JP2017060341A - Controller for open wiring system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a controller for open wiring system, capable of suppressing generation of switching noise at a carrier midpoint while maintaining high-response performance of current control.SOLUTION: The controller for open wiring system detects a current detection value iof an inverter at a peak point of a carrier signal, updates a voltage command value von the basis of the current detection value i, updates the voltage command value von the basis of a current prediction value i^at a midpoint of a carrier signal and updates the voltage command value vat a 1/4 period of the carrier signal.SELECTED DRAWING: Figure 6

Description

  The present invention relates to a control method for an inverter that drives an electric motor, and more particularly to a control device for an open winding system.

  In an inverter that drives a load such as an electric motor, it is desirable that the response performance of current control is high in order to drive at high speed and high accuracy. In order to improve the current control response performance of the inverter, it is necessary to shorten the current control cycle.

  The voltage source inverter generates a switching command signal (gate signal) of a switching device in the inverter by PWM control that compares the voltage command value calculated by the current controller with a triangular wave carrier signal, and controls the current. . Therefore, in order to shorten the control cycle, it is necessary to increase the switching frequency of the switching command signal (gate signal).

  However, an increase in the switching frequency leads to an increase in the loss of the switching device, which causes an increase in the temperature of the switching device and a decrease in the efficiency of the inverter.

  The inverter configuration includes an H bridge circuit and a half bridge circuit. FIG. 11A shows a configuration example of an H bridge circuit that drives a winding load, and FIG. 11B shows a configuration example of a half bridge circuit that drives the winding load. In FIGS. 11A and 11B, the IGBT is applied to the switching device, but other switching devices may be used.

  The H bridge circuit controls the current of each phase with twice as many switching devices as the half bridge circuit. Therefore, it is possible to double the switching frequency of each phase without increasing the loss of each switching device. When the H bridge circuit and the half bridge circuit are controlled by the same carrier signal, the switching frequency of each switching device is the same, but the switching frequency of the winding load voltage vs (the number of times the pulsed voltage waveform changes) is: The H bridge circuit is twice the half bridge circuit.

  In the half bridge circuit, the voltage command value is updated in a half cycle of the carrier signal, whereas in the H bridge circuit, the voltage command value is updated in a quarter cycle of the carrier signal. In the half-bridge circuit, the current does not include a switching ripple at the time of the peak of the carrier signal, and in the H-bridge circuit, the current does not include the switching ripple at the time of the peak and intermediate point of the carrier signal. The current value fed back to the current control is sampled at a point not including these switching ripples.

  From the above, the current control cycle is a half cycle of the carrier signal in the half bridge circuit, whereas the H bridge circuit is a quarter cycle of the carrier signal, and the control cycle of the H bridge circuit is the half bridge circuit. The length is reduced to 1/2 of the length.

  Non-patent document 1 is known as a prior art for improving single-phase current control response using an H-bridge circuit. As shown in FIG. 1, a circuit in which an H bridge circuit is provided in three phases (a phase, b phase, and c phase) is also called an open winding system. Patent Documents 1 and 2 are disclosed as the prior art of the open winding system, and Non-Patent Document 2 is disclosed as the prior art for improving the current control response using the open winding system.

  FIG. 1 shows a configuration in which the induction machine IM is driven by a three-phase H-bridge circuit (open winding system). The H-bridge circuit includes a three-phase first inverter Inv1 and a three-phase second inverter Inv2. The first and second inverters Inv1, Inv2 have three switching devices connected in series. Between the neutral point of the two switching devices of each phase of the first inverter Inv1 and the neutral point of the two switching devices of each phase of the second inverter Inv2, the induction machine ("Open-end winding" in FIG. 3) IM Windings for each phase are provided. The induction machine IM has connection terminals at both ends of the winding of each phase (a phase, b phase, c phase). The voltage vs = [vsa, vsb, vsc] applied to the windings of each phase is the difference between the output voltages of the legs of the first inverter Inv1 and the second inverter Inv2. Therefore, when the output voltages of the first inverter Inv1 and the second inverter Inv2 are vi1 = [vi1a, vi1b, vi1c], vi2 = [vi2a, vi2b, vi2c], the winding voltage vs is expressed by the following equation (1). Become.

