| The development and production capacity of large civil aircraft (CA) is an important manifestation of a country’s comprehensive strength, and it plays a major role in driving industrial upgrading, promoting scientific and technological progress,and forming a national economic growth point. CA assembly is one of the critical production processes and an important manifestation of the core competitiveness of the manufacturers. Therefore, it is necessary for China to construct CA assembly lines.As the structural complexity and performance requiements of modern CA increase,and the assembly technology upgrade, the cost of the assembly line is also increasing.It is imperative to study how to control the cost. Especially in the light that construction investment decisions in the early stage not only directly determines the initial investment in the project construction, but also largely determines the cost of latter operation and maintenance, even demolition. Generally, the pre-project study of feasiblity and optimization is much cost-effective than latter revisions, it is important that China speed up the cost management research systematically and thouroughly at the investment decision-making stage of CA assembly line project.After reviewing related domestic and international research, from the perspectives of life cycle cost (LCC) management and significant cost management theories, with the utilization of various basic methods such as Particle Swarm Optimization, RBF Neural Network and Petri Net, and expert interviews and questionnaires etc., the related issues of cost management of the CA assembly line investment are studied. Firstly, based on the theory of LCC management, the composition of LCC of CA assembly line project is analyzed and an LCC model is constructed. On this basis, the content of cost management of CA assembly line project in decision-making stage is proposed and the key factors of cost management,which is extracted through questionnaires, include scientific program selection and accurate investment estimation. Secondly, the scheme comparison and selection method of CA assembly line project are put forward from three aspects: production mode, process arrangement and process optimization. Thirdly, the significant cost theory is used to extract the significant items and their key influencing factors, and the investment estimation model is constructed to make the investment estimation more accurate.The innovation of the paper is mainly reflected in the following aspects:(1) The LCC model of the CA assembly line project is built and the key factors of cost management in the decision-making stage are extracted. Based on the content and composition of CA assembly line, the composition of its LCC is identified, and the LCC model is constructed by discount method. Next, from the perspective of the life cycle, the content of the cost management in the investment decision stage of the CA assembly line project is put forward and the key factors, which are determined through questionnaire analysis, include construction scheme comparison and investment estimation.(2) A method based on PSO for balancing assembly line and configuring assembly station are put forward. Targeting at optimal LCC, Given the number of jobs,the time of assembly and the priority, a method based on PSO for balancing assembly line and configuring assembly station are put forward, which provides a scientific method for the process layout plan of the assembly line of civil aircraft.(3) A Petri net model is constructed by the combination of Top-Down and Hierarchical Decomposition to optimize the production process. the combined Petri net method of Top - Down and hierarchical decomposition is built and the method of state equation is used to judge whether the Petri net model meet the accessibility,boundedness and activity of the basic system requirements, and through the improvement and optimization of the simulation process.(4) The investment estimation model of civil aircraft assembly line based on RBF neural network is constructed, and an example is given. The significant project items and the key influencing factors are extracted by using the significant cost theory.The investment estimation model of CA assembly line based on RBF neural network is constructed. In order to verify the applicability of the model, 26 samples were are selected for analysis, the results show that the relative errors of the test samples are less than 10%, which means that the prediction accuracy is rather high. |