Synthesis of control structures by singular value analysis: Dynamic measures of sensitivity and interaction

Abstract

Abstract A strategy for synthesizing regulatory control structures is developed within the framework provided by singular value decomposition (SVD). Quantitative measures of interaction and sensitivity for the nodes, as well as for the entire system, are established with this method. The approach enables analysis of the control struture over a range of frequencies which are of practical importance for a particular processing unit, thus insuring that both static and dynamic effects are encompassed. In addition, this method allows the designer to identify the modeling aspects which are important in determining the performance of the synthesized control structures. A series of examples, presented in order of increasing complexity, demonstrate the usefulness and versatility of this procedure. The technique is implemented as an interactive computer package.