The inverter current (induction machine current) i _S = [i _Sa , i _Sb , i _Sc ] of each phase and the detected angular velocity value ωr of the induction machine are fed back to the control device 1, and the first and second inverters Inv1 from the control device 1 are fed back. , Inv2 is sent a gate signal for switching the switching device.

JP 2008-5688 A JP2013-55886A Japanese Patent Laid-Open No. 3-89885

H. Fujita, ‘‘ A single-phase active filter using an h-bridge pwm converter with a sampling frequency of the EfficientEffluentElectricsFrequencyEffects 24, no. 4, pp. 934-941, April 2009. H. Kubo, Y. et al. Yamato, K .; Kondo, K .; Rajashkara, and B.R. Zhu, ‘’ High bandwidth current control for open-end winding instruction motor, ‘in Industrial Electronics Society, IECON 2014-40th annualEffort. 607-613.

  In the current control of 1/4 cycle of the carrier signal in the H-bridge circuit, the current is sampled at the apex and the midpoint of the carrier signal. However, when the voltage command value crosses zero, the switching device switches at the time of the intermediate point of the carrier signal, so that the current measurement at the intermediate point of the carrier signal is affected by switching noise.

  FIG. 12 shows the waveform of each part when the voltage of the H-bridge circuit is controlled by the method of comparing the voltage command value and the triangular wave carrier signal.

The current i _S flowing through the winding load detects a value that does not include a switching ripple at the timing of the peak and intermediate points of the carrier signal (+ timing in the figure). On the other hand, switching noise is added to the output signal sis of the current sensor for detecting the current immediately after the switching device is switched as shown in FIG. When the voltage command values v ^* i1, v ^* i2 approach zero, switching occurs near the sampling point at the carrier intermediate point (eg, points a and b in FIG. 12), so that the current detection value input to the control device 1 Switching noise also appears on (is _{, abc in} FIG. 1).

  As a result, the control performed by the control device 1 may malfunction, and the open winding system may not operate normally.

  As described above, in the control device of the open winding system, it becomes a problem to suppress switching noise at the carrier midpoint while maintaining high response performance of current control.

  The present invention has been devised in view of the above-described conventional problems. One aspect of the present invention includes a first inverter having three phases of two switching devices connected in series and two switching devices connected in series. Each phase winding between a neutral point of the two switching devices of each phase of the first inverter and a neutral point of the two switching devices of each phase of the second inverter Is a control device for an open winding system, which calculates a current command value and a slip command value based on a torque command value and a magnetic flux command value, a current detection value, an angular velocity detection value, and a current A command value, a current control unit that calculates a voltage command value based on the slip command value, and a command value conversion unit that generates a voltage command value for the first inverter and a voltage command value for the second inverter based on the voltage command value The first inverter that compares the voltage command value of the first inverter with the carrier signal to generate the gate signal of the first inverter, the voltage command value of the second inverter and the carrier signal, A second comparator for generating a gate signal of the inverter, wherein the current control unit calculates a current predicted value based on the current detection value, the load voltage estimated value, and the voltage command value at the top of the carrier signal, At the intermediate point of the carrier signal, a current prediction unit that calculates a current prediction value based on the previous current prediction value, load voltage estimation value, and voltage command value, rotor magnetic flux estimation value or magnetic flux command value, angular velocity detection value, A load voltage estimation unit that calculates a load voltage estimated value based on the current predicted value, and a voltage command value calculation that calculates a voltage command value based on the current command value of one sampling destination, the current predicted value, and the load voltage estimated value Part It has, and updates a voltage command value in quarter period of the carrier frequency.

  Further, as one aspect thereof, the voltage command value calculation unit calculates a voltage command value based on the equation (9).

Note that v ^* _{s, k + 1} is the voltage command value, i ^* _{s, k + 2} is the current command value of one sampling destination, i _{s, k + 1} is the current detection value, and v ^ _{l, k + 1} is the load The estimated voltage value, L is the leakage inductance of the winding load, and Ts is the current control period.

