KR101073197B1 - Power converter - Google Patents

Abstract

For the purpose of precisely correcting the voltage error caused by the dead time near the zero cross of the output current, the power conversion means can obtain a power converter capable of outputting AC power like an AC voltage command. The first time when the voltage command correction means provided between the command calculation means provides a predetermined current range including the zero level with respect to the detected output current, and the value of the output current enters from the outside from the outside of the current range. Is based on the first time and frequency (also the output current), and the timing of the zero cross of the output current is a time for switching the polarity of the correction voltage for correcting the AC voltage command that the voltage command calculation means calculates and outputs. Set, that is, recall the polarity switching of the correction voltage to correct the voltage error caused by the dead time By applying force at the current zero-cross, before and after the zero-crossing of the output current is negative and donggeuk correction voltage resistance of the output current was a structure capable of performing the correction of the AC voltage command.

Description

Power converter {POWER CONVERTER}

The present invention relates to a power converter for supplying AC power to an AC load.

In the power converter, the upper and lower arm power device elements constituting the power conversion means provided at the output stage perform an switching operation based on a voltage command to generate an alternating voltage and output the alternating voltage to the alternating current load. For the purpose of preventing short circuit due to conduction, a period for simultaneously controlling the power device elements of the upper and lower arms in an off operation state is provided. Although this period is called dead time, it is known that the dead time causes an error between the voltage command received by the power converting means and the voltage actually output to the load by the power converting means. The error voltage resulting from this dead time becomes a voltage of a polarity opposite to that of the output phase current of the power conversion means.

In order to correct the voltage error caused by this dead time, the output phase current of the power conversion means is detected by the current detecting means provided in the power converter, and the voltage having the same polarity as that of the detected output phase current is applied. A method of applying the voltage command corrected in addition to the command to the power converting means is known (for example, Non-Patent Document 1). Moreover, as a method of current detection, besides directly detecting the output phase current of a power conversion means, the method of detecting from the DC link current of a power converter is also known (for example, nonpatent literature 2).

However, in the vicinity of the current zero cross where the polarity of the output phase current is switched, since the absolute value of the output phase current is small, it is difficult to accurately detect the polarity of the output phase current by the current detecting means. Further, in the vicinity of the current zero cross, the output phase current chatters, so in the correction method of applying a voltage having the same polarity as the polarity of the output phase current to the voltage command, a polarity error of the correction voltage applied to the voltage command is caused. It occurs, or the phenomenon that the polarity of the correction voltage is continuously switched alternately occurs.

Therefore, in order to avoid the problem of the dead time correction in the vicinity of current zero cross where the absolute value of the output phase current is small, various proposals have been made (for example, Patent Documents 1 to 4 and the like).

That is, in patent document 1, the output voltage error correction apparatus of the inverter apparatus which correct | amends the output voltage error resulting from the short circuit prevention period of the output voltage of the inverter apparatus is a current detection means which detects the output current of an inverter apparatus, and this current. Current polarity discriminating means for discriminating the polarity of the output current detected by the detecting means, and output voltage according to the determined polarity of the output current when the output current of the inverter device is out of the threshold with respect to the threshold set for the output current. An example is disclosed in which a voltage error correction means that corrects an error and corrects an output voltage in accordance with the polarity of a voltage command when within a threshold is disclosed.

In addition, Patent Document 2 configures an inverter that bridges an arm, which is a semiconductor switching element, and converts direct current into alternating current, and describes a pulse width modulated voltage signal obtained by comparing the magnitude order between the voltage command signal and the carrier signal. When controlling the inverter by applying to the upper arm element or the lower arm element of the bridge connection, the control circuit of the pulse width modulation control inverter is provided with an on delay time to prevent the simultaneous turning on of the upper arm element and the lower arm element, Current polarity discriminating means for outputting a signal for discriminating the polarity of the output current when the output current of the inverter exceeds a positive or negative predetermined value, and voltage polarity discriminating means for outputting a signal for discriminating the polarity of the voltage command signal; And the error voltage generated in the output voltage of the inverter due to the on delay time. First compensation amount calculation means for calculating an amount, compensation amount distribution means for outputting the polarity of the first compensation amount calculation value in correspondence with the current polarity and voltage polarity, and the compensation amount distribution means for the voltage command signal. An example is disclosed in which an adding means for setting a value obtained by adding an output value to a new voltage command signal is provided.

In Patent Document 3, current detecting means for detecting a current flowing through an AC motor, deviation current calculating means for calculating a deviation current from a command current and a detected current, a current control unit for calculating a command voltage from the deviation current, A current addition determining means for determining a current polarity, a dead time compensation means for outputting a dead time compensation voltage, and a voltage addition operation for calculating a final command voltage by applying the dead time compensation voltage to a command voltage calculated from the current controller; Means for converting a DC voltage into an AC voltage using information of a final command voltage obtained by said voltage adding and calculating means, wherein said dead time compensation means comprises: Command current polarity determining means for determining the command current polarity from the command current; An example is disclosed in which the detection current polarity determination means for determining the detection current polarity from the detection current and the final current polarity determination means for determining the final current polarity from the information of the command current polarity and the detection current polarity are disclosed.

