Research On Modeling And Optimization Of Remanufacturing System Scheduling Problem Considering Energy Consumption

Manufacturing industry is the pillar industry of China’s national economy.While bringing great convenience to people’s lives,it has also led to a surge in the number of end-of-life products.Improper disposal of increasingly end-of-life(EOL)products will bring a series of serious problems such as environmental pollution,waste of resources and land occupation.Meanwhile,energy problems in China are particularly prominent,and a series of policies and regulations aimed at energy-saving and consumption-reducing have been introduced.As a natural extension of the manufacturing industry,the remanufacturing industry has great development potential.Remanufacturing industry changes product lifecycle from an open-loop to a closed-loop,significantly reducing the negative effects on the environment.Remanufacturing has gradually become an important part of the circular economy and green economy,which is the most ideal approach to deal with the growing number of EOL products.Remanufacturing system scheduling is an important way to realize the digitization and greening of remanufacturing.Therefore,this dissertation takes the disassembly-reprocessingreassembly three-stage remanufacturing system scheduling problem as the main research line,and deeply studies the model establishment and solution strategies of remanufacturing system scheduling problem from the single objective of minimizing the total energy consumption to multi-objective of minimizing the total energy consumption and the maximum completion time.Different from the traditional remanufacturing system scheduling problem,which only considers one or two production stages,the remanufacturing system studied in this dissertation includes three production stages of disassembly-reprocessing-reassembly,and consists of disassembly shop,reprocessing shop and reassembly shop.Specifically,the system configuration of the remanufacturing system studied in this dissertation is as follows:the disassembly shop is composed of unrelated parallel disassembly workstations,the reprocessing shop is composed of parallel dedicated flow-shop-type reprocessing lines,and the reassembly shop is composed of unrelated parallel reassembly workstaions.Firstly,the three-stage remanufacturing system scheduling problem considering energy consumption is studied.Based on the detailed discussion of the remanufacturing system configuration studied in this dissertation,the components of the total energy consumption of the remanufacturing system are analyzed one by one and its corresponding energy consumption modeling is carried out.Considering the actual production constraints,a remanufacturing system scheduling optimization mathematical model is established to minimize the total energy consumption.Combined with the characteristics of the scheduling problem and optimization objective,an improved genetic algorithm(IMGA)is therefore proposed.IMGA uses a doublelayer encoding mechanism based on the machine allocation layer-priority layer to encode chromosome individuals,and adopts a combination of random strategy and heuristic strategy to initialize the population to improve the quality of initial population.According to the encoding characteristics,two different kinds of crossover operators and three different kinds of mutation operators are designed to promote the high-quality evolution of chromosome populations.In order to improve the convergence effect of the algorithm,IMGA also introduces an elite strategy based on chromosome replacement mechanism.The results of test instances and comparison with other algorithms show that the IMGA can effectively solve the three-stage remanufacturing system scheduling problem that minimizes the total energy consumption.Secondly,the remanufacturing system scheduling problem considering Turn Off and On strategy is studied.The Turn Off and On strategy is briefly described at first,and the necessary conditions for the machine to execute Turn Off and On strategy are quantitatively described with mathematical formulas.The total energy consumption model of the remanufacturing system is the remodeled,so that the idle energy consumption in the original total energy consumption model can be further reduced by Turn Off and On strategy.A hybrid genetic algorithm based on variable neighborhood search(GAVNS)is proposed to solve the scheduling optimization mathematical model.In GAVNS,chromosome individuals are formed by sequence encoding mechanism,and three different decoding methods are designed.The algorithm uses an improved variable neighborhood strategy to enhance its local search ability.The effectiveness of GAVNS is verified through test instances analysis and algorithm comparison.Research results show that the Turn Off and On strategy can effectively reduce the total energy consumption of the remanufacturing system;the decoding method based on minimizing total energy consumption of the system has better performance than the decoding method based on minimizing the maximum completion time and the decoding method based on minimizing the processing energy consumption.Then,the energy-efficient remanufacturing system scheduling problem is studied.A multi-objective optimization mathematical model is established to minimize the total energy consumption and the maximum completion time.To solve this model,a multi-objective invasive weed optimization algorithm(MOIWO)is proposed.According to the characteristics of the scheduling problem,a weed individual in MOIWO is composed of a workstation assignment vector and a job scheduling vector;the initial population is generated by a hybrid method;the quantitative evaluation of individuals is realized by the Sigma method;the generation of offspring is completed by a spatial diffusion mechanism based on mutation distance and multiple mutation combinations.An external archive is employed to continuously store and update Pareto optimal solutions.Finally,an example is used to analyze the solution results in detail,and the conflict relationship between the two optimization objectives of maximum completion time and total energy consumption is explained.The comparison experiments with the classical multi-objective algorithms show that MOIWO can find Pareto optimal solutions with better convergence and distribution.Meanwhile,it is also found that the increase in the number of disassembly workstations/reassembly workstations can not only increases the flexibility of the remanufacturing system,but also further obtains a better Pareto optimal solution set,which can provide reference ideas for enterprise managers to implement energy-efficient scheduling.Next,the energy-efficient remanufacturing system scheduling problem considering lotstreaming is studied.In the problem,jobs enter the remanufacturing system in the form of lotstreaming.The lots adopt the inequal consistent splitting strategy and sublots cannot be mixed.Due to the fact that basic scheduling unit changes into several sublots of the lot and the introduction of Turn Off and On strategy,two optimization objectives of total energy consumption and maximum completion time are remodeled,and then a multi-objective optimization mathematical model is established to minimize the maximum completion time and the total energy consumption of the remanufacturing system.According to the characteristics of the studied problem,a fruit fly optimization algortithm incorporating simulated annealing(FFO-SA)is proposed.In FFO-SA,fruit fly individuals adopt a three-layer coding mechanism of lot order-workstaion allocation-lot size splitting;the lot order parts of initial population are generated by random strategy,the workstaion allocation parts of initial population are generated by a combination of random strategy and heuristic strategy based on the idea of workstation load balancing,while the lot splitting parts of initial population are generated by independently designed rules;According to different subproblems,several reasonable neighborhood structures and genetic operators are designed;in order to maintain the diversity of the population and avoid the algorithm from falling into a local optimum,a neighborhood solution replacement mechanism based on the idea of simulated annealing is introduced.Finally,the experimental analysis is carried out around the example from the literature,and experimental results show that:the "Lot preemption" decoding strategy is more suitable for the studied problem than the "Sublot preemption" decoding strategy;FFO-SA has good stability in addressing the energy-efficient remanufacturing system scheduling problem considering lotstreaming;regarding lot-stream scheduling,lot splitting strategy performs better than non-lot splitting strategy;the Turn Off and On strategy can effectively reduce the total energy consumption of the remanufacturing system.Finally,based on the above theoretical research results,a remanufacturing system scheduling management platform for remanufacturing enterprises is designed and developed,and its composition architecture and functional modules are introduced.Feasibility and validity of the theoretical research in this dissertation is verified by a real case from an automobile remanufacturing enterprise.