WO2018179822A1

WO2018179822A1 - Induction motor drive device

Info

Publication number: WO2018179822A1
Application number: PCT/JP2018/003473
Authority: WO
Inventors: 直希國廣; 岩路　善尚; 徹郎児島
Original assignee: 株式会社日立製作所
Priority date: 2017-03-28
Filing date: 2018-02-01
Publication date: 2018-10-04
Also published as: JP2018166350A

Abstract

The purpose of the invention is to improve an initial speed estimation method in sensorless vector control of an induction motor, thereby reducing time required for estimation and estimating initial speed with high-speed and a reduced torque shock in restart from stop or a free-running state. When the initial speed of an induction motor is estimated, simultaneously with the start of energization to the induction motor, the drive frequency of energization current is modified in accordance with an energization current value to estimate the initial speed, and the change rate of the drive frequency during the modification is also changed in accordance with the energization current value.

Description

Induction motor drive device

The present invention relates to an induction motor drive device or a railway vehicle, and more particularly to an induction motor drive device equipped with a control device capable of reducing vehicle body vibration when the railway vehicle is stopped or restarted from coasting.

Most industrial equipment is driven by motors, and the percentage of domestic power consumption occupied by motors has reached about 50%. To reduce this power consumption, development of highly efficient motors has been underway. Further, since a large energy saving effect can be obtained by variable speed driving of a motor using a power conversion device such as an inverter, the power conversion device has been widely used.

Motor control software is installed in the power converter, and vector control that can control the motor with high accuracy is the mainstream as a control method. This vector control requires motor speed information. However, in recent years, the estimated speed is used in place of detection by the speed sensor for the purpose of saving power, improving reliability, and improving output by increasing the motor capacity. Speed sensorless vector control has been developed that eliminates the need for sensors.

JP 2013-106461 A International Publication No. 2010/125637

The inventor of the present application diligently studied to reduce the time required for initial speed estimation in sensorless vector control of an induction motor and to greatly reduce torque shock at start-up. As a result, the following knowledge was obtained. .

In speed sensorless vector control, the control software has a speed estimation algorithm to estimate the motor rotation speed. If the estimated speed has an error with respect to the actual speed, the control axis (dq axis) recognized by the vector control is shifted from the dq axis of the actual motor, and torque deviation or torque shock is detected. Will cause.

When the motor not driven by the power converter is stopped or restarted from the free-run state, initial speed estimation is performed to estimate the speed at the same time as the energization is started. Here, in addition to reliably estimating the initial speed without failing, it is required not to give a torque shock to the load side to which the motor is connected, reducing the time taken to estimate the initial speed. Reducing torque shock is an issue.

The initial speed estimation method has different problems for each motor because the principle of obtaining rotational force differs depending on the type of motor such as a permanent magnet synchronous motor or induction motor.

If it is a motor that always has a magnetic flux, such as a permanent magnet synchronous motor, an induced voltage is generated when the magnetic flux is linked to the stator coil at the same time as the gate of the power converter is turned on, and the stator coil Current flows. Various techniques have been studied in order to suppress excessive generation of the induced voltage. However, since the induced voltage can be used immediately after the start of energization, the initial speed estimation can be performed in a relatively short time.

On the other hand, in an induction motor, since the field source is an induced current that flows to the secondary side, it is necessary to estimate the speed from the state where the magnetic flux is not established immediately after the start of energization, and the initial speed is higher than the permanent magnet synchronous motor. There is a problem that takes time to estimate. If it takes time to estimate the initial speed, the time during which the control axis has an error also becomes longer. Therefore, when there is no axis error, a q-axis side secondary magnetic flux that becomes zero is generated, and the q-axis side secondary magnetic flux and d A torque shock proportional to the product of the shaft current is generated.

Japanese Patent Laid-Open No. 2013-106461 (Patent Document 1) describes a frequency sweep when a motor control unit performs a speed search in order to accurately identify the rotational speed of a free-running motor in speed sensorless vector control. When the polarity of the detected q-axis current Iq is reversed during the frequency sweep, if the value of the detected d-axis current Id is equal to or less than a predetermined determination threshold, the rotational speed (actual speed) of the motor And the command rotational speed (sweep frequency) are disclosed to have a function of determining that they match. [0041] From [0041] to [0043] of Patent Document 1 disclose that an optimum rate is set by a teaching function because the frequency sweep rate becomes unstable when it is set too high.

However, in Patent Document 1, the success or failure of the initial speed estimation is greatly influenced not only by the control response but also by the time constant of the motor, and therefore changes dynamically even if the optimum rate is determined and teaching is performed under specific conditions. When the temperature condition and the magnetic flux density condition of the motor change, the time constant also changes, so there is a possibility of destabilization.

International Publication No. 2010/125637 (Patent Document 2) includes a first calculation unit that calculates a free-run speed for the purpose of further reducing torque shock at the time of restarting an AC rotary machine in a free-run state, It is disclosed that a second calculation unit is provided that calculates and outputs a voltage command and a drive speed, with at least one of the speed calculated by the first calculation unit or the magnetic flux as an initial value. In the first calculation unit, an observer method based on a voltage equation is applied. Further, in [FIG. 2] of Patent Document 2, the estimated initial value is 0 Hz. Therefore, the estimation of the low speed region (near 0 Hz) can be performed at high speed and low torque shock.

