A MIMO periodic ARX identification algorithm for the Floquet stability analysis of wind turbines

Abstract

The paper presents a new stability analysis approach applicable to wind turbines. At first, a reduced order periodic model is identified from response time histories, and then stability is assessed using Floquet theory. The innovation of the proposed approach is in the ability of the algorithm to simultaneously consider multiple response time histories, for example in the form of measurements recorded both on the rotor and in the stand still system. As each different measurement carries a different informational content on the system, the simultaneous use of all available signals improves the quality and robustness of the analysis.