Identification of defects in linear systems based on optimal control methods

Abstract

The problem of identifying defects in dynamic systems described by linear differential equations is considered. Observers constructed on the basis of optimal control methods are used to solve this problem. The proposed method to synthesize such observers is based on the idea of introducing a new subsystem, in which the role of an unknown vector function describing emerging defects is played by some auxiliary control. This control is proposed to be formed in such a way as to provide a minimum of the residual functional. Unlike well-known approaches to solving the problem of defect identification based on observers operating in a sliding mode, the proposed method allows to expand the class of systems for which the identification problem can be solved, since the method of constructing sliding observers imposes certain restrictions on the systems under consideration. Results of simulation are presented to confirm the operability of the diagnostic observers synthesized using the described method.