Research On Quality Control Technology Of Key And Impotant Parts

As an important content in quality control, Quality control technology could ensure that the production process run successfully and the product quality is good. This paper researched the pattern recognition system of quality control charts of key and important parts, this system consist of data collection and quality control. Firstly, the process data of key and important parts are transferred to the computer, after dealt by the computer, the process data quality control charts are established, the pattern of quality control charts is identified, at last recognition result is output.Firstly, this paper studied the process data collection of key and important parts and use PCI bus to realize data collection, because PCI bus has the advantages of independent of processor and high transmission speed. This paper studied the structure and characteristic of PCI bus, designed the hardware of data collection, in order to shorten development time, adopted PCI9054 which is the special interface chip of PCI bus. Data collection card consist of three parts as following: interface module of singlechip and tri COMERO, configuration module of EEPROM, interface module of singlechip and PCI bus. In the windows enviroment, it is necessary to provide device driver in order to operate the hardware device, so the paper designed the device driver of data collection card, designed the communication program of the singlechip, and used Verilog HDL language to design the logic control program of communication of PCI bus and singlechip.Then, this paper researched quality control of process data of key and important parts, studied the application of neural network in pattern recognition. In the Matlab entironment, this paper designed neural network quality control chart pattern recognition module and parameter estimation module. Using actual process data to test the module, this paper got the result of that the accuracy of the pattern recognition module is 95.3% and the parameters estimation precision of parameter estimation module is 5%. The paper analyzed and studied the process capability, calculated the process capability index. The paper researched the eight judge abnormal rules of national standards, designed corresponding judge abnormal program. This designed program could accurately judge whether or not the quality control chart meet the eight judge abnormal rules of national standards. At last, the paper designed friendly GUI interface. The sample data and corresponding parameters could be input by users, after dealt by the computer, corresponding results is output by this interface.