JP7382864B2 - Magnetic bearing device and low temperature liquefied gas pump - Google Patents

Description

The present invention relates to a magnetic bearing device and a low-temperature liquefied gas delivery pump equipped with the same.

A magnetic bearing is a bearing that levitates and supports a rotating body using magnetism, and is used in rotating machines under special environments. Magnetic bearings provide non-contact support, so no lubricating oil is required, and the life of the bearing is semi-permanent. Since no lubricating oil is required, it is possible to operate in vacuum environments and ultra-low temperature environments, which are difficult to do with conventional bearings.

Such a magnetic bearing is disclosed in, for example, Japanese Unexamined Patent Publication No. 2013-57250.

JP2013-57250A

Rolling bearings such as ball bearings require bearing maintenance at regular intervals, and cannot be used in low-temperature liquefied gas pumps used for cooling superconducting equipment, which require continuous operation for long periods of time. Therefore, Japanese Unexamined Patent Publication No. 2013-57250 discloses a device that employs a magnetic bearing that does not require regular maintenance as a bearing of a low-temperature liquefied gas pump.

In order for magnetic bearings to cope with disturbances occurring in the impeller, it is generally necessary to always apply a constant current (bias current) corresponding to the maximum load to the electromagnetic coil of the magnetic bearing. In a high discharge pressure pump, a large load is applied in the radial direction of the impeller due to the pressure balance around the impeller, so it is necessary to increase the load capacity. For this reason, it was necessary to increase the bias current of the magnetic bearing, and this resulted in increased power consumption and heat generation in the electromagnet section.

The present invention was made to solve the above problems, and its purpose is to provide a magnetic bearing device that rotates stably while reducing the amount of heat generated, and a low-temperature liquefied gas pump equipped with the magnetic bearing device. That's true.

In summary, the present invention is a magnetic bearing device that supports a rotating shaft, and includes a first electromagnet and a second electromagnet arranged to sandwich the rotating shaft, a sensor that detects displacement of the rotating shaft, and a first electromagnet. and a calculation unit that calculates the current flowing through the second electromagnet. The calculation unit (a) calculates control currents of the same value with opposite signs for the first electromagnet and the second electromagnet based on the displacement detected by the sensor, and (b) calculates control currents of the same value with opposite signs for the first electromagnet and the second electromagnet according to the increase/decrease in the control current. A bias current having the same sign and value is calculated for the second electromagnet, and (c) the control current and the bias current are added to calculate a coil current to be applied to each of the first electromagnet and the second electromagnet.

Preferably, a lower limit value is defined for the bias current. The calculation unit increases the bias current and decreases the DC component of the control current when the DC component of the control current increases and the DC component of the control current becomes larger than a first threshold value set in advance. However, when the DC component of the control current becomes smaller than a preset second threshold, the bias current is reduced within a range that does not fall below the lower limit.

More preferably, the first threshold is set larger than the second threshold, and the bias current increase/decrease characteristic has hysteresis.

Preferably, a lower limit value is defined for the bias current. The calculation unit increases the bias current and reduces the AC component of the control current when the amplitude of the AC component of the control current increases and the amplitude of the AC component of the control current becomes larger than a preset third threshold. When the amplitude of the AC component of the control current decreases and the amplitude of the AC component of the control current becomes smaller than a preset fourth threshold value, the bias current is decreased within a range that does not fall below the lower limit.

More preferably, the third threshold is set larger than the fourth threshold, and the bias current increase/decrease characteristic has hysteresis.

Preferably, the arithmetic unit changes the bias current stepwise in a preset increase/decrease range.

Preferably, the calculation unit calculates the control current by performing PID control (proportional-integral-derivative control), and updates the control parameters of the PID control at the same time as updating the bias current.

More preferably, the control parameter is a gain of a loop transfer function of PID control. When the bias current exceeds a preset judgment value and the bias current is increased, the calculation section increases the gain and increases the bias current when the bias current exceeds a preset judgment value and increases the bias current. If it is decreased, the gain is decreased.

More preferably, the control parameter is an integral gain of PID control. If the bias current exceeds a preset judgment value and the bias current is increased, the calculation unit decreases the integral gain, and if the bias current exceeds a preset judgment value and increases the bias current, If the value is decreased, increase the integral gain.

In another aspect, the present invention is a low-temperature liquefied gas liquid pump including any of the above magnetic bearing devices.

According to the present invention, since the bias current is controlled to an appropriate value, stable rotation can be achieved while suppressing the amount of heat generated.

