Optimal sensor placement for modal identification using system-realization methods

It is shown analytically that the Effective Independence (EfI) sensor placement method proposed by Kammer (1991) also enhances the modal identification process itself using system realization methods. It is shown that maximizing the independence of the target modes minimizes the size of the observability matrix required to render the system observable. This also minimizes the size of the required block Hankel data matrix used in the modal identification process. By maximizing the determinant of the Fisher information matrix, the EfI method maximizes the observability of the system and maximizes the singular values of the data matrix used in the eigensystem realization algorithm method to differentiate between target modes and computational modes. It also enhances the estimation of the discrete system matrix. It is believed that the EfI method provides an efficient technique for selecting a sensor configuration that not only provides the modal data required by structural dynamicists, but also produces an enhanced modal identification.