Control of a nonlinear flexure-jointed X-Y positioning stage using LTV-FIR command prefiltering for finite-time error cancellation. (April 2021)

• Record Type: Journal Article
• Authors: Kuresangsai, Pongsiri; Cole, Matthew O.T.
• Is Part Of: Mechanical systems and signal processing. Volume 151(2021)
• Publication Date: 2021-04
• Subjects: Motion control -- Vibration reduction -- Feedforward control -- Linear time-varying system -- Compliant mechanism; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2020.107349
• Languages: English
• ISSNs: 0888-3270

Abstract

Highlights: New FIR command prefilter for linear-time-varying systems introduced. Offers significant performance improvements compared with robust LTI prefiltering. Can be applied with LPV model construction for nonlinear mechanical systems. Applied to 2 d.o.f. compliant mechanism with voice coil actuators. Can reduce settling time by factor > 10 for arbitrary motions. Abstract: This paper presents a new command input prefilter that achieves exact residual vibration cancellation for a single degree-of-freedom oscillator with linear time-varying (LTV) dynamics. The prefilter has a time-varying finite impulse response (FIR) function that is constructed using a time-warping technique and can be convolved with any command input signal to achieve finite settling time. Numerical results confirm that a discrete-time implementation of the prefilter can suppress vibration more effectively than robust linear time-invariant prefilter designs. The LTV-FIR prefilter was applied to a flexure-based X-Y micro-positioning platform having nonlinear dynamics and cross-coupling between motion axes. A configuration-dependent LTV model was constructed by using polynomial interpolation of a set of experimentally identified LTI models. Prefiltering was applied separately to X and Y axis command variables and a cross-coupling compensation technique used to minimize dynamic interaction between the motion axes. The LTV-FIR prefilter is shown to be effective for settling-time reduction in point-to-point motions (reduced from > 6 to ≈ 0.4 seconds). Moreover, improved motion performance could be achieved over a large operating space without the use of high-gain feedback control, which is inappropriate in this case due to noise excitation and the risk of instability.