FIG. 2 is a diagram showing the configuration of a first example of a pump device that uses a magnetic bearing to pump low-temperature liquefied gas. It is a figure showing the composition of the pump device of the second example which sends low-temperature liquefied gas using a magnetic bearing. FIG. 3 is a diagram showing the structure of an axial magnetic bearing in a cross section including a rotating shaft. FIG. 2 is a diagram showing the structure of a radial magnetic bearing (homopolar magnetic bearing) in a cross section perpendicular to the rotation axis. FIG. 2 is a diagram showing the structure of a radial magnetic bearing (homopolar magnetic bearing) in a cross section including a rotating shaft. FIG. 2 is a diagram showing the structure of a radial magnetic bearing (heteropolar magnetic bearing) in a cross section perpendicular to the rotation axis. FIG. 2 is a diagram showing the structure of a radial magnetic bearing (heteropolar magnetic bearing) in a cross section including a rotating shaft. FIG. 3 is a diagram showing the relationship between coil current and magnetic force of an electromagnet. It is a figure which shows the relationship between the coil current of an electromagnet and magnetic force when the electromagnet is made to oppose on both sides of a rotating shaft. FIG. 2 is a control block diagram for explaining control using a bias current method of the magnetic bearing device. It is a graph for explaining a state in which the relationship between coil current and magnetic force is linearized by bias magnetic flux. It is a graph for explaining the relationship between coil current and magnetic force when changing bias current. It is a graph showing the relationship between bias current and rigidity of a magnetic bearing. 7 is a graph showing that the transfer function of the controlled object changes depending on the magnitude of the bias current. FIG. 1 is a block diagram of a magnetic bearing system. 16 is a Bode diagram showing the frequency characteristics of the open loop transfer function of the magnetic bearing system of FIG. 15. FIG. 17 is a Bode diagram showing the frequency characteristics of the open loop transfer function when the bias current is increased from the state of FIG. 16. FIG. 18 is a Bode diagram showing the frequency characteristics of the open-loop transfer function when the overall gain is increased from the state shown in FIG. 17. FIG. 18 is a Bode diagram showing the frequency characteristics of the open loop transfer function when the integral gain is lowered from the state of FIG. 17 to reduce the influence on the phase on the low frequency side. FIG. It is a flowchart of the main processing of a magnetic bearing. 21 is a flowchart of a process for determining a current command value in S2 of FIG. 20. 22 is a flowchart showing a first example of a process for determining a bias current command value in S12 of FIG. 21. FIG. 22 is a flowchart showing a second example of the process of determining the bias current command value in S12 of FIG. 21. FIG. FIG. 7 is a block diagram corresponding to a second example of processing for determining a bias current command value.

Embodiments of the present invention will be described below with reference to the drawings. In the following drawings, the same or corresponding parts are given the same reference numerals, and the description thereof will not be repeated.

[Overall configuration of pump device]
FIG. 1 is a diagram showing the configuration of a first example of a pump device that uses a magnetic bearing to pump low-temperature liquefied gas. FIG. 2 is a diagram showing the configuration of a second example of a pump device that uses a magnetic bearing to pump low-temperature liquefied gas.

The low-temperature liquefied gas is, for example, liquid nitrogen. The pump 10 shown in FIG. 1 has a configuration in which the electromagnet part is separated from the impeller 40, and the pump 10A shown in FIG. 2 has a configuration in which the electromagnet part is close to the impeller 40. The pump 10A is a type of pump called a submerged pump in which the pump housing is immersed in a low-temperature liquid. The pump 10 and the pump 10A are attached to the top plate or lid of a container storing low-temperature liquefied gas, and suck up and send out the liquid in the container.

Referring to FIGS. 1 and 2, pump 10 and pump 10A include an impeller 40, a pump casing 50, a drive unit (motor) 30, a rotating shaft 20, and magnetic bearings 60, 70, and 80.

Impeller 40 is housed in pump casing 50. The pump casing 50 is provided with a suction port at the bottom and a discharge port on the side of the rotating shaft. The impeller 40 is attached to the lower end of the rotating shaft 20. The rotating shaft 20 is rotationally driven by a driving section 30.

The magnetic bearings 60, 70, and 80 support the rotating shaft 20 in a non-contact state by magnetic force. The magnetic bearing 60 is an axial bearing that supports the axial position, the magnetic bearing 70 is a radial bearing on the anti-pump side, and the magnetic bearing 80 is a radial bearing on the impeller side.

Both of the configurations shown in FIGS. 1 and 2 are configured to prevent heat from the motor from being transmitted to the impeller. The pump is a centrifugal pump type, and when the impeller 40 rotates, it sucks in liquid from below and discharges it from a discharge port on the outside in the radial direction.

In a pump that pumps low-temperature liquefied gas as shown in FIGS. 1 and 2, it is desirable to minimize the intrusion of heat into the liquefied gas section. Therefore, it is important to improve heat insulation performance and reduce heat generation in the bearing and motor.

When a magnetic bearing is used as a bearing, it is common to set the magnitude of the bias current based on the expected maximum load. At high flow rates, the fluid force generated on the impeller is different. If the bias current is determined according to high discharge pressure and low flow rate, at low discharge pressure and high flow rate, the difference in pressure balance in the pump chamber is small and the disturbance applied to the impeller is small, so an extra bias current will be passed. .

Since the bias current increases the copper loss of the magnetic bearing, it is desirable that the bias current be small in order to reduce heat generation. However, reducing the bias current reduces the support rigidity of the magnetic bearing. Conventionally, optimization according to the operating point of the pump was not possible, but in this embodiment, the state of disturbance is detected from the DC component and AC component of the displacement and current command value, and the bias current is adjusted according to the disturbance. Stabilize control by changing the In addition, since changing the bias current changes what is being controlled, it is possible to maintain stability over a wide operating range by determining the bias current control range in advance and switching control system parameters as necessary within the bias current control range. Levitation control becomes possible.

