assume that the jobs have been prepared in advance and can be processed at any time.In fact,the start time of one job should be determined by its preorder workshop.Hence,collaborative scheduling optimization among related workshops is apt to achieve better overall optimization effect.Compared with single workshop scheduling,the variables of multi-workshop collaborative scheduling are multiplied,and the feasible solutions are the product of the feasible solutions in each workshop which increase the difficulty to solve the problem.This thesis considers the production scheduling in the manufacture of metal structural parts as the research object,studies the single-workshop scheduling and multi-workshop collaborative scheduling problems,and develops the production scheduling optimization platform which is suitable for the metal structural parts workshop.The research contents are as follows in detail:First,in view of the situation that the identical parallel machine and unrelated parallel machine scheduling coexist in the cutting workshop,and the characteristics that the processing time is determined by both the machine and the job,a mixed integer programming model(MIPM)for complex parallel machine scheduling problem(CPMSP)is established.The problems including plate allocation to the cutting machine reasonably and the process sequence of each plate on any machine should be solved at the same time.The minimization of maximum makespan and total tardiness are realized.The non-dominated sorting genetic algorithm III is modified by local search strategy(LNSGAIII)to solve this problem.Numerical examples of 20 different scales are designed and compared.The experiment results show that LNSGAIII is superior to NSGAII and NSGAIII in both exploration and exploitation capacity.The effectiveness of elite selection operator based on Pareto non-dominant sorting and reference point set is fully testified.In addition,the scatter chart of the Pareto solutions shows that the results by the new coding method perform better,and the distribution is more concentrated.Compared with the deterministic ?-constraint method to solve the scheduling numerical example of ? =0.2,? =0.4,the proposed algorithm is more effective especially on solving large-scale problems.Meanwhile,according to the characteristics of multi-machine operation to a same job in the welding process of large structural parts,the welding flow scheduling problem(WFSP)is proposed.A bi-objective mathematical model is established with the total tardiness and the total machine load as the optimization objectives.This model is designed to solve the problem of welding sequence of each job and the number of welding machines used at each stage.For this model,the algorithm NSGAIII with a restart strategy(RNSGAIII)is designed,and a local search which can reduce the machine load is proposed to improve the algorithm performance.The numerical examples are run and compared.It can be seen that the improved RNSGAIII is apt to search for more and better Pareto solutions than other algorithms.Modeling,algorithm design and solution of the bi-objective PMSP and WFSP show that the elite selection strategy based on the reference point can not only accelerate the solving speed,but also increase the number of Pareto solutions and improve the distribution performance of the solutions.Secondly,by analyzing the fractal manufacture characteristics of cutting workshop and the process correlation between plate cutting and machining of parts,the mathematical model of collaborative scheduling between cutting workshop and machining workshop is established.It is used to determine the corresponding machine to cutting plates,the plate cutting sequence on each machine,and the machining production sequence of the parts after cutting.An improved decimal Grey Wolf Optimizer with collaborative attacking strategy(CDGWO)is designed to solve this problem.Experiments demonstrate that the improved CDGWO can obtain excellent solutions with high solving efficiency.On the basis of the welding process BOM of the structural parts and kitting characteristics,the machining and welding shop collaborative scheduling problem is proposed which should determine both the production sequence of parts in the machining shop and the operation sequence of welding parts in the welding shop.The algorithm MGWOII based on scheduling rules is designed to solve this problem.Numerical experiments are run by 8 compared algorithms,the results show that the designed MGWOII algorithm based on first come(included parts)first process(welding part)scheduling rule performs best.The evolution speed of MGWOII is the fastest and the solve efficiency is the highest.All the above confirm the effectiveness of the local search and the attacking strategy.Then,the characteristics that the process object changing with the process and the inclusion relation are studied,and the collaborative scheduling problem of the cutting,machining and welding,where fractal manufacturing and assembly production coexist,is proposed.The bi-objective optimizing mathematical model is established.This problem should confirm the corresponding cutting machine of each plate and the cutting sequence,the parts sequence on each machining machine,and the production sequence of welding parts in the welding workshop simultaneously,aiming to minimize the total flow time and the total completion time.Based on the advantages of NSGAIII and GWO algorithms,a co-evolutionary bi-objective GWO algorithm(EBGWO)based on the reference point is designed.Considering the coevolution characteristic of this problem,a three-layer coding scheme is designed,and the scheduling rule of first come(included parts)first process(welding part)is designed to arrange the weld assembly scheduling.In this paper,a local search strategy for machining workshop is proposed,which is likely to reduce the maximum machining completion time without changing the total makespan,then lessen the total flow time.Examples with 36 different scales are designed and solved by the compared algorithms.Several box figures of IGD are drawn and analyzed.It can be seen that the EBGWO performs the best and it can obtain the Pareto solutions with good distribution for most questions in a limited time.At last,based on the above theoretical research,the workshop scheduling optimization system of metal structural part is designed and developed.The production data collection,scheduling optimization,process traceability and corresponding control functions in the manufacturing of metal structural part are realized.The application is verified in the structural part workshop of a shield mechanism manufacturing company.