However, in Patent Document 2, it is expected that the higher the actual initial speed, the later the rise of the magnetic flux, the longer the required time for initial speed estimation, and the occurrence of torque shock.

An object of the present invention relates to improving the initial speed estimation method in sensorless vector control of an induction motor, reducing the time required for estimation, and greatly reducing torque shock at startup.

When estimating the initial speed of the induction motor, the present invention corrects the AC frequency to be energized according to the value of the flowing AC current simultaneously with the start of energization of the induction motor, and changes the rate of change of the AC frequency according to the current value. Relating to adjusting.

According to the present invention, the initial speed can be estimated at a high speed and a low torque shock when the induction motor driven by the sensorless vector control is stopped or restarted from the free-run state.

For example, according to the present invention, unlike in Patent Document 1, the rate of change of the drive frequency is changed according to the value of the energization current, so that the speed estimation is speeded up under the condition that the estimated speed is greatly different from the actual speed. When the estimated speed approaches the actual speed, the speed estimation gain can be reduced and is naturally stabilized. The frequency of the motor is not forcibly swept at a constant rate, but the control system makes the speed estimation response variable based on detection information (current detection value) that includes changes in the motor status. Even when conditions and magnetic flux density conditions change, stable initial speed estimation can be realized without failure of initial speed estimation.

In addition, according to the present invention, unlike Patent Document 2, the property that the current detection value of the induction motor is increased according to the difference between the estimated speed value and the actual speed is used, and the rate of change of the drive frequency is calculated as the current flow rate. Since the speed estimation response is made variable and the speed is increased according to the value, even if the initial speed estimated value and the actual speed are far from each other, the initial speed can be estimated at high speed and low torque shock can be realized. .

Functional block diagram of the induction motor driving apparatus according to the first embodiment. T-type equivalent circuit diagram of the induction motor according to the first embodiment Diagram showing the relationship between slip frequency and phase current in induction motors Comparison chart of speed estimation results in initial speed estimation period according to embodiment 1 Comparison diagram of torque shock reduction effect in initial speed estimation period according to embodiment 1 Functional block diagram of the induction motor driving apparatus according to the second embodiment. Functional block diagram of the induction motor driving apparatus according to the third embodiment. Functional block diagram of the induction motor driving apparatus according to the fourth embodiment. The figure which shows the initial speed estimation result by Example 4. Functional block diagram of the induction motor driving apparatus according to the fifth embodiment. Schematic explanatory diagram showing d-axis secondary magnetic flux command calculation according to the fifth embodiment Functional block diagram of an induction motor drive device according to Embodiment 6 Schematic configuration diagram showing a part of the drive device for an induction motor according to the seventh embodiment. Schematic block diagram showing a part of a railway vehicle having an induction motor drive device according to Embodiment 8

In the embodiment, the induction motor drive device drives the induction motor by sensorless vector control, and the control device that controls the induction motor based on a command from the outside induces the drive frequency at the start of energization to the induction motor. Having an initial speed estimating section for correcting the current speed detected by the current detection means of the motor to estimate the initial speed, and further changing the change rate of the driving frequency at the time of the correction according to the detected current value Is disclosed.

Further, in the embodiment, the method of estimating the initial speed in the sensorless vector control of the induction motor, the control device of the induction motor corrects the drive frequency at the start of energization of the induction motor by the value of the current flowing through the induction motor, Furthermore, the thing which changes the change rate of the drive frequency at the time of correction according to an electric current value is disclosed.

Also, the embodiment discloses that the deviation between the q-axis current command value and the q-axis current detection value is controlled to be zero.

Further, in the embodiment, as the d-axis current detection value at the start of energization at the time of estimating the initial speed becomes larger than the d-axis current command value, the change rate of the drive frequency is made higher than at the start of energization. Is disclosed.

In the embodiment, the second d-axis current command value is added by adding the d-axis current compensation amount to the first d-axis current command value when the estimated initial speed is equal to or lower than a predetermined threshold value. Is disclosed.

In the embodiment, the control gain of the initial speed estimation unit includes the d-axis secondary magnetic flux command value, and the d-axis secondary magnetic flux command value is transiently changed according to the waveform or application time of the d-axis current command. It is disclosed that a calculation command value of a changing d-axis secondary magnetic flux is used. The induction motor control device includes a d-axis secondary magnetic flux command value as a control gain, and changes transiently according to the waveform or application time of the d-axis current command as the d-axis secondary magnetic flux command value. The use of the calculation command value of the d-axis secondary magnetic flux is disclosed.

Also, in the embodiment, it is disclosed that the estimated initial value is set based on speed information input from the outside of the control device.

Also, the embodiment discloses that the power converter that drives the induction motor is a SiC inverter. Moreover, it discloses that the control apparatus of an induction motor controls a SiC inverter.

Also, in the embodiment, an electric vehicle equipped with an induction motor drive device is disclosed. Also disclosed is a method of driving an induction motor of an electric vehicle by using an initial speed estimation method.

Hereinafter, the above and other novel features and effects of the present invention will be described with reference to the drawings. It should be noted that the examples and the drawings are used exclusively for understanding the invention and do not limit the scope of rights.

FIG. 1 is a functional block diagram of the induction motor drive device according to the present embodiment, and shows the configuration of the drive device of the induction motor that drives the induction motor according to the present embodiment by sensorless vector control. .

