Automatic chatter monitoring and modal analysis of robotic milling based on multi-joint servo current signals

In comparison to traditional machine tools, milling robots offer the advantages of a larger workspace and greater flexibility, which makes them better suited for machining complex, large-scale surfaces. However, due to their relatively lower stiffness, milling robots are more susceptible to vibrations, which can adversely affect their operational accuracy, motion stability, and structural integrity. At present, chatter monitoring methods primarily rely on external sensors, such as accelerometers, which face challenges in terms of installation, maintenance, and adaptability, thereby limiting their applicability in industrial environments. Therefore, this study presents a novel approach for chatter monitoring and modal parameter identification using motor current signals, facilitating chatter monitoring without the need for sensors. Modal parameters of joint motor current signals were identified through milling experiments under milling excitation. Furthermore, automatic chatter monitoring was implemented using the power spectral entropy difference method, combined with variational mode decomposition and adaptive filtering to automatically select decomposition layers. The effectiveness of this approach was validated through both stable and chatter milling experiments, demonstrating its potential for advancing milling robots in manufacturing.