Development of machine tool condition monitoring system using pattern recognition

The design and development of a condition monitoring and diagnosis system for a CNC milling machine is discussed in this paper. Various techniques are used for the condition monitoring of the machine which include the DC servo feed drive, spindle system, gearbox and machine table, as well as the cutting process. Several signal classification methods are implemented for automatic diagnosis, which include simple comparators, Baye's processor, and a linear machine.