An induction motor 3 is connected to the induction motor drive device 1 shown in FIG. A control device 2 and a power converter (main circuit) are incorporated in a drive device 1 for an induction motor. The power converter is composed of a power device composed of a transistor such as an IGBT (Insulated Gate Bipolar Transistor) or a diode, and converts a DC voltage or an AC voltage into an arbitrary AC voltage, and is connected to the power converter. Drive the induction motor. The control device 2 is mounted with a control program for vector control of the induction motor 3 without a speed sensor. In FIG. 1, the power converter is not shown.

The induction motor 3 generates a rotational torque by an interaction between a rotating magnetic field generated by applying a three-phase AC voltage to the stator winding and an induced current flowing through the rotor conductor when the rotor slips. .

The voltage command calculation unit 11 outputs a d-axis voltage command Vd ^* for the d-axis that is an excitation axis (magnetic flux axis) and a q-axis voltage command Vq ^* in a direction that is electrically orthogonal to the d-axis. The coordinate converter 7 generates three-phase voltage commands Vu ^* , Vv ^* , Vw ^* from the d-axis voltage command Vd ^* and the q-axis voltage command Vq ^* , and the PWM calculation unit 6 (Pulse Width Modulation) performs PWM control. Thus, the three-phase voltages Vu, Vv, and Vw are output from the induction motor drive device 1 to the induction motor 3. In the coordinate converter 5 and the coordinate converter 7, a phase θ obtained by integrating the frequency ω1 is required for coordinate conversion, but the illustration is omitted in FIG. Actually, ω is an angular frequency, but in the present embodiment, it is simplified and used as a speed or a frequency.

The current flowing through the induction motor 3 is detected by the current detector 4. The current detector 4 is composed of a Hall CT (Current Transformer) or the like, and detects currents Iu, Iv, and Iw of three phases of U phase, V phase, and W phase. However, the current detector 4 is not necessarily required to detect the currents of all three phases, and any two phases are detected, and the other one phase is obtained by calculation on the assumption that the three-phase current is in an equilibrium state. But it ’s okay.

The detected current is converted from the three phases to the vector-controlled dq coordinate by the coordinate converter 5 and taken into the induction motor drive device 1 as Id and Iq.

The current controller 15 performs feedback control so that the deviation amount of the detected current value is zero with respect to the d-axis current command value Id ^* and the q-axis current command value Iq ^* .

The initial speed estimation calculation unit 10 takes in the detected current value and calculates the estimated speed value ωr ^ using a speed estimation model based on information such as the equivalent circuit constant of the induction motor 3. This speed estimation model may be a method of controlling the q-axis current deviation amount to zero, a model-based observer method, or the like, and is not necessarily limited to the method described here.

Also, in the control method in each feedback control system, PI control (proportional / integral control), PII control (proportional / double integral control), etc. are used as appropriate.

The speed estimation change rate calculation unit 12 will be described. In this embodiment, the induction motor 3 uses the property that the impedance of the induction motor 3 changes according to the “slip frequency” which is the difference between the frequency of the drive voltage applied to the induction motor 3 and the rotor frequency, Speed estimation is accelerated. This focuses on the property of the induction motor 3 that obtains rotational force by causing an induction current to flow on the secondary side, and is therefore suitable for a linear induction motor or the like in which a conductor is disposed on the secondary side.

FIG. 2 is a T-type equivalent circuit diagram of the induction motor according to the present embodiment, and shows a T-type equivalent circuit of the induction motor 3. When the phase voltage vector V1 and the phase current vector I1 are defined as in the equations (1) and (2), the combined impedance Z is expressed by the equation (6).

Here, r1 is the primary resistance, r2 is the secondary resistance, x1 is the primary leakage reactance, x2 secondary leakage reactance, s is slip, Z1 primary side impedance, Z2 is the secondary side impedance, Y ₀ is the excitation admittance, Xm is an excitation reactance. The current vector I1 of the induction motor 3 is expressed by Equation (7).

FIG. 3 is a diagram showing the relationship between the slip frequency and the phase current in the induction motor, and shows the relationship between the slip frequency fs (= speed estimation error) according to the equation (7) and the absolute value | I1 | of the phase current vector I1. . Under the condition that the estimated speed value matches the actual speed, the slip frequency is zero and the secondary impedance Z2 is infinite, so the phase current is minimized.

Conversely, when the slip is excessively positive or negative, the secondary impedance Z2 is small, and the current is large compared to the condition where the slip is zero. In other words, the current is minimized under conditions where the estimated speed matches the actual speed, and the detected value | I | of the phase current is large under conditions where the estimated speed is significantly different from the actual speed.

In the present embodiment shown in FIG. 1, the phase current command value | I * ^| and the phase current detected value | I | was input to the speed estimated change rate arithmetic unit 12, the phase current command value | I * ^| phase with respect to the current As the detected value | I | increases, an operation for increasing the variable gain Kx is performed. The variable gain Kx calculated by the speed estimation change rate calculation unit 12 is integrated with the control gain of the initial speed estimation calculation unit 10 to speed up the speed estimation response of the initial speed estimation calculation unit 10 when the speed estimation deviation amount is large. To play a role.

