Methods Of Multi-criteria Decision MakingOriented Towards Mechanical Manufacturing Process

The mechanical manufacturing process is the whole process of transforming raw materials into the final products which can be directly used for customers, and of putting them into markets, in which process involve many decision-making problems. In fact, decision-making is one of the basic activities existing in any aspects of life and production. This paper, aiming at material selection, process parameter optimization, and agile supply chain configuration in the product manufacturing process, implemented researches on the methods of multi-criteria decision making. It has important theoretical and practical significances for improving the scientificalness, accuracy, and reliability of decision making in the product manufacturing process, and further for improving the economic benefits of any manufacturing enterprises and winning in the fierce marketplace.First of all aiming at green manufacturing, implemented decision-making on green material selection. Researched into the relationships of six types of criteria (beneficial type, cost type, fixation type, deviation type, interval type, and deviated interval type), based on which a method was proposed to normalize raw ratings with reference to the equation of calculating the distance of any two interval numbers. Employed to identify the criterion weights was the ANP (Analytic Network Process) for which we presented a method of directly constructing a consistence comparison matrix by interactive method with no need to carry out consistency check. In view of the decision-making compensation effect in multi-attribute utility theory, the PROMETHEE (the Preference Ranking Organization Method for Enrichment Evaluations) in the presence of dependences, one of the outranking relation method, was used to aggregate the attribute ratings to obtain the overall ratings of each alternative. Taking the material selection for a journal bearing as example, the decision-making procedure is enunciated.Secondly taking the fused deposition modeling (FDM) as an application engineering background, carried out the theory and practice researches into process parameter optimization. With experiments the effects of process parameters on the fabrication process and the part precision have been analyzed, and then selected the four parameters having very significant impact on the part precision, like line width compensation, extrusion velocity, filling velocity, and layer thickness as control factors, input variables, and dimensional error, warp deformation, and built time were selected as output responses, evaluation indexes. The magnitude relation between the input variables and the output responses were obtained through uniform experiment design with the experiment design table of U (171617). The three output responses were converted with fuzzy inference system (FIS) to a single comprehensive response (CR). The relation between the comprehensive response and the four input variables was derived with second-order response surface methodology (RSM), the correctness of which is further validated with artificial neural network (ANN). Fitness function was created using penalty function, solved with genetic algorithm (GA) toolbox in Matlab software, and consequently the optimized process parameters were obtained. With confirmation test, the results are proved correct.Thirdly taking as an application engineering background a three-level agile supply chain including supplies, manufacturers, distributors, and customers, carried out the researches into group decision making on primary and final partner selection, and the optimal task allocation. Due to the decision maker’s knowledge field, attitudes, motivations and personality and the nature of evaluated attributes, the decision makers may provide the assessments with different formats such as real numbers, intervals, trapezoidal fuzzy numbers, and linguistic variables possibly with different cardinalities. Such a type of MADM (Multi-Attribute Decision Making) problems is called the fuzzy heterogeneous MADM problems with which seldom literature deals. Analyzed the four kinds of computing modeling widely used in dealing with linguistic ratings and the methods of unifying them with different cardinalities into the linguistic terms in the standard linguistic set. Since ANP can only solve this type of MADM with eliminable interdependences, but cannot solve it with no eliminable ones. Fuzzy measures and fuzzy integrals are effective means to address the problems of interdependences. The basic concept of fuzzy measures and fuzzy integrals, and the conversions among fuzzy measures, M bius representations, and interaction indexes were elaborated. Analyzed all sorts of linguistic aggregation operators based on Choquet integrals, and, based on this, A linguistic hybrid weighted geometric averaging with interaction (LHWGAI) operator was proposed and used for partner primary selection. As for the partner final selection and the optimal task allocation, an interactive two-level fuzzy programming was proposed, the solution method of which used the LINGO13software. Since the method gives consideration to both upper level and lower level interests, an optimal compromise solution can be obtained through repeated negotiation between the decision makers. The decision making procedure was illustrated with an example, and showed the advantages.Finally, summarized the main contents researched and innovations proposed, and something to be desired as well as the further research directions were pointed out.