Further, in Patent Document 4, the voltage correction means for correcting the inverter output voltage by the upper and lower arm short circuit prevention period and the error voltage calculated using the PWM carrier frequency and the DC voltage has a larger absolute value than the predetermined value. In the PWM control inverter device to correct the inverter output voltage according to the polarity of the inverter output current, and to correct the inverter output voltage according to the polarity of the inverter output voltage when the absolute value of the inverter output current is less than a predetermined value, the PWM carrier The PWM carrier setting means for setting the frequency maintains the PWM carrier frequency at that value when the error voltage is less than the inverter output voltage, and maintains a constant ratio between the error voltage and the inverter output voltage when the error voltage is greater than the inverter output voltage. An example is disclosed in which the PWM carrier frequency is configured to be maintained to change.

Patent Document 1: Japanese Patent No. 2756049 (16 pages, Fig. 11)

Patent Document 2: Japanese Patent No. 3245989 (page 12, Fig. 7)

Patent Document 3: Japanese Patent Laid-Open No. 2004-112879 (6 pages, FIG. 2)

Patent Document 4: Japanese Patent No. 3287186 (8 pages, Fig. 3)

[Non-Patent Document 1] Sugimoto, Koyama, Tamai: Theory and Design of AC Servo Systems: A Comprehensive Electronic Publisher (Page 55, Lines 9 to 57, Line 5)

Non-Patent Document 2: "Three-Phase Current-Waveform-Detection on PWM Inverters from DC Link Current-Steps" IPEC-Yokohama'95 p.p.271-275)

DISCLOSURE OF INVENTION

Problems to be Solved by the Invention

However, in such a conventionally proposed correction method, the polarity of the correction voltage cannot be reliably switched at zero cross of the output phase current, and the instant of inversion of the polarity shifts before and after the zero cross of the output phase current. In the vicinity of zero cross, a voltage error occurs between the corrected voltage command input to the power converting means, that is, the voltage command plus the corrected voltage and the output voltage from the power converting means.

In particular, this voltage error becomes remarkable when the AC power of low frequency is output from the power conversion means. When the rotor is connected, the rotational unevenness increases due to this voltage error, and the driving performance is lowered. In addition, there is a problem such that the rotational unevenness becomes remarkable during low speed driving.

This invention is made | formed in view of the above, Comprising: The voltage error resulting from the dead time in the zero cross vicinity of an output current is correct | amended precisely, and the power conversion means obtains the power converter which can output AC power like a voltage command. For the purpose of

Means to solve the problem

In order to achieve the above object, the present invention calculates the power conversion means for generating AC power supplied to the AC load in accordance with the input AC voltage command and the AC voltage command of the frequency f to be applied to the power conversion means. A power converter comprising: voltage command calculation means; and current detection means for obtaining an output current which is a current component in AC power supplied from the power conversion means to the AC load, wherein the power conversion means and the voltage command calculation are performed. Between means, the correction voltage which correct | amends the AC voltage command calculated | required by the voltage command calculation means based on the said output current is calculated based on the voltage error resulting from the dead time in the said power conversion means, and the said correction voltage is mentioned above. The voltage command applied to the power conversion means by adding to the AC voltage command obtained by the voltage command calculation means. Providing a correction means, wherein the voltage command correction means provides a predetermined current range including the zero level with respect to the output current, and a first time at which the value of the output current enters from outside of the current range. The zero cross timing of the output current is obtained using the first time and the frequency f, and the obtained zero cross timing is set to a time for switching the polarity of the correction voltage.

According to the present invention, since the polarity of the correction voltage for correcting the AC voltage command can be switched at the moment when the detected output current becomes zero, that is, the polarity and the same polarity of the output current before and after the zero cross of the output current. Since the AC voltage command is corrected by the correction voltage of the castle, it is possible to reduce the voltage error caused by the dead time between the corrected AC voltage command applied to the power conversion means and the voltage output by the power conversion means near the zero cross of the output current. The power conversion means can output the AC power like the AC voltage command. Therefore, when the output of the power conversion means is applied to the rotating machine, the rotation unevenness of the rotating machine can be reduced, and the driving performance can be improved.

Effects of the Invention

According to the present invention, since the voltage error caused by the dead time in the zero cross vicinity of the output current can be corrected precisely, the power conversion means has an effect that the AC power can be output like the voltage command.

1 is a block diagram showing a configuration of a power converter according to Embodiment 1 of the present invention;

2 is a flowchart for explaining a correction operation procedure of the voltage command correction means shown in FIG. 1;

3 is a view for explaining the processing contents in ST10 to ST16 shown in FIG. 2;

4 is a diagram showing an example of a relationship characteristic table between a frequency and a correction coefficient K1 used in the processing of ST16 shown in FIG. 2;

5 is a diagram illustrating an example of a relationship characteristic table between an output current and a correction coefficient K2 used in the processing of ST16 shown in FIG. 2;

FIG. 6 is a view showing an example of a relationship characteristic table between frequency and time variation ta used when the load is an induction machine in the processing of ST16 shown in FIG. 2;

FIG. 7 is a time chart illustrating collectively the correction operation of the voltage error due to the dead time realized by the correction processing shown in FIG. 2;

8 is a flowchart for explaining a correction operation procedure of the voltage command correction means included in the power converter according to the second embodiment of the present invention;

9 is a flowchart for explaining a correction operation sequence of the voltage command correction means included in the power converter according to the third embodiment of the present invention;

10 is a block diagram showing a configuration of a power converter according to Embodiment 4 of the present invention;

11 is a flowchart for explaining a correction operation procedure of the voltage command correction means shown in FIG. 10;

12 is a flowchart for explaining a correction operation sequence of the voltage command correction means included in the power converter according to the fifth embodiment of the present invention;

FIG. 13 is a time chart illustrating the correction operation of the voltage error due to the dead time realized in the correction processing shown in FIG. 12;

14 is a flowchart for explaining a correction operation sequence of the voltage command correction means included in the power converter according to the sixth embodiment of the present invention;

FIG. 15 is a flowchart for explaining a correction operation sequence of the voltage command correction means included in the power converter according to the seventh embodiment of the present invention; FIG.

