Abstract
Purpose
A precise identification and control method for the 6D micro-vibration exciting system is proposed in this paper. The controlled object is a Stewart configuration platform, which can provide six degree-of-freedom (6DOF) micro-vibration excitation excited by six actuating legs.
Methods
In view of the amplitude nonlinearity in this kind of exciting system, a nonlinearly iterative identification method is introduced to establish an inverse model to decouple the system in different levels of magnitude. For the decoupled system, a multi-input multiple-output (MIMO) adaptive disturbance cancellation (ADC) control method is proposed for the time domain control of the exciting system. Furthermore, the feedforward control loop is added to MIMO ADC control method to improve the rapidity and stability of the control system.
Results
Experiments show that the nonlinearly iterative identification method can greatly improve the decoupling accuracy of the system, and then improve the accuracy of sinusoidal vibration control.
Conclusion
The precise identification and control method can achieve sinusoidal vibration control of high precision in the control frequency band up to 300 Hz.
Similar content being viewed by others
References
Steier F, Runte T, Monsky A et al (2019) Managing the microvibration impact on satellite performances. Acta Astronaut 162:461–468. https://doi.org/10.1016/j.actaastro.2019.06.027
Dyne SJC, Tunbridge DEL, Collins PP (1993) The vibration environment on a satellite in orbit. IEE Colloquium on High Accuracy Platform Control in Space 12:1–6
Sun X, Yang B, Hu W, Bai Z (2020) Simultaneous precision positioning and vibration control for on-orbit optical payloads: an integrated actuator development and analysis. J Vib Eng Technol. https://doi.org/10.1007/s42417-020-00244-z
Araki K (2002) In-orbit measurements of short term attitude and vibrational environment on the Engineering Test Satellite VI using laser communication equipment. Opt Eng 40:827. https://doi.org/10.1117/1.1355976
Whiteman WE, Berman MS (2002) Fatigue failure results for multi-axial versus uniaxial stress screen vibration testing. Shock Vib 9:319–328. https://doi.org/10.1155/2002/109715
Peres MA, Kallmeyer C, Shop TM (2015) Advantages of Multiple-Input Multiple-Output Testing. 8–12
Dasgupta B, Mruthyunjaya TS (2004) The Stewart platform manipulator: a review. Mech Mach Theory 35:15–40. https://doi.org/10.1016/S0094-114X(99)00006-3
Preumont A, Horodinca M, Romanescu I et al (2007) A six-axis single-stage active vibration isolator based on Stewart platform. J Sound Vib 300:644–661. https://doi.org/10.1016/j.jsv.2006.07.050
Kong Y, Huang H (2018) Vibration isolation and dual-stage actuation pointing system for space precision payloads. Acta Astronaut 143:183–192. https://doi.org/10.1016/j.actaastro.2017.11.038
Peterson R, Novokov M, Hsu J et al (2018) 6 DOF high-frequency motion simulator phase II. https://doi.org/10.1117/12.474707
Idan M, Sahar D (1996) A robust controller for a dynamic six degree of freedom flight simulator. Flight Simul Technol Conf 1996:53–60. https://doi.org/10.2514/6.1996-3476
Stewart D (2007) A platform with six degrees of freedom. Proc Inst Mech Eng 180:371–386. https://doi.org/10.1243/pime_proc_1965_180_029_02
Yang XL, Wu HT, Chen B et al (2019) Dynamic modeling and decoupled control of a flexible Stewart platform for vibration isolation. J Sound Vib 439:398–412. https://doi.org/10.1016/j.jsv.2018.10.007
