Robust Optimal Influence-Coefficient Control of Multiple-Plane Active Rotor Balancing Systems

Abstract

Rotating mass imbalance causes harmful vibration of high-speed machine tools, turbomachinery, etc. Constant speed, steady-state influence coefficient control allows active balancing systems to suppress this vibration if the influence matrix is estimated accurately. An optimal strategy for multiple-plane active balancing control is presented here that improves control robustness to modeling and estimation errors. The vibration controller objectively trades off residual vibration, control effort, and control rate of change. Penalizing control effort and rate of change is shown to enhance control stability margin, with certain performance trade-offs. Experimental results illustrate the improvement in control robustness compared with traditional weighted least squares optimal control.