Tool-Workpiece Contact Detection Using Adaptive Kalman Filtering and Spectral Flux

Detection of tool-workpiece contact is crucial for ensuring the machining accuracy and health management of CNC machines. However, traditional detection methods require complex geometric modeling or multiple sensors, leading to high costs and poor real-time performance. To address these issues, a tool-workpiece contact detection approach utilizing adaptive Kalman filtering and spectral flux is proposed. The adaptive Kalman filter is employed to denoise the raw signals, thereby improving signal quality. Subsequently, the most impactful frequency band is extracted from the denoised signal, and its spectral flux is calculated to achieve real-time detection of the tool-workpiece contact state. Furthermore, actual engineering data is analyzed to evaluate the performance of the proposed method. The results show that this method offers significant advantages in terms of noise resistance, detection accuracy and real-time performance.