|
It is a well known fact that many chemical processes are inhere- tly nonlinear.And the need for nonlinear modeling and control ha- s recognized as well.Inspite of this , majority of chemical proc- esses continue to bo controled vialinear system analysis and des- ign. An attempt has been made in this research to develop a nonlinear gain Hammersteinmodel for a simulated high - purity distillation column . An extension of SISO modelingtechnique is made to themu- lti - input / multi - output system in order to reduce thedegree of complexity involved . Subsequent design of diagonal PI andful- l PI controllersis also presented. Test results reveal the factt hat even with the simplified Hammersteinmodel improved controlis obtained when compared with the linear model based design. It is believed that much better control is achievable with a full- scale nonlinear Hammerstein model.
|