Optimality and computational feasibility in transient control: Part I. A modified criterion for optimality

Publisher Website
Google Scholar
Abstract
The notion of optimality with regard to transient control is critically examined with a particular view toward the computational difficulties of solving the optimal control problem, and the arbitrary aspects of the usual objective functions. Realizing a need for effective approximations, a criterion for transient control is developed which requires a minimum of computational effort to apply in practice. Optimality is achieved in an instantaneous sense, and it is argued that overall optimality is well approximated for many cases of practical interest. To demonstrate its use, the criterion is applied to the transient control of a stirred‐tank reactor. Numerical examples are given and the results are discussed and compared with those obtained by alternative methods.