Balancing Optimization And Dynamic Simulation Of Stochastic U-shaped Partial Disassembly Line

The disassembly line is an important link to realize the recycling and remanufacturing of waste mechanical and electrical products.The large-scale and automatic disassembly of parts and raw materials are realized through assembly line operation.In-depth research on the balancing of the disassembly line is to ensure the economic benefits of the enterprise while reducing the environmental impact influence and promote the development of green manufacturing in China.In view of the uncertainty of the work time of the disassembly task caused by the complex factors such as the changeable structure of the product to be disassembled and the skill proficiency of the worker,the random time of the work time is represented by the interval number which is easy to obtain the upper and lower boundary information because of the inability to obtain its specific probability distribution and membership function.The randomness of time,combined with the U-shaped layout,which has the characteristics of high production flexibility and small footprint,considering that enterprises pay more attention to the complete disassembly of hazardous and demanding parts,the stochastic U-shaped partial disassembly line balancing problem is proposed.A multi-objective optimization mathematical model that minimizes the number of workstations,balances indicators of idle time,maximizes disassembly profit and reliability of disassembly lines.Aiming at the characteristics of the problem,an adaptive opposition-based learning multiobjective wolf pack algorithm is proposed to solve the problem.In the process of discretizing the algorithm,consider removing the head wolf effect and retaining other algorithm operations.Using the "roulette" selection operation,the wolves are divided into elite wolves and ferocious wolves,and the elite wolves scouting behavior is executed based on random perturbation.As the number of iterations increases,the number of scouting search directions is reduced,taking into account the optimization ability of the algorithm in the early stage and the stability in the later stage.Based on the cross operator and mutation operator to execute the calling behavior and enhance the cooperative communication between wolves,the optimal solution of each subgoal is used as the prey target to perform the besieging behavior.The opposition-based learning strategy is introduced to generate the reverse solution population to avoid falling into the local optimal solution,and the global non-inferior solution is recorded and output through the elite retention strategy.The proposed algorithm is applied to a random incomplete disassembly example of a printer and 19 benchmark examples.Orthoplan experiments are designed to determine the best parameter combinations for problems of different scales,and they are applied to example problems to verify the excellent performance of the algorithm in comparison with existing solutions.Since the disassembly line balancing problem is a typical combination optimization problem and has a wide classification,a disassembly line design and simulation system is designed,which can realize the static design of the disassembly line scheme and the simulation in the dynamic complex environment,and combine the output results of the two modules,the optimal solution is selected for decision makers,and put forward targeted improvement opinions,to further improve the efficiency of disassembly.Taking a CRT TV and printer as disassembly examples of different scales,the system algorithm optimization module is called,enter the problem parameters and design parameters at different levels of confidence to solve each to obtain 10 sets of feasible solutions,the system dynamic simulation module is called,set the production simulation parameters of the enterprise,automatically build a simulation model for each solution and realize the simulated disassembly,obtaining the statistical data of the expected monthly disassembly output and workstation working condition statistical data of each solution.Then the optimal scheme was selected for the decision makers,and the improvement suggestions were put forward,which increased the monthly disassembly output by more than11%,and the work proportion of each workstation was more than 79.58%.