System and corresponding method for controlling active magnetic bearings of rotating machinery of equipment
By using an extended Kalman filter in conjunction with a radial position signal in rotating machinery, the reliance on an additional speed sensor in the AMB system is eliminated, enabling lower-cost and higher-accuracy speed measurement and synchronous filtering control.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SKF MAGNETIC MECHATRONICS SAS
- Filing Date
- 2021-03-02
- Publication Date
- 2026-06-26
AI Technical Summary
In the prior art, rotating machinery using active magnetic bearings (AMB) requires additional speed sensors or measuring devices to activate the synchronous filtering system, resulting in additional costs and potential sources of failure, while also lacking accuracy.
An extended Kalman filter is used to determine the rotor's rotational speed by measuring the radial position signal of the rotor relative to the stator, combined with at least two position sensors, thus eliminating the dependence on additional sensors. The radial position signal and the Kalman filter are used for synchronous filter control.
It reduces the control cost of rotating machinery, provides more reliable speed measurement, improves the accuracy of synchronous filtering systems, and reduces the risk of failure.
Smart Images
Figure CN113346817B_ABST
Abstract
Description
Technical Field
[0001] The technical field of this invention is the control of rotating machinery, and more specifically, the control of machinery equipped with magnetic bearings. Background Technology
[0002] Most industrial applications using active magnetic bearings (AMB) require rotor rotation speed to activate synchronous filtering systems, such as Active Balancing System (ABS), Adaptive Vibration Reduction System (AVR), and Optimal Damping Control System (ODC).
[0003] These filters are an essential part of controlling the magnetic bearing MBC (short for "magnetic bearing controller").
[0004] The rotational speed is determined by involving other devices, such as speed sensors, motor power signal measurement graphs, or induction-based measurement systems.
[0005] However, these estimation devices have the following drawbacks: additional cost, becoming a source of failure, and insufficient accuracy for certain applications.
[0006] There is a problem with obtaining the rotational speed of the rotor guided by the AMB solely based on the internal signals of the AMB without any additional devices.
[0007] This information is necessary for users for process control and safety reasons.
[0008] This information is also necessary for the AMB control loop in order to perform the synchronous filtering (ABS, AVR, and ODC) as described above. Summary of the Invention
[0009] The subject of this invention is a system for controlling at least one active magnetic bearing in a rotating machine comprising a rotor and a stator. The control system includes: at least one means for measuring the radial position of the rotor relative to the stator based on signals from at least two position sensors; and at least two loops for controlling the active magnetic bearing based on the radial position of the rotor, each control loop of the magnetic bearing being provided with at least one synchronization filter based on the rotor's rotational speed. The system includes means for determining the rotor's rotational speed relative to the stator, which receives a measurement of the rotor's radial position as input from a measuring device and controls the measurement of the rotor's radial position within a predetermined time while the rotor's rotational speed is zero.
[0010] The device used to determine the rotational speed may include an extended Kalman filter.
[0011] The state vector can depend on the rotor's angular position, rotor speed, offset value, and the radius of the track described by the position sensor.
[0012] The extended Kalman filter can be initialized based on the state vector of the average deviation of the rotor's radial position at zero rotational speed and the covariance matrix of the standard deviation of the rotor's radial position at zero rotational speed over a predetermined measurement time.
[0013] In the case where a second active magnetic bearing is provided in the rotating machinery, the device for determining the rotational speed of the rotor can receive a measurement of the radial position of the rotor as input from the measuring device of the second active magnetic bearing, and perform the measurement of the radial position of the rotor within a predetermined time while the rotational speed of the rotor is zero. The device for determining the rotational speed of the rotor is configured to determine the rotational speed of the rotor based on the measurements received from the measuring devices of the two active magnetic bearings.
[0014] In this embodiment, the means for determining the rotational speed may include an extended Kalman filter.
[0015] The extended Kalman filter can be initialized based on the state vector of the average deviation of the rotor's radial position at zero rotational speed and by the covariance matrix of the standard deviation of the rotor's radial position at zero rotational speed within a predetermined measurement time determined by each measuring device.