16 is a view for explaining the necessity of adjusting the amplitude of the correction voltage as the eighth embodiment of the present invention;

Fig. 17 is a diagram showing an example of a relationship characteristic table between the amplitude of the correction voltage and the value of the output current used in the amplitude adjustment of the correction voltage.

Explanation of the sign

1a, 1b: power converter

2a: AC load

2b: Rotator as an example of AC load

3: power conversion means

4: current detection means

5a, 5b: control unit

6a, 6b: voltage command calculation means

7a, 7b: voltage command correction means

8: Torque command setting means

Best Mode for Carrying Out the Invention

Hereinafter, with reference to the drawings, a preferred embodiment of the power converter according to the present invention will be described in detail.

(Embodiment 1)

1 is a block diagram showing a configuration of a power converter according to Embodiment 1 of the present invention. In addition, although this embodiment demonstrates the power converter which outputs 3-phase AC power as an example in the Embodiment 1 shown below, the content is applicable also to the power converter which outputs single-phase AC power similarly.

The power converter 1a shown in FIG. 1 controls the power conversion means 3, the current detection means 4, and the power conversion means 3 to which the three-phase AC load 2a is connected, and the power conversion means 3 ) Is provided with a control device 5a for controlling supply of AC power from the AC load 2a to the AC load 2a, and the control device 5a is composed of a voltage command calculation means 6a and a voltage command correction means 7a. . The voltage command correction means 7a is provided between the output side of the voltage command calculation means 6a and the input side of the power conversion means 3.

The power conversion means 3 is an alternating voltage command vu **, vv which the power device elements of the upper and lower arms input from the control apparatus 5a (in FIG. 1, the voltage command correction means 7a outputs). **, vw **) performs a switching operation to generate an AC voltage and output it to the AC load 2a, but the control device 5a aims to prevent short circuits due to simultaneous conduction of the power device elements of the upper and lower arms. As an AC voltage command (in Fig. 1, AC voltage commands vu **, vv **, vw ** output by the voltage command correcting means 7a) applied to the power conversion means 3, the power of the upper and lower arms is shown. It includes a dead time voltage command that simultaneously controls the device device in the off operation state.

The current detecting means 4 is an example of the current detecting means in the present invention, which obtains current information necessary when the power converter 1a (control device 5a) implements the present invention. In this specification, since the main purpose is to describe the voltage command correction means 7a provided in the control device 5a, the output (three-phase output currents iu, iv, iw) of the current detection means 4 is a voltage command. It is supposed to be applied to the correction means 7a.

Hereinafter, the specific content of the current detection means in this invention which obtains the electric current information required when the power converter 1a (control apparatus 5a) implements this invention is shown. In FIG. 1, as a specific example, the current detection means 4 which directly detects the output current from the power conversion means 3 is shown. And in the case of direct detection, in FIG. 1, the current detection means 4 is arrange | positioned in all of a three phase output line so that the output current iu, iv, iw of three phases from the power conversion means 3 may be detected. Although the case is shown, it is not necessary to arrange all the 3-phase output lines, and since there is a relationship of iu + iv + iw = 0, the current detection means 4 is arrange | positioned at the 2-phase output line of 3 phases, and the 2 A method of detecting the output current of the phase and calculating the output current of the remaining one phase may be used.

In addition, although FIG. 1 shows the case where the current of each phase is detected directly, besides, the method of detecting the said output current from the DC link current of the power conversion means 3 introduce | transduced in Nonpatent literature 2 is used, for example. Also good. In addition, since the electric current information required when the power converter 1a (control device 5a) implements the present invention has various forms in connection with the configuration method of the voltage command calculating means 6a, it is shown.

First, the voltage command calculating means 6a calculates the (three phase) alternating voltage commands vu *, vv *, vw * of the frequency f to be applied to the power converting means 3 inherently, but this AC voltage command vu *, vv *, vw * with the three-phase output currents iu, iv, iw determined by the current detection means 4 and the current commands iu *, iv *, iw * to be output from the power conversion means 3; The structure which amplifies and outputs a deviation can be employ | adopted. In this case, the current information required when the power converter 1a (control device 5a) implements the present invention includes any of output currents iu, iv, iw, or current commands iu *, iv *, iw *. It may be a side.

When the AC load 2a is a three-phase rotating machine, the voltage command calculating means 6a configures a magnetic flux observer therein and converts the output using the estimated magnetic flux inside the rotating machine and the constant of the rotating machine. The structure which calculates AC voltage command vu *, w *, vw * using the current iu, iv, iw can also be employ | adopted. The output currents iu, iv, and iw obtained by this calculation are also included in the current information required when the power converter 1a (control device 5a) implements the present invention.

