Parameter estimation for MIMO systems

In the design of adaptive control systems, it is sometimes necessary to estimate parameters for systems represented by state-space models. A direct approach using those state-space models usually leads to a very complicated and noneffective parameter estimation algorithm. However, if a relationship between the state-space model and an input-output model is used it is possible to estimate the parameters for a state-space model by first estimating the parameters for the corresponding input-output model. The authors present such models, such a relationship, and a general recursive parameter estimation algorithm based on some ideas from the prediction error method and the method of stochastic approximation. Our parameter estimation technique is then developed using this general recursive algorithm. A two-input two-output system example is used to illustrate the technique