Control system identification using additive test signals

System characteristic identification using normal operating signals and additive test signals is discussed. The advantages of additive test signals are enumerated. A system which senses and adapts to changes in a nondominant pole was simulated on an analog computer. A two-parameter adaptive control using a single high-frequency sinusoidal test signal to detect changes in gain and pole locations is proposed. When a random probe is used, it is possible to obtain a set of coefficients theoretically capable of characterizing nonlinear systems. The design problems encountered in applying this method to simple linear passive networks are considered. Crosscorrelation identification of linear systems using periodic discrete interval binary noise is discussed.Electrical and Computer Engineerin