Gao C, Yang Z, Zheng S, Cong D (2018) An algorithm for real-time forward kinematics of 6-degree-of-freedom parallel mechanisms. Proc Inst Mech Eng Part I J Syst Control Eng 232:109–120. https://doi.org/10.1177/0959651817739793
He JF, Jiang HZ, Tong ZZ (2017) Modal control of a hydraulically driven redundant actuated fully parallel mechanism. JVC/J Vib Control 23:1585–1592. https://doi.org/10.1177/1077546315596661
Vaes D, Engelen K, Anthonis J et al (2007) Multivariable feedback design to improve tracking performance on tractor vibration test rig. Mech Syst Signal Process 21:1051–1075. https://doi.org/10.1016/j.ymssp.2005.07.004
Smallwood DO (1982) A random vibration control system for testing a single test item with multiple inputs. SAE Trans 1982:4571–4577. https://doi.org/10.4271/821482
Underwood M, Ayres R, Keller T (2011) Some aspects of using measured data as the basis of a multi-exciter vibration test. Structural dynamics, vol 3. Springer, New York, pp 939–954
De Coninck F, Desmet W, Sas P (2004) Increasing the accuracy of MDOF road reproduction experiments: calibration, tuning and a modified TWR approach. Proc 2004 Int Conf Noise Vib Eng ISMA 709–721
Vaes D, Souverijns W, De Cuyper J, et al (2002) Decoupling feedback control for improved multivariable vibration test rig tracking. Proc 2002 Int Conf Noise Vib Eng ISMA 525–534
Bhattacharya A, Malik N, Jindal S (2017) Optimization of simulation channels for inverse FRF calculation on 6-axis road load simulator: an experimental approach. SAE Tech Pap 2017–Janua. https://doi.org/10.4271/2017-26-0303
Phillips AW, Allemang RJ, Zucker AT (1998) An overview of MIMO-FRF excitation/averaging techniques. Proc 23rd Int Conf Noise Vib Eng ISMA 655–666
Schoukens J, Swevers J, Pintelon R, Van Der Auweraer H (2004) Excitation design for FRF measurements in the presence of non-linear distortions. Mech Syst Signal Process 18:727–738. https://doi.org/10.1016/S0888-3270(03)00084-0
Verboven P, Guillaume P, Vanlanduit S, Cauberghe B (2006) Assessment of nonlinear distortions in modal testing and analysis of vibrating automotive structures. J Sound Vib 293:299–319. https://doi.org/10.1016/j.jsv.2005.09.039
Gatto M, Peeters B, Coppotelli G (2010) Flexible shaker excitation signals for improved FRF estimation and non-linearity assessment. Proc ISMA 2010—Int Conf Noise Vib Eng Incl USD 2010 2475–2488
Gloth G, Sinapius M (2004) Analysis of swept-sine runs during modal identification. Mech Syst Signal Process 18:1421–1441. https://doi.org/10.1016/S0888-3270(03)00087-6
Pauwels S, Michel J, Robijns M (2006) A new MIMO sine testing technique for accelerated, high quality FRF measurements. Conference Proceedings of the Society for Experimental Mechanics Series
Smolders K, Swevers J (2006) Nonparametric MIMO FRF matrix estimation using a single periodic broadband excitation. Proc IEEE Int Conf Control Appl. https://doi.org/10.1109/CACSD-CCA-ISIC.2006.4777045
De Cuyper J, Verhaegen M (2002) State space modeling and stable dynamic inversion for trajectory tracking on an industrial seat test rig. JVC/Journal Vib Control 8:1033–1050. https://doi.org/10.1177/107754602029580
Guillaume P, Pintelon R, Schoukens J (1992) Nonparametric frequency response function estimators based on nonlinear averaging techniques. IEEE Trans Instrum Meas 41:739–746. https://doi.org/10.1109/19.199393
Cornelis B, Toso A, Verpoest W, Peeters B (2013) Adaptive modelling for improved control in durability test rigs. 20th Int Congr Sound Vib 2013, ICSV 2013 1:145–152
GUAN G (2009) Application of FRF estimator based on errors-in-variables model in multi-input multi-output vibration control system. Chinese J Mech Eng English Ed 20:101. https://doi.org/10.3901/cjme.2007.04.101