[Configuration of each magnetic bearing]
FIG. 3 is a diagram showing the structure of the axial magnetic bearing in a cross section including the rotation axis. Referring to FIG. 3, magnetic bearing 60 includes stator cores 62 and 64 that sandwich thrust disk 61 fixed to rotating shaft 20, stator coils 63 and 65, and a gap sensor 66. A magnetic flux is generated as shown by the broken line by the current flowing through the stator coil 63 and the current flowing through the stator coil 65, and the balance of the magnetic force that attracts the thrust disk 61 can be changed. By controlling the current flowing through the stator coil 63 and the current flowing through the stator coil 65, the axial position of the thrust disk 61, that is, the axial position of the rotating shaft 20 is controlled.

FIG. 4 is a diagram showing the structure of a radial magnetic bearing (homopolar magnetic bearing) in a cross section perpendicular to the rotation axis. FIG. 5 is a diagram showing the structure of a radial magnetic bearing (homopolar magnetic bearing) in a cross section including the rotating shaft. Referring to FIGS. 4 and 5, magnetic bearing 70 includes stator cores 72 and 74 sandwiching rotating shaft 20, stator coils 73 and 75, and a gap sensor (not shown). Stator core 72 and stator coil 73 constitute a first electromagnet M1. Stator core 74 and stator coil 75 constitute a second electromagnet M2. A magnetic flux is generated as shown by the broken line by the current flowing through the stator coil 73 and the current flowing through the stator coil 75, and the balance of the magnetic forces of the first electromagnet M1 and the second electromagnet M2 that attract the rotating shaft 20 can be changed. By controlling the current flowing through stator coil 73 and the current flowing through stator coil 75, the radial position of rotating shaft 20 is controlled. In addition, the third electromagnet M3 and the fourth electromagnet M4 are similarly arranged at positions rotated by 90 degrees from the first electromagnet M1 and the second electromagnet M2, and the position of the rotating shaft 20 in the radial direction rotated by 90 degrees is controlled.

Note that a heteropolar magnetic bearing may be used instead of the magnetic bearing 70. FIG. 6 is a diagram showing the structure of a radial magnetic bearing (heteropolar magnetic bearing) in a cross section perpendicular to the rotation axis. FIG. 7 is a diagram showing the structure of a radial magnetic bearing (heteropolar magnetic bearing) in a cross section including the rotating shaft. Referring to FIGS. 6 and 7, magnetic bearing 70A includes stator teeth 72A and 74A that sandwich rotating shaft 20, stator coils 73A and 75A, and a gap sensor (not shown). A magnetic flux is generated as shown by the broken line by the current flowing through the stator coil 73A and the current flowing through the stator coil 75A, and the balance of the magnetic force that attracts the rotating shaft 20 can be changed. The radial position of rotating shaft 20 is controlled by controlling the current flowing through stator coil 73A and the current flowing through stator coil 75A.

FIG. 8 is a diagram showing the relationship between the coil current of the electromagnet and the magnetic force. As shown in FIG. 8, the magnetic force F that attracts the rotating shaft increases in proportion to the square of the coil current i. The magnetic force F is a quadratic function of the electromagnet coil current i, as shown in equation (1).

In equation (1), B represents the magnetic flux density, S represents the cross-sectional area of the magnetic path, N represents the number of coil turns, i represents the coil current, x represents the gap length, and μ ₀ represents the vacuum Indicates magnetic permeability.

FIG. 9 is a diagram showing the relationship between the coil current and magnetic force of the electromagnet when the electromagnets are opposed to each other with the rotation axis in between. The configuration shown in FIG. 8 is made into a facing type as shown in FIG. In the range of i > 0, ic2 = 0, in the range of i < 0, ic1 = 0, and if current is applied only to the coil of one of the opposing electromagnets, the composite magnetic force of magnetic forces F(ic1) and F(ic2) becomes non-linear, and even if the current is increased, the magnetic force F does not increase much in the vicinity of i=0. Therefore, in the very vicinity of i=0, even if a current is applied when the rotational axis is deviated from the center, the magnetic force is small, and the force for correcting the deviation is weak, that is, the rigidity is low.

In the configuration shown in FIG. 9, the slope near zero current is small, resulting in low rigidity, and due to nonlinearity, a Bode diagram used to verify the stability of the control system cannot be used. Furthermore, in this case, when a load occurs and a control current flows through the coil, the controlled object, which consists of the magnetic bearing and the rotating shaft supported by the magnetic bearing, changes significantly, making it difficult to ensure stability. It is.

[Control of magnetic bearing device]
In order to increase the rigidity near zero current, a bias current is passed through the coil.

FIG. 10 is a control block diagram for explaining control using a bias current method of a magnetic bearing device. Referring to FIG. 10, the magnetic bearing device 300 includes a first electromagnet M1 and a second electromagnet M2 arranged to sandwich the rotating shaft 20, displacement sensors 121 and 122 that detect the displacement of the rotating shaft 20, and a second electromagnet M2. The electromagnet M1 includes a calculation unit 100 that calculates the current flowing through the first electromagnet M1 and the second electromagnet M2. The calculation unit 100 (a) calculates control currents ic with opposite signs and the same value for the first electromagnet M1 and the second electromagnet M2 based on the displacements detected by the displacement sensors 121 and 122, and (b) calculates the control current ic of the control current ic. According to the increase/decrease, calculate the bias current ib with the same sign and the same value for the first electromagnet M1 and the second electromagnet M2, (c) add the control current ic and the bias ib current, and calculate the bias current ib of the first electromagnet M1 and the second electromagnet M2 Coil currents ic1 and ic2 to be applied to each of M2 are calculated.

