Control system for controlling a magnetic levitation system

Abstract

A control system for controlling a magnetic levitation system comprises a sensor (101) configured to generate a position signal indicative of a position of an object (102) to be magnetically levitated, and a controller (103) configured to control a current supplied to a magnetic actuator of the magnetic levitation system in dependence on the position signal to magnetically levitate the object. The control system comprises a computing system (104) configured to maintain and update a computational model of the magnetic levitation system based on an identification run in which an identification run signal is supplied to the magnetic actuator of the magnetic levitation system and a response to the identification run signal is detected from the sensor and / or the magnetic actuator. The computing system is configured to compare quantities related to the computational model and quantities related to the magnetic levitation system to reveal deviations from expected operating conditions of the magnetic levitation system.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to a control system for controlling a magnetic levitation system, for example, but not necessarily, an active magnetic bearing "AMB" system. Furthermore, the present disclosure relates to a magnetic levitation system. BACKGROUND

[0002] Magnetic levitation systems, for example active magnetic bearing "AMB" systems, are commonly used to levitate (e.g. rotate or oscillate) an object. A typical application is the levitation of a rotor of an electric machine (e.g. a high speed electric machine). In many cases, the levitation is accomplished by balancing the attractive forces of counteracting magnets, at least one of which is a controllable electromagnet, and other forces acting on the object to be levitated. In principle, it is also possible to balance the attractive force of one controllable electromagnet against other forces (e.g. gravitational forces) acting against the attractive force of the electromagnet. There are several different types of magnetic levitation systems. Some systems use permanent magnet material to generate a bias magnetic flux, while others use a direct bias current to generate a bias magnetic flux. The bias serves to linearize the operation of the system and to improve the control dynamics of the system.

[0003] Due to the inherent instability of magnetic levitation, active control of the magnetic forces acting on the degrees of freedom of the levitated object (e.g. a rotor of an electric machine) is required. This instability is caused by the fact that the magnetic attractive force acting between a magnet and an object made of e.g. ferromagnetic material increases when the air gap between the magnet and the object decreases. The control system ensures the stability of the magnetic levitation system (e.g. an active magnetic bearing "AMB" system) by providing an appropriate control algorithm. In combination with the AMB system, the control system captures the position of the rotor in three dimensions by a set of sensors at each end of the rotor. With this information, the control system estimates the necessary forces to be exerted on the rotor by supplying appropriate currents to the coils of the AMB system. Furthermore, angle sensors can be used to estimate the rotor angle and / or the rotational speed around the geometric axis of rotation.

[0004] There are several challenges for rotating machines, for example unbalance, changes in rotor dynamics and changes in air gap values due to heating and rotational expansion. Furthermore, it can be necessary to monitor the system integrity and health. One straightforward approach is for the manufacturer to select certain parameters to be monitored, which are typically defined by a vibration level, a maximum current level and / or a maximum unbalance. If the magnetic levitation system exceeds one or more of the allowed boundaries, a fault alarm or warning is issued. However, this information is very limited and only provides a notification when the magnetic levitation system has already exceeded the one or more allowed boundaries. SUMMARY

[0005] The following presents a simplified summary of some aspects of the various inventive embodiments in order to provide a basic understanding of such aspects. This summary is not an extensive overview of the application. It is intended to neither identify key or critical elements of the application nor delineate the scope of the application. Its sole purpose is to present some concepts of the application in a simplified form as a prelude to the more detailed description of the exemplary embodiments of the application.

[0006] In this document, the word "geometric" when used as a prefix refers to geometric concepts that are not necessarily part of any physical object. For example, a geometric concept can for example be a geometric point, a geometric straight line or a geometric curve, a geometric plane, a non-planar geometric surface, a geometric space, or any other geometric entity of zero, one, two or three dimensions.

[0007] According to the present invention, there is provided a new control system for controlling a magnetic levitation system, for example but not necessarily a so-called active magnetic bearing "AMB" system for levitating a rotating element, for example a rotor of an electric machine.

[0008] The control system according to the present invention comprises:

[0009] - a sensor configured to generate a position signal indicative of a position of an object to be magnetically levitated,

[0010] - a controller configured to supply a current to coils of magnetic actuators of the magnetic levitation system in dependence on the position signal to magnetically levitate the object, and

[0011] - a computing system configured to maintain and update a computational model of the magnetic levitation system based on an identification run in which an identification run signal is supplied to the coils of the magnetic levitation system and a response to the identification run signal is detected from at least one of the sensor and the coils, wherein the computing system is configured to compare quantities related to the computational model and quantities related to the magnetic levitation system to reveal differences of the magnetic levitation system with respect to the computational model.