[0016] The state vector can depend on the rotor's angular position, rotor's rotational speed, offset value, and the radius of the track described by the position sensor, the measuring device of the first active magnetic bearing, as well as the rotor's angular position, rotor's rotational speed, offset value, and the radius of the track described by the position sensor received from the measuring device of the second active magnetic bearing, and the value of the angular phase shift between the two active magnetic bearings, in order to account for possible rotor motion.
[0017] The measuring device can be configured to determine the radial position value of the rotor at the sampling frequency of the MBC control loop, and the device for determining the rotational speed performs extended Kalman filter prediction and updates the operation at the same frequency as the sampling frequency of the control loop of the magnetic bearing.
[0018] Another subject of the present invention is a method for controlling at least one active magnetic bearing of a rotating machine comprising a rotor and a stator. The control method includes at least one step of measuring the radial position of the rotor relative to the stator based on signals from at least two position sensors, and at least one step of controlling the active magnetic bearing based on the radial position of the rotor. The step of controlling the magnetic bearing includes at least one step of synchronous filtering based on the rotational speed of the rotor. The method further includes the steps of: determining the rotational speed of the rotor relative to the stator based on measurements performed during the measurement step, and determining the radial position and radial angular position of the rotor performed within a predetermined time while the rotational speed of the rotor is zero.
[0019] The control system and method according to the present invention have the following advantages: by eliminating the devices traditionally used to determine the rotational speed, the control cost of rotating machinery can be reduced.
[0020] The control system and method according to the invention have the following advantages: in addition to the conventionally used means for determining the rotational speed, a second measurement signal is provided during use. Attached Figure Description
[0021] Other objects, features, and advantages of the invention will become apparent from the following description, given purely by way of non-limiting example, and with reference to the accompanying drawings, in which:
[0022] - Figure 1 The radial position of a point rotated in a fixed orthogonal reference frame is shown, and
[0023] - Figure 2 A control system according to the present invention is shown. Detailed Implementation
[0024] The control system according to the invention processes signals from sensors indicating the radial position of the rotor relative to the stator to determine the rotational speed. Such position sensors are inherently and necessarily part of the magnetic bearing AMB. The control loop uses these signals to control the bearing AMB.
[0025] The control system according to the present invention uses statistical data on the radial position signal of the rotor, combined with an extended Kalman filter, to determine the rotational speed of the rotor.
[0026] Magnetic bearings allow for modification of the shaft's position in each of two directions contained within a plane perpendicular to the shaft. For each direction, the desired displacement is determined and then translated into a command transmitted to two actuators (electromagnets). Each actuator generates an attractive magnetic force on the shaft, drawing it in. The displacement is then obtained based on the difference in attractive forces between the two actuators.
[0027] For magnetic bearings, it is necessary to determine the displacement in one direction and the displacement in another direction, which are usually orthogonal. These displacements are translated into commands for four actuators, two actuators in each direction.
[0028] For the rest of the description, radial position is defined as the coordinates (v, w) of a point rotated in a fixed orthogonal reference frame.
[0029] Values α, β, a, v, w in Figure 1 As shown in the image.
[0030] Figure 2 Rotor 1 is shown, which is based on Figure 1 The angular position of the reference frame (x, y) in the first direction is determined by the measuring device 3a based on a signal from the radial position sensor 2a. The displacement of the magnetic bearing supporting the rotor 1 in the first direction of the reference frame (x, y) is controlled by a setpoint received by the first subtractor 5a. The first subtractor 5a generates an error signal based on the difference between the setpoint and the position measurement determined by the measuring device 3a. The error signal is then filtered by at least one synchronization filter 6a before being transmitted to the two control loops 70a and 71a. Each control loop 70a and 71a determines a position command based on the first direction of the reference frame (x, y) intended for the actuators 8a and 9a of the magnetic bearings, respectively.
[0031] It can also be seen that the rotor 1 is based on Figure 1 The angular position in the second direction of the reference frame (x, y) is determined by the measuring device 3b based on a signal from the radial position sensor 2b. The displacement of the magnetic bearing supporting the rotor 1 in the second direction of the reference frame (x, y) is controlled by a setpoint received by the second subtractor 5b. The second subtractor 5b generates an error signal based on the difference between the setpoint and the position measurement determined by the measuring device 3b. The error signal is then filtered by at least one synchronization filter 6b before being transmitted to the two control loops 70b and 71b. The control loops 70b and 71b determine position commands based on the second direction intended for the actuators 8b and 9b of the magnetic bearing, respectively.