Displacement from the center of the rotating shaft 20 supported by the magnetic bearing 70 is detected by displacement sensors 121 and 122 and input to the sensor circuit 120. The sensor circuit 120 outputs a control amount corresponding to the deviation of the rotation axis from the center to the calculation section 100. Upon receiving this, the calculation section 100 outputs coil current command values to the amplifiers 111 and 112, respectively. The amplifier 111 supplies a coil current ic1 to the stator coil 73 of the magnetic bearing, and the amplifier 112 supplies a coil current ic2 to the stator coil 75 facing the stator coil 73 of the magnetic bearing.

The calculation unit 100 receives the output of the sensor circuit 120, performs gain adjustment, phase compensation, and filter processing, and outputs a control current command value Vic, and a controller 101 that outputs a control current command value Vic. It includes a bias current setting section 102 that outputs Vib, an adder 103, and a subtracter 104.

Adder 103 outputs coil current command value Vic1, which is the sum of control current command value Vic and bias current command value Vib, to amplifier 111, and amplifier 111 outputs coil current ic1 (=ib+ic) to stator coil 73. Subtractor 104 outputs coil current command value Vic2 obtained by subtracting control current command value Vic from bias current command value Vib to amplifier 112, and amplifier 112 outputs coil current ic2 (=ib-ic) to stator coil 75. .

In a typical magnetic bearing, the bias current ib is constant, and the shaft is supported at the center of rotation by flowing a control current ic according to disturbance.

FIG. 11 is a graph for explaining a state in which the relationship between coil current and magnetic force is linearized by bias magnetic flux. When bias current ib is applied to the coil from the state shown in FIG. 9, graph F(ic1) shifts to the left and graph F(ic2) shifts to the right. Therefore, the relationship between the coil current and the composite magnetic force F is linearized by the bias magnetic flux. If the magnetic force generated by the electromagnet M1 is F1, and the magnetic force generated by the electromagnet M2 is F2, F1 and F2 are in opposite directions, and the composite magnetic force Ft is expressed as follows from equation (1). Strictly speaking, the gap x is x0 equally for electromagnets M1 and M2 at the balanced position, and if the deviation from the balanced position is Δx, then x = x0 + Δx for electromagnet M1 and x0 - Δx for electromagnet M2. However, in the region near the origin, Δx=0, and the gap lengths were the same for the electromagnets M1 and M2.
Ft=F1-F2=k(ib+ic) ² /x0 ² -k(ib-ic) ² /x0 ²
Ft=k・4ib・ic/x0 ² ∝ic
Therefore, the resultant force F becomes a linear function proportional to the control current ic in the range -ib<ic<ib.

In the calculation unit 100 of this embodiment shown in FIG. 10, the bias current setting unit 102 changes the bias current command value Vib according to the control current command value Vic. Therefore, the state of linearization of magnetic force also changes.

FIG. 12 is a graph for explaining the relationship between coil current and magnetic force when changing the bias current. By changing the bias current to ib1, ib2, and ib3, the bias magnetic flux also changes. When the bias magnetic flux is increased, the linear range of the generated force with respect to the coil current expands, as shown in FIG. 12, and the slope of the force with respect to the coil current also increases. When the magnitude of the bias current is set to ib1<ib2<ib3, the magnitude of the composite magnetic force Ft also increases as Ft1<Ft2<Ft3. The larger the bias current ib, the stronger the rigidity, and the greater the force that pulls the rotating shaft back to the center of rotation when displacement occurs.

As can be seen from FIGS. 11 and 12, it is possible to linearize the relationship between the coil current and magnetic force up to twice the bias current. In the bias current method, which applies a bias magnetic flux by applying the same current to the coils of opposing electromagnets, the bias current is varied according to the magnitude of the load generated on the rotating shaft. Generally, the bias current is set to a fixed value that is predetermined in consideration of the maximum load, but in this embodiment, it is changed according to the load, so the bias current becomes smaller under operating conditions with a small load, which reduces the copper loss of the coil. Iron loss of the rotating shaft during rotation can be reduced and heat generation can be suppressed.

The load (for example, fluid force) applied to the rotating shaft 20 can be estimated from the coil currents ic1 and ic2 of the electromagnets and the outputs of the displacement sensors 121 and 122. Since the coil currents ic1 and ic2 are equivalent to the coil current command values Vic1 and Vic2 which are the outputs of the calculation section 100, the calculation section 100 can know the coil currents ic1 and ic2. Since the electromagnet specifications of the magnetic bearings 70 and 80 are known, the load on the magnetic bearings can be estimated if the current value is known. Furthermore, since the arrangement of the bearings and the position of the impeller 40 are known, the radial load applied to the impeller 40 can also be estimated. The bias current ib may be varied at the same value for each radial magnetic bearing (one set of four electromagnets (M1 to M4)), or may be varied for each X-axis and Y-axis (two electromagnets (M1, M2) facing each other). 1 set, or 2 opposing electromagnets (M3, M4) may be used as 1 set).

Furthermore, even when the load is large, the current can be increased up to a preset maximum current range, and improvements in control stability and rigidity can be expected by expanding the linear range.

