Optimum robust detection of changes in the AR part of a multivariable ARMA process

We investigate the theoretical properties of new instruments-based test statistics recently proposed [3] for detection and diagnosis of changes in the AR part of a multivariable ARMA process. The design flexibilities are analyzed, and the optimum design of the test is exhibited. The connection with the accuracy of the I.V. identification method [14] is established, and the comparison with the local likelihood ratio tests is done. These tests have been developed as a solution to the problem of vibration monitoring for offshore platforms.