Sparse Identification of Nonlinear Delayed Dynamics

A data-driven approach to identify governing equations in turning processes by incorporating time-delay effects essential for capturing regenerative cutting phenomena.

• Standard system identification techniques fail with turning processes due to time-delay
• Custom libraries incorporating delay terms and modulated cutting coefficients
• Systematic delay and frequency identification through parameter sweeping
• Two-model approach separating acceleration and force dynamics
• High prediction accuracy for cutting force dynamics