System for controlling at least one active magnetic bearing equipping a rotating machine comprising a rotor and a stator, at least one means for measuring the radial positions of the rotor as a function of the signal from at least one position sensor, and at least two control loops of the active magnetic bearing as a function of the radial positions of the rotor, each control loop of the magnetic bearing being provided with at least one synchronous filter as a function of the rotation speed, and an extended Kalman filter for determining the rotation speed of the rotor with respect to the stator receiving as input, from position sensors, measurements of radial position of the rotor and as a function of measurements of radial position of the rotor performed over a predetermined time at zero rotor rotation speed.

A drive apparatus includes a DC link; a rectifier to convert power of an external power supply into predetermined DC power to be supplied to the DC link; a motor to rotationally drive a compression mechanism; a supporter to magnetically support a rotating shaft of the compression mechanism; a first driver to drive the motor with power from the DC link, to cause the motor to execute regenerative operations of converting power from the rotating shaft into electric energy, and to output regenerative power to the DC link; and a second driver to drive the supporter with the power from the DC link, wherein the first driver causes the motor to execute the regenerative operations so that a voltage of the DC link becomes higher than a voltage of the DC link when the power of the external power supply is normally supplied.

A machine learning device learns a control condition for a magnetic bearing device that includes a magnetic bearing having a plurality of electromagnets that apply an electromagnetic force to a shaft. The machine learning device includes a learning unit, a state variable acquisition unit, an evaluation data acquisition unit, and an updating unit. The state variable acquisition unit acquires a state variable including at least one parameter correlating with a position of the shaft. The evaluation data acquisition unit acquires evaluation data including at least one parameter selected from a measured value of the position of the shaft, a target value of the position of the shaft, and a parameter correlating with a deviation from the target value. The updating unit updates a learning state of the learning unit by using the evaluation data. The learning unit learns the control condition in accordance with an output of the updating unit.

A method of detecting a vibration node between a non-collocated sensor-actuator pair of a rotatable component includes applying an excitation signal to an actuator of the sensor actuator pair. The method also includes obtaining frequency response data from the sensor-actuator pair. The method further includes analyzing the frequency response data to ascertain a resonant frequency of the rotatable component. The method includes identifying a resonance/anti-resonance peak pair in the frequency response data for the non-collocated sensor-actuator pair. Furthermore, the method includes determining whether the vibration node is located between a sensor and the actuator of the non-collocated sensor-actuator pair based on the resonance/anti-resonance peak pair.

A method for constructing an active magnetic bearing controller based on a look-up table method includes: building finite element models of an active magnetic bearing to obtain two universal Kriging prediction models in X-axis and Y-axis directions about actual suspension forces being in association with actual displacement eccentricities and actual control currents in the X-axis and Y-axis directions of the active magnetic bearing based on a universal Kriging model; creating two model state tables in the X-axis and Y-axis directions about the actual suspension forces being in association with the actual displacement eccentricities and the actual control currents to construct two look-up table modules with the two built-in model state tables, respectively; and constructing an active magnetic bearing controller by using two fuzzy adaptive PID controllers, two amplifier modules in the X-axis and Y-axis directions, the two look-up table modules, and two measurement modules in the X-axis and Y-axis directions.

A control apparatus includes a constant storage portion that stores constant values of an electromagnet coil including a resistance value Rm, an inductance Lm, a sampling time Ts, etc. A current storage portion stores previous current command values Ir having been regularly sampled by a microcomputer inside a current control circuit. A low-frequency feedback circuit generates a signal for suppressing an error between DC components and low-frequency components of an input current command value Ir and a detected current value IL and outputs the signal. An output voltage computing circuit calculates, based on the input current command value Ir[n+1], a stored value Ir[n] of the current storage portion, a stored value of a constant storage portion, and the signal of the low-frequency feedback circuit, a voltage for suppling the electromagnet coil with a current in accordance with a command, and outputs the calculated voltage.

A magnetic bearing controller for controlling a magnetic levitation motor, the magnetic levitation motor including: a rotor; a pair of electromagnets that causes the rotor to levitate by electromagnetic force; an auxiliary bearing that supports a rotating shaft of the rotor when the rotor is stopped; and a rotor position detector that detects the rotor's position in a levitation direction. The magnetic bearing controller includes an operation current generator that generates an operation current value corresponding to a deviation between a position command value and the rotor's position detected by the rotor position detector. The operation current generator is configured to give a predetermined initial value greater than 0 to the operation current value at a start of levitation for causing the rotor in a state where the rotating shaft of the rotor is supported by the auxiliary bearing to levitate and be positioned at a predetermined target position.

An electric motor system includes a rotary shaft having an axis line displaceable relative to a rotation center, a magnetic bearing for supporting the rotary shaft, a permanent magnet mounted on the rotary shaft and having a plurality of magnetic poles arranged in a circumferential direction around the axis line of the rotary shaft, three detection elements arranged in the circumferential direction around the rotation center for detecting a magnetic flux generated from the permanent magnet, and a coordinate detection section for determining coordinates of the axis line of the rotary shaft based on output values of two detection elements selected out of the three detection elements in accordance with a rotation angle of the rotary shaft.

A method for constructing an active magnetic bearing controller based on a look-up table method includes: building finite element models of an active magnetic bearing to obtain two universal Kriging prediction models in X-axis and Y-axis directions about actual suspension forces being in association with actual displacement eccentricities and actual control currents in the X-axis and Y-axis directions of the active magnetic bearing based on a universal Kriging model; creating two model state tables in the X-axis and Y-axis directions about the actual suspension forces being in association with the actual displacement eccentricities and the actual control currents to construct two look-up table modules with the two built-in model state tables, respectively; and constructing an active magnetic bearing controller by using two fuzzy adaptive PID controllers, two amplifier modules in the X-axis and Y-axis directions, the two look-up table modules, and two measurement modules in the X-axis and Y-axis directions.

An electric motor system includes a rotary shaft having an axis line displaceable relative to a rotation center, a magnetic bearing for supporting the rotary shaft, a permanent magnet mounted on the rotary shaft and having a plurality of magnetic poles arranged in a circumferential direction around the axis line of the rotary shaft, three detection elements arranged in the circumferential direction around the rotation center for detecting a magnetic flux generated from the permanent magnet, and a coordinate detection section for determining coordinates of the axis line of the rotary shaft based on output values of two detection elements selected out of the three detection elements in accordance with a rotation angle of the rotary shaft.