Parameter identification for a triaxial shaker system

Different approaches to parameter identification for a triaxial shaker system are examined. One approach uses an output error formulation of a cost function, which is then minimized using a standard algorithm. Two other approaches use a cost function defined through an input error formulation. The cost function is minimized through, alternately, a standard function minimization algorithm and a linearized least-squares algorithm. For the case of fewer measurements than degrees of freedom, a generalized technique for reconstructing the states is proposed. This technique does not make any assumptions regarding the absence of external forces on unmeasured degrees of freedom. The three approaches to parameter identification are compared for two different systems