  In another aspect, each phase of the first inverter includes: a first inverter having three phases of two switching devices connected in series; and a second inverter having three phases of two switching devices connected in series. A control device for an open winding system in which a winding of each phase is provided between a neutral point of each of the two switching devices and a neutral point of the two switching devices of each phase of the second inverter, the torque command A voltage command value is calculated based on the FOC unit for calculating the current command value and the slip command value based on the value and the magnetic flux command value, the detected current value, the detected angular velocity value, the current command value, and the slip command value. A current control unit; a command value conversion unit for generating a voltage command value for the first inverter and a voltage command value for the second inverter based on the voltage command value; a voltage command value for the first inverter and a carrier signal; The first comparator that generates the gate signal of the first inverter, and the second comparator that generates the gate signal of the second inverter by comparing the voltage command value of the second inverter and the carrier signal The current control unit calculates a current predicted value based on the current detection value, the load voltage estimated value, and the voltage command value at the apex of the carrier signal, and the previous current predicted value at the intermediate point of the carrier signal. A current prediction unit that calculates a current predicted value based on the load voltage estimated value and the voltage command value, a rotor magnetic flux estimated value or magnetic flux command value, an angular velocity detection value, and a current predicted value based on the load voltage estimated value A voltage command value calculation unit that calculates a voltage command value by performing PI control on the difference between the current command value one sampling period ahead and the current prediction value. The career And updates a quarter cycle of the wave number.

  As one aspect, the current prediction unit calculates a current prediction value based on Equation (5) at the top of the carrier signal, and calculates a current prediction value based on Equation (6) at an intermediate point of the carrier signal. It is characterized by doing.

In equation (5), i ^ _{S, k + 1} is a current predicted value, v ^* _{s, k} is a voltage command value, v ^ _{l, k} is a load voltage estimated value, i _{s, k} is a current detection value, ( 6) i ^ _{S, k} is the predicted current value, v ^* _{s, k-1} is the voltage command value, v ^ _{l, k-1} is the estimated load voltage, and i ^ _{s, k-1} is the previous current. Indicates the predicted value. L represents the leakage inductance of the winding load, and Ts represents the current control period.

  As one aspect thereof, the load voltage estimation unit calculates a load voltage estimated value based on the equation (7).

Here, v ^ _{l, k} is the estimated load voltage, i ^ _{s, k} is the predicted current value, ωr is the angular velocity detection value, Rs is the stator winding resistance of the winding load, Lr is the rotor inductance of the winding load, Lm is the mutual inductance of the winding load, and φ ^ r, k is the estimated value of the rotor magnetic flux.

  According to the present invention, in a control device for an open winding system, it is possible to suppress switching noise at a carrier midpoint while maintaining high response performance of current control.

Schematic which shows an example of an open winding system. The figure which shows the control apparatus of the open winding system in Embodiment 1. FIG. The figure which shows the conversion from a winding voltage command value to an inverter voltage command value. The timing chart of a current control part. The timing chart of the conventional electric current control part. FIG. 3 is a block diagram illustrating a current control unit according to the first embodiment. The block diagram which shows the conventional electric current control part. The figure which shows Embodiment 1 and the conventional current waveform. Bode diagram of current control. FIG. 4 is a block diagram illustrating a current control unit according to a second embodiment. Schematic which shows an H bridge circuit and a half bridge circuit. The figure which shows the comparison of a voltage command value and a carrier signal.

[Embodiment 1]
FIG. 1 is a diagram showing a power converter of a three-phase H-bridge circuit (open winding system) that drives the induction machine IM. FIG. 2 is a block diagram showing a control device 1 of a three-phase H-bridge (open winding system) that drives the induction machine IM.

  The control device 1 has both a speed control mode for controlling the rotational speed of the induction machine IM to a command value and a torque control mode for controlling the torque of the induction machine IM to a command value.

As shown in FIG. 2, in the speed control mode, the speed control unit 2 performs PI control, PID control, etc., based on the detected angular speed value ωr of the induction machine IM and the angular speed command value ωr ^* , and the torque command value T ^* _e. Is generated.

The switching unit 3 inputs a control mode (speed control mode or torque control mode). In the speed control mode, the value calculated by the speed control unit 2 is input. In the torque control mode, the value input from the outside is input. The torque command value T ^* _e is output.