That is, when the estimated speed value ωr ^ by the initial speed estimation calculation unit 10 is greatly deviated from the actual speed ωr, the variable gain Kx is increased, and the estimated speed is increased by increasing the response speed. ωr ^ can be rapidly converged to the actual speed ωr.

It is conceivable that the speed estimation system becomes unstable by increasing the speed estimation gain, but when the speed estimation value ωr ^ approaches the actual speed ωr as shown in FIG. Since the difference between the current command | I ^* | and the phase current detection value | I | is also small as shown in FIG. 3, the variable gain Kx is naturally reduced and the speed estimation system is stabilized.

FIG. 4 is a comparison diagram of speed estimation results in the initial speed estimation period according to the present embodiment, and shows an example of calculation results comparing the presence / absence of application of the present embodiment. FIG. 4 shows that in the free-run state of the rotor frequency fr = 30 Hz, a predetermined current command is given simultaneously with the start of the initial speed estimation, and the estimated initial value is set to 200 Hz based on the information of the detected current value flowing through the induction motor 3. Is the result of estimating the initial speed.

FIG. 4A shows the initial speed estimation result when the present embodiment is applied. The detected current value increases after the initial speed estimation starts, and the variable gain Kx that is the output of the speed estimation change rate calculation unit 12 is It can be seen that it has increased. Since the variable gain Kx is integrated with the speed estimation gain, the rate of change of the speed estimation value fr ^ changes according to Kx. Thus, the speed estimated value fr ^ can be converged at high speed toward the actual speed fr.

FIG. 4B shows the result of the initial speed estimation by the conventional technique (Patent Document 1). As described in [0015] of Patent Document 1, the frequency sweep at the time of performing the speed search is changed from the highest frequency at a constant rate, and the drive frequency does not depend on the current detection value, but has a constant change rate (Kx = 1). In this case, although the estimated speed value fr ^ is toward the convergence direction with respect to the actual speed fr, the estimated speed at the initial speed estimation start point such as the estimated initial value (200 Hz) and the actual speed (30 Hz). Under conditions where the deviation is large, a huge amount of time is required.

FIG. 5 is a comparison diagram of the torque shock reduction effect during the initial speed estimation period according to the present embodiment, and shows a comparison of torque shock between the present embodiment and the prior art (Patent Document 1) at the time of initial speed estimation. Comparison is made at a ratio when the torque shock in the prior art (Patent Document 1) is 100%. According to the present embodiment, since the initial speed estimation can be speeded up, it is possible to reduce the torque shock by 96% compared to the prior art.

If the estimated speed does not match the actual speed, a large current flows to generate a q-axis secondary magnetic flux Φ2q, which leads to a torque shock. In the present embodiment, as compared in FIG. 4, in order to converge the speed estimation value ωr ^ to the actual speed ωr at a high speed before a large current flows, the generation of the q-axis secondary magnetic flux Φ2q is suppressed, and the torque Shock can be greatly reduced.

Furthermore, in recent years, the induction motor 3 has been made more efficient by meeting the demand for reduction in power consumption and further reducing the secondary resistance r2 of the rotor (sliding) more than before. The lower the slip, the longer the electrical and magnetic time constants and the slower the response, so that it takes time to estimate the initial speed. In this embodiment, since the speed estimation change rate calculation unit 12 can speed up the speed estimation, the initial speed can be reduced in a short time and with a low torque shock even for a highly efficient induction motor 3 with a rated slip of 1% or less. Can be estimated.

As described above, according to this embodiment, by using the impedance change according to the slip frequency of the induction motor 3 and making the speed estimation response variable based on the detected current value, the initial speed estimation can be made faster. It is possible to provide a drive device for an induction motor that realizes a low torque shock.

In this embodiment, unlike the first embodiment, the speed estimation system and the current control system are integrated. Hereinafter, the difference from the first embodiment will be mainly described.

FIG. 6 is a functional block diagram of the induction motor driving apparatus according to the present embodiment, and shows the configuration of the induction motor driving apparatus for driving the induction motor by sensorless vector control using functional blocks.

In this embodiment, for the purpose of improving robustness against fluctuations in motor constants, the initial speed estimation calculation unit 10 performs speed estimation calculation based on the q-axis current deviation amount, and the d-axis current command Id ^* is fed forward. Thus, the voltage command calculation unit 11 is different from the first embodiment in that it is input.

The initial speed estimation calculation unit 10 calculates a speed estimated value ωr ^ by adding the proportional gain KP8 and the integral gain KI9 to the q-axis current deviation amount and adding them together. The primary frequency command ω1 ^* is calculated by adding the slip frequency command value ωs ^* to the estimated speed value ωr ^ calculated by the initial speed estimation calculation unit 10. However, in this embodiment, since the q-axis current command Iq ^* , which is the torque current, is controlled to zero, the slip frequency command ωs ^* is also zero, and the estimated speed value ωr ^ is the same value as the primary frequency command value ω1 ^*. It becomes.

Hereinafter, the derivation of the proportional gain KP8 and the integral gain KI9 in the present embodiment will be described. The voltage equation of the induction motor is expressed by equation (8).

