Towards an Improved Eigensystem Realization Algorithm for Low-Error Guarantees

Eigensystem realization algorithm (ERA) is a tool that can produce reduced order model (ROM) from just input–output data of given system. ERA creates the ROM while keeping number internal states to minimum level. This was first implemented by Juang and Pappa (1984) analyze vibration aerospace structures impulse response. We reviewed tested it on single input output (SISO) system as well multiple inputs (MISO) prediction agreed with actual data. Unlike other reduction techniques (balanced truncation, balanced proper orthogonal decomposition), works fine without need adjoint system, makes promising, completely data-driven, thrifty method. In this work, we propose modified eigensystem relies upon an optimally chosen time resolution for used also checks good performance through frequency analysis. Four examples are discussed: two confirm generating ability last illustrate its capability low-dimensional (for large-scale system) much more accurate than one produced traditional ERA.