[0032] The determining device 10 enables the determination of a rotor rotation speed signal intended for use in the synchronization filters 6a and 6b based on the radial position determined by the measuring devices 3a and 3b.
[0033] Also seen in the control system according to the invention is an extended Kalman filter (EKF) type determining device 10 receiving, as input, the estimated radial position of the rotor according to the axis x of the reference system connected to the stator and the estimated position of the rotor according to the axis y of the reference system connected to the stator, received from the measuring devices 3a, 3b. It should be noted that the axis z of the orthogonal reference system z(x, y, z) coincides with the rotation axis of the rotor.
[0034] The extended Kalman filter then determines the estimated rotational speed, which is transmitted to each synchronization filter in the control loop of the active magnetic bearing.
[0035] In the first method, to reduce the processing time of EKF, it is recommended to choose the state vector x, which is simplified to its simplest expression:
[0036]
[0037] in:
[0038] rotation angle about axis z
[0039] Rotation speed
[0040] α: Offset based on axis X
[0041] β: Based on the offset of the y-axis
[0042] a: Track radius described by the radial position sensor
[0043] The state model of EKF is simple and linear. It can be represented as:
[0044]
[0045] in:
[0046]
[0047] The desired value of the selective filter used in ABS, AVR, and ODC AMB control modes is in the following form:
[0048]
[0049] The EKF observer model is as follows:
[0050]
[0051] Instead of making assumptions about the dynamic behavior of Ω, α, β, and a, their initial values and the characteristics of noise affecting some of them can be determined when the rotor is at rest (suspended and not rotating).
[0052] It is then suggested that signals v and w be measured for a period of time T in a static state to initialize the state vector.
[0053] Then define the following matrices x0 and P0:
[0054] x0 = [0 0 vw 0] T
[0055]
[0056] in:
[0057] μ v The average deviation of signal v
[0058] μ w The average deviation of signal w
[0059] σ v Standard deviation of signal v
[0060] σ w Standard deviation of signal w
[0061] The mean deviation and standard deviation are determined at the sampling frequency of the MBC control loop.
[0062] It is proposed to perform filter prediction and update operations at the same frequency as the sampling frequency of the MBC control loop.
[0063] In another embodiment, an EKF filter is used to receive, as in the first embodiment, the estimated angular position of the rotor based on the reference frame x connected to the stator, the estimated position of the rotor based on the reference frame y connected to the first AMB measuring device, and the estimated position of the rotor based on the reference frame x connected to the stator, and the estimated position of the rotor based on the reference frame y connected to the second measuring device (which is connected to the second AMB) as input. In practice, typically two AMBs are used to guide the rotating machinery rotor. Therefore, it is suggested that two available measuring devices be redundantly used to supply the EKF filter.
[0064] For all the remaining descriptions, indices 1 and 2, which identify relative values, are defined by the measuring devices of the first AMB and the second AMB, respectively.
[0065] In such an embodiment, the state vector becomes:
[0066]
[0067] Since there is usually a phase shift between the two AMB measurement devices, the phase shift variable is... Introduced into the state vector.
[0068] The state model is as follows:
[0069]
[0070] in:
[0071]
[0072] The observation model is as follows:
[0073]
[0074] For the first embodiment, the state vector and covariance matrix must be initialized using data derived from position measurements in a static state, as well as data that determines their static properties (mean and standard deviation).
[0075] Then, the following initial state vector x0 and covariance matrix P0 are obtained:
[0076] x0 = [0 0 v1 w1 0 v2 v2 0 0] T
[0077]
[0078] in:
[0079] μ v1 The average deviation of signal v1 from the first AMB measuring device
[0080] μ w1 The average deviation of signal w1 from the first AMB measuring device
[0081] σ v1 Standard deviation of signal v1 from the first AMB measuring device
[0082] σ w1 Standard deviation of signal w1 from the first AMB measuring device
[0083] μ v2 The average deviation of signal v2 from the second AMB measuring device
[0084] μ w2 The average deviation of signal w2 from the second AMB measuring device
[0085] σ v2 Standard deviation of signal v2 from the second AMB measuring device
[0086] σ w2 Standard deviation of signal w2 from the second AMB measuring device
[0087] The average value and standard deviation are determined at the sampling frequency of the MBC control loop.
