| The two-sided assembly line is widely used in the assembly and production of large-scale mechanical products such as diesel engines and loaders.Compared with the traditional one-sided assembly line,it has the advantages of short line length,high resource utilization rate,low production time and cost,etc.In recent years,the upgrading of the equipment manufacturing industry has intensified the changes in market demand,forcing companies to continuously use the new process and technology to rebuild the original assembly line.How to optimize the task assignment based on the original balancing scheme,complete the assembly line reconstruction with the minimum cost,and ensure the efficient operation of the assembly line after the reconstruction is an urgent and difficult problem to be solved.The task reassignment on the two-sided assembly line needs to consider the correlation effect of the task movement,as well as the limitation and influence of the assembly process change,the task reassignment constraints and the space constraint,so as to effectively control the assembly line reconstruction cost while improving the assembly line operation efficiency as much as possible.Therefore,it is urgent to carry out research on multi-objective two-sided assembly line rebalancing methods under multiple constraints.Specific research contents are as follows:(1)Considering the task reassignment of the two-sided assembly line is affected by the precedence constraint and the left-right side allocation mechanism,a mathematical model of rebalancing two-sided assembly line with the objectives of optimizing the assembly line capacity and rebuilding cost after reconstruction is established.The correlation effect of task movement in the task reassignment process is studied,and the inhalation,push and single-workstation-based task reassignment strategies are proposed.The multi-objective two-sided assembly line rebalance heuristic algorithms based on ?-constraint method are designed.The differences,advantages and disadvantages of the three task reassignment strategies are compared and analysed through internationally open standard cases and an engineering case,and results show that single-workstation-based task reassignment strategies have evident advantages.(2)An improved multi-objective genetic algorithm is proposed for the problem of assembly line rebalance caused by changes of assembly tasks,the precedence relationship and assembly line structure under the condition of process change.The method of rebalance cost calculation based on task movement inside and outside workstation and the cycle time estimation method based on idle time in the station are studied.The crossover and mutation operators based on the control of rebalancing cost and estimated cycle time is designed to obtain the feasible solution set under multiple constraints.A decoding method based on estimated cycle time is proposed to quickly search for the better solution set in the solution space.The Pareto optimal solution set update mechanism based on nondominated sorting is utilized to obtain the Pareto solution set of two-sided assembly line rebalancing problem with the objectives of optimizing the production capacity and rebalancing cost.Finally,the effectiveness of the algorithm is verified by the international standard case set and engineering case of the two-sided assembly line.(3)Considering the influence and limitation of the limited resources on the task reassignment,four types of task reassignment constraints are proposed,and a rebalancing model is established to optimize the station's completion time balance,cycle time and reconstruction cost.A multi-objective optimization algorithm based on imperialist competitive algorithm is proposed.The population initialization method based on multi-variable weight is designed to ensure the feasibility and diversity of the solution.The cycle time estimation method is improved,the pre-assignment mechanism of restricted tasks and the task assignment mechanism based on side are proposed.A decoding method combining pre-assignment and assignment is designed to ensure that the task assignments of left and right station are balanced and effective.A heuristic assimilation method that satisfies the task assignment constraints is proposed,which increases the local optimization ability of the algorithm and ensures the feasibility of the solution.The international standard case set of the two-sided assembly line is used to test and verify the validity of the algorithm from the aspects of algorithm framework,improved assimilation method and decoding method,etc.Among 34 cases,the optimal solutions of 28 cases are obtained,and3 of them obtained the best cycle time so far.The final engineering case results verified the efficient solving ability of the algorithm again.(4)As for the increased feasible solution search difficulty of task reassignment caused by the interaction of time-space constraints,the influence of the limited work space of a station on task reassignment is analyzed based on the UML models of assembly line and assembly task,and the mathematical model is constructed.An optimization algorithm for task reassignment under space-time constraints is proposed,the upper limit estimation method of station length based on task length is designed and the task assignment and reassignment mechanisms based on estimated cycle time and station length are proposed.Finally,the proposed algorithm is tested on the standard case and an engineering instance including spatial attributes and constraints,the performance of the algorithm is verified and analyzed from four aspects: solving ability,performance stability,operating efficiency and computing time.In addition to prove the effectiveness of the algorithm,the engineering case results also show that the existence of spatial constraints reduces the number of the explored Pareto solutions,and results in more rebalancing costs are spent to achieve cycle time optimization and production balance,which prove the validity and feasibility of the model. |
PDF Full Text Request