A procedure for the identification of industrial processes is discussed, which has been succesfully applied for the modelling of dynamic properties of various industrial processes. The procedure conforms to industrial constraints and in this respect differs in some respects from conventional theories. The central model set used for the identification is known as the common denominator form of the minimom polynomial form. Parameter estimation is performed in several steps. The final step implies minimization of an output error criterion in least squares sense. As an illustrative example the identification of a glass feeder is described.
|Title of host publication||Application of multivariable system techniques (AMST 90), University of Bradford, UK, 18-19 April 1990|
|Place of Publication||London|
|Publisher||Elsevier Applied Science Publishers|
|Publication status||Published - 1990|