FIG. 13 is a graph showing the relationship between bias current and rigidity of a magnetic bearing. FIG. 13 shows the bearing rigidity when a displacement sensor is attached to the magnetic bearing and the position is controlled by a feedback loop. The stiffness on the high frequency side is determined by the mass of the shaft, so the influence of the bias current is small. The stiffness on the low frequency side is influenced by the bias current, and the larger the bias current, the higher the stiffness. The stiffness formula is expressed by the following formula (2).

In equation (2), Fd is the disturbance, x is the displacement, m is the mass of the rotor, K _IF is the current-force conversion constant (gain indicating the relationship between the electromagnet current and the generated force), and K _GF is, Gap-force conversion constant (gain indicating the relationship between gap and force), Gs is sensor gain, Ga is amplifier gain, and Gc is control parameter (PID).

FIG. 14 is a graph showing that the transfer function of the controlled object changes depending on the magnitude of the bias current. The object to be controlled here is a magnetic bearing. Here, the vertical axis K of the transfer function of the controlled object in FIG. 14 is expressed by the following equation (3).

In Equation (3), K _IF represents a current-force conversion constant (a gain that represents the relationship between the electromagnet's current and the generated force), and K _GF represents a gap-force conversion constant (a gain that represents the relationship between the gap and force). ), and m indicates the mass of the rotor. Both K _IF and K _GF are parameters whose values change depending on the bias current.

As shown in FIG. 14, on the high frequency side, the magnitude of the bias current does not affect the transfer function, but on the low frequency side, the smaller the bias current, the higher the transfer function, and the larger the bias current, the lower the transfer function. The following equation (4) shows the equation of the loop transfer function of the feedback system.

In equation (4), m is the mass of the rotor, K _IF is the current-force conversion constant (gain indicating the relationship between the electromagnet's current and the generated force), and K _GF is the gap-force conversion constant (the gap and Gs indicates a sensor gain, Ga indicates an amplifier gain, and Gc indicates a control parameter (PID).

In other words, the open loop transfer function is the product of the transfer function of the controlled object multiplied by the sensor gain, amplifier gain, and control gain. To judge the stability of a system, the phase must not lag by -180 degrees at the crossover frequency (the frequency that crosses 0 dB (zero decibel) from top to bottom in the Bode diagram) in the Bode diagram of the open-loop transfer function ( phase margin).

The crossover frequency of the loop transfer function changes depending on changes in the controlled object. In some cases, the crossover frequency may move to an area where there is no phase margin, or a new crossover frequency may occur, resulting in instability.

FIG. 15 is a block diagram of the magnetic bearing system. In FIG. 15, amplifier 204 outputs current according to a command from controller 203. The displacement of the controlled object 205 (rotary axis) is detected by a displacement sensor 206, and the difference from the target value (center position) is calculated by a subtracter 202. The controller 203 performs PID control (gain adjustment, phase compensation, filter processing).

The coil current of the magnetic bearing has a current feedback loop as shown in FIG. Since the control current ic follows the current command value up to a predetermined frequency band, the magnitude of the actual control current can be estimated from the current command value. Furthermore, by adding a bias current to this control current, the bias magnetic flux can linearize the generated force with respect to the control current of the magnetic bearing (FIGS. 8 to 11).

FIG. 16 is a Bode diagram showing the frequency characteristics of the open-loop transfer function of the magnetic bearing system of FIG. 15. It is known that the system will oscillate unless the phase is above -180° at the frequency at which the gain crosses 0 dB (crossover frequency fc). In FIG. 11, it is stable when the phase φ>−180° at fc.

FIG. 17 is a Bode diagram showing the frequency characteristics of the open loop transfer function when the bias current is increased from the state shown in FIG. 16. Increasing the bias current increases rigidity against shaft vibration. However, the gain decreases and a new crossover frequency fc1 is acquired. In FIG. 17, the phase is unstable when the phase φ<-180° at the crossover frequency fc1.

FIG. 18 is a Bode diagram showing the frequency characteristics of the open-loop transfer function when the overall gain of the open-loop transfer function is increased from the state shown in FIG. 17. In FIG. 18, by increasing the gain of the open loop transfer function, fc1 in FIG. 17 is no longer the crossover frequency. At the crossover frequency fc, it is stable when the phase φ>-180°.

FIG. 19 is a Bode diagram showing the frequency characteristics of the open loop transfer function when the integral gain is lowered from the state shown in FIG. 17 to reduce the influence on the phase on the low frequency side. In FIG. 19, the phase at fc1 changes to φ>-180° by lowering the integral gain, so it is stable. In this embodiment, either the process in FIG. 18 or the process in FIG. 19 may be executed, or the process in FIG. 19 may be executed in addition to the process in FIG. 18.

The stability of the magnetic bearing is verified in advance, and control parameters such as the overall gain and integral gain are varied at the same time as the bias current increases or decreases. The calculation unit 100 executes PID control to calculate the control current command value Vic, and updates the control parameters of the PID control at the same time as updating the bias current.

As shown in Fig. 14, when the bias current is increased, the gain on the low frequency side decreases (Fig. 17) in the frequency characteristics of the controlled object (Bode plot: Fig. 16), so judging from the gain margin, the overall PID control Increase the gain.

