A Study On Balancing And Sequencing Problems Of Mixed-model Assembly Lines And System Development

Mixed-model assembly line is complex relative to single variety assembly line and could improve resource utilization and production efficiency, having been accepted by manufacturing industry. In this paper, the research is carried out from two aspects as balancing and sequencing problems of the mixed-model assembly line. Improved intelligent optimization algorithm is designed combining with the multi-objective function and the constraint condition, and the balancing and sequencing of the mixed-model assembly line system software is developed by directly outputting optimization scheme and improvement project at last.Firstly, the multi-objective optimization models on balancing and sequencing are built by theoretical analysis. The main idea of solving the balancing problem is that mixed-model assembly line is transformed into single variety assembly line on the basis of interference model and precedence matrix, establishing an integrated objective by considering production cycle and station workload. The main idea of solving the sequencing problem is that actual demand of these varieties is simplified into the demand in the time of MPS, and then the step above is repeated until demand is reached. The multi-objective function is established by normalizing minimal production set and inputting the balancing result of station workload.Secondly, genetic algorithm flow is deeply improved to avoid precocious phenomena existing in solving the balancing and sequencing problem. In the process of solving balancing problem, product initial population by the means of random topological sort to meet the constraint conditions. In the early term of evolution, choose cross generational elitist selection as selection method to ensure the diversity of samples. In the later term of evolution, mixed sampling selection method is used to close to optimal solution. Further, use immigration operator to import new individuals in each generation of evolution. The searching ability of this algorithm becomes stronger because of inserting local search operator into algorithm flow of sequencing problem. Mutation operator chooses the algorithm based on neighborhood search to get more excellent progenies. The optimization results indicate that the improved algorithm is more effective than the old algorithm through the actual case analysis in the last part.Finally, this paper presents an automobile mixed-model assembly line in case to test the system's reliability. The result shows that the system can be used to control the essential information related to the assembly line, calculate the balancing and sequencing optimization, and output the visualization results (histogram, work station distribution diagram, optimization algorithm iteration curve) and data table.