Subspace identification using dynamic invariance in shifted time-domain data

A novel subspace identification method is presented which uses a shift-invariant property of the output data to estimate system dynamics. It is shown that the algorithm may be used with correlation function estimates in addition to input-output data. The algorithm is compared to other subspace methods in a simulation study based on an existing benchmark problem. The results show that the proposed method used with correlation function data achieves consistent system estimates in the presence of highly-colored noise.