Adaptive vibration suppression perfect tracking control for linear time-varying systems with application to ball-screw feed drives

Abstract

In high-precision positioning of feed drive systems, mechanical resonances are one of the main limiting factor for fast and precise motion. Vibration suppression tracking controllers are employed to improve positioning performance, although precise plant models are generally required. Hence, in systems with time-varying modes, the increasing modeling errors often result in the loss of the needed performance. This paper presents an adaptive tracking controller for vibration suppression of linear systems with time-varying dynamics. The technique is based on multirate feedforward, alleviating the unstable-zeros problem of discrete time plant models, for which the nominal plant model is adjusted using recursive estimation. A novel estimation scheme is presented to estimate the modal parameters of the plant based on the relative motion between the mechanical components. The proposed control scheme is applied to ball-screw feed drives with varying dynamics during the cutting process and the advantages demonstrated through simulations and experiments.