Zhang J, Zhang F, Jiang J (2020) Identification of multi-point dynamic load positions based on filter coefficient method. J Vib Eng Technol. https://doi.org/10.1007/s42417-020-00248-9
Underwood MA (1994) Adaptive control method for multiexciter sine tests.U.S. Patent 5,299,459
Chen J, Chen Z, He H, Zhou J (2012) Optimum control strategy study for multi-exciter sine test control system. Jixie Gongcheng Xuebao/Journal Mech Eng 48:159–166. https://doi.org/10.3901/JME.2012.08.159
Owens DH, Chu B, Songjun M (2012) Parameter-optimal iterative learning control using polynomial representations of the inverse plant. Int J Control 85:533–544. https://doi.org/10.1080/00207179.2012.658867
Wang X, Cong D, Yang Z et al (2019) Modified Quasi-Newton optimization algorithm-based iterative learning control for multi-axial road durability test rig. IEEE Access 7:31286–31296. https://doi.org/10.1109/ACCESS.2019.2897711
Wang X, Cong D, Yang Z et al (2019) Iterative learning control with complex conjugate gradient optimization algorithm for multiaxial road durability test rig. Proc Inst Mech Eng Part C J Mech Eng Sci 233:2349–2360. https://doi.org/10.1177/0954406218786981
Li C, Chen Z, Zu H, Zhao Y (2018) An improved optimal adaptive control method for MIMO sine vibration control of a multichannel coupled system. ASME Int Mech Eng Congr Expo Proc. https://doi.org/10.1115/IMECE2018-86983
Tang Y, Shen G, Zhu ZC et al (2014) Time waveform replication for electro-hydraulic shaking table incorporating off-line iterative learning control and modified internal model control. Proc Inst Mech Eng Part I J Syst Control Eng 228:722–733. https://doi.org/10.1177/0959651814536553
Eksteen JJA, Heyns PS (2016) An alternative update formula for non-linear model-based iterative learning control. Inverse Probl Sci Eng 24:860–888. https://doi.org/10.1080/17415977.2015.1088536
Müller T, Endisch C (2017) An estimation based iterative learning control approach for cross-coupled vehicle test systems. IEEE/ASME Int Conf Adv Intell Mechatronics, AIM. https://doi.org/10.1109/AIM.2017.8014210
Wang H, Huang H, Zhang Z, Li W (2015) Multiple-degree-of-freedom sinusoidal vibration generation based on a hexapod platform. Proc Inst Mech Eng Part I J Syst Control Eng 229:139–148. https://doi.org/10.1177/0959651814552718
Zheng Y, Zhou Z, Huang H (2020) A multi-frequency MIMO control method for the 6DOF micro-vibration exciting system. Acta Astronaut 170:552–569. https://doi.org/10.1016/j.actaastro.2020.02.033
Vold H, Crowley J, Rocklin GT (1984) New ways of estimating frequency response functions. J Sound Vib 18:34–38
Zhou H , Hai H, et al (2015) Structural optimization design of 6-DOF shaker's table. Spacecraft Environ Eng. CNKI:SUN:HTHJ.0.2015-05-021
Peeters B, Van Der Auweraer H (2012) Application of multisine excitation to aircraft Ground Vibration Testing. J IFAC. https://doi.org/10.3182/20120711-3-BE-2027.00369
Acknowledgements
The authors would like to thank for the support by the Beijing Institute of Spacecraft Environment Engineering.
Funding
The authors would like to thank for the support by the Beijing Institute of Spacecraft Environment Engineering.
Author information
Authors and Affiliations
Contributions
HH put forward the research problems and some guiding opinions on the research ideas of this paper. YZ performed the experiment and the analysis of the experimental data, and then wrote the manuscript. ZZ did some auxiliary work on the analysis of the experimental results.