[0012] In many cases, the control system described above can already produce an indication of an anomaly in the magnetic levitation system when the anomaly has not yet progressed to a significant impact on the operation of the magnetic levitation system, because even if the anomaly has not yet had an impact on the operation, the differences of the magnetic levitation system with respect to the computational model can typically be discerned. Thus, an early warning is available.

[0013] Furthermore, the algorithm used in the control system described above can be further developed to produce balancing data for the automatic or semi-automatic balancing of a rotor of an active magnetic bearing "AMB" system, for example. The algorithm can be configured to propose, for example, the best combination of weights in the rotor and holes into which these weights should be placed for the next advancement in the balancing process.

[0014] According to the present application, there is also provided a new magnetic levitation system comprising:

[0015] - an object to be magnetically levitated,

[0016] - a magnetic actuator configured to magnetically levitate the object, and

[0017] - a control system according to the present application for controlling the electric current supplied to the coils of the magnetic actuator.

[0018] Exemplary and non-limiting embodiments are described in the appended dependent claims.

[0019] Various exemplary and non-limiting embodiments in relation to structure and method of operation, and additional objects and advantages thereof, will be best understood from the following description of specific exemplary and non-limiting embodiments, when read in connection with the accompanying drawings.

[0020] The verbs "comprise" and "include" are used in this document as open limitations that neither exclude nor require the existence of also un-recited features. The features recited in the dependent claims are mutually freely combinable. Moreover, it should be understood that the use of "a" or "an", e.g. "a feature" or "an element" in the description indicates there can be one, but there can also be more than one. BRIEF DESCRIPTION OF DRAWINGS

[0021] The exemplary and non-limiting embodiments and the advantages thereof, together with additional objects, will be best understood from the following description, when read in connection with the accompanying drawings, wherein:

[0022] Figure 1a 、 Figure 1b and Figure 1c a magnetic levitation system comprising a control system according to one exemplary and non-limiting embodiment is shown, and

[0023] Figure 2 an equivalent circuit of a part of a magnetic levitation system according to one exemplary and non-limiting embodiment is shown. DETAILED DESCRIPTION

[0024] The specific examples provided in the following description should not be construed as limiting the scope and / or applicability of the appended claims. Lists and groups of alternatives provided in the description are not exhaustive unless otherwise explicitly stated.

[0025] Figure 1a A magnetic levitation system comprising a control system is shown according to one example and non-limiting embodiment. The magnetic levitation system comprises magnetic actuators 105 and 106 configured to magnetically levitate an object 102. In this example case, the magnetic levitation system is an active magnetic bearing "AMB" system, and the object 102 is a rotating element, e.g. a rotor of an electric machine. The magnetic actuators 105 are radial magnetic bearings, and the magnetic actuators 106 are axial magnetic bearings.

[0026] The control system comprises sensors 101 for producing position signals indicative of the position of the object 102 relative to a reference position of the object 102. Figure 1b An example sensor arrangement around the object 102 is shown. For example, the sensors 101 can comprise inductive sensors, where the inductance of each inductive sensor depends on the distance from the considered inductive sensor to the surface of the object 102. There can also be means for forming the position signals based on the difference between the inductances of the coils of the magnetic actuators 105 and / or 106. The inductance of each coil can be estimated, e.g. based on the rate of change of the current di / dt when the voltage directed to the considered coil is changed stepwise. In this example case, no separate sensors are needed, but the sensors are implemented by means of the magnetic actuators 105 and / or 106.

[0027] The control system comprises a controller 103 configured to supply currents to the coils of the magnetic actuators 105 and 106 in dependence on the position signals to magnetically levitate the object 102. The control system comprises a computing system 104 configured to maintain and update a computational model of the magnetic levitation system based on identification runs in which identification run signals are supplied to the coils of the magnetic levitation system and the responses to the identification run signals are detected from the sensors and / or the coils of the magnetic actuators. The computing system 104 is configured to compare the real-time output of the magnetic levitation system with the expected output of the computational model to assess deviations. In this way, information is obtained that can give insight into the root cause of possible problems. In many cases, the computing system 104 can already produce an indication about an anomaly in the magnetic levitation system when the anomaly has not yet developed into a significant impact on the operation, since the difference of the magnetic levitation system from the computational model can typically be discerned even if the anomaly has not yet had an impact on the operation. Thus, early warnings are available.