The FOC (Field oriented control) unit 4 converts the torque command value T ^* _e and the magnetic flux command value φr ^* into the current command value i ^* _{s, dq0} (i ^* _{s, d0} i ^* _{s, q0} ) and the slip command ω ^* _sl . Converted and output to the current control unit 6. The conversion equation is the following equation (2).

In the equation (2), φr ^* is a magnetic flux command, LM is an exciting inductance of the induction machine IM, POLE is the number of poles of the induction machine IM, and Rr is a secondary resistance of the induction machine IM.

The coordinate converter 5 converts the detected current value _{s, abc} into the detected current value _{s, αβ0} on the αβ coordinate _{, and outputs} it to the current control unit 6. The current control unit 6 determines the voltage command value v ^* _s, based on the current detection value is _{, αβ0} on the αβ coordinate, the angular velocity detection value ωr, the current command value i ^* _{s, dq0,} and the slip command value ω ^* _{sl .} Calculate _abc . Details of the configuration of the current control unit 6 will be described later.

The voltage command value v ^* _{s, abc} output from the current control unit 6 is converted into a voltage command value v ^* _{i1, abc} of the first inverter Inv1 and a voltage command v ^* _{i2, abc} of the second inverter Inv2 by the command value converter 8. Converted.

FIG. 3 shows conversion from the voltage command value v ^* _s of the windings to the voltage command values v ^* _{i1, abc} , v ^* _{i2, abc of} the first and second inverters Inv1, Inv2. The conversion is performed by the following equation (3). Equation (3) satisfies the condition of Equation (1).

Finally, the voltage command values v ^* _{i1, abc} , v ^* _{i2, abc of} the first and second inverters Inv1, Inv2 and the triangular carrier signal output from the carrier signal generator 7 are compared by the comparators 9, 10. In comparison, gate signals of the first and second inverters Inv1, Inv2 are generated.

FIG. 4 shows a timing chart of the current control unit 6. The period of the current control is a quarter period of the carrier signal, and the sampling points are the vertex and the middle point of the carrier signal. The subscript k means the kth sampling time. The gate signal of each inverter is controlled so that the average voltage v￣s, k between sampling points approximates the voltage command value v ^* _{s, k} .

Note that the solid line of Average Voltage in FIG. 4 indicates the average voltage v ￣ _{s, k} , and the dotted line indicates the voltage command value v ^* _{s, k} . The current sampled current (current at the + position in FIG. 4) measured at the sampling point is a value that does not include switching ripples.

  Also, the current value sampled at the carrier apex is not affected by switching noise. On the other hand, the current value sampled at the carrier middle point may be affected by switching noise and is not used for control. The control at this time is performed based on the plant model discretized at the current control period Ts in the following equation (4). The current value at the next sample time is predicted from (4).

Here, L is a leakage inductance of the induction machine IM (that is, winding load), and vl is a load voltage composed of a voltage drop due to the counter electromotive force and winding resistance. The squares of “Current Prediction”, “Load voltage estimation”, and “Reference voltage calculation” indicate the calculation process of the current predicted value is, the load voltage estimated value vl, and the voltage command value v ^* _{s, k} , respectively, and the arrows indicate their calculation processes. Indicates the input and output of the calculation process.

In the section starting at the carrier vertex (between k-2 and k-1 and between k and k + 1 in FIG. 4), the voltage command value v ^* _{s, k} , the load voltage estimated value v ^ _{l, k} , and the current detection value i _{From s and k} , the predicted current value i ^ _{s, k + 1} is predicted by the following equation (5).

In the section starting at the carrier intermediate point (between k-1 to k and k + 1 to k + 2 in FIG. 4), the voltage command value v ^* _{s, k} , the estimated load voltage v ^ _{l, k} , and the previous current prediction value From i ^ _{s, k + 1} , the predicted current value i ^ _{s, k} is predicted by the following equation (6).

  Further, the estimated load voltage v ^ l, k is predicted from the following equation (7).

   The calculations of equations (4) to (7) are performed in a stationary coordinate system consisting of an α-axis component and a β-axis component. Therefore, the equations (4) to (7) are vector equations having an α-axis component and a β-axis component.

