The main objective of the work reported here was to develop an intelligent condition monitoring system that was able to detect when a cutting tool was worn out. To accomplish this objective the use of a hybrid intelligent system, based on an expert system and two neural networks, was investigated. The neural networks were employed to process data from sensors and the classifications made by the neural networks were combined with information from the knowledge base to make an estimate of the wear state of the tool. The novelty of this work is mainly associated with the configuration of the developed system that estimates tool wear in a new way. The combination of sensor-based information and inference rules results in an online system that can be updated when the cutting conditions fall outside of the trained zone of the neural networks. The neural networks resolved the problem of interpreting the complex sensor inputs while the expert system, by keeping track of previous success, estimated which of the two neural networks was more reliable. Misclassifications were filtered out through the use of a rough but approximate estimator Taylor's tool life model. The use of Taylor's tool life model, although weak as a tool life estimator, proved to be crucial in achieving higher performance levels. The application of the self-organizing map to tool wear monitoring proved to have a slightly larger zone of influence and make slightly more accurate estimates of tool wear than the adaptive resonance theory neural network and overall the system made reliable, accurate estimates of the tool wear. © IMechE 2006.
|Number of pages||14|
|Journal||Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture|
|Publication status||Published - 2006|
- Expert system
- Neural network
- Tool wear