[0028] To keep the above-mentioned computational model up to date, the computing system 104 applies an identification algorithm by exciting the magnetic levitation system via the magnetic actuators 105 and 106 and collecting output information via the sensors 101. It is also possible for the computing system 104 to collect output information via the coils of the magnetic actuators. The excitation advantageously takes place through several predefined patterns and for a number of operating points and conditions. The model parameters of the computational model under the respective conditions are fitted to the various acquired data to keep the computational model in sync with the magnetic levitation system. The identification runs for the parameter fitting can take place, for example, during initial commissioning, then periodically during the service life, or upon occurrence of a specified event and triggering of the identification runs. The identification runs for the parameter fitting can be triggered by a set of events or when operating parameters, for example, the current of the magnetic levitation system, exceed a specified range.

[0029] Advantageously, in addition to the above-described tasks for generating early warnings about possible abnormalities, the above-mentioned computational model can also be used for several other purposes. For example, a high-speed balancing procedure is discussed below. Figure 1a The object 102 shown in the middle has several balancing planes, which are perpendicular to the z-axis of the coordinate system 199 and are provided by inserting additional weights into specially prepared holes 107. Angular position sensors are advantageously synchronized with the circumferential spacing of the holes 107, and a balancing algorithm within the computing system 104 can be configured to estimate the next balancing step in the balancing procedure. The next balancing step can be estimated based on the unbalance vector, the previous balancing state, and the angular position of the object 102 with the computational model containing the balancing plane positions. The algorithm proposes the optimal combination of weights and holes into which the weights should be placed for the next advance in the balancing process. In this way, balancing can be performed in an automated manner, with reduced personnel involvement, thus eliminating the risk of human error. The computational model can be arranged to keep stored balancing steps so that if, for some reason, rebalancing is needed after a period of time, rebalancing can be done not only by adding weights but also by removing weights added earlier.

[0030] The control system according to exemplary and non-limiting embodiments comprises a differential sensor arrangement for both radial and axial measurements. The differential sensor arrangement provides the possibility to estimate the expansion and contraction of the object 102 due to rotational speed and / or temperature variations. Information from temperature sensors within the magnetic levitation system provides input to the computational model to assess the temperature of the object 102 as well as the magnetic actuators 105 and 106. Thus, in this exemplary case, the effects of expansion / contraction from two sources (rotation and temperature variations) can be taken into account in the control algorithm. As the length and diameter of the object 102 varies, the effective length of the magnetic air gap changes accordingly, affecting the force gain coefficient. Consequently, the gain of the feedback law is also advantageously adjusted, respectively, to maintain the robustness and stability of the magnetic levitation system. Figure 1b The expansion and contraction of the object 102 in the radial direction is shown, the change in radius in the x-direction of the coordinate system 199 is denoted by Ax. Figure 1c The expansion and contraction of the object 102 in the axial direction is shown, the change in axial length is denoted by Az.

[0031] In the control system according to exemplary and non-limiting embodiments, the computing system 104 is configured to estimate the forces acting on the magnetic levitation object 102 based on the computational model and data measured from the magnetic levitation system. The computing system 104 can be configured to recognize one or more predetermined behavior patterns of the estimated forces and to predict maintenance needs depending on the recognized behavior patterns of the estimated forces. Based on the computational model and the measured data, the forces acting on the object 102 can be estimated as well as their frequency and magnitude. This information can be used to assess the health status of the magnetic levitation system and of the process it is involved in. A given force pattern can be indicative of, for example, mechanical wear of the impeller and / or the presence of contamination in the process. The control algorithm can be configured to distinguish between predefined behaviors and to predict maintenance needs and their schedule based on the computational model.

[0032] In the control system according to exemplary and non-limiting embodiments, the computational model of the magnetic levitation system comprises a first model parameter modeling the inductance of the coils as a function of the position signal. In the control system according to exemplary and non-limiting embodiments, the computational model of the magnetic levitation system comprises a second model parameter modeling the resistance of the coils. In the control system according to exemplary and non-limiting embodiments, the computational model of the magnetic levitation system comprises a third model parameter modeling eddy current and hysteresis losses in the ferromagnetic parts of the magnetic levitation system.

[0033] Figure 2 An equivalent circuit of a part of the magnetic levitation system according to exemplary and non-limiting embodiments is shown. The circuit 210 models a part of the magnetic actuators, and the circuit 220 models the object 102. Figures 1a to 1cA portion of the object 102 is modeled. The mutual inductance M(5) depends on the air gap 5 between the portion of the magnetic actuator and the portion of the object 102. Li is the inductance of the respective coil of the magnetic actuator, and L2 models the ability of the induced current in the object 102 to generate a magnetic flux. The resistance Rs describes the resistance of the respective coil of the magnetic actuator, the resistance Rmi describes eddy current and hysteresis losses in the ferromagnetic portion of the magnetic actuator, and the resistance Rm2 describes eddy current and hysteresis losses in the object 102. The magnetic flux linkage across the air gap can be estimated as M = M(5)(Ii - I2), and the magnetic force acting between the portion of the magnetic actuator and the portion of the object 102 is proportional to M 2 If the object 102 is made of an ideal, lossless material, Rm2 is infinite, and M = M(5)Ii.