[0088] The filter prediction and update operations are performed at the same frequency as the sampling frequency of the MBC control loop.
Claims
1. A control system for controlling at least one active magnetic bearing in rotating machinery comprising a rotor (1) and a stator, The control system includes: at least one device (3a, 3b) for measuring the radial position of the rotor relative to the stator based on signals from at least two position sensors (2a, 2b); and at least two control loops (7a, 7b) for an active magnetic bearing based on the radial position of the rotor, each control loop of the magnetic bearing being provided with at least one synchronization filter (6a, 6b) based on the rotor rotation speed. Its features are, It includes a device (10) for determining the rotational speed of the rotor relative to the stator, which receives a measurement of the radial position of the rotor as input from a measuring device (3a, 3b) and is based on the measurement of the radial position of the rotor performed within a predetermined time while the rotational speed of the rotor is zero. The rotating machinery is equipped with a second active magnetic bearing. The device for determining the rotor's rotational speed receives a measurement of the rotor's radial position as input from the measuring device of the second active magnetic bearing, and performs the measurement of the rotor's radial position within a predetermined time while the rotor's rotational speed is zero. The device for determining the rotor's rotational speed is configured to determine the rotor's rotational speed based on measurements received from measuring devices of the two active magnetic bearings.
2. The control system according to claim 1, wherein, The device for determining the rotational speed includes an extended Kalman filter.
3. The control system according to claim 2, wherein, The state vector depends on the angular position of the rotor, the rotational speed of the rotor, the offset value, and the radius of the track described by the position sensor.
4. The control system according to any one of claims 2 and 3, wherein, The extended Kalman filter is initialized based on the state vector of the average deviation of the rotor's radial position at zero rotational speed and the covariance matrix of the standard deviation of the rotor's radial position at zero rotational speed over a predetermined measurement time.
5. The control system according to claim 1, wherein, The device for determining the rotational speed includes an extended Kalman filter.
6. The control system according to claim 5, wherein, The extended Kalman filter is initialized by a state vector based on the average deviation of the rotor's radial position at zero rotational speed and by a covariance matrix based on the standard deviation of the rotor's radial position at zero rotational speed over a predetermined measurement time determined by each measuring device.
7. The control system according to claim 6, wherein, The state vector depends on the angular position of the rotor, the rotational speed of the rotor, the offset value, and the radius of the track described by the position sensor, the measurement value of the measuring device of the first active magnetic bearing in the at least one active magnetic bearing, as well as the offset value, the angular position of the rotor, the rotational speed of the rotor, and the radius of the track described by the position sensor received from the measuring device of the second active magnetic bearing in the at least one active magnetic bearing, and the value of the angular phase shift between the two active magnetic bearings, in order to take into account the possible motion of the rotor.
8. The control system according to any one of claims 2 and 3, wherein, The measuring device is configured to determine the value of the radial position of the rotor at the sampling frequency of the control loop, and the device for determining the rotational speed performs extended Kalman filter prediction and updates the operation at the same frequency as the sampling frequency of the control loop of the magnetic bearing.
9. A method for controlling at least one active magnetic bearing in rotating machinery comprising a rotor and a stator. The control method includes at least one step of measuring the radial position of the rotor relative to the stator based on signals from at least two position sensors, and at least one step of controlling the active magnetic bearing based on the radial position of the rotor, wherein the step of controlling the magnetic bearing includes at least one step of performing synchronous filtering based on the rotor's rotational speed. The method also includes the following steps: Based on the measurements performed during the measurement step, the rotational speed of the rotor relative to the stator is determined, and the radial position and radial angular position of the rotor are determined within a predetermined time while the rotor's rotational speed is zero. The rotating machinery is equipped with a second active magnetic bearing. The device for determining the rotor's rotational speed receives a measurement of the rotor's radial position as input from the measuring device of the second active magnetic bearing, and performs the measurement of the rotor's radial position within a predetermined time while the rotor's rotational speed is zero. The device for determining the rotor's rotational speed is configured to determine the rotor's rotational speed based on measurements received from measuring devices of the two active magnetic bearings.