The control parameter updated at this time is the overall gain of the open loop transfer function of the PID control. When the bias current exceeds a preset judgment value and the bias current is increased, the calculation unit 100 increases the gain and increases the bias current when the bias current exceeds a preset judgment value and increases the bias current. If , decrease the gain.

Specifically, the controller 203 adjusts the overall gain (FIG. 18) and changes the phase state on the low frequency side using the integral gain (FIG. 19). Since the state in FIG. 19 is unstable with respect to changes in the controlled object, the process in FIG. 19 is performed in conjunction with the overall gain adjustment in FIG. 18.

The control parameter that is updated at this time is the integral gain of PID control. When the bias current exceeds a preset determination value and the bias current is increased, the calculation unit 100 decreases the integral gain, and when the bias current exceeds the preset determination value and the bias current is increased. If the current is decreased, increase the integral gain.

FIG. 20 is a flowchart of the main processing of the magnetic bearing. Referring to FIGS. 10 and 20, in step S1, calculation unit 100 acquires signals from displacement sensors 121 and 122 via sensor circuit 120. Then, in step S2, the calculation unit 100 determines a current command value for each of the opposing electromagnets M1 and M2. Then, in step S3, coil currents ic1 and ic2 are applied from the amplifiers 111 and 112 to the coils 73 and 75 according to command values from the calculation unit 100. In the main process, steps S1, S2, and S3 are repeatedly executed.

FIG. 21 is a flowchart of the process of determining the current command value in S2 of FIG. Referring to FIGS. 10 and 21, in step S11, calculation unit 100 performs calculations such as phase compensation, PID processing, gain adjustment, filter processing, etc. related to control and determines control current command value Vic. In parallel with determining the control current command value Vic, the calculation unit 100 determines the bias current command value Vib in step S12. Furthermore, in step S13, the calculation unit 100 adds the control current command value Vic to the bias current command value Vib on the one hand, and subtracts the control current command value Vic from the bias current command value Vib on the other hand, so that the coil currents ic1, ic2 Coil current command values Vic1 and Vic2 corresponding to are determined and output. The magnetic bearing 70 includes a pair of electromagnets M1 and M2 facing each other with the rotating shaft in between, and supports the shaft in a non-contact manner by the combined magnetic force generated by the coil currents ic1 and ic2 of the respective electromagnets.

FIG. 22 is a flowchart showing a first example of the process of determining the bias current command value in S12 of FIG. 21. As shown in FIG. 22, when the DC component Vicdc of the control current command value increases and the DC component Vicdc of the control current command value becomes larger than a preset first threshold value Vth1, , the bias current is increased, the DC component Vicdc of the control current command value decreases, and when the DC component Vicdc of the control current command value becomes smaller than the preset second threshold Vth2, the bias current decrease. Preferably, the first threshold Vth1 is set larger than the second threshold Vth2, and the increase/decrease characteristic of the bias current ib has hysteresis.

As shown in the flowchart of FIG. 22, the DC component Vicdc of the control current command value Vic of the opposing electromagnet is monitored, and when the load becomes large (YES in S23), the DC component of the control current command value Vic of the opposing electromagnet is monitored. At the timing when Vicdc becomes larger than the determination value Vth1 (NO in S24), the bias current ib is increased (S25). When the load becomes small (NO in S23), the DC component Vicdc of the control current command value Vic of the opposing electromagnet becomes small, so the bias current ib is changed at the timing when it becomes smaller than the judgment value Vth2 (NO in S26). decrease (S27). By setting the judgment value Vth1 and the judgment value Vth2 to different values and providing hysteresis in the threshold value for increasing/decreasing the bias current ib, unnecessary switching of the bias current is suppressed in the alternating current component of load fluctuation.

The initial value of the bias current ib is set to a value ibL that allows the rotating shaft 20 to float stably under no load, and the set value of the bias current ib is not made smaller than this initial value. That is, the initial value is set as the lower limit value ibL (minimum value) of the bias current ib. Further, the upper limit value ibU (maximum value) of the bias current is determined from the constraints of the electromagnet, amplifier, and power supply. When the bias current is determined by the processing in steps S21 to S27, in the guard processing in step S28, if the bias ib exceeds the upper limit value ibU, the bias current ib is set as ibU, and conversely, the bias ib is set to be lower than the lower limit value ibL. If the bias current is smaller, the bias current is set to ib=ibL.

When the bias current ib is increased or decreased, the bearing rigidity changes (FIG. 13). As the bias current ib increases, the bearing rigidity increases, and as the bias current ib decreases, the bearing rigidity decreases. Therefore, when the DC component Vicdc of the control current command value Vic becomes higher than a constant value (Vth1) and the bias current ib is increased accordingly, the rigidity becomes higher. When the rigidity increases, the DC component Vicdc of the control current command value Vic becomes smaller for the same load. Similarly, when the DC component Vicdc of the control current command value Vic becomes lower than a constant value (Vth2) and the bias current ib is reduced, the rigidity decreases. When the rigidity decreases, the DC component Vicdc of the control current command value Vic increases if the load is the same.

The calculation unit 100 may continuously change the bias current ib using PID control or the like. Alternatively, the calculation unit 100 may change the bias current ib in steps with a preset increase/decrease width (for example, 0.1 A, etc.).

