MODEL ACCURACY FOR ECONOMIC OPTIMIZING CONTROLLERS - THE BIAS UPDATE CASE

In many advanced control applications the numbers of manipulated and c ontrolled variables are not the same. A combination of steady-state ec onomic optimization and model predictive control has been employed to exploit the optimization opportunities presented by such ''non-square' ' systems. The economic optimization subsystem uses a process model an d is usually coupled with a model updating scheme to compensate for pl ant/model mismatch. Care must be exercised during the design of such a system to ensure the model-based optimization yields operating condit ions which are optimal for the true plant. This paper presents a rigor ous criterion for determining under what conditions the model-based op timization embedded within the model predictive controller, using bias update, is capable of finding the plant optimum despite plant/model m ismatch. The model accuracy criterion can be applied by solving an app ropriate nonlinear programming problem. Further, it is shown that when the embedded model-based optimization problem has linear constraints, the bias update method will ensure convergence to the plant optimum, providing the model accuracy criterion is satisfied and an appropriate filter is included in the feedback path. Discussions are concluded wi th a demonstration of the methods on a real-time gasoline blending con trol and optimization problem.