Filtered Closed Loop Output Error Method
This method has been developed in order to identify without bias "plant + disturbance" models corresponding to closed loop output error structure . It is similar to the closed loop output error method, except the fact that a filtered regressor vector is used. It means that the regressor vector f(t) is replaced by ff(t) in the parameter adaptation algorithm.
ff(t) = S/P0 f(t)
where S is the denominator of the controller (see: RST controller ) and P0 is an initial estimation of the closed loop characteristic polynomial:
P0 = A0(q-1)S(q-1)+B0(q-1)R(q-1)
A0(q-1) and B0(q-1) are the initial estimations of the plant model denominator and numerator, respectively, obtained for example by an open loop type identification method.
See: Closed loop identification module .
(file fcloe.htm)