The output currents iu, iv, iw and the current commands iu *, iv *, iw * obtained by the various methods demonstrated above represent the current component in the AC power which the power conversion means 3 supplies to the AC load 2a. The information detecting means according to the present invention includes all of the means for obtaining the output currents iu, iv, iw and the current commands iu *, iv *, iw * described above. This is the same even if the power converter 1a outputs single-phase AC power. In the following, for ease of understanding, the three-phase output currents iu, iv, iw directly detected by the current detection means 4 are applied to the voltage command correction means 7a according to the configuration shown in FIG. 1. do.

Next, the frequency f used by the voltage command calculating means 6a is the frequency of the AC voltage commands vu *, vv *, and vw *, that is, the frequency of the AC power output from the power converting means 3. In the steady state, it is equal to the frequencies of the output currents iu, iv, iw, current commands iu *, iv *, iw *, and output voltages vu, vv, vw from the power converting means 3. When the AC load 2a is a synchronizer, the rotation frequency (electric angle) is also the same.

When the AC load 2a is an inductor, the frequency of the AC power output from the power converting means 3 and the rotation frequency (electric angle) of the inductor are different by the slip frequency. The slip frequency may be calculated using a known technique, and the slip frequency may be added to the rotational frequency to determine the frequency of the AC power output from the power conversion means 3, and may be used as the frequency f.

By the way, the voltage command correction means 7a calculates correction voltages Δvu, Δvv, and Δvw for correcting the AC voltage commands vu *, vv *, and vw * obtained by the voltage command calculation means 6a, and the correction voltages Δvu, AC voltage commands vu **, vv **, vw **, which are corrected by adding Δvv and Δvw to the phases corresponding to the AC voltage commands vu *, vv *, and vw *, are output to the power conversion means 3. That is, AC voltage command vu ** is vu ** = vu * + Δvu, AC voltage command vv ** is vv ** = vv * + Δvv, and AC voltage command vw ** is vw ** = vw * + Δvw.

At that time, the voltage command correction means 7a is, for example, in the order shown in FIG. 2, at a dead time set for the purpose of preventing short circuits by simultaneous conduction of the power device elements of the upper and lower arms constituting the power conversion means 3. The process of precisely reducing the voltage error in the vicinity of the zero cross of the resulting output currents iu, iv and iw is performed based on the output currents iu, iv and iw.

Hereinafter, with reference to FIGS. 1-7, the correction operation | movement of the voltage error resulting from the dead time which the voltage command correction means 7a which concerns on this Embodiment 1 performs is demonstrated. However, since the correction operation is performed for each of the u, v, and w phases, in order to facilitate the explanation, each phase of the u, v, w is denoted by x and described. That is, the current of each phase is expressed by ix, and the voltage of each phase is described by noting vx.

2 is a flowchart for explaining a correction operation procedure of the voltage command correction means 7a. In FIG. 2, the step which shows a process sequence is abbreviated as ST. It is a figure explaining the process content in ST10-ST16 shown in FIG. FIG. 4 is a diagram illustrating an example of a relationship characteristic table between the frequency and correction coefficient K1 used in the processing of ST16 shown in FIG. 2. FIG. 5 is a diagram showing an example of a relationship characteristic table between the output current and the correction coefficient K2 used in the processing of ST16 shown in FIG. 2. It is a figure which shows an example of the relationship characteristic table of the frequency and time change amount ta used when a load is an induction machine in the process of ST16 shown in FIG. FIG. 7 is a time chart illustrating the correction operation of the voltage error due to the dead time realized by the correction processing shown in FIG. 2.

In FIG. 2, in ST10, the voltage command correction means 7a adopts the output current ix of one phase among the output currents of the three phases detected by the current detection means 4, and detects the current value. In FIG. 3, the horizontal axis represents time t and phase θ, and the vertical axis represents amplitude (effective value Irms) of the detected one-phase output current ix, while reference numeral 29 shown in FIG. 3 indicates the one-phase output current ix thus detected. The waveform in one cycle of is shown.

In ST11, the voltage command correction means 7a uses the upper limit of the predetermined current range 30 including the zero level from the zero level to the anode side in one cycle of the output current 29 of the one-phase output shown in FIG. It is determined by adjusting and setting the lower limit value -ΔI2 from ΔI1 and the zero level to the cathode side, respectively. In addition, you may set the value (absolute value) of the upper limit (DELTA) I1 and the lower limit -ΔI2 to the same value. In addition, in consideration that the offset may be added to the output current ix depending on the performance of the current detecting means 4, the upper limit value ΔI1 and the lower limit value −Δ2 may be adjusted to different values, respectively. . In any case, the upper and lower limits of the current range 30 are determined by a phenomenon in which the offset is superimposed on the output current ix by the current detecting means 4 or by a current chattering phenomenon near the zero cross of the current. It is important to adjust and set ix so that the detection of the polarity is not wrong. Since there are various types of specific adjustment and setting methods of the predetermined current range 30 including the zero level, these are collectively described later.

In ST12, the voltage command correction means 7a determines whether the value of the output current ix is greater than or equal to the upper limit value ΔI1 or less than or equal to the lower limit -ΔI2, that is, the value of the output current ix detected at the sample timing is equal to the current range ( It is determined whether it is outside the range of 30). As a result, when the value of the output current ix is out of the range of the current range 30 (ST12: YES), the polarity of the output current ix is recorded in the polarity setting signal signx (ST13), and each process of ST21 to ST23 is performed. Then, the value of the output current ix detected at the previous sample timing is updated to the value of the output current ix detected at the sample timing, and the process shifts to ST10 at the next sample timing.