Where r1 is a primary resistance, r2 ^′ is a primary converted secondary resistance, Lσ is a primary converted leakage inductance, M is a mutual inductance, L2 is a secondary self-inductance, T2 is a secondary time constant, and Φ2d is a d-axis. Secondary magnetic flux, Φ2q is a q-axis secondary magnetic flux, and s is a Laplace operator. Here, the primary self-inductance L1, the secondary self-inductance L2, the primary conversion leakage inductance Lσ, the secondary time constant T2, and the primary conversion secondary resistance r2 ^′ are as follows.

When the expression relating to Vq in the second row in Expression (8) is expanded, the q-axis currents Iq and Iq ^* in the induction motor 3 and the induction motor drive device 1 are respectively expressed by the following expressions.

Here, assuming that the constant of the induction motor 3 and the constant of the control device 2 match, and assuming that Id, Φ2d, and Φ2q match the command value, the following equation is obtained.

That is, the transfer function representing the deviation of the rotor frequency from the deviation of the q-axis current is as follows.

Therefore, the control equation for estimating the speed of the induction motor 3 by the first-order lag system from the deviation amount of the q-axis current is the following equation. However, ωacr is a response frequency of current control.

From the above, the proportional gain KP8 and the integral gain KI9 in FIG.

Then, with respect to the final speed estimation formula of this embodiment, the output Kx of the speed estimation change rate calculation unit 12 is integrated to obtain the following formula.

As described above, in this embodiment, as shown in FIG. 1, the current controller 15 and the initial speed estimation calculation unit 10 are not separately provided, and the speed estimated value ωr ^ is set so that the q-axis current deviation amount is zero. By adopting a configuration that controls the speed, the speed estimation system and the current control system are integrated.

In addition, regarding the robustness against the constant fluctuation, in this embodiment, when the constant fluctuation occurs due to the temperature rise of the induction motor 3 or the like, the initial speed estimation time changes, but the q-axis current deviation is zero (the slip frequency is reduced). The initial speed can be estimated with high accuracy with respect to speed estimation accuracy.

Taking other examples as examples, the observer method based on the voltage equation of the induction motor 3 may be easily affected by constant fluctuations. For example, in the first calculation unit of Patent Document 2, the initial speed is calculated from an equation of (Vq ^* −R1 · Iq) / Φ1. In particular, in the low speed range, Vq ^* becomes small. Therefore, when the R1 constant error occurs due to the temperature fluctuation of the motor, the effect of the R1 error becomes relatively large. It may be connected.

In this embodiment, the speed estimation response can be made the same as the current control response by integrating the speed estimation system and the current control system with a configuration that estimates the speed so as to control the q-axis current deviation to zero. In addition to high speed estimation, even when constant fluctuation occurs, the feedback control is performed so that the Iq deviation amount becomes zero, so that the robustness against the constant fluctuation is also high.

Regarding the fluctuation of the initial speed estimation time at the time of the constant fluctuation described above, a margin is provided in advance so that the initial speed estimation can be completed even under the condition that the resistance value of the induction motor 3 is minimum (lower limit value of the motor temperature range). The control device 2 is designed.

In this embodiment, it is possible to provide a drive device for an induction motor that achieves higher speed and lower torque shock for initial speed estimation than in Embodiment 1 and further enhances robustness against constant fluctuation.

In this embodiment, the calculation of the speed estimation change rate is simplified compared to the first and second embodiments. Hereinafter, the difference from the first and second embodiments will be mainly described.

FIG. 7 is a functional block diagram of the induction motor driving apparatus according to the present embodiment, and shows the configuration of the induction motor driving apparatus for driving the induction motor by sensorless vector control using functional blocks.

In the present embodiment, for the purpose of simplifying the calculation of the speed estimation change rate calculation unit 12 and reducing the calculation load in the control device 2, the speed estimation change rate calculation unit 12 uses the d-axis current command values Id ^* and d. The shaft current detection value Id is input to calculate the variable gain Kx.

Since the q-axis is controlled to make the current deviation zero, and the d-axis is configured to be feedforward control, assuming that Iq = Iq ^* , the speed estimation deviation appears in the d-axis current. That is, since | I | ≈Id can be considered, the magnitude of the speed estimation deviation amount can be determined only from the d-axis current deviation amount, and the input of the speed estimation change rate calculation unit 12 can be Id ^* and Id.

In the second embodiment, the process of calculating the absolute values of the current detection values Id and Iq is performed. However, in this embodiment, this calculation process is not required, the calculation process is simplified, and the load factor of the calculation of the control device 2 is increased. Can be reduced.

The calculation process G14 in the speed estimation change rate calculation unit 12 may be a simple ratio such as Id / Id ^* , but a variable gain table for the d-axis current deviation amount is prepared in advance and the d-axis deviation amount is input. A variable gain calculation method may be used. As in the first embodiment, the speed estimation response may be made variable by utilizing the property that the induction motor 3 changes the current according to the slip frequency.

In the present embodiment, the calculation of the speed estimation change rate calculation unit 12 can be simplified as compared with the second embodiment, and an induction motor drive device can be provided in which the calculation load is further reduced.

In this embodiment, the d-axis current command Id ^* is increased in the low speed range. Hereinafter, the difference from the first to third embodiments will be mainly described.

FIG. 8 is a functional block diagram of the induction motor driving apparatus according to the present embodiment, and shows the configuration of the induction motor driving apparatus for driving the induction motor by sensorless vector control using functional blocks.

