The Car-motor Mixed Model Assembly Line Sequencing Research Based On Intelligent Optimization Algorithms

Along with the social economy development and the progress of science andtechnology level, the car industry rapid development and the national economicdevelopment plays an increasingly heavy role. As the heart of cars, no matter from thelevel of technology, cost, personnel, engines are required to have high level and position,the engine manufacturing in the whole process of automobile production has the vital role.With the improvement of people’s living standard, the personalized demand for cars isbecoming more and more high, the current domestic automobile engine manufacturingmostly adopts hybrid varieties of assembly line production mode. To organized theproduction of multi-species production, the key is the stabilization and heijunka assemblyline is the core of products that optimize the production sequence. Optimize theproduction sequence to ensure a balanced production, shorten the delivery time, reduceinventory and improve the competitiveness of enterprises, to better adapt to current marketdemand. Therefore, mixed-model through the production line to sort the problem, could bebetter way to play a mixed-flow assembly advantages.In this paper, based on the enterprise production operation and purpose, combinedwith the actual engine production line, a multi-objective optimization function on mixedmodel assembly line is proposed. This function is minimizing idle workstations andauxiliary working time, parts consumption rate and minimize the homogenization productsswitch adjustment costs. And establish the corresponding mathematical model.A long time ago, assembly sequencing problem has been proven to be NP-hard. Inthis paper, an improved NSGA-Ⅱ algorithm is proposed to solve the established model ofmulti-objective optimization. On the one hand the improved NSGA-Ⅱ algorithmpreserves the fast non-dominated sorting method, the crowding comparison mechanismand elitist strategy of NSGA-Ⅱ algorithm. On the other hand,the improved NSGA-Ⅱalgorithm improves the generation type of first sub-populations and implements sortingrank according to the demand strategy. Finally in this paper, two groups of problems areused for simulation. The results show that the solutions of the improved NSGA-Ⅱisbetter in comparison of set coverage, spacing and maximum spread than other algorithms.