  When the expression (7) is divided into an α-axis component and a β-axis component, the following expression (8) is obtained.

  Here, Rs is a stator winding resistance, Lr is a rotor inductance, Lm is a mutual inductance, and φr ^ is an estimated value of the rotor magnetic flux. A method for estimating the rotor magnetic flux estimated value φr ^ is described in Patent Document 3.

  In Equations (13), (15), (17), and (18) of Patent Document 3, equations for estimating and calculating the magnetic flux from the current and rotation speed of the induction machine IM are shown. The method of Patent Document 3 may be used in the first embodiment, or another method may be used. Further, φr ^ in the equations (7) to (8) may be replaced with a magnetic flux command value. FIG. 6 is a control block diagram when φr ^ is replaced with a magnetic flux command value φr *. In this case, the magnetic flux command value φr * is input to the current control unit 6 in FIG.

Finally the current command value i ^* _{s, k + 2,} the current estimated value i ^ _{s, k + 1} and the load voltage estimated value vl ^, the voltage command value of the next interval from _{^{k + 1 v * s, k}} + 1 is It is calculated by the following equation (9).

  For comparison, FIG. 5 shows a timing chart of the conventional method. In the conventional method, current estimation is always predicted from the detected current value using equation (5). On the other hand, the first embodiment is characterized in that the equation (5) and the equation (6) are switched depending on whether the vertex of the carrier signal or the intermediate point of the carrier signal.

FIG. 6 shows a block diagram of the current control unit 6. Also, the magnetic flux command value φ ^* _rd of the d-axis component of the synchronous coordinate system is converted by the converter 14 from the synchronous coordinate system dq coordinate to the stationary coordinate system α-β coordinate value φ ^* _rαβ . (Here, the q-axis component magnetic flux command value φ ^* _rq = 0 in the synchronous coordinate system is not shown in FIG. 6).
The electrical angle calculation unit 12 uses the slip command value ω _sl ^* and the rotor rotation speed detection value ωr of the induction machine IM to calculate the electrical angles θ ^ e, k + 1, θ ^ e, necessary for coordinate conversion in the converters 11 and 14. k + 2 is calculated. The electrical angle calculation unit 12 has an integration function, and its output is calculated by the following equation (10).

The calculation of the current control unit 6 is performed in the stationary coordinate system α-β coordinate. The current command value i ^* _{sdq, k + 2} is converted by the converter 11 into a value i ^* _{sαβ, k + 1} from the synchronous coordinate system dq coordinate to the stationary coordinate system α-β coordinate.

When the sampling point is the peak of the carrier signal, the switch 13 sets the input to the current prediction unit 15 as the current detection value i _{sαβ, k,} and when the sampling point is the carrier intermediate point, sets the previous current prediction value i ^ _{sαβ, k} . .

The current predicting unit 15 predicts the current using the equations (5) and (6) based on the output of the switch 13, the load voltage estimated value v ^ _{slαβ, k,} and the voltage command value v ^* _{sαβ, k.} The value i ^ _{sαβ, k + 1} is calculated. The buffer 16 outputs to the switch 13 the value i ^ _{sαβ, k} one control period before the current predicted value i ^ _{sαβ, k + 1} .

Based on the magnetic flux command value φ ^* _rαβ , the detected angular velocity value ωr, and the predicted current value i ^ _{sαβ, k + 1} , the load voltage estimated value 17 is calculated by the equation (7) according to the load voltage estimated value v ^ _{αβ, k + 1} is calculated. The buffer 18 outputs the value v ^ _{lαβ, k} one control cycle before the estimated load voltage value v ^ _{αβ, k + 1} to the current prediction unit 15.

Based on the current command value i ^* _{sαβ, k + 2} , the current predicted value i ^ _{sαβ, k + 1,} and the load voltage estimated value v ^ _{lαβ, k + 1} , 9) The voltage command value v ^* _{sαβ, k + 1} is calculated from the equation (9). Buffer 19 outputs the voltage command values _v ^* _{sαβ, k + 1} of one control period previous value _v ^* _sαβ, a _k to the current prediction unit 15.