In this embodiment, for the purpose of preventing failure of initial speed estimation in a low speed range, d-axis that adds ΔId ^* to d-axis current command Id ^* when speed estimated value ωr ^ is equal to or less than a predetermined determination threshold value. A boost control calculation unit 13 is provided.

As shown in Expression (16), the initial speed estimation method of the second embodiment and the like linearizes the relationship between the speed electromotive force and the q-axis current deviation amount. As is clear from Expression (16), when the speed electromotive force on the right side of Expression (16) becomes zero, this speed estimation method is not established. Since the right side of the equation (16) is proportional to the product of the rotor speed ωr of the induction motor 3 and the d-axis secondary magnetic flux Φ2d, it decreases as the speed decreases. That is, in order to prevent the speed estimation from failing in the low speed range, the d-axis secondary magnetic flux command Φ2d ^* is increased, and the d-axis secondary magnetic flux Φ2d in the induction motor 3 is increased, so that the speed is also increased in the low speed range. A decrease in power can be suppressed.

When the d-axis boost control calculation unit 13 of this embodiment determines that the low-speed region is based on the estimated speed value ωr ^, the d-axis boost control calculation unit 13 outputs a d-axis current command compensation amount ΔId ^* of an appropriate magnitude, The second current command Id ^** is calculated by adding ΔId ^* to the current command Id ^* .

As a result, only when it is determined that the speed is low, the speed electromotive force can be increased and speed estimation failure can be prevented. The d-axis current command compensation amount ΔId ^* is appropriately designed according to the circuit constants of the induction motor 3 and the specifications of the control device 2.

In the present embodiment, the initial value of estimation is set to 200 Hz (upper limit value of driving frequency), but is not limited to this value. Although it is not always necessary to set the upper limit value, it is desirable to set a value higher than the median value of the driving speed range. If the estimation initial value is set to the low speed range and the high speed range is estimated, the d-axis boost control calculation unit 13 may operate immediately after the estimation is started, and an unnecessary torque shock may be generated.

In the control configuration of the present embodiment, the d-axis boost control calculation unit 13 is operated only when the high-speed range where the speed electromotive force is large and the initial speed estimation is easy is first searched, and then determined as the low-speed range, and the d-axis The secondary magnetic flux Φ2d is increased. In this control configuration, in order to prevent the speed electromotive force from becoming excessively small, a current command having a required magnitude is provided in a required speed range. As a result, it is possible to estimate to a low torque shock without failing estimation over the entire speed range.

FIG. 9 is a diagram showing an initial speed estimation result according to the present embodiment, and shows a comparison of speed estimation results with and without a d-axis boost control calculation unit. FIG. 9 compares the initial speed estimation results at the rotor frequency fr = −10 Hz (during reverse rotation), which is particularly difficult to estimate the initial speed. In this comparison, as an example, the d-axis boost control calculation unit 13 is operated when the estimated speed value fr ^ becomes 3 Hz or less. The determination value of the estimated speed value is appropriately set according to the specifications of the induction motor 3. In FIG. 9A, when the estimated frequency fr ^ becomes 3 Hz or less, the d-axis boost control operates, the second d-axis current command Id ^** increases stepwise, and the d-axis secondary magnetic flux is increased. When Φ2d rises and a predetermined speed electromotive force is secured, the speed estimated value fr ^ converges to the actual speed fr. On the other hand, in FIG. 9B, although the estimated frequency fr ^ goes to the actual frequency fr (-10 Hz), the d-axis secondary magnetic flux Φ2d does not rise, and fr ^ converges to zero (the speed electromotive force becomes zero). Convergence), resulting in failure of initial velocity estimation. The phenomenon in which the speed electromotive force becomes zero is due to the fact that the q-axis secondary magnetic flux Φ2q is more likely to be generated and the d-axis is easily affected as the speed is lower in the principle of the induction motor 3. In this embodiment, the d-axis secondary magnetic flux Φ2d is increased to prevent the d-axis secondary magnetic flux Φ2d from becoming zero due to the interference of the q-axis secondary magnetic flux Φ2q. The d-axis current command compensation amount ΔId ^* is added stepwise, but may be any shape as long as it increases the d-axis current.

In addition, since the highly efficient induction motor 3 has low slip, the secondary time constant T2 (secondary magnetic flux time constant) is long, it is difficult to establish the d-axis secondary magnetic flux, and initial speed estimation failure occurs. It tends to be easy to do. By applying this embodiment to such a low slip induction motor 3 as well, the initial speed can be estimated without causing a speed estimation failure.

As described above, in this embodiment, the d-axis boost control for increasing the d-axis current command Id ^* in the low speed range is provided, so that the accuracy of the initial speed estimation in the low speed range is further improved than in the third embodiment. A drive device can be provided.

In this embodiment, a transient value is used for the d-axis secondary magnetic flux command Φ2d ^* . Hereinafter, the difference from the first to fourth embodiments will be mainly described.

FIG. 10 is a functional block diagram of the induction motor driving apparatus according to the present embodiment, and shows the configuration of the induction motor driving apparatus for driving the induction motor by sensorless vector control using functional blocks.

In this embodiment, for the purpose of further shortening the initial speed estimation and reducing the torque shock, a transient value is used instead of a steady value for the d-axis secondary magnetic flux command Φ2d ^* used in the initial speed estimation calculation unit 10.

