Title :
Parameter estimation of dual-rate stochastic systems by using an output error method
Author :
Ding, Feng ; Chen, Tongwen
Author_Institution :
Control Sci. & Eng. Res. Center, Southern Yangtze Univ., Wuxi, China
Abstract :
In this note, a new gradient-based recursive algorithm for estimating parameters of dual-rate sampled-data systems is presented. The basic idea is, by using an output error method, to identify the single-rate models and the unknown noise-free outputs (true outputs) directly from dual-rate input-output data. Further, convergence properties of the proposed algorithm in parameter and output estimation are analyzed; and the result is illustrated and tested with an experimental water-level system.
Keywords :
gradient methods; recursive estimation; sampled data systems; stochastic processes; dual-rate input-output data; dual-rate sampled-data systems; dual-rate stochastic system; gradient-based recursive algorithm; martingale theory; multirate system; output error method; output estimation; parameter estimation; recursive identification; single-rate model; stochastic gradient; Algorithm design and analysis; Convergence; Estimation error; Parameter estimation; Polynomials; Recursive estimation; Stochastic processes; Stochastic resonance; Stochastic systems; System testing; Convergence properties; martingale theory; multirate systems; output error method; output estimation; recursive identification; stochastic gradient;
Journal_Title :
Automatic Control, IEEE Transactions on
DOI :
10.1109/TAC.2005.854654
