Analysis of estimation errors in the identification of non-linear systems
- 1 January 1981
- journal article
- research article
- Published by Taylor & Francis in International Journal of Systems Science
- Vol. 12 (2) , 205-225
- https://doi.org/10.1080/00207728108963740
Abstract
Estimation errors introduced in the identification of non-linear systems are analysed. The influence of record length, mean level, power and bandwidth of the input excitation, the effects of input and output noise and errors introduced by the decomposition techniques are investigated. Experimental design and input selection are discussed and the results are illustrated by simulated examples,Keywords
This publication has 4 references indexed in Scilit:
- Theory of separable processes with applications to the identification of nonlinear systemsProceedings of the Institution of Electrical Engineers, 1978
- Effects of nonlinearities on the measurement of weighting functions by crosscorrelation using pseudorandom signalsProceedings of the Institution of Electrical Engineers, 1973
- A New Method for Minimising a Sum of Squares without Calculating GradientsThe Computer Journal, 1970
- High-order autocorelation functions of pseudorandom signals based on m sequencesProceedings of the Institution of Electrical Engineers, 1970