FIG. 23 is a flowchart showing a second example of the process of determining the bias current command value in S12 of FIG. FIG. 24 is a block diagram corresponding to a second example of processing for determining a bias current command value. Note that the process in FIG. 22 is the main process, and when the process in FIG. 23 is performed, it is applied simultaneously with the process in FIG. As shown in FIGS. 23 and 24, the calculation unit 100A increases the amplitude of the AC component Vicac of the control current command value and sets the amplitude of the AC component Vicac of the control current command value to a preset third threshold Vth3. When the bias current becomes larger, the bias current is increased, the amplitude of the AC component Vicac of the control current command value is decreased, and the amplitude of the AC component Vicac of the control current command value is smaller than a preset fourth threshold value Vth4. If this happens, reduce the bias current. Preferably, the third threshold Vth3 is set larger than the fourth threshold Vth4, and the bias current increase/decrease characteristic has hysteresis. In this case, the bias current setting unit 102 is configured to receive signals from two places: the output of the sensor circuit 120 and Vic. The alternating current component is weighted into gain according to frequency in processing such as phase compensation of the control unit 100A. Therefore, mainly for high frequency components, Vib is generated from a value obtained by filtering the output of the sensor circuit 120 within the bias current setting section 102. Furthermore, it may be easier to determine the low frequency component of the AC component when Vic is used, so both outputs are used.

As with the DC component, as for the AC component, as shown in the flowchart of FIG. 23, the amplitude value of the AC component Vicac of the sensor signal or the control current command value Vic of the opposing electromagnet is monitored and extracted (S31, S32). If the load is increasing (YES in S33), the bias current is increased at the timing when the sensor signal or the amplitude of the AC component Vicac of the control current command value Vic of the opposing electromagnet becomes larger than the judgment value Vth3 (NO in S34). (S35). If the load is decreasing (NO in S33), the amplitude of the AC component Vicac of the control current command value Vic of the sensor signal or the opposing electromagnet becomes smaller, so the amplitude of the AC component Vicac of the control current command value Vic is determined. At the timing when the bias current becomes smaller than the value Vth4 (NO in S36), the bias current is decreased (S37). By setting the determination value Vth3 and the determination value Vth4 to different values and providing hysteresis in the threshold value for increasing or decreasing the bias current, unnecessary switching of the bias current is suppressed in the alternating current component of load fluctuation.

To extract the AC component Vicac, the rotation synchronous component and the low frequency region are separated. When accelerating or decelerating the rotational speed of the rotating shaft, the vibration of the rotation synchronization component is monitored, and if the vibration due to unbalance of the rotating shaft is large, the bias current is increased to increase the rigidity. In addition, since vibrations caused by the installation environment of the rotating equipment or disturbances such as earthquakes are in the low frequency range (for example, 50Hz or less), the presence or absence of disturbances can be determined from the sensor signal or the amplitude value of the alternating current component Vicac of the control current command value Vic of the opposing electromagnet. However, when a disturbance occurs, the bias current is increased to increase rigidity.

The initial value of the bias current is a value that allows the rotating shaft 20 to float stably under no load, and the set value of the bias current ib is not made smaller than this initial value. That is, the initial value is set as the lower limit value ibL (minimum value) of the bias current ib. Further, the upper limit value ibU (maximum value) of the bias current is determined from the constraints of the electromagnet, amplifier, and power supply. When the bias current is determined by the processing in steps S31 to S37, in the guard processing in step S38, if the bias ib exceeds the upper limit value ibU, the bias current ib is set as ibU, and conversely, the bias ib is set to be lower than the lower limit value ibL. If the bias current is smaller, the bias current is set to ib=ibL.

Increasing or decreasing the bias current changes the bearing stiffness (Figure 13). As the bias current increases, the bearing rigidity increases, and as the bias current decreases, the bearing rigidity decreases. Therefore, the amplitude of the AC component Vicac of the signals of the displacement sensors 121, 122 or the control current command value Vic becomes larger than a constant value (Vth3), and when the bias current is increased accordingly, the bearing rigidity increases. When the bearing stiffness increases, the amplitude of the AC component Vicac of the sensor signal or control current command value Vic becomes smaller if the load is the same. Similarly, when the amplitude of the alternating current component Vicac of the sensor signal or control current command value Vic becomes smaller than a constant value (Vth4) and the bias current is reduced accordingly, the bearing rigidity decreases. When the bearing rigidity decreases, the amplitude of the AC component Vicac of the sensor signal or control current command value Vic increases if the load remains the same.

The bias current ib may be continuously varied using control such as PID. Alternatively, it may be changed stepwise at a preset value, for example 0.1A.

As explained above, in the magnetic bearing device according to the present embodiment, the bias current of the electromagnet is varied according to the load generated on the rotating shaft. It is possible to reduce core loss of the shaft and suppress heat generation. Furthermore, even when the load is large, the current can be increased up to a preset maximum current range, and improvements in control stability and rigidity can be expected by expanding the linear range.

The embodiments disclosed this time should be considered to be illustrative in all respects and not restrictive. The scope of the present invention is indicated by the claims rather than the description of the embodiments described above, and it is intended that all changes within the meaning and range equivalent to the claims are included.

