Tool Wear State Identification Based on SVM Optimized by the Improved Northern Goshawk Optimization

Area of Science:

• Manufacturing Engineering
• Mechanical Engineering
• Signal Processing

Background:

• Tool wear significantly impacts equipment downtime and machining precision.
• Accurate tool wear state identification is crucial for optimizing manufacturing processes.
• Existing methods may lack the required accuracy and efficiency.

Purpose of the Study:

• To develop a novel and highly accurate tool wear state identification technique.
• To improve the efficiency and stability of tool wear monitoring systems.
• To enhance machining precision and reduce operational costs.

Main Methods:

• Wavelet packet thresholding denoising for multi-source signal processing and feature extraction.
• Support Vector Machine Recursive Feature Elimination (SVM-RFE) for selecting relevant features.
• An improved northern goshawk optimization (INGO) algorithm to optimize Support Vector Machine (SVM) parameters, forming the INGO-SVM model.

Main Results:

• The INGO algorithm demonstrated superior convergence efficacy and stability in simulation tests.
• The proposed INGO-SVM model achieved a remarkable recognition accuracy rate of 97.9% in milling wear experiments.
• This approach outperformed five other comparative methods in terms of recognition accuracy.

Conclusions:

• The INGO-SVM model offers a highly accurate and reliable solution for tool wear state identification.
• This technique can significantly contribute to reducing equipment downtime and improving machining precision.
• The study validates the effectiveness of combining advanced signal processing with optimized machine learning for industrial monitoring.