Model predictive force control in milling based on an ensemble Kalman filter

• DOI: 10.1007/s10845-022-01931-2
• Publication Date: 2022-04-02T17:02:16Z
• Authors: Max Schwenzer; Sebastian Stemmler; Muzaffer Ay; Adrian Karl Rüppel; Thomas Bergs; Dirk Abel
• Keywords: Model Predictive Control; Machine tool; Online model; Identification

Abstract Process force determines productivity, quality, and safety in milling. Current approaches of process design often focus on a priori optimization. In order to enable online optimization, the establishment active controllers is required. Due fast-changing engagement conditions tool conjunction with slower machine dynamics, classic control not suited. A promising approach application model predictive (MPC) for control, which proposed this contribution. The controller (MPFC) explicitly takes into account predict immediate future. It consists separate process. describes relation between feed velocity tool, force, geometric properties such as radial deviation, tool/workpiece engagement. feedback loop closed by an identification changes material or wear state. For ensemble Kalman filter (EnKF) applied. MPFC solves optimization problem future behavior each sampling step determine optimal output enabling high dynamic control. system validated experimentally compared conventionally designed constant feed. can be shown that manufacturing time reduced 50%. enables paradigm shift milling processes operating at its technological limit.