Balance Platform Model Parameter Identification

Abstract:

In the first stage of the identification process was to build a model describing the system, using the measurement data recorded on input and output system [. The model was built without the knowledge of the mechanisms that occur in the process, only based on the relationship between the collected data. As an input parameter the defined type of motion was used, next converted to digital form by the real-time computer and sent to the axes control card. While the output signal was registered as a change of encoders position. Sampling time 20 ms was adopted. For each axis separately recorded measurements (signals were with different amplitudes and frequencies of motion tape device) [. In the next step the estimation was performed for the selection suitable algorithm, and the determination the parameters of model selected in the previous step [. The last step was to verify, to check the results of identification process. By comparing the signals obtained in response to a signal given the model of a registered object actual output signal, it is possible to visually estimate the accuracy of the identification and designation of the error. The last part of the article shows the results of various tests.

• Periodical: Solid State Phenomena (Volume 198)
• Pages: 439-444
• DOI: https://doi.org/10.4028/www.scientific.net/SSP.198.439
• Online since: March 2013
• Authors: Agata Nawrocka, Andrzej Kot
• Keywords: ARMAX Model, Balance Platform, Identification Process

References

[1] A. Bielińska, D. Bismor, J. Figwer, J. Kasprzyk, T. Legierski, Z. Ogonowski, M. Pawełczyk, Identyfikacja procesów. Wydawnictwo Politechniki Śląskiej, Gliwice 1997, 56- 59, 62-70.

[2] Z. Hendzel, M. Nawrocki, Identyfikacja parametrów modelu robota, Acta Mechanica et Automatica, vol. 4 no. 2, pp.69-73, Oficyna Wydawnicza Politechniki Białostockiej, (2010).

[3] A. Kot, A. Nawrocka, Balance Platform System Modeling and Simulation 12th International Carpathian Control Conference ICCC 2011, Czech Republic, 2011, IEEE Xplore Digital Library, DOI 10. 1109/CarpathianCC. 2011. 5945851.

DOI: 10.1109/carpathiancc.2011.5945851

[4] K. Janiszowski, Identyfikacja modeli parametrycznych w przykładach. Warszawa: Akademicka Oficyna Wydawnicza EXIT, 2002, I. 1-I. 50.