Quality improvement of tandem hot metal strip rolling using an augmented state-dependent Riccati equation technique

The tandem rolling of hot metal strip is a highly complex nonlinear large-scale industrial process with significant time delays plus major uncertainties and disturbances. The control challenge also is heightened by a hostile environment that precludes the location of certain sensors to measure variables that are important for control. Based on the results of our previous work using a modified state-dependent Riccati equation technique for control of the tandem cold metal rolling process, it is considered that a similar basis could be expanded upon to realize an improved method for the control of this more complex application. In this paper we present a comprehensive process model and expand on the results of our first efforts in the development of a suitable controller for this process. The results of simulations of the controller coupled to the model show a strong potential for improvement in the quality of the final product. The methodology used also has the potential for expansion to improve the control of other large nonlinear systems.