Eventos Anais de eventos
DINAME2019
DINAME2019
Electrical Fault Detection in a Rotor Supported by Active Magnetic Bearing
Submission Author:
Gilberto Machado da Silva , SP
Co-Authors:
Gilberto Machado da Silva
Presenter: Gilberto Machado da Silva
doi://10.26678/ABCM.DINAME2019.DIN2019-0019
Abstract
The problem of failure detection applied to rotational system with active magnetic bearing (AMB) has a special difficulty because the mechanical system equations are associated to the system control structure. The measurement of all state variables normally is not possible in real systems and the knowledge of stiffness and damping values are also difficult to be identified. In order to avoid this practical characteristic, we propose a methodology that works with the structure of the model and generate compatibility equations involving correlations between a reduced numbers of state variables. These relations are obtained by the matrix equation of Ljapunov. The faults in the system are detected by monitoring the change in the physical parameters and comparing the theoretical and estimated correlation functions. Artificial neural networks are used to map the correlations involving states that are not measured. The proposed method is applied numerically to detect and locate sensor and actuator faults.
Keywords
Fault Diagnosis, Magnetic Bearings, Rotor Dynamics, neural networks
