FIELD: control systems.
SUBSTANCE: invention relates to automatic control. Method of identification of a linear dynamic system includes the initial assignment of the transfer function of the a priori type of system in a fractionally rational form. Set the normalized duration of unsteady motion, then the arrays of samples of the input and output signals are normalized by time depending on the measured duration of unsteady motion of the system. Set the criterion for the significance of the coefficients of the normalized model, then check the coefficients of the numerator and denominator of the model according to a given criterion sequentially from the highest to the lowest until the detection of the first significant coefficient. Insignificant coefficients of the numerator and denominator are excluded from the model. Significant coefficients of the model lead to real values depending on the measured duration of the unsteady motion and the normalized duration of the unsteady motion of the system and obtain a model of a linear dynamic system.
EFFECT: system identification is simplified.
4 cl, 3 dwg
Authors
Dates
2018-12-18—Published
2018-03-28—Filed