| In the increasingly competitive 21 st century,the quality has become an important factor when measuring the development of a company,the level of quality determines the economic strength of an enterprise.In the development process of many enterprises,especially in manufacturing business,they gradually attach much importance to the production quality management and control.Taking the automobile parts manufacturing industry as an example,more and more enterprises use the Statistical Process Control(SPC)and Process Capability Analysis(CPK)to carry out quality control in the production of auto parts,thereby improving the quality of their products.However,in the modern auto parts manufacturing industry,due to the requirements of some high-precision parts manufacturing,the process parameters do not obey the normal distribution,such as the size of the injection nozzle pinhole and leakage rate in the process of the spark plug.For these processes,the process parameters can not meet the normal distribution,if some enterprises still take the traditional SPC technology and CPK index for quality management,they can not achieve the desired quality management effectiveness and can not accurately find the problem in the production process.Therefore,we need to study the new algorithm under the non-normal distribution,using the appropriate and accurate SPC control charts and CPK process capability index.In this paper,it will use the traditional control chart of normal statistical process(including the selecting control chart)and process capability analysis to study the basic theory as the theoretical basis,firstly introducing when the characteristic value of process data obeys normal distribution,what is the basic concept,the criterion and control limits of the traditional control chart and selecting control chart,as well as what is the basic concept,the condition of evaluation index and the criterion of process capability.Based on the related Chinese and foreign literature,it will select auto parts as the research object,and make the data of auto parts production obeying non-normal distribution to be the background.Then,aiming at the automobile parts manufacturing industry,we use Minitab to simulate data to propose a false result which is based on the traditional SPC control chart analyzing the quality control.Therefore,it is important to analyz the process of Box-Cox transformation function and Box-Cox transformation in detail.Based on Box-Cox transformation function,anew transformation method of anti-normal data is put forward.The two data transformation methods will be compared and analyzed,and the Box-Cox conversion method is selected as the main data conversion method in this paper.According to the transformed data,we use Minitab to draw relevant conventional control charts and selecte control charts to evaluate the process capability and complete the quality monitor and diagnosis process of the whole process.Finally,the paper will do empirical research with M automobile parts manufacturing company,collecting relevant data of M company's injector shaft pin aperture,testing normality of data,and illustrating the quality monitoring and diagnosis process while data characteristic is under non-normal distribution by examples.On the basis of 5M1 E quality management factors,combining with fishbone diagram,it analyzes the problems of quality management in automotive parts manufacturing process,and puts forward proposals for improving quality in the field of automotive parts.This paper analyzes the manufacturing process of automobile parts from the aspects of theory and application to analysis quality monitoring and diagnosis,and puts forward the suitable monitoring and diagnosis methods of process data under non-normal distribution.The quality management research will greatly improve the efficiency of auto parts manufacturing process management,and improve product quality to reduce the loss of quality.it is of great significance to strengthen the market competitiveness of enterprises. |
PDF Full Text Request