Minimum Variance Unbiased State Estimation,
Abstract
Implementation of an optimal control developed from a quadratic performance function for a linear problem requires in general the feedback of the plant state vector multiplied by a computed gain matrix. However, if the accessible plant output consists only of noisy linear combinations of the individual states, then it is necessary to reconstruct from these measurements some estimate of the actual state vector. Using the separation theorem, it can be shown that the optimal state estimate for this purpose is the one obtained using the Kalman-Bucy algorithm. It is the purpose of this tutorial paper to first derive and then to discuss the useage of the Kalman-Bucy equations for state estimation. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1970
- Accession Number
- AD0717103
Entities
People
- Howard Kaufman
Organizations
- Rensselaer Polytechnic Institute