A multi-sensor monitoring methodology for grinding wheel wear evaluation based on INFO-SVM

Abstract

It’s a significant challenge to accurate and efficient evaluation of grinding wheel wear. The evaluating grinding wheel wear traditional evaluation model has several weaknesses, including low accuracy, poor efficiency, and the need for a large database. To address these issues, an evaluating grinding wheel wear optimize model method is proposed based on weIghted meaN oF vectOrs optimized Support Vector Machine (INFO-SVM), and an data processing method is proposed based on Whale Optimization Algorithm to optimize Variational Mode Decomposition (WOA-VMD). Firstly, the grinding wheel wear was analyzed by grinding wheel and workpiece topography images. Secondly, the WOA-VMD data processing method has distinguished frequency bands between the grinding process and environmental noise signal, the method thereby eliminating environmental noise to enhance the signal-to-noise ratio in evaluating grinding process signals. Based on ReliefF algorithm established dataset, finally, the INFO-SVM algorithm method to evaluate grinding wheel wear has verified the robustness, effectiveness, and computational efficiency. The experimental results demonstrate the method's effectiveness in noise reduction, high accuracy, fast recognition speed, and strong robustness. Therefore, multi-sensor monitoring holds promising potential for application in the field of grinding wheel wear evaluation.