In ST12 via ST10 and ST11 executed at a predetermined sample timing thereafter, when the value of the output current ix detected at the sample timing is within the range of the current range 30 (ST12: NO), the voltage command correction means. (7a) indicates, in ST14, whether the value of the detected value ix [N-1] of the output current ix before one sampling is equal to or greater than the upper limit value ΔI1 or less than or equal to the lower limit -ΔI2, that is, the output current before one sampling. Determine whether the value of ix is outside the 위 ² range of the current range 30.

That is, in ST14, when the value of the output current ix detected at the sample timing is within the range of the current range 30 (ST12: NO), the voltage command correcting means 7a is the first in the sample timing. Whether it means that it is detected that it has entered the range from outside the range of the current range 30, or has already entered the range from outside the range of the current range 30 at the timing of one sample, It is examined whether or not it means that the detection of keeping the state within the range of (30) is meant.

Then, the voltage command correction means 7a determines that the value of the output current ix detected at the sample timing has entered the range from outside the range of the current range 30 for the first time at the sample timing (ST14: YES). Detects the time which entered the range from outside the range of the current range 30 as the first time (symbols 32 and 34 shown in FIG. 3), and executes the respective processes of ST15 and ST16. Although the processing order of ST15 and ST16 may be reversed, in ST15, the polarity of the detected value ix [N-1] of the output current ix before one sampling is recorded in the polarity setting signal signx, and in ST16, the range of the current range 30 At the first time (symbols 32 and 34 shown in Fig. 3) which is the time from the outside into the range, the time for switching the polarity of the correction voltage Δvx for correcting the AC voltage command vx * is set as follows.

The voltage error between the AC voltage command vx * and the output voltage vx from the power converting means 3 due to the dead time occurring near the zero cross of the output current ix mainly occurs at a voltage having a polarity opposite to that of the output current ix. Therefore, in order to precisely perform voltage correction, in Fig. 3, before and after the zero cross 31 of the output current ix, the correction voltage? Vx having the same polarity as the polarity of the output current ix is applied to the AC voltage command vx *. It is good to add. That is, the zero cross 31 of the output current ix is the ideal correction voltage polarity switching time. In ST16, based on this, at the first time 32, 34, the time when the polarity of the output current ix is switched, that is, the zero cross 31 of the output current ix, is converted into the first time 32 and the frequency f. And the time at which the polarity of the correction voltage Δvx is switched so as to be obtained based on the output current ix and to match the timing of the current zero cross 31 in which the polarity of the output current ix is switched.

In setting the time for switching the polarity of the correction voltage Δvx, first, it is necessary to accurately determine the amount of time change or phase change from the first time 32, 34 to the current zero cross 31 in which the polarity of the output current ix is switched. Since there is, it is obtained as a theoretical value as follows.

In Fig. 3, when the output current ix is bipolar, the current whose polarity of the output current ix is switched from the first time 32 when the value (effective value Irms) falls within the range from outside the range of the current range 30. The amount of phase change θa up to the zero cross 31 is obtained by applying the value of the output current ix (effective value Irms) and the upper limit value ΔI1 to equation (1). In addition, in order to obtain Formula (1), it is assumed that the period from the first time 32 to the current zero cross 31 in which the polarity of the output current ix is switched is sufficiently short compared with the period of alternating current.

Then, converting from the phase change amount θa obtained using Equation (1) to the time change amount ta that varies from the first time 32 at the frequency f to the time 31 when the polarity of the output current ix is switched, (2) is obtained. Is the circumference.

Further, when the output current ix is negative, the current zero cross at which the polarity of the output current ix is switched from the first time 34 when the value (effective value Irms) falls within the range from outside the range of the current range 30. The formula for calculating the phase change amount θb to (31) is such that the upper limit value ΔI1 in the formula (1) is replaced with the lower limit value -ΔI2, and the current zero cross at which the polarity of the output current ix is switched from the first viewing angle 32 ( The formula for calculating the time variation tb up to 31) is a formula in which θa in Formula (2) is replaced by θb, and the upper limit value ΔI1 is replaced by the lower limit value -ΔI2.

The time for switching the polarity of the correction voltage Δvx can be set using equation (1) or equation (2). First, the setting method using Formula (2) is demonstrated. That is, when output current ix is bipolar, the 1st time 32 is made into time t1, and the time t1 + ta after the time change amount ta calculated | required by Formula (2) based on this time t1 is the polarity of correction voltage (DELTA) vx. Is set as time t2 for switching. When the output current ix is negative, the first time 34 is set as time t1, and time t1 + tb after the time variation amount tb is set as time t2 for switching the polarity of the correction voltage Δvx. By setting the time for switching the polarity of the correction voltage Δvx in this manner, the time for switching the polarity of the correction voltage Δvx can be precisely matched to the time of the zero cross 31 in which the polarity of the output current ix is switched. In the following description, the first time is indicated only by the reference numeral 32 unless it is necessary to distinguish it.

Next, the setting method using Formula (1) is demonstrated. In this case, it converts time t1 which is the 1st time 32 into arbitrary phases, iteratively performs grasping time progress as a phase change, and in the phase which can confirm the phase change amount (theta) a obtained by Formula (1). The time is set as the time for switching the polarity of the correction voltage Δvx. The same applies to the case where the phase change amount θb is used.

