Research On Intelligent Quality & Diagnosis Technologies And The Application In CIMS

Day by day, the intense market competition and the complex production process, promoted the enterprise to bring advanced technologies as information technology, computer technology, intelligent control and so on into manufacture quality control system, which laid foundation of realizing modern quality control and quality guarantee for them. Realizing enterprises' total process quality control includes two aspects: one, guarantee process is optimized, improving the product quality; two is guarantee the stable invariable process environment, obtaining a dispersion degree minimum product quality. Developing research for construction of new enterprise modern quality control model and effectively application in CIMS(Contemporary Integrated Manufacturing System) ,taking the computer technology application as the foundation, the information integration as the core, unifying the advanced intelligence process control technologies, the information processing technology and the modern quality control methods, has vital significance.The thesis briefly introduced process definition and control diagram in quality control system and clarified it's firstly. Along with the process environment unceasing transformation, has produced the massive non-normal distribution data, thus does not satisfy the premise of control charts-- normal distribution data, in statistics process control. The thesis has studied the common four kind of non-normal distributions, and calculated their influence to control charts, under different skewness and kurtosis, pointed out it is very important to strictly judge data distribute characteristics in real practice. The complex process modeling always is the research hot spot, The thesis proposed a process modeling technology based on the SOM (self organization mapping), using SOM characteristic to auto-adapted to probability distribution of the input the space, the district linear model is foundation of diagnosis technologies in later period, like PCA (Principle Component Analysis) and FA (Factor Analysis) and so on. Quick and objective identification to process variation source is Essential technology of quality control, in the thesis, designed an algorithm based on EDA (Exploratory Data Analysis), which presented a better effect in the practical application. In the process diagnosis, to identify the variation source according to the sample variance variation, but also may according to the geometry arrangement pattern of samples. The thesis proposed the process diagnosis technology which unify spectral analysis based on FFT and fuzzy inference technology, in view of the existing nerve network's insufficiency , like the typical sample gain difficulty, can not recognize new patterns and so on weakness, Finally introduced three essential technologies in CIMS, three systems overhead construction, analyzed several steps to implement quality control and the invariable scale application system solution of SOAD(Service Oriented Distributed Architecture).