Biased estimation for dynamic systems.Optimization and regulation of static and dynamic systems require good estimates of the states of the system model and the parameters of the model in the presence of input and measurement noise. A biased estimator is proposed as an alternative to the familiar best linear unbiased estimator. Depending on the value of a constant, this biased estimator can be made unbiased and is then identical to the best linear unbiased estimator, thus permitting additional freedom of choice in design and application. The sum of squared errors of this bs tmator can be less than the best linear unbiased case. An existence theorem is established and various properties are discussed.
Document ID
19720037186
Acquisition Source
Legacy CDMS
Document Type
Conference Proceedings
Authors
Chang, J. W.
Leathrum, J. F.
Hoerl, A. E. (Delaware, University Newark, Del., United States)
Date Acquired
August 6, 2013
Publication Date
January 1, 1972
Subject Category
Electronics
Meeting Information
Meeting: Symposium on Nonlinear Estimation Theory and Its Applications