| Tool wear condition monitoring constitutes an important part of intelligentmanufacturing technology, the quality of the monitoring system for enhancing theutilization of equipment, improve production quality, reduce waste, save processingcosts, reduce labor intensity and reduce the consumption of raw materials to ensuresafety in production has an important significance. Therefore, the study of thispaper have broad application prospects.In This paper introduces an intelligent tool condition monitoring system foronline research and experimental, application of multi-sensor fusion technology,collection machine tool variety of signals, including vibration signal, acousticemission signals, cutting force signal, power signal, wavelet analysis and neuralnetwork technology used in the multi-sensor fusion line tool wear monitoring.Variety of sensor signal acquisition via wavelet analysis to extract its eigenvalues,and the characteristic value as a neural network input, identified by comparing thestate of wear of the tool to determine the status of tool wear, the experimentalverification, based on wavelet analysis and neural network multi-sensor fusiontechnology can effectively identify tool wear.This paper used multi-resolution wavelet decomposition of various signals andcalculate the corresponding frequency band energy, these energies as input BPneural network, the final output state of the tool, highly nonlinear neural networks,long continuous computing, large-scale grid energy analysis and calculation data,with excellent stability, reliability, self-learning, this paper based on this technologyto create the tool condition monitoring system to ensure the system can be appliedto complex environments, enabling monitoring. BP neural network model forcomputing processing experimental data, and get better results, finally labview and matlab mixed programming based development tool condition monitoring systembasically enables CNC machine tool wear on-line monitoring. |
PDF Full Text Request