System modeling and optimization under vector-valued criteria

Abstract

The integrated problem of optimization and parameter estimation of a given system is addressed in the context of vector-valued optimization techniques. The validity of the existing convergence properties of the basic two-step approach for solving the joint problem of scalar optimization and parameter estimation is extended for the vector-valued case under adequate conditions. Some elements of decision analysis are incorporated into the joint problem in order to reflect conflicting situations derived from the simultaneous consideration of multiple performance criteria. To illustrate the main aspects of the approach proposed, the problem of optimization and parameter estimation of a nonlinear dynamic batch reactor is considered. The results attest the computational feasiblity of the procedure and its usefulness as a qualitative tool for the synthesis of integrated control policies.