Hierarchical least squares identification for periodically and non-uniformly sampled multirate systems

For the systems which the control input and output are updated and sampled non-uniformly, the lifted state-space models are derived, and then the corresponding input-output relationships are obtained. According to the hierarchical identification principle, the obtained identification model is decomposed into two fictitious subsystems, one containing the parameter vector only, and the other containing the parameter matrix. In order to take into account the causality constraint in the lifted system model, the subsystem with the parameter matrix is also decomposed into several sub-submodels, and the hierarchical least squares identification method is developed for non-uniformly sampled systems. The simulation results indicate that the proposed algorithm is effective.