Abstract
The problems and algorithms in question are applied to the parameter design of automatic control systems. The generality of (1-1) permits the definition of a wide range of constraints in the time and frequency domains, as well as on the eigenvalues and singular values of the system matrices. The designer can set up and satisfy direct requirements to the form of the output reaction, the locus of the return difference, the frequency characteristics, the degree of stability etc. Examples and graphics illustrate the characteristic features of the methods discussed. The algorithms are implemented in a software system for vector optimization developed by the authors