Parameter estimation and statistical analysis on frequency-dependent active control forces

Over the years, methods have been developed to estimate the dynamic coefficients of bearings based on parameter identification techniques. The stage where the estimated coefficients obtained from these methods need to be validated on actual operating machines, especially those operating at hypercritical speeds, has been reached. This article extended the frequency domain parameter estimation technique normally used on bearings to extract the dynamic coefficients of the bearing pedestals in isolation. It is demonstrated here that by using only a single forcing mechanism located at an angle to the bearing pedestal, the eight possible stiffness and damping coefficients can be estimated accurately. An on-line frequency domain algorithm for estimating these coefficients was developed and tested in the laboratory for this purpose. Its ability to reduce testing time and to produce reliable coefficient estimates from structural spurious resonance is observed. Validations of the dynamic coefficients of the bearing pedestals were carried out by comparing the experimental and computer simulated responses and resonance frequencies of the arrangement of two multi-mode rotor-bearing-system's configurations. By using ANSYS Inc. finite element software as a comparison tool, the results indicated that a nine stations rotor model formulated by the stiffness coefficient method is highly accurate at predicting the rotor-bearing system's responses and resonance frequencies. The results also confirmed that rotor bearing's synchronous response is insensitive with respect to individual estimated parameters that have been determined by statistical hypothesis test to be insignificant.

show abstract

“…The advantages of SPHS over other multi-frequency test signals (i.e. step, impulse, white noise, and PRBS) can be summarized as follows [25].…”

Section: Parameter Estimation Procedures and Resultsmentioning

confidence: 99%

Validating the dynamic coefficients of bearing pedestals in a multi-mode rotor—bearing system

Lim

Chai

2009

Proceedings of the Institution of Mechanical Engineers, Part C:

View full text Add to dashboard Cite

show abstract

“…These five factors are significant contributions toward achieving the estimated dynamic support's coefficients of the system with minimum variance. A multifrequency signal, such as the SPHS (26), which was successfully employed by Lim and Cheng (20,24) to persistently excite a simple second‐order model test rig, will be adopted as the excitation force signal in the following parameter estimation procedures.…”

Section: System's Dynamic Model and Parameter Estimation Proceduresmentioning

confidence: 99%

“…By expressing in the form of frequency domain, the complex transfer function which is defined as the displacement/force, , can be separated into real and imaginary parts. And rearranging the coefficients to be estimated on the left and the known parameters on the right (24), we can obtain the least square estimator for the coefficient matrix, given by Gujarati (29) as:…”

Section: System's Dynamic Model and Parameter Estimation Proceduresmentioning

confidence: 99%

“…However, the lower G.o.F of 0.65 to 0.95 in the 0–40 Hz range is attributed to the integral feedback constant used in Table 1 of the HMBs controller. This is because integral feedback is well known to eliminate low frequency position offset error, thus acting as a low‐pass filter in the control loop (24).…”

Section: Radial Experimental Parameter Estimation Of Hmbsmentioning

confidence: 99%

“…Hence, the theoretical frequency‐dependent controller's transfer function has been used in simulation or experimentally to obtain the stiffness and damping coefficients of magnetic bearings (21–23). However, this method excludes the system lags due to the digital signal processing (DSP), amplifiers, feedback sensor, actuator's cross‐axis capabilities, and electromagnetic losses (24,25).…”

mentioning

confidence: 99%

See 2 more Smart Citations

Parameter Estimation and Actuator Characteristics of Hybrid Magnetic Bearings for Axial Flow Blood Pump Applications

Lim¹,

Cheng²,

Chua³

2009

Artificial Organs

Self Cite

View full text Add to dashboard Cite

Axial flow blood pumps are generally smaller as compared to centrifugal pumps. This is very beneficial because they can provide better anatomical fit in the chest cavity, as well as lower the risk of infection. This article discusses the design, levitated responses, and parameter estimation of the dynamic characteristics of a compact hybrid magnetic bearing (HMB) system for axial flow blood pump applications. The rotor/impeller of the pump is driven by a three-phase permanent magnet brushless and sensorless motor. It is levitated by two HMBs at both ends in five degree of freedom with proportional-integral-derivative controllers, among which four radial directions are actively controlled and one axial direction is passively controlled. The frequency domain parameter estimation technique with statistical analysis is adopted to validate the stiffness and damping coefficients of the HMB system. A specially designed test rig facilitated the estimation of the bearing's coefficients in air-in both the radial and axial directions. Experimental estimation showed that the dynamic characteristics of the HMB system are dominated by the frequency-dependent stiffness coefficients. By injecting a multifrequency excitation force signal onto the rotor through the HMBs, it is noticed in the experimental results the maximum displacement linear operating range is 20% of the static eccentricity with respect to the rotor and stator gap clearance. The actuator gain was also successfully calibrated and may potentially extend the parameter estimation technique developed in the study of identification and monitoring of the pump's dynamic properties under normal operating conditions with fluid.

show abstract

Development of Hybrid Magnetic Bearings System for Axial-Flow Blood Pump

Meng

Cheng

2008

Lecture Notes in Electrical Engineering

View full text Add to dashboard Cite

Parameter estimation and statistical analysis on frequency-dependent active control forces

Cited by 17 publications

References 7 publications

Validating the dynamic coefficients of bearing pedestals in a multi-mode rotor—bearing system

Validating the dynamic coefficients of bearing pedestals in a multi-mode rotor—bearing system

Parameter Estimation and Actuator Characteristics of Hybrid Magnetic Bearings for Axial Flow Blood Pump Applications

Development of Hybrid Magnetic Bearings System for Axial-Flow Blood Pump

Contact Info

Product

Resources

About