10, 10A pump, 20 rotating shaft, 30 drive unit, 40 impeller, 50 pump casing, 60, 70, 70A, 80 magnetic bearing, 61 thrust disk, 62, 64, 72, 74 stator core, 63, 65, 73, 73A , 75, 75A Stator coil, 66 Gap sensor, 72A, 74A Stator teeth, 100, 100A Arithmetic unit, 101, 203 Controller, 102 Bias current setting unit, 103 Adder, 104, 202 Subtractor, 111, 112, 204 Amplifier, 120 Sensor circuit, 121, 122, 123, 206 Displacement sensor, M1 First electromagnet, M2 Second electromagnet.

Claims (11)

A magnetic bearing device that supports a rotating shaft,
a first electromagnet and a second electromagnet arranged to sandwich the rotating shaft;
a sensor that detects displacement of the rotating shaft;
a calculation unit that calculates a current flowing through the first electromagnet and the second electromagnet,
The arithmetic unit is
(a) based on the displacement detected by the sensor, calculate control currents with opposite signs and the same value for the first electromagnet and the second electromagnet;
(b) calculating bias currents with the same sign and the same value for the first electromagnet and the second electromagnet in accordance with the increase/decrease in the control current;
(c) adding the control current and the bias current to calculate a coil current to be applied to each of the first electromagnet and the second electromagnet ;
In the magnetic bearing device, the calculation unit executes PID control to calculate the control current, and updates the control parameters of the PID control at the same time as updating the bias current. A lower limit value is defined for the bias current,
The calculation unit increases the bias current and increases the control current when the DC component of the control current increases and the DC component of the control current becomes larger than a first threshold value set in advance . 2. When the DC component of the control current decreases and the DC component of the control current becomes smaller than a preset second threshold, the bias current is reduced within a range that does not fall below the lower limit value. 1. The magnetic bearing device according to 1. 3. The magnetic bearing device according to claim 2, wherein the first threshold is set larger than the second threshold, and the increase/decrease characteristic of the bias current has hysteresis. A lower limit value is defined for the bias current,
The calculation unit increases the bias current when the amplitude of the AC component of the control current increases and the amplitude of the AC component of the control current becomes larger than a preset third threshold; When the amplitude of the AC component of the control current decreases and the amplitude of the AC component of the control current becomes smaller than a preset fourth threshold, reduce the bias current within a range that does not fall below the lower limit value. The magnetic bearing device according to claim 1. 5. The magnetic bearing device according to claim 4, wherein the third threshold is set larger than the fourth threshold, and the bias current increase/decrease characteristic has hysteresis. The magnetic bearing device according to any one of claims 1 to 5, wherein the calculation unit changes the bias current stepwise in a preset increase/decrease range. The control parameter is a gain of the open loop transfer function of the PID control,
When the bias current exceeds a preset determination value and the bias current is increased, the calculation unit increases the gain and causes the bias current to exceed a preset determination value; The magnetic bearing device according to claim 1 , wherein the gain is decreased when the bias current is decreased. The control parameter is an integral gain of the PID control,
When the bias current exceeds a preset determination value and the bias current is increased, the calculation unit reduces the integral gain and determines whether the bias current exceeds the preset determination value. 2. The magnetic bearing device according to claim 1 , wherein the integral gain is increased when the bias current is decreased. A magnetic bearing device that supports a rotating shaft,
a first electromagnet and a second electromagnet arranged to sandwich the rotating shaft;
a sensor that detects displacement of the rotating shaft;
a calculation unit that calculates a current flowing through the first electromagnet and the second electromagnet,
The arithmetic unit is
(a) based on the displacement detected by the sensor, calculate control currents with opposite signs and the same value for the first electromagnet and the second electromagnet;
(b) calculating bias currents with the same sign and the same value for the first electromagnet and the second electromagnet in accordance with the increase/decrease in the control current;
(c) adding the control current and the bias current to calculate a coil current to be applied to each of the first electromagnet and the second electromagnet;
A lower limit value is defined for the bias current,
The calculation unit increases the bias current and increases the control current when the DC component of the control current increases and the DC component of the control current becomes larger than a first threshold value set in advance. a magnetic bearing that reduces the bias current within a range that does not fall below the lower limit when the DC component of the control current decreases and the DC component of the control current becomes smaller than a preset second threshold. Device. A magnetic bearing device that supports a rotating shaft,
a first electromagnet and a second electromagnet arranged to sandwich the rotating shaft;
a sensor that detects displacement of the rotating shaft;
a calculation unit that calculates a current flowing through the first electromagnet and the second electromagnet,
The arithmetic unit is
(a) based on the displacement detected by the sensor, calculate control currents with opposite signs and the same value for the first electromagnet and the second electromagnet;
(b) calculating bias currents with the same sign and the same value for the first electromagnet and the second electromagnet in accordance with the increase/decrease in the control current;
(c) adding the control current and the bias current to calculate a coil current to be applied to each of the first electromagnet and the second electromagnet;
A lower limit value is defined for the bias current,
The calculation unit increases the bias current when the amplitude of the AC component of the control current increases and the amplitude of the AC component of the control current becomes larger than a preset third threshold; When the amplitude of the AC component of the control current decreases and the amplitude of the AC component of the control current becomes smaller than a preset fourth threshold, reduce the bias current within a range that does not fall below the lower limit value. magnetic bearing device. A low temperature liquefied gas liquid delivery pump comprising the magnetic bearing device according to any one of claims 1 to 10 .

Images