Abstract:

In order to improve the production quality and increase mean-time-between-failure (MTBF) of equipment, the remote diagnosis system has got wide attention in recent years. By monitoring process variables on-line, the degradation of equipment performance can be predicted dynamically. Therefore, engineers can insert maintenance arrangements into schedule before the failures occur.

 

Firstly, a remote diagnosis architecture is developed. The previous researches in our laboratory are further integrated into the architecture. A remote diagnosis server is used to diagnose the equipment in the remote side and a diagnosis knowledge database is constructed. If a failure is detected, the generic message passing platform (GMPP) module will inform the related operators.

 

Secondly, a diagnosis method of CNC punch press machine is proposed. The statistical process control module with Nelson violation rules is used to monitor the variance of historical data. In the mean time, the correlation between the violated rules and equipment status is recorded as well. A neural network is used to construct the classification model from the collected data.

Thirdly, a flexible manufacturing system (FMS) model is built based on distributed colored timed Petri net (DCTPN). The proposed remote monitoring and diagnosis system are integrated into the FMS system. Furthermore, this system also integrates with a cluster tool controller.