For example, when the output current ix is bipolar, and the phase at time t1 which is the first time 32 is set to 0, the phase corresponding to the time t2 for switching the polarity of the correction voltage Δvx is expressed by the formula (1). Is the amount of phase change θa obtained. Therefore, if any time between time t1 and time t2 is tx, the phase θx at time tx is expressed using the frequency f, and θx = 2πf · (tx-t1). It is also assumed that the frequency f is constant. Gradually obtaining the phase θx at this arbitrary time tx to grasp the passage of time as the phase change is repeated until the phase θx and the phase change amount θa coincide. When the phase θx and the phase change amount θa coincide, it can be judged that the time t2 at which the polarity of the correction voltage Δvx is switched has been reached. Therefore, the time t2 at which the phase θx and the phase change amount θa coincide is switched between the polarities of the correction voltage Δvx. Set as time.

Thus, in the setting method using Formula (1), the conversion by Formula (2) is not necessary, so that the time for switching the polarity of the correction voltage Δvx can be set with the same precision as the setting method using Formula (2). have. In addition, the phase at the first time point 32 may not be set to 0 but may be set to the same value as the phase of the output current ix or the voltage command vx *. In this case, the time change is understood as a change in the phase of the output current ix or a change in the phase of the voltage command vx *.

The above is the basic setting method of the time of switching the polarity of the correction voltage (DELTA) vx, but two examples of the specific setting method at the time of using Formula (2) are shown. First, in the calculation of formula (2), since the denominator includes the value of the frequency f and the output current ix (effective value Irms), calculation using division is necessary. Therefore, for the purpose of reducing the amount of calculation in the control device 5a, the voltage command correction means 7a has a characteristic table (see FIG. 4) of the correction coefficient K1 which changes according to the frequency f and the value of the output current ix. The characteristic table (refer to FIG. 5) of the correction coefficient K2 which changes according to (effective value Irms) is memorized, and both or one of the characteristic tables are applied to equation (2), and the correction voltage? You may use the method of setting the time to switch polarity.

4 is an example of the characteristic table of the correction coefficient K1 which changes according to the frequency f. 5 is an example of the characteristic table of the correction coefficient K2 which changes with the value (effective value Irms) of the output current ix. From equations (2) and (3), since the relationship between the frequency f and the correction coefficient K1 becomes "K1 = 1 / f", the curve of the theoretical value shown in FIG. 4 is drawn. In addition, since the relationship between the value of the output current ix (effective value Irms) and the correction coefficient K2 becomes "the value of K2 = 1 / the value of the output current ix (effective value Irms)", the curve of the theoretical value shown in FIG. 5 is drawn. In order to reduce the amount of data stored in the voltage command correction means 7a, the characteristic tables relating to the correction coefficients K1 and K2 may be simplified and replaced with an approximate straight line as shown in FIG. 4 or 5.

Next, a method of simplifying the setting of the time for switching the polarity of the correction voltage Δvx will be described. This is a method of obtaining the time variation amount ta from the frequency f without using the value of the output current ix (effective value Irms) for setting the time for switching the polarity of the correction voltage Δvx.

For example, the output current ix (three-phase current iu, iv, iw) is used to generate magnetic flux based on the frequency f and the phase of the three-phase current iu, iv, iw for the voltage commands vu *, vv *, vw *. When the required excitation current id (d-axis current) and the torque current iq (q-axis current) proportional to the load torque when the AC load 2a is a rotor are divided using a known vector operation, the value of the output current ix The relationship between (effective value Irms), the excitation current id, and the torque current iq is represented by Formula (4).

When the AC load 2a is an inductor, the exciting current id may be set to a constant current value (this value is id0) flowing in the case of no load operation. This constant current value id0 is an integer known by a value unique to each rotor (induction machine). In this case, since the exciting current id becomes the constant current value id0, the value of the output current ix (effective value Irms) is changed by the torque current iq flowing in accordance with the torque generated from the AC load (induction apparatus) 2a. However, here, the change in the torque current iq is regarded as a value depending on the frequency f, and information about the current obtained from the output current ix such as the value of the output current ix (effective value Irms), the excitation current id, and the torque current iq is obtained. The method of obtaining the said time change amount ta based on the frequency f without using it is shown below.

Fig. 6 shows an inductor having a rated capacity of 3.7 kW and a rated slip frequency of 3 Hz, in which no-load (id = id0, iq = 0) shown in the table and rated load (id = id0, iq = rated torque shown in the table). It is a figure which plotted the value of the time change amount ta with respect to the frequency f in two driving conditions of electric current (retro)). In the drawing of FIG. 6, although the upper limit value ΔI1 (lower limit value -ΔI2) was 6% (constant value) of the rated current, it does not necessarily have to be a constant value, and the value of the upper limit value ΔI1 (lower limit value -ΔI2) is the AC load (induction period) ( In the case where it is gradually changed depending on the output state of 2a), the value of the time change amount ta may be adjusted according to the upper limit value ΔI1 and the value of the time change amount tb according to the lower limit value -ΔI2.

In the case of the rated load shown by the table | surface in FIG. 6, the range in which the frequency f becomes the rated slip frequency 3 Hz or less is also plotted, but in an actual induction machine, if the rated slip frequency is 3 Hz, the frequency f at the rated retrograde load is 3 Hz or more. Since the frequency f is within the range of the rated slip frequency (3 Hz) or less, the rated retrograde load driving is impossible. Therefore, in the range where the frequency f is equal to or less than the rated slip frequency (3 Hz), the load is small, and the torque current iq is also smaller than the rated torque current. That is, at the time of load driving, since the load becomes smaller than the rating, the amount of time change ta is larger than the characteristic at the rated load shown in the table in Fig. 6, and is closer to the characteristic at no load shown in the table.

