Multi-valued logic and its application in the fault diagnosis of the sensors of magnetic bearings

Magnetic bearings are accurate and complex electromechanical devices. The sensors of a magnetic bearing are important to the bearing. Only when the sensor work normally makes the rotor of the magnetic bearing rotate stably. Since the sensors of a magnetic bearing are invisible in the device with very high positioning precision, the traditional ways of fault diagnosis of the sensors are difficult to be used in magnetic bearings. Considering the characteristics of the sensors of magnetic bearings, a multi-valued logic algebra method based on sequential variables is presented for the fault diagnosis of the sensors of magnetic bearings. The definition of the logic algebra is given, and some theorems are proved, which are usually used in deduction. According to the presented method, the corresponding experiments should be done, in which only the states of the magnetic bearings are measured by processing the signals from the sensors. The states of the sensors can be deduced with the multi-valued logic algebra based on sequential variables. It is significant to know the states of the sensors before fixing or compensating the sensors.