Research On Key Technology Of Magnetic Bearing With High Reliability

Active Magnetic Bearing (AMB) has gradually been applied to actual industrial area from laboratory. Nonexistent friction makes AMB own longer work life and high reliability compared to conventional bearings. However, the reliability still can not satisfy the requirements of some special equipment, such as more-electric gas turbine engines and nuclear generator equipped with AMB. Therefore, methods to enhance the reliability of AMB have recently been the research hotspots at home and abroad. Key technologies of AMB with high reliability are studied in the paper.Two displacement sensor failure diagnosis methods are presented according to the active magnetic bearing adopting differential method to measure the displacement of rotor. The relationship between the output signal of controller, the output of displacement sensors and the disturbed signal applied to displacement sensor probes is analyzed. Characteristics of displacement sensor's output and controller's output of the failed sensor are analyzed when different displacement sensor probe failed. Based on the principle, fault diagnosis methods for the displacement sensor probes based on adaptive filtering method and discrete Fourier transform are presented. The methods are firstly studied by simulation using MATLAB tool and then the experiments are made on a five degree of freedom active magnetic bearing test rig. The result shows that both of the methods can identify the fault of the displacement sensors when the rotor rotates at high speed and stays stationary.The coordinate transformation-based actuator fault-tolerant control algorithm on the basis of current distribution matrix reconstruction is presented. Aiming at the five failure modes of a six-core radial AMB with coupled magnetic path, the corresponding coordinate transformation matrix as well as the coil re-mapping matrix for each mode are derived. Suspending and failover characteristics of the rotor in each failure mode are simulated, and then the relative experiments are taken out on a two degree of freedom active magnetic bearing text rig.The displace sensors and actor fault-tolerant control method based on coordinate transformation is presented,which can ensure the stability of the rotor suspension in the condition that the sensors and the excitation coils fail at the same time. This algorithm is simulated using MATLAB and studied on a two degree of freedom active magnetic bearing text rig. The result shows that the fault-tolerant control algorithm can achieve the stability of the rotor suspension in the state that one displacement sensor and three excitation coils at most failed synchronously. The fault-tolerant control algorithm based on coordinate transformation of the radial magnetic bearing installed with three displacement sensors is studied and simulated using MATLAB. The experiment is also made on a two degree of freedom active magnetic bearing text rig. The result shows that the fault-tolerant control algorithm can achieve the stability of the rotor suspension in the state that any displacement sensor failed.The fault-tolerant control method based on current distribution matrix reconstruction is studied according to a six-core radial AMB with coupled magnetic path and distribution matrixes of normal state and two fault states are derived. Suspending and failover characteristics of the rotor in the three kinds of states are simulated, and then the relative experiments are carried out on a two degree of freedom active magnetic bearing text rig. The result shows that this fault-tolerant control method can achieve stable suspension in fault condition and can switch to normal station stably when the failure occurs.In the end, the influence of mechanical error on the performance of AMB is studied. Common mechanical errors of a classic 5-DOF AMB are introduced. The influence mechanism of each mechanical error on the working state of AMB is analyzed in detail. The relationship between machining errors and the working parameters of AMB is quantitatively analyzed. The results are used to identify the problems caused by cutting and assembling so that the assembly accuracy of mechanical part and the workability can be improved.