Therefore, in a specific implementation, the approximate broken line shown in the table in Fig. 6 is set so that the error from each characteristic at substantially no load and rated load is small, and the voltage command correction means 7a is compared with the frequency f. The characteristic characteristic table of the time variation amount ta to be changed is stored, and the time variation amount ta is obtained from the characteristic table according to the frequency f.

In this way, the time change amount ta can be obtained using the frequency f without using information on the current obtained from the output current ix, such as the value of the output current ix (effective value Irms), the excitation current id, the torque current iq, and the like. The time for switching the polarity of the correction voltage Δvx can be set.

In addition, although the approximate broken line shown in the table | surface in FIG. 6 was set to the error from each characteristic with no load shown by the table, and the rated load time shown by the ○ table, the reference | standard which always obtains the said approximate broken line is, At no load or at rated load, the reference for obtaining the approximate broken line may be changed depending on the load range. The approximate curved line may be an approximation curve by increasing the data score in accordance with the storage capacity of the voltage command correction means 7a. In addition, when setting the time to switch the polarity of correction voltage (DELTA) vx using phase change amount (theta) a using Formula (1), it is good to convert it using the relationship of (theta) a = 2 (pi) ta from the time change amount ta calculated | required from the said approximate broken line. .

Then, the time change amount ta is obtained by using the frequency f without using information on the current obtained from the output current ix, such as the value of the output current ix (effective value Irms), the excitation current id, the torque current iq, and the like. Since the amount of calculation performed by the apparatus 5a can be reduced, the effect which can reduce the load of the calculator, including the microcomputer stored in the control apparatus 5a, is acquired. The above is the setting method of the time which switches the polarity of the correction voltage (DELTA) vx in ST16. When the setting is finished, the respective processes of ST21 to ST23 are executed to execute the necessary update processing and the process shifts to the processing at the next sample timing.

On the other hand, when it is determined in ST14 that the value of the output current ix detected at the sample timing is kept in the state within the range of the current range 30 at the sample timing (ST14: NO), the voltage command correction means ( Since the time for switching the polarity of the correction voltage Δvx has already been set as described above, 7a) determines whether the time for switching the polarity of the correction voltage Δvx has passed (ST17).

As a result, when the time for switching the polarity of the correction voltage Δvx has not elapsed (ST17: NO), the polarity of the correction voltage Δvx is not changed. Therefore, the polarity of the correction voltage Δvx is the same as before the first time 32. Set to polarity. That is, the polarity recorded in the polarity setting signal signx in ST15 is matched with the polarity of the current correction voltage Δvx (ST20). When it finishes, each process of ST21-ST23 is performed, a necessary update process is performed, and it transfers to the process in the next sample timing.

Then, when it is determined that the time for switching the polarity of the correction voltage Δvx has elapsed, at ST17 when the period from the first time 32 immediately before the time for switching the polarity of the correction voltage Δvx has passed (ST17: Yes) ), The voltage command correction means 7a executes respective processes of ST18 and ST19. Although the processing order of ST18 and ST19 may be reversed, in ST18, the polarity of the correction voltage Δvx is set in ST15 as the processing in the period until the value (effective value Irms) of the output current ix goes out of the current range 30. The polarity recorded in the signal signx and the inverted polarity are set. In ST19, the time for switching the polarity of the correction voltage? Vx set above is reset. When it finishes, each process of ST21-ST23 is performed, a necessary update process is performed, and it transfers to the process by the next sample timing.

In ST21 to ST23, the following processing is performed. That is, in ST21, the amplitude | Δvx | of the correction voltage Δvx is set. The magnitude (absolute value) of the error caused by the dead time between the AC voltage command vx * and the output voltage vx from the power conversion means 3 is, in theory, the carrier frequency [Hz] and the DC of the power conversion means 3. This is obtained by multiplying the link voltage [V] by the dead time [sec]. Based on this, if the value obtained from these three products is | Verr |, the amplitude | Δvx | of the correction voltage may be set based on the value | Verr |.

In ST22, the AC voltage command vx ** is obtained by adding the correction voltage Δvx of the voltage amplitude | Δvx | having the polarity set in the polarity setting signal signx to the AC voltage command vx *. In ST23, the added AC voltage command vx ** is output to the power conversion means 3.

The voltage command correction means 7a sequentially performs a series of processes described above for each of steps ST10 to ST23 for each of the u, v, and w phases, and performs the power conversion means 3 to correct the voltage error caused by the dead time. Output voltage commands vu **, vv **, vw ** respectively.

Next, the scene in which the voltage command correction means 7a switches the polarity of the correction voltage will be described in detail with reference to FIG. 7. In Fig. 7, the output current ix of one phase received from the current detection means 4 in ST10, the AC voltage command vx * of one phase output by the voltage command calculation means 6a, and the voltage command correction means 7a are described above. The relationship with the polarity of the correction voltage determined in order is shown. There is a phase difference between the output current ix and the AC voltage command vx *.