[0034] For example, Figure 1a The illustrated controller 103 can comprise transistors, e.g. insulated gate bipolar transistors “IGBTs” or metal oxide semiconductor field effect transistors “MOSFETs”, configured to constitute, e.g., a full or half H-bridge, and driver circuits for driving the transistors. Furthermore, the controller 103 can comprise analog and / or digital circuits configured to constitute a regulator for controlling the driver circuits based on the position signal. Figure 1a The illustrated computing system 104 can comprise one or more processor circuits, each of which can be a programmable processor circuit, e.g. a programmable digital signal processor “DSP” or a programmable microcontroller unit “MCU”, provided with appropriate software, a dedicated hardware processor, e.g. an application specific integrated circuit “ASIC”, or a configurable hardware processor, e.g. a field programmable gate array “FPGA”. Furthermore, the computing system 104 can comprise one or more memory devices, e.g. random access memory “RAM” devices.

[0035] The specific examples provided in the description given above should not be construed as limiting the scope and / or applicability of the appended claims. Lists and groups of examples provided in the description given above are not intended to be exhaustive unless otherwise explicitly stated.

Claims

1. A control system for controlling a magnetic levitation system, the control system comprising: - Sensor (101), the sensor being configured to generate a position signal indicating the position of the object to be magnetically levitized, and - A controller (103), configured to supply current to the coils of the magnetic actuators (105, 106) of the magnetic levitation system according to the position signal, so as to magnetically levitate the object. The control system is characterized in that it includes a computing system (104) configured to maintain and update a computational model of the magnetic levitation system based on an identification run, wherein an identification run signal is supplied to the coils of the magnetic levitation system and a response to the identification run signal is detected from at least one of the sensors and the coils, wherein the computing system is configured to compare quantities related to the computational model with quantities related to the magnetic levitation system to reveal differences between the magnetic levitation system and the computational model.

2. The control system according to claim 1, wherein, The computing system (104) is configured to periodically update the computing model.

3. The control system according to claim 2, wherein, The computing system (104) is configured to update the computing model in response to predetermined events during the operation of the magnetic levitation system or based on parameters measured on the magnetic levitation system.

4. The control system according to claim 1, wherein, The computational model of the magnetic levitation system includes a first model parameter, which depends on the position signal modeling the inductance (L1, L2, M) of the coil.

5. The control system according to claim 2, wherein, The computational model of the magnetic levitation system includes a first model parameter, which depends on the position signal modeling the inductance (L1, L2, M) of the coil.

6. The control system according to claim 3, wherein, The computational model of the magnetic levitation system includes a first model parameter, which depends on the position signal modeling the inductance (L1, L2, M) of the coil.

7. The control system according to any one of claims 4 to 6, wherein, The computational model of the magnetic levitation system includes a second model parameter, which models the resistance (Rs) of the coil.

8. The control system according to claim 7, wherein, The computational model of the magnetic levitation system includes a third model parameter (Rm1, Rm2), which models the eddy currents and hysteresis losses in the ferromagnetic part of the magnetic levitation system.

9. The control system according to any one of claims 1 to 6, wherein the computing system is configured to estimate the force acting on the magnetically levitated object based on the computing model and data measured from the magnetic levitation system, the computing system is configured to identify one or more predetermined behavior patterns of the estimated force, and the computing system is configured to predict maintenance needs depending on the identified behavior patterns of the estimated force.

10. A magnetic levitation system, comprising: -The object to be magnetically levitated (102), -Magnetic actuators (105, 106), said magnetic actuators being configured to magnetically levitate the object, and - Control system (101, 103, 104), the control system being used to control the current supplied to the coil of the magnetic actuator. The control system is characterized in that it is a control system according to any one of claims 1 to 6.

11. The magnetic levitation system according to claim 10, wherein, The magnetic actuator (105) includes a radial magnetic bearing.

12. The magnetic levitation system according to claim 10, wherein, The magnetic actuator (106) includes an axial magnetic bearing.

13. The magnetic levitation system according to claim 11, wherein, The magnetic actuator (106) includes an axial magnetic bearing.

Images