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no known conflicts of interest/competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
A.1 Transient Response of LMS ADC Control Method
Taking a single-DOF linear system with a transfer function of \(H = Ae^{j\varphi }\) at frequency \(\omega_{{0}}\) as an example, the transient response of the LMS ADC control method with feedforward is derived, which reflects the characteristics of the method in the convergence process.
To reflect the superiority of the method proposed in this paper over the traditional control method without feedforward, the transient response of the traditional LMS ADC method without feedforward is derived first.
In the traditional LMS ADC control method, the weight coefficient is generally defined as \({\varvec{w}}(n) = [w_{1} ,w_{2} ]^{T}\), the reference signal as \({\varvec{x}}(n) = [\cos (\omega_{0} n),\sin (\omega_{0} n)]^{T}\), the error signal as \(e(n) = d(n) - y(n)\), the expected signal as \(d(n) = D\cos (\omega_{0} n + \gamma )\), and the control signal as \(u(n) = {\varvec{w}}^{T} (n){\varvec{x}}(n)\).
According to the literature [43, 44], the weight coefficient updating formula is as follows:
The output response signal is as follows:
By substituting control signal into the Eq. (53), the Eq. (54) is obtained as follows:
Since \(Ae^{j\varphi } {\varvec{x}}(n)\) is constant, the response is only related to the weight coefficient. Therefore, it is more intuitive to analyze the transient response of the output signal with the convergence process of the weight coefficient. By substituting Eq. (54) into Eq. (52), the Eq. (55) is obtained as follows:
To simply represent the process of transient response, the system is normalized. Make \(A = D = 1\), and \(\gamma = 0\). Then the invariant term in Eq. (55) can be expressed as:
By substituting Eq. (56) into Eq. (55) and taking mathematical expectation on both sides of Eq. (55), the Eq. (57) is obtained as follows:
Since it can be assumed that \({\varvec{w}}(n)\) is independent of \({\varvec{x}}(n)\), Eq. (57) can be simplified as follows:
The steady-state value can be obtained by calculating the limit on both sides of Eq. (58) as follows:
It is assumed that the initial value is \({\varvec{w}}(0) = {\varvec{w}}_{0}\), and the Z-transform is used to solve Eq. (59). The solution of Eq. (59) can be obtained as follows:
It is known that \(E[{\varvec{x}}(n){\varvec{z}}^{T} (n)]\) in Eq. (60) is the correlation matrix of vector \({\varvec{x}}(n)\) and vector \({\varvec{z}}(n)\), which can be easily obtained by the theory of matrix and linear algebra as follows:
For a convergent LMS ADC control method, the weight coefficients eventually converge to the optimal value. Obviously, the relationship between the steady-state value and the optimal value of the weight coefficients is as follows:
Therefore, by substituting Eq. (61) and Eq. (62) into the Eq. (60), the general term expression of transient response can be obtained as follows:
For the control system without feedforward, the initial value of weight coefficient is \({\varvec{w}}_{0} { = 0}\).
For the LMS ADC control method with feedforward, \({\varvec{w}}^{T} (n){\varvec{x}}(n) + {\varvec{d}}(n)\) is used as control signal. Since the control signal is always sinusoidal in the iterative process, it can be transformed into \({\varvec{w}}_{e}^{T} (n){\varvec{x}}(n)\), where \({\varvec{w}}_{e}^{T} (n)\) is the equivalent weight coefficient by trigonometric function theory. Compared with the weight coefficient of the system without feedforward, the initial value of the equivalent weight coefficient is different, but the steady-state solution is the same.
Therefore, the general term expression in Eq. (63) can be applicable to the LMS ADC control method with and without feedforward, and the difference only lies in the initial values of the weight coefficients.
Rights and permissions
About this article
Cite this article
Zheng, Y., Zhou, Z. & Huang, H. A Precise Identification and Control Method for the 6D Micro-Vibration Exciting System. J. Vib. Eng. Technol. 9, 1511–1531 (2021). https://doi.org/10.1007/s42417-021-00311-z
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s42417-021-00311-z