A System Based On Multi Sensors For Monitoring Parameters Of Machine Tool State And Identification Of Machine Tool State

At present, many enterprises in China have machine tools of many types. Low degree of automation and slow development has made it unable to adapt to China's current needs for the manufacturing industry. For many companies, the underlying device's state information cannot completely reach the upper. This problem has a great impact on enterprise's information construction or further development of intelligent manufacturing. Therefore, this paper studies the key technologies for the system of machine tool condition monitoring and identification, realizes the development of prototype system, and successfully applies the system to the enterprise equipment workshop.Firstly, this paper expounds the significance of the research on the state parameter monitoring and identification technology from the research background and purpose. The domestic and foreign research status of the related technology is analyzed, including the research of the state parameter collection technology for the machine tool, the research of the signal processing technology, the research of the data transmission technology and the research of the state identification method for the machine tool. According to the requirement analysis, status recognition content and machine tool's monitoring scheme is presented. According to the analysis above, the overall framework of the monitoring and identification system of the machine state parameters, the software architecture of the host computer and the system architecture of the lower computer system are designed.Then, the key technologies involved in the monitoring and identification system of machine tool state parameters based on multi sensors are studied, including: 1) Research on signal processing technology. In this paper, the sampling clock generated by the timer is calculated and analyzed. The conclusion is that the SNR becomes smaller with the increase of the sampling frequency. To reduce this problem, a multi-channel sampling scheduling method based on clock jitter is designed, and the calculation formula of the time interval calculation for SNR is given; 2) Research on multi-channel data transmission technology based on USB. The multi-channel data transmission of the upper and lower computer is realized by using the characteristics of USB transmission's fast speed, easy expansion and so on. A data frame filling method is designed, which can solve the problem of multi-channel, multi-precision and multi-rate data's uploading. The algorithm flow chart of signal acquisition data upload is given based on the advantages of double buffering transmission; 3) Technology based on fusion of multi sensors for machine tool's state recognition. The performance and application of several common recognition algorithms are analyzed and compared. A method based on fusion of multi sensors and fuzzy recognition classifier for recognition of machine running state is designed. In this method, the membership function is used to process combination feature to get the membership degree matrix. Then, the method based on multi sensor parameters for identification of machine tool running state is realized.Finally, with the use of the above theory and research, a prototype system based on multi sensors is developed. In order to improve the stability of the system, this system uses ARM+Linux as the operating platform for the embedded terminal architecture. The system has been successfully applied in the enterprise, and it provides a solid foundation for the enterprise to realize the automation management.