A phase constraints tuning approach for complex object identification under uncertainty

The following approach is presented for an arbitrary dynamic system, which has structure and parameters uncertainties. This approach handles these uncertainties during the system functioning period. It is assumed that the system to be identified is exposed to disturbances, their information being incomplete. Under these conditions, the acceptable accuracy of parameters and structure should be provided (i.e. robust system identification should be performed). The presented approach utilizes the continuousness and smoothness properties formation of a certain phase domain with respect to the current trajectory corresponding to a given control function. The boundaries of the phase domain can be approximately described by varying phase constraints. By making the phase domain narrow enough (i.e. tuning it for varying phase constraints approach), we obtain the algebraic inequalities, under identification would satisfy to. These inequalities may be sowed using various approaches. The approach itself can be used for special tests with systems to be identified, including real-time scale, in case a system is under external disturbances. The parameters and structure uncertainty is taken away while the system's functioning.