摘要
Single-point diamond turning (SPDT) is a widely used process in the production of high-quality, high-precision optoelectronic components. Surface roughness is a key indicator of the surface quality of workpieces produced by SPDT. During the turning process, factors such as tool-tip vibrations, tool conditions, and the machining environment may cause discrepancies between theoretically predicted and actual surface roughness values. In this study, acoustic emission (AE) sensors were used to monitor machining anomalies and predict surface roughness during the SPDT of optical-grade polymethyl methacrylate. Material removal energy theory was adopted to obtain typical AE signal diagrams under tough cutting conditions, thereby enabling real-time monitoring. To predict surface roughness, the eigenvalues of AE signals and theoretical surface roughness values were extracted and used as inputs for a support vector regression model. This model had a root mean square error of 1.76 nm and an average error rate of 14.8 %, indicating its ability to both accurately predict surface roughness and describe the on-machine metrology in SPDT. In summary, combining anomaly monitoring and surface roughness prediction during machining enables the rapid identification and evaluation of workpiece surface quality in SPDT. This integrated approach may result in enhanced line yields and a major reduction in inspection costs.