Grey Prediction of CBN Grinding Process

Surface grinding is a machining process with unstable quality which is usually deteriorated as the process proceeds. If grinding can be forecast to alarm before unsatisfactory, the process could be controlled better. The purpose of this paper is to construct a grey model for CBN grinding based upon acoustic emission (AE) energy extracted from the AE grinding signal to reflect ground roughness variation. A grey model from the conducted experiment was found to be well correlated with the grinding trends. The prediction accuracy, inor out- of- sample, exceeds 90%, making grey prediction suitable for prognostic monitoring of grinding.