Integrated model based control of multivariable nonlinear systems

Abstract A new algorithm, called Integrated Model Based Control, is proposed for the control of complex unit operations. The algorithm utilizes a process model based on first principles as opposed to mathematically convenient forms. All available process measurements are utilized in an on-line identification step of the algorithm. The algorithm allows us to approach the control problem in an integrated fashion taking into account overall control objectives, process operating constraints and the inherent interaction among the process variables. Most importantly, the algorithm allows the integration of process control with the economics of process operation. A feasibility study was conducted using simulated implementation of the algorithm on a distillation column. The results are very encouraging and suggest further exploration of the concepts.