An investigation of the total least-squares criteria in time domain based, parameter identification for flight flutter testing

Pinkelman, J. K.; Batill, Stephen M.; Kehoe, Michael W.

doi:10.2514/6.1995-1247

Cited by 5 publications

(2 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Perangelo and Waisanen (31) examined the capability of the maximumlikelihood technique to determine the modal frequency and damping estimates from noisy flutter response signals by comparison with least-squares flutter analysis procedures. Batill et al (32) , Pinkelman et al (33,34) used parameter identification time series models for linear dynamic structural systems. Total least squares provides an approach that appears to significantly reduce the bias error in the parameter estimates.…”

Section: Literature Reviewmentioning

confidence: 99%

A comparative assessment of flutter prediction techniques

2020

View full text Add to dashboard Cite

To establish flutter onset boundaries on the flight envelope, it is required to determine the flutter onset dynamic pressure. Proper selection of a flight flutter prediction technique is vital to flutter onset speed prediction. Several methods are available in literature, starting with those based on velocity damping, envelope functions, flutter margin, discrete-time Autoregressive Moving Average (ARMA) modelling, flutterometer and the Houbolt–Rainey algorithm. Each approach has its capabilities and limitations. To choose a robust and efficient flutter prediction technique from among the velocity damping, envelope function, Houbolt–Rainey, flutter margin and auto-regressive techniques, an example problem is chosen for their evaluation. Hence, in this paper, a three-degree-of-freedom model representing the aerodynamics, stiffness and inertia of a typical wing section is used(1). The aerodynamic, stiffness and inertia properties in the example problem are kept the same when each of the above techniques is used to predict the flutter speed of this aeroelastic system. This three-degree-of-freedom model is used to generate data at speeds before initiation of flutter, during flutter and after occurrence of flutter. Using these data, the above-mentioned flutter prediction methods are evaluated and the results are presented.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

A comparative assessment of flutter prediction techniques

2020

View full text Add to dashboard Cite

show abstract

“…Compared with the frequency domain method, usually, the time domain methods are more direct. Many time domain procedures have been surveyed [58,59,60,61,62,63], and two of them have been evaluated in this study, which are the Hilbert envelope method and the frequency sweeping method.…”

Section: Time Domain System Identificationmentioning

confidence: 99%

Modeling and real-time feedback control of MEMS device

Wang¹

View full text Add to dashboard Cite

Modeling and Real-Time Feedback Control of MEMS Device Limin Wang Applying closed-loop control to a MEMS devices not only can handle the abnormal behaviors caused by manufactory imprecision or device failure, enabling MEMS devices to survive in critical conditions, but also can increase the application where MEMS devices are used to drive components under varying load conditions. This study mainly focuses on the effort of closed-loop control on the Lateral Comb Resonator (LCR) MEMS device. The success of closed-loop control has been achieved on lateral comb resonator with novel integrated through wafer optical monitoring technique [1]. Availability of a system model and feedback signals are mandatory conditions for the implementation of closed-loop control. Because of the fabrication process tolerance, the parameters of the Lateral Comb Resonator (LCR)'s model, especially the damping parameterβ, cannot be determined accurately based merely on theoretical analysis. Therefore, performing system identification through experiments can be a tool to verify the system model. Three different system identification methods in both the time domain and frequency domain have been implemented, and the results agree. Noise analysis on the optical monitoring signal shows that at least 90% of the noise in the signal is due to the optical monitoring setup, and the simulation shows that both a wavelet thresholding method and frequency domain low pass filter are efficient in removing this Gaussian distributed noise. Different designs are developed to monitor the LCR, both for single opening and grating structure LCRs, resulting in monitoring signals that are very different in nature. The optical monitoring signal for single opening device is highly correlated with the LCR's shuttle position. After removing the noise, this signal can be used directly as a feedback signal to perform closed-loop control on the shuttle for damping shock effect or to perform stroke-length control on the shuttle [2]. The optical monitoring signal for the grating structure LCR is a kind of frequency modulation of the shuttle's displacement. Based on this signal, both position and velocity signals can be reconstructed in real time [1,3]. Even though the reconstructed position and velocity signals will inherit and even amplify all the noises that exists in the optical monitoring signal, with this method acceptable performance has been achieved in the tracking control experiment of the LCR's shuttle. In this experiment, the newly designed model reference adaptive fuzzy sliding controller (MFSC) effectively minimized the side effects caused by either signal noise or imprecision of the system model. Furthermore, experimental success on force estimation has been achieved based on this signal reconstruction method [4]. The signal reconstruction method, which can decouple the noise from the optical signal, has been implemented. 2µm's resolution can be achieved with the current single beam optical monitoring method. The resolution can be improved several times with...

show abstract