On the use of constraints in least squares estimation and control (Q1614351): Difference between revisions

Using quadratic penalty functions, the authors examine the effect of constraints on estimation and control methods. The effect of such constraints on the statistical properties of least squares estimators is investigated, and the role of constraints in the formulation of maximum likelihood and maximum a posteriori estimators is explored. The potential for improved performance over unconstrained least squares is demonstrated by an example involving parameter estimation.