Application of Generalized Sensitivity Functions in Estimating the Real-Time Computational Expense Necessary for Stabilizing the Dynamics of Drifting Parameter Plants by Process Computer

Abstract

In the hierarchic control of a technological process with time-variable parameters the reduction of parameter sensitivity of the individual partial processes represents one of the tasks of the lower control level. The design of a DDC-system for this purpose leads to the question whether the admissible dynamic subsystem tolerances are to be guaranteed by use of adaptive or nonadaptive control algorithms and what real-time computational expense will be caused by the different solutions. From the design engineer’s point of view it is desirable, for planning the DDC-computer, to despose of computational expense estimations already in the state of predesign where only rough process data are available. For obtaining such estimations certain presumptions are to be made concerning the character of the control algorithm, the subsystem input signal, the process and the changements of its parameters.