Services on Demand
Journal
Article
Indicators
- Cited by SciELO
- Access statistics
Related links
- Cited by Google
- Similars in SciELO
- Similars in Google
Share
Iteckne
Print version ISSN 1692-1798
Abstract
ORTIZ BRAVO, Víctor Alfonso; NIETO ARIAS, Manuel Antonio and QUINTERO SALAZAR, Edwin Andrés. Methodology for Real-Time Parameter Estimation Through Kalman Filter and Least-Squares. Iteckne [online]. 2013, vol.10, n.1, pp.37-44. ISSN 1692-1798.
This paper presents a methodology for realtime estimation of the parameters of unknown nonlinear dynamic systems by Kalman filtering and the least squares estimation method. The algorithms resulting from the application of the proposed methodology are implemented in real time on a TMS320F2812 digital signal processor (DSP) from Texas Instruments©. In order to evaluate the effectiveness of the method, estimation of a dynamic plant second order filter with stable low-pass behavior, simulated by an electronic circuit containing operational amplifiers, operating under laboratory conditions that generate a high disturbance in the input and output functions was performed. These signals are acquired by the digital signal processor via its embedded analogue-digital (A/D) converter, to implement the real-time parameter identification algorithm proposed in the methodology. Finally, we obtained the estimated transfer function of the system with an error of 0.0019% in the cutoff frequency, 10.34% in the overshoot, and 14.28% in the settling time, demonstrating the effectiveness of the real-time implementation of the estimation strategy proposed.
Keywords : Estimation of parameters; Kalman filter; system identification; least squares; digital signal processor; real time processing.