Research On The Fault Diagnosis System Of Cnc Manufacturing Unit Based On Multi-Sensor Information Fusion

CNC machine tool is a process control equipment, consist in mechanical, electrical and hydraulic subsystems. Due to its advancement, complexity and higher intelligence, fault diagnosis for it is complicated and special. Traditional diagnosis methods are over dependent on the engineering experience of skilled technicians, including man’s subjective factors. These methods are usually for qualitative analysis, and the diagnosis work is inefficient. While for the method by using single sensor to detect and diagnosis the machine tool, the accuracy and reliability of the diagnosis decision is low, and the anti-jamming capability is also weak.According to this background, regarding typical CNC machine tools of AVIC Chengdu Aircraft Industrial (Group) CO. LTD as the monitoring object, a fault diagnosis system model based on multi-sensor information fusion is proposed in this paper. And on this basis, the fault diagnosis software for CNC manufacturing unit is developed. The software has the efficient and accurate diagnostic performance, with good usability and application value. The main work of this paper is as follows:1. The basic structures of CNC machine tool are shown. The structure, fault phenomena and fault reasons of the key mechanical components of CNC machine tool are analyzed and studied, and the corresponding diagnosis strategy is put forward. Meanwhile, the common fault diagnosis theory and methods are listed in detail and introduced systematically.2. The foundation of multi-sensor information fusion theory is reviewed. Then a fault diagnosis system model based on multi-sensor information fusion is proposed for the fault diagnosis of the key mechanical parts of Changzheng 718 machine tool, including the first-stage fusion center based on adaptive weighted algorithm, the second-stage fusion center based on the improved D-S evidence theory and the third-stage expert diagnostic system.3. Based on C# and Matlab mixed programming technology and B/S and C/S combined network architecture, a fault diagnosis software for CNC manufacturing cell is developed. The software consists of two pars, the client and Web client. It is able to pre-process and calculate the multi-sensor data to analysis and diagnosis the CNC machine tool fault, and upload the data to the server’ database, so as to LAN users to query and manage anytime and anywhere.