FIG. 11 is a schematic explanatory diagram showing the d-axis secondary magnetic flux command calculation according to the present embodiment, and is a comparison between the present embodiment and the d-axis secondary magnetic flux command Φ2d ^{* in} the prior art. In the prior art, based on the d-axis current command Id ^* , the steady-state d-axis secondary magnetic flux command Φ2d ^* is output simultaneously with the start of the initial speed estimation. On the other hand, in the present embodiment, by passing an LPF (Low Pass Filter) having a secondary time constant T2 based on the d-axis current command Id ^* , the d-axis secondary command Φ2d ^* is expressed by the following equation: A transient value corresponding to the elapsed time t from the start of initial speed estimation is output. However, the equation (23) is an equation indicating a transient value of the d-axis secondary magnetic flux Φ2d when the d-axis current command Id ^* is step-inputted. However, the d-axis current command Id ^* is a step shape and is filtered. The first-order lag shape or a forcing shape that flows rapidly and immediately attenuates immediately after the start of energization may be used, and is not limited.

Immediately after the start of the initial speed estimation, since the speed estimation values do not match and the control axes do not match, the d axis actually generated in the induction motor 3 as shown in FIG. The secondary magnetic flux Φ2d is a value that is substantially close to zero. That is, when the steady value is taken into consideration for the d-axis secondary magnetic flux command Φ2d ^* of the control gain in the initial speed estimation calculation unit 10, the gain of the speed estimation system becomes insufficient as shown in Expression (22). In this embodiment, the calculation command value of the d-axis secondary magnetic flux command that changes transiently according to the waveform and application time of the d-axis current command Id ^* is added to the d-axis secondary magnetic flux command Φ2d ^* of the speed estimation system gain. By using it, the value can be made closer to the actual Φ2d of the induction motor 3, the accuracy of the estimation model can be improved, and this lack of gain can be eliminated.

In addition, in the low slip induction motor, the secondary time constant T2 becomes longer and the rise of the d-axis secondary magnetic flux command Φ2d ^* becomes gentler. However, in this embodiment, as the induction motor 3 becomes lower slip, Acts in the direction of increasing the gain of the initial speed estimation system.

In the present embodiment, the d-axis secondary magnetic flux command Φ2d ^* used in the initial speed estimation calculation unit 10 is calculated by using a calculation command value that changes transiently according to the waveform of the d-axis current command Id ^* and the application time. The initial speed can be estimated at a higher speed than in the fourth embodiment, and torque shock can be reduced.

In this embodiment, the speed information of the drive system is used as an initial value. Hereinafter, the difference from the first to fifth embodiments will be mainly described.

FIG. 12 is a functional block diagram of the induction motor driving apparatus according to the present embodiment, and shows the configuration of the induction motor driving apparatus for driving the induction motor by sensorless vector control using functional blocks.

In this embodiment, for the purpose of further shortening the initial speed estimation and lowering the torque shock, the speed information of the drive system such as a railway vehicle is taken into the control device 2 and the initial value of the integrator of the initial speed estimation calculation unit 10 is acquired. Set to.

The speed information Vcar of the drive system is taken into the control device 2, converted into the electrical angular frequency ωrini ^ by the electrical angular frequency conversion unit 15, and used as the estimated initial value of the initial speed estimation calculation unit 10. The speed information Vcar is not necessarily an accurate value, and may be an approximate value such as a relay signal of a speed stage used in a railway vehicle. By applying the configuration according to the first to fifth embodiments with the approximate value of the drive frequency as the initial value, it is possible to realize the initial speed estimation of higher speed and lower torque shock. This embodiment can be applied to a system that can incorporate speed information of the drive system into the drive device 1 of the induction motor.

In this embodiment, since the estimated initial value used for the initial speed estimation calculation unit 10 can be brought close to the actual speed, it is possible to realize an initial speed estimation that is faster and lower in torque shock than the fifth embodiment.

In this embodiment, a SiC element is used as a switching element or a diode. Hereinafter, the difference from the first to sixth embodiments will be mainly described.

FIG. 13 is a schematic configuration diagram showing a part of the drive device for the induction motor according to the present embodiment, and is a schematic configuration diagram showing a part of the drive device for the induction motor that drives the induction motor by sensorless vector control.

FIG. 13 shows an example of a part of a railway vehicle drive system. Based on the vector control calculation of the control device 2, the gate driver 104 controls the switching operation of the semiconductor module 103. Further, on the input side, the LC reactor circuit is configured by the filter reactor 105 and the filter capacitor 106, and the noise flowing in the overhead line 107 is cut.

In the present embodiment, for the purpose of further shortening the initial speed estimation and lowering the torque shock, in the semiconductor module 103 of the induction motor driving apparatus 1, a SiC element is used for one or both of the switching element 101 and the diode 102. .

In the present embodiment, the carrier frequency can be improved and the current control response ωacr in the control device 2 can be improved by using a SiC inverter capable of high-speed switching with low loss. As described in the second embodiment, since the current control system and the speed estimation system are integrated, when the current control response ωacr is improved, the speed estimation system can obtain the same estimation response.

According to this embodiment, due to the synergistic effect with the SiC inverter, it is possible to realize the initial speed estimation with a higher speed and lower torque shock than in the sixth embodiment.

In this embodiment, the induction motor drive device is used in a railway vehicle. Hereinafter, the difference from the first to seventh embodiments will be mainly described.

