Research On Performance Analysis And Evaluation Method Of Manufacturing System

Although our country has become a big manufacturing country,but she is facing "big but not strong" situation.In view of the existing problems,in 2013,the Chinese Academy of engineering carried out a major consulting project named "Powerful Manufacturing Country strategy research".After a six-month study,more than 150 academicians and experts proposed four characteristics of the "Powerful Manufacturing Country " and comprehensive index evaluation system of manufacturing industry.This achievement points the milestones out for each stage of the development of China's manufacturing industry,and becomes the theoretical basis of the strategy named "China's manufacturing 2025" in 2015.However,how to measure the competitiveness of the basic unit of the manufacturing industry——manufacturing system is still an urgent problem to be solved at present.As the positions of the existing analysis and evaluation methods is not clear,the system is not perfect,the standard setting is not reasonable,in this paper,the performance analysis and evaluation method which is effective and feasible of the manufacturing system is studied,in order to realize scientific and reasonable analysis and evaluation,make the manufacturing enterprises recognize the achievements and shortcomings in the development process,and improve the operation level of manufacturing system.(1)By analyzing the actual operation process of the manufacturing system and the existing problems of the performance analysis system,the performance analysis index system of manufacturing system is constructed from internal process and external performance.Meanwhile,using the analytic hierarchy process(AHP)to analyze the internal and external capability of manufacturing system and calculate the contribution matrix,determine the relative importance of the internal capacity to the external capability.(2)Learning from the concept of relative performance evaluation,the establishment method of dynamic evaluation criteria for the performance of manufacturing system is proposed.As the operation data of the reference manufacturing system collection is difficult,the automatic data acquisition system of the reference manufacturing system based on Python is compiled.Taking machinery equipment manufacturing industry as the research object,8050 operation data of 175 typical domestic manufacturing enterprises are collected.Based on principal component analysis,the data were analyzed,and the dynamic evaluation criteria of the performance of the manufacturing system was established,as a basis for the evaluation of the performance of the manufacturing systems.(3)Based on the frontier theory of machine learning,including adaptive boosting method(Adaboost)and random forest algorithm(Randomforest),the performance evaluation model is built by using dynamic evaluation criteria for the performance of the manufacturing system of machine equipment manufacturing industry as data base.Using the about 25% of the sample data set as the forecast set,the performance of the model is studied,and the validity and feasibility of the method is verified.