Process Identification Utilizing a Sequential Instrumental Variable Regression Algorithm.
Abstract
Instrumental variable (IV) regression is applied to the estimation of the parameters of a difference equation model of a process subject to noise. The technique preserves the simplicity of least-squares estimation, and is shown to significantly reduce the bias on the parameter estimates caused by measurement noise. A second-order example is used to illustrate the performance of the IV estimator, and to study the selection of sample time and initialization parameters. Demonstration is given on the parameter tracking capability of the dynamic form of the algorithm. The dynamic algorithm is important for use in adaptive control. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1976
- Accession Number
- ADA032847
Entities
People
- A. B. Corripio
- A. T. Touchstone
Organizations
- Louisiana State University