FIG. 14 is a schematic configuration diagram showing a part of a railway vehicle having a drive device for an induction motor according to the present embodiment.

Electric power is supplied to the drive device of the railway vehicle from the overhead line 107 via the current collector, and AC power is supplied to the induction motor 3 via the drive device 1 of the induction motor, whereby the induction motor 3 is driven. To do. The induction motor 3 is connected to the axle of the railway vehicle via a reduction gear, and the railway vehicle travels by a tangential force generated between the wheel 109 and the rail 108 connected to the axle.

When the induction motor drive device 1 of the first to seventh embodiments is applied to a railway vehicle, the initial speed estimation time can be shortened and the torque shock can be greatly reduced. Therefore, when the railway vehicle is stopped or restarted from coasting It can reduce the body vibration of railway vehicles and improve passenger comfort.

DESCRIPTION OF SYMBOLS 1 ... Drive apparatus of induction motor, 2 ... Control apparatus, 3 ... Induction motor, 4 ... Current detector, 5 ... Coordinate converter, 6 ... PWM calculating part, 7 ... Coordinate converter, 8 ... Proportional gain, 9 ... Integral Gain 10, initial speed estimation calculation unit 11 voltage command calculation unit 12 speed estimation change rate calculation unit 13 d axis boost control calculation unit 14 gain 15 current controller 16 d axis 2 Next magnetic flux command calculation unit, 100 ... power converter, 101 ... switching element, 102 ... diode, 103 ... semiconductor module, 104 ... gate driver, 105 ... filter reactor, 106 ... filter capacitor, 107 ... overhead wire, 108 ... rail, 109 …Wheel

Claims

A drive device for an induction motor that drives the induction motor by sensorless vector control,
A control device that controls the induction motor based on a command from the outside estimates the initial speed by correcting the drive frequency at the start of energization of the induction motor with the current value detected by the current detection means of the induction motor. Further, the induction motor drive device further comprises an initial speed estimation unit that changes the rate of change of the drive frequency at the time of the correction in accordance with the detected current value.
In the induction motor drive device according to claim 1,
The induction motor drive device, wherein the initial speed estimation unit controls the deviation between the q-axis current command value and the q-axis current detection value to be zero.
In the induction motor drive device according to claim 1 or 2,
The initial speed estimator increases the rate of change of the drive frequency higher than that at the start of energization as the d-axis current detection value at the start of energization at the time of initial speed estimation becomes larger than the d-axis current command value. A drive device for an induction motor.
In the induction motor drive device according to any one of claims 1 to 3,
The initial speed estimation unit adds a d-axis current compensation amount to the first d-axis current command value when the estimated initial speed is equal to or less than a predetermined threshold value, thereby generating a second d-axis current command. An induction motor drive device characterized by calculating a value.
In the drive device of the induction motor according to any one of claims 1 to 4,
The control gain of the initial speed estimator includes a d-axis secondary magnetic flux command value, and the d-axis secondary magnetic flux command value changes transiently according to the waveform or application time of the d-axis current command. A drive device for an induction motor, wherein a calculation command value of a secondary magnetic flux is used.
In the induction motor drive device according to any one of claims 1 to 5,
An induction motor drive device, wherein an estimated initial value of the initial speed estimator is set based on speed information input from the outside of the control device.
The induction motor drive device according to any one of claims 1 to 6,
The induction motor drive device, wherein the power converter for driving the induction motor is a SiC inverter.
An electric vehicle equipped with the induction motor drive device according to any one of claims 1 to 7.
An initial speed estimation method in sensorless vector control of an induction motor,
The control device for the induction motor corrects the drive frequency at the start of energization of the induction motor based on the current value flowing through the induction motor, and further changes the rate of change of the drive frequency at the time of correction according to the current value. An initial speed estimation method characterized by changing.
The initial speed estimation method according to claim 9,
An initial speed estimation method, wherein the induction motor control device controls the deviation between the q-axis current command value and the q-axis current detection value to be zero.
In the initial speed estimation method according to claim 9 or 10,
As the d-axis current detection value at the start of energization at the time of estimating the initial speed increases with respect to the d-axis current command value, the control device for the induction motor increases the rate of change in the drive frequency from that at the start of energization. An initial speed estimation method characterized by increasing the speed.
The initial speed estimation method according to any one of claims 9 to 11,
When the estimated initial speed falls below a predetermined threshold, the control device for the induction motor adds a d-axis current compensation amount to the first d-axis current command value, thereby obtaining a second d-axis current. An initial speed estimation method characterized by calculating a command value.
The initial speed estimation method according to any one of claims 9 to 12,
The induction motor control device includes a d-axis secondary magnetic flux command value as a control gain, and the d-axis secondary magnetic flux command value changes transiently according to the waveform or application time of the d-axis current command. An initial speed estimation method using an operation command value of a shaft secondary magnetic flux.
The initial speed estimation method according to any one of claims 9 to 13,
An initial speed estimation method, wherein an estimated initial value is set based on speed information input from outside the control device for the induction motor.
The initial speed estimation method according to any one of claims 9 to 14,
An initial speed estimation method, wherein the induction motor control device controls a SiC inverter.
A method for driving an induction motor of an electric vehicle by using the initial speed estimation method according to any one of claims 9 to 15.