Fault Prognosis with Stochastic Modelling on Critical Points of Discrete Processes

The primary role of a machine tool is produce the good quality parts, but a machine tool goes always through a process of degradation and wear which will affect the accuracy and precision of machining and the quality of products. Therefore, monitoring the degradation of machine tool and quantifying its health is very important. The degradation level of a machine can be qualified by an index which is called health indicator (HI). Based on the HI, fault prognosis can provide the Remaining Useful Life (RUL) of machine which is useful for an effective maintenance policy, thus, that helps to increase efficiency of operations and manufacturing. However, the HI is not usually predetermined in most Discrete Manufacturing Processes (DMP). This paper presents a new method of HI extraction based on the degradation reconstruction. The HI is then modeled with a stochastic process. For the online supervision, the RUL is estimated for each inspection time.