Identification of objects with unknown bounded disturbances

An identification problem for linear objects with unknown bounded disturbances in both input and output channels is solved. The experimental information regarding the identified object has an interval character: only the lower and upper boundaries for values of the measured signals are known. The region of possible values of the model coefficients that do not contradict experimental information is constructed. This region is described by a system of linear inequalities, which is investigated by linear programming. Some additional possibilities for the investigation and control of objects in such situations are discussed.