For comparison, FIG. 7 shows a current control unit of a conventional method. The conventional method uses the detected current value i _{sαβ, k} for the calculation of the current prediction unit 15 regardless of whether it is the apex or the intermediate point of the carrier signal, so the input to the current prediction unit 15 is switched depending on whether it is the apex of the carrier signal. There is no function.

Although not shown in FIGS. 6 and 7, the current control unit 6 uses the three-phase voltage command values v ^* _sa , v ^* _sb , v as the voltage command values v ^* _sα and v ^* _sβ in the stationary coordinate system. ^* There is a two-phase three-phase converter that converts to _sc . The output of the current control unit 6 shown in FIG. 2 is v ^* _sa , v ^* _sb , and v ^* _sc of the output of the two-phase / three-phase converter.

  With the above configuration, the influence of switching noise at the carrier midpoint can be avoided while controlling the current at a quarter of the carrier signal.

  FIG. 8 shows that “hh” is obtained when the current is controlled in a 1/2 cycle of the carrier signal, and “qq” is obtained when the current is controlled in a 1/4 cycle of the carrier signal using the current value sampled in the 1/4 cycle of the carrier signal. ”, And the current waveform of“ qh ”when the current value sampled at the carrier apex of the first embodiment and the current prediction value are switched and the current is controlled in a quarter cycle of the carrier signal. These current waveforms are currents flowing through the winding load of the open winding system.

  Since “qq” uses a current detection value including switching noise for current control, the current waveform is disturbed. The current waveform is not disturbed in “qh” of the first embodiment.

  FIG. 9 shows a Bode diagram of current control of “hh”, “qq”, and “qh” obtained by inputting a high frequency from 200 Hz to 3000 Hz as a current command value. From the Bode diagram of Phase (phase), “qh” in the first embodiment, in which the current is controlled in a quarter cycle of the carrier signal, is compared with “hh” in which the current is controlled in a half cycle of the carrier signal. It can be seen that the phase delay is small.

  As described above, according to the first embodiment, when the current of the H-bridge circuit is controlled at a quarter cycle of the carrier signal, the current detection value at the carrier peak and the current prediction of the previous control cycle at the carrier middle point. By using the value, it is possible to avoid the influence of the switching noise at the middle point of the carrier while maintaining the high response performance of 1/4 current control of the carrier signal of the H bridge circuit.

  Thereby, malfunction of control performed by the controller can be suppressed, and the reliability of the open winding system can be improved.

[Embodiment 2]
In the second embodiment, instead of obtaining the voltage command value v ^* _{s, k + 1} by the equation (9) by the voltage command value calculation unit 20, the current command value i ^* _{s, k + 2} and the predicted current value i ^ _{s , k + 1} is obtained by PI control.

  FIG. 10 shows a current control unit when the current is PI-controlled in the second embodiment. The calculation of the current control unit 6 is performed in the synchronous coordinate system dq coordinates.

  With the above configuration, the influence of switching noise at the carrier midpoint can be avoided while controlling the current at a quarter of the carrier signal.

  In the first and second embodiments, the load of the open winding system is not limited to the induction machine IM shown in FIG.

Inv1, Inv2 ... first and second inverters i _s ... current detection value i ^* _s ... current command value i ^ _s ... current prediction value v ^ _s ... load voltage estimation value T ^* _e ... torque command value φr ^* ... magnetic flux command Value ω ^* _sl ... slip command value v ^* _s ... voltage command value 5 ... coordinate converter 6 ... current control unit 7 ... carrier signal generation unit 8 ... command value conversion unit 9, 10 ... comparator

Claims (5)

