Hybrid machine learning-enabled multi-information fusion for indirect measurement of tool flank wear in milling

Abstract

To increase the utilization efficiency of multi-sensor signals and improve the measurement accuracy of tool wear, this study proposes an indirect tool wear measurement method based on multi-information fusion by hybrid machine learning technologies. In this method, triaxial cutting forces and vibration signals are collected and then preprocessed by wavelet packet denoising. The time, frequency, and time–frequency domain features are extracted to reduce the information redundancy, and the kernel principal component analysis is applied to fuse the most sensitive characteristics. A fusion model of least squares support vector regression and the particle swarm optimization algorithm is established to learn the dependency relationship between fused features and tool flank wear. Milling experiments under multiple working conditions are implemented to verify the effectiveness of the proposed method. Experimental results sufficiently demonstrated overall performance of the proposed method is better than that of other comparison methods.