| The economic globalization and growth of diversified demands of customers motivate manufacturing enterprises to meet the customers’demands rapidly without holding large end product inventory. Without any substantial modification to existing production facilities, mixed-model assembly lines (MMALs) are capable of producing a variety of product models similar in construction and configuration concurrently, provided that MMALs are well balanced and sequenced. MMALs yield customized products and services with the manufacturing cost and response speed which are usually required under mass production. With the diversification and subdivision of market demand, automakers choose to use MMALs to sharpen their competitive edge. Balancing and sequencing involve the quantitative process aiming at improving the performance of MMALs. This paper addresses the balancing, sequencing and other key technical matters in MMALs.Most of the presented approaches are based on reducing the multiple models into a single one by combining the precedence relationships and adjusting the related operation time in the literature on mixed-model assembly line balancing. However, the adjusting operation time will lead to workstation workload in real production. A mathematical model was presented to solve mixed model assembly line balancing problem of type II (MMALBP-II) as a multiple-objective optimization problem with the objectives to minimize the cycle time’s variation, the smoothness index and minimize the processing time’s variation. These three objectives are typically inversely correlated with each other, and therefore, simultaneously optimization of three is challenging. In this paper, a non-dominated sorting particle swarm is designed. The proposed non-dominated sorting particle swarm is tested on representative instance and the results are compared with those of other algorithms.Based on the effect analysis of the launching sequence on material flow, setup time, workload and minimal production cycle of MMALs, we treated the sequencing task in MMALs as a multiple-objective optimization problem with the objectives to minimize the fluctuation of material consumption, the total setup cost and total task overlapped time. These three objectives are typically inversely correlated to each other, and the parallel optimization of the three is challenging. To achieve this, a multi-objective optimization algorithm with non-dominated sorting particle swarm (NSPSO) is designed. Through an extensive experiment studies, the performance of the proposed NSPSO was compared against best known algorithms reported in the literature. The results verified the effectiveness of the proposed algorithms. Order-oriented products assembly sequence among different assembly lines becomes a critical problem for mass customization manufacturing systems, we pose products assembly sequencing in multi-mixed model assembly lines (MMMALs) as a multiple-objective optimization problem with the objectives to minimize consumption waviness of each material in the lines, total setup cost and finished product inventory cost. The multi-objective optimization algorithm based on NSGAII is designed. Computational experiment has been demonstrated to the applicability of using NSGAII to solve the problem and effectiveness of the proposed approach. By means of this research, the valid solutions for products assembly sequence can be offered to the decision makers effectively.Next, a method for predicting the workstation overlapped task time in MMALs using autoregressive moving average (ARMA) model was proposed. The workstation overlapped task time series were sampled. According the characteristics of the autocorrelation and partial correlation coefficient of overlapped task time series, the ARMA (p, q) model was built to fit the overlapped task time series. Finally, compare the residual sum of square (RSS) using F-criteria, an ARMA (4,3) model was determined. Next-steps prediction can be performed with the determined ARMA (4,3) model. The resultant model offers a prerequisite to implement the utility work needed to avoid line stoppages.For practical applications, the existing problems and the demand in the mixed-model production plan management in a sample enterprise was analyzed. A mixed-model balancing and sequencing software package is developed. The balancing and sequencing model proposed in this thesis are embedded in the software package and has been applied to the real production. The practical and successful operations of production planning, workshop management and system management are introduced. The results validate the theories forwarded in the thesis. |