A first inverter having three phases of two switching devices connected in series, and a second inverter having three phases of two switching devices connected in series, wherein two switching devices of each phase of the first inverter A control device for an open winding system in which a winding of each phase is provided between a sex point and a neutral point of two switching devices of each phase of a second inverter,
An FOC unit that calculates a current command value and a slip command value based on the torque command value and the magnetic flux command value;
A current control unit that calculates a voltage command value based on a current detection value, an angular velocity detection value, a current command value, and a slip command value;
A command value conversion unit that generates a voltage command value of the first inverter and a voltage command value of the second inverter based on the voltage command value;
A first comparator that compares the voltage command value of the first inverter with the carrier signal and generates a gate signal of the first inverter;
A second comparator that compares the voltage command value of the second inverter with the carrier signal and generates a gate signal of the second inverter;
The current controller is
At the top of the carrier signal, the current predicted value is calculated based on the current detection value, the load voltage estimated value, and the voltage command value, and at the intermediate point of the carrier signal, the previous current predicted value, load voltage estimated value, and voltage command value are calculated. A current prediction unit for calculating a current predicted value based on the current value;
A load voltage estimation unit that calculates a load voltage estimation value based on a rotor magnetic flux estimation value or a magnetic flux command value, an angular velocity detection value, and a current prediction value;
A voltage command value calculation unit that calculates a voltage command value based on a current command value of one sampling destination, a current prediction value, and a load voltage estimation value;
A control device for an open winding system, wherein a voltage command value is updated at a quarter cycle of a carrier frequency. The voltage command value calculator is
2. The control device for an open winding system according to claim 1, wherein the voltage command value is calculated based on the equation (9).
Note that v ^* _{s, k + 1} is the voltage command value, i ^* _{s, k + 2} is the current command value of one sampling destination, i _{s, k + 1} is the current detection value, and v ^ _{l, k + 1} is the load The estimated voltage value, L is the leakage inductance of the winding load, and Ts is the current control period. A first inverter having three phases of two switching devices connected in series, and a second inverter having three phases of two switching devices connected in series, wherein two switching devices of each phase of the first inverter A control device for an open winding system in which a winding of each phase is provided between a sex point and a neutral point of two switching devices of each phase of a second inverter,
An FOC unit that calculates a current command value and a slip command value based on the torque command value and the magnetic flux command value;
A current control unit that calculates a voltage command value based on a current detection value, an angular velocity detection value, a current command value, and a slip command value;
A command value conversion unit that generates a voltage command value of the first inverter and a voltage command value of the second inverter based on the voltage command value;
A first comparator that compares the voltage command value of the first inverter with the carrier signal and generates a gate signal of the first inverter;
A second comparator that compares the voltage command value of the second inverter with the carrier signal and generates a gate signal of the second inverter;
The current controller is
At the top of the carrier signal, the current predicted value is calculated based on the current detection value, the load voltage estimated value, and the voltage command value, and at the intermediate point of the carrier signal, the previous current predicted value, load voltage estimated value, and voltage command value are calculated. A current prediction unit for calculating a current predicted value based on the current value;
A load voltage estimation unit that calculates a load voltage estimation value based on a rotor magnetic flux estimation value or a magnetic flux command value, an angular velocity detection value, and a current prediction value;
A voltage command value calculation unit that calculates a voltage command value by performing PI control on the difference between the current command value one sampling period ahead and the current prediction value;
A control device for an open winding system, wherein a voltage command value is updated at a quarter cycle of a carrier frequency. The current prediction unit calculates a current prediction value based on Equation (5) at the top of a carrier signal, and calculates a current prediction value based on Equation (6) at an intermediate point of the carrier signal. Item 4. The control device for an open winding system according to any one of Items 1 to 3.
In equation (5), i ^ _{S, k + 1} is a current predicted value, v ^* _{s, k} is a voltage command value, v ^ _{l, k} is a load voltage estimated value, i _{s, k} is a current detection value, ( 6) i ^ _{S, k} is the predicted current value, v ^* _{s, k-1} is the voltage command value, v ^ _{l, k-1} is the estimated load voltage, and i ^ _{s, k-1} is the previous current. Indicates the predicted value. L represents the leakage inductance of the winding load, and Ts represents the current control period. 5. The control device for an open winding system according to claim 1, wherein the load voltage estimation unit calculates a load voltage estimated value based on an expression (7). 6.
Here, v ^ _{l, k} is the estimated load voltage, i ^ _{s, k} is the predicted current value, ωr is the angular velocity detection value, Rs is the stator winding resistance of the winding load, Lr is the rotor inductance of the winding load, Lm is the mutual inductance of the winding load, and φ ^ r, k is the estimated value of the rotor magnetic flux.