02B Estimation of Sys Paras - SYS635 Adaptive Control...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon
SYS635 Adaptive Control Systems KaCC 02B Estimation of Sys Paras 09/10/05 1 CHARACTERIZATION OF 1 st ORDER SYSTEM SYSTEM IDENTIFICATION - Parameter estimation Assumption. We will assume that the input u(k) and output y(k) for the system are accessible for measurements. That is we can acquire sample data {u(k) ,k = 1, 2, … N} {y(k), k = 1, 2, … N} where N = the # of samples taken or samples of interest to be considered. Batch Mode Parameter Estimation. From the ARMA model, we obtain a set of time series relationship ya y b u b u y b u b u y b u b u yN a buN buN () ( ) ( ) 21 2 1 32 3 2 43 4 3 11 01 =− + + + + + + + + M which can be vectorized as { 0 1 (2) (1) (2) (3) (2) (3) (2) (4) (3) (4) (3) ( 1 ) () ( 1 ) yy u u a u u b u u b uN uN YM θ   = −−  14243 1444442444443 which can be abbreviated as * = We can solve for the parameters using left pseudo inverse solution (Least Squared Error Estimate) = (' *) ' * MM MY 1
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
SYS635 Adaptive Control Systems KaCC 02B Estimation of Sys Paras 09/10/05 2 Real-Time Parameter Estimation using Recursive Least Square Algorithm At time t = NT, we may compute the batch estimate for the parameters [] 1 ( ) '* covariance of estimation errors ( )= ( ) the estimate computed using N samples PN M M NP NMY θ == = The parameter estimate can also be updated at each sampling instance using a recursive algorithm. For instance, a new set of samples y(N+1) and u(N+1), obtained after N data has been collected, can be used to update the already computed parameter estimate as follows: [ ] () 3 (1 ) )(1 )( ) i n f o r m a t i o n ) ' ) 1(1 ) ( ) ' ) ( 1) ( ) ( 1)( ( ( 1) ( ) new estimate with the N+1 th samples ) ) ' ) ( ) / N y N u N u N newupdate KN N N N NN K N y N I N ϕ ϕϕ θθ +=− + + += + + + + + + +− + += − + + NOTE: The above solutions (batch and recursive) are mathematically solutions to the system of linear algebraic equation under precise modeling and measurement conditions.
Background image of page 2
Image of page 3
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 04/17/2011 for the course SYS 635 taught by Professor Re during the Spring '11 term at Albany College of Pharmacy and Health Sciences.

Page1 / 9

02B Estimation of Sys Paras - SYS635 Adaptive Control...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online