Modeling And Optimization Of Variety-Induced Manufacturing Complexity In Mixed-Model Assembly Systems

The number of product varieties offered by manufacturers has increased drastically as a result of the paradigm change from mass production to mass customization. As variety increases, the assembly system may become quite complex, which, in turn, has significant negative impact on the system performance. This has become one of the biggest challenges for achieving mass customization. By now, the mainstream of related research has been focusing on cost or productivity oriented technologies and methodologies, such as modular technology and grouping technology. However, the effectiveness of these methods are limited by the difficulties of accurately calculating manufacturing cost and process time for different product varieties. As an important application of complexity science in the manufacturing field, manufacturing complexity has been recognized as a brand new method for handling variety-induced manufacturing problems. However, the existing assembly system complexity model is only a basic one with certain limitations. In addition, the research on methodologies of reducing manufacturing complexity is very limited.To solve above problems, the thesis tries to set up a general variety-induced complexity model in mixed-model assembly systems (MMAS), and study the approaches of reducing complexity through the processes of product family design, assembly system design, and production scheduling. The main contents and conclusions of the thesis are as follows:(1) Modeling of variety-induced manufacturing complexity in MMASThe existing complexity model only fits for independent modules assembly and the system configuration needs to be serial. Thus, the representation of complexity in parallel configurations is firstly discussed, as well as the r epresentation of complexity for dependent modules assembly. Based on above discussions, a general complexity model is then established, which is more comprehensive and accurate in reflecting the impact of product variety on system performance. In addition, an index of relative complexity is proposed, which is the division of normal complexity by maximum complexity of the system. The new index is not only more intuitive, but also fits for complexity-included multi-objective optimization problems.(2) Multi-objective optimization of manufacturing complexity and product variety in MMAS through product family design The mechanism of how product variety selection process impacts complexity is analyzed, then the trade-off between market share and complexity is discussed. Based on above discussions, a multi-objective optimization model is formulated by using the "utopia point method". The model uses relative complexity as the index and aims at selecting the best combination of product varieties to the market that achieves high market share and low manufacturing complexity at the same time through the product variety selection process. Generic algorithm is applied to solve the problem and an numerical example is provided to illustrate the application of the proposed approach. The result shows that manufacturing complexity can be reduced by selecting proper provided product variants. Finally, the impact of product competitiveness and market volatility on the optimization result is analyzed.(3) A ssembly sequence planning to minimize manufacturing complexity in MMAS with hybrid configurationsThe mechanism of how assembly sequence planning process impacts complexity is analyzed. It it pointed out that the key fator is if the dependence information between modules are effectively used. Based on above discussions, an non-liner optimization model for assembly sequence planning is formulated which aims at minimizing system complexity. Then, dynamic programing is applied to solve the problem and an numerical example is provided to illustrate the application of the proposed approach. The result shows that manufacturing complexity can be greatly reduced by carefully design the assembly sequence.(4) Migitation of manufacturing complexity in MMAS with hybrid configurations throught product variant differentiationThe mechanism of how product variant differentiation impacts complexity is analyzed. It is pointed out that the key factor is the changing of number and mix ratios of varieties in sub-stations. Based on above discussion, an non-liner optimization model for product variant differentiation is formulated which aims at minimizing system complexity with balancing efficiency as one of the constrains. Generic algorithm is applied to solve the problem and two numerical examples are provided to illustrate the application of the proposed approach. The result shows that system complexity could be greatly reduced by proper differentiation of product variants. In addition, the proposed approach is applied in the system configuration selection process to find the best configuration with lowest complexity.(5) Optimization of complexity in a high-speed-train sidewall assembly systemThe proposed complexity model and optimization approaches are applied in a high-speed-train sidewall assembly system to reduce its system complexiy. First, the variety-induced manufacturing problems of the original system are summarized. Then, through product design, assembly system design, and production scheduling processes, the sidewall assembly system is optimized using proposed optimization approaches. The complexity of the system is greatly reduced after the optimization process. The application result shows that the variety-induced negative impacts are greatly reduced and the system preference is significantly improved.In summury, the thesis tries to describe and minimize the negative impact of product variety on assembly system performance through manufacturing complexity point of view. A general mathematical model of variety-induced manufacturing complexity in MMAS is established. Then, complxity optimization approaches are studied. Fianlly, the proposed model and approaches are applied in a CRH3 high-speed-train sidewall assembly system. The research is an successful application of complexity science in manufacturing, which provides an brand new way of handing variety-induced manufacturing problems in MMAS.