Although the current range 30 including the zero level is set in ST11 for the output current ix, the times 35, 32, 36, when the value of the output current ix and the upper limit value ΔI1 and the lower limit value -ΔI2 of the current range 30 coincide with each other. Among the 34 and 37, the time 35 and 37 is the time when the value of the output current ix exceeds the upper limit value ΔI1 and goes out of the upper side of the current range 30, and the time 36 shows the value of the output current ix as the lower limit value −. It is the time to go beyond ΔI2 to the outside of the lower side of the current range 30. In addition, the time 32 is the 1st time when the value of the output current ix enters into the current range 30 from the outside of the upper side of the current range 30, and the time 34 is the value of the output current ix is the current range. It is a 1st time entering into the electric current range 30 from the outside of the downward side of (30).

The period 1 is a period in which the value of the output current ix is located outside the upper limit value ΔI1 and the lower limit value -ΔI2 of the current range 30. This period (1) is based on the time values 35, 32, 36, determined based on the upper limit value ΔI1 and the lower limit value -ΔI2 of the current range 30 set in ST11 and the value at the time of sampling of the output current ix used in ST12. 34), the polarity of the correction voltage [Delta] vx in this period (1) is set to the same polarity as the value at the time of the current sampling of the output current ix used in ST12 in ST13.

The period 2 is a period until the value at the time of sampling of the output current ix reaches from the first time 32 just before the time 38 at which the polarity of the correction voltage Δvx is switched, and the period of the correction voltage Δvx When the time 38 for changing the polarity is reached, the polarity switching of the correction voltage Δvx is performed. In this period (2), each process of ST12, ST14, ST15, and ST16 is performed. That is, the polarity of the correction voltage Δvx in this period 2 was determined that the value at the time of the current sampling of the output current ix in ST12 was inside the current range 30, but one sampling of the output current ix in ST14. Since it was determined that the previous value did not fall into the current range 30, it is set to the same polarity as the value at the time of sampling of the output current ix used in ST12 as in the period 1 (ST15). In parallel, the time 38 for switching the polarity of the correction voltage Δvx is based on the first time 32 and the frequency f (not necessarily required, but the value at the time of the current sampling of the output current ix). ) Is set to a time very close to (). In Patent Documents 1 and 2 described above, since this period 2 does not exist, the accuracy of error correction was not good.

The period 3 is a time 35 when the value at the time of sampling of the output current ix goes out of the current range 30 from the current zero cross 31 after the time 38 when the polarity of the correction voltage Δvx is switched. 36, 37). In this period (3), when the passage of the period (2) is confirmed in ST17, the polarity of the correction voltage (DELTA) vx is set to the polarity opposite to the period (1) (ST18), and in parallel, of the correction voltage (DELTA) vx previously determined The time 38 at which the polarity is switched is reset (ST19).

As described above, according to the first embodiment, the voltage command correcting means 7a provided between the power converting means 3 and the voltage command calculating means 6a for calculating and outputting the AC voltage command to the power converting means 3. Is for the purpose of correcting an error between the voltage command generated due to the dead time voltage applied to the power device elements of the upper and lower arms constituting the power conversion means 3 and the output voltage from the power conversion means 3. Provide a predetermined current range including a zero level with respect to the detected output current, and at a first time when a value of the output current enters from outside of the current range, based on the first time and frequency f The timing (time or phase) of the zero cross of the output current is set at the time of switching the polarity of the correction voltage for correcting the AC voltage command, that is, the dead time The polarity switching of the correction voltage for correcting the voltage error caused is performed at zero cross of the output current, and correcting the AC voltage command with the correction voltage of the polarity and the same polarity of the output current before and after the zero cross of the output current. Therefore, the voltage error caused by the dead time in the vicinity of zero cross of the output current can be precisely reduced. As a result, when the rotor is connected to the power converter, the rotational unevenness in the vicinity of the zero cross of the output current can be reduced, thus exhibiting a remarkable effect that has not been conventionally improved in driving performance.

If the time for switching the polarity of the correction voltage is also obtained by applying the output current, the timing of the zero cross of the output current can be directly obtained, so that the zero cross of the output current can be obtained more precisely. The effect which can further reduce the voltage error in the vicinity of the zero cross of the said output current is acquired.

The current range is set as a predetermined current range including a zero level with respect to the detected output current, that is, as a current range surrounded by the upper limit value on the anode side and the lower limit value on the cathode side from the detected zero level of the output current. In this way, the effect of accurately setting the polarity of the correction voltage is obtained without being affected by the current polarity detection error due to the chattering phenomenon occurring near the current zero cross.

(Embodiment 2)

Fig. 8 is a flowchart for explaining the corrective operation procedure of the voltage command correction means included in the power converter according to the second embodiment of the present invention. In addition, in FIG. 8, the same code | symbol is attached | subjected to the process sequence same or equivalent to the process sequence shown in FIG. Here, it demonstrates centering around the part which concerns on this Embodiment.

As shown in FIG. 8, the voltage command correcting means included in the power converter according to the second embodiment has a negative determination in ST14 during the transition process from ST14 to ST17 in the processing procedure shown in FIG. 2. In this case, the process proceeds to ST17 after the process of ST30 is performed.

Referring to FIG. 7, in ST30, in the period 2 from the first time 32 to the correction voltage polarity switching time 38, the correction voltage polarity switching time 38 depends on the value of the output current ix. Correction processing is performed. As the value of the output current ix, any one of an effective value, an amplitude value, and an instantaneous value may be used. In addition, the above-mentioned current command ix * may be used instead of the output current ix.

When the output current ix is not constant and its value (effective value Irms) changes, the value (effective value Irms) is a function of time t. That is, Irms = Irms (t).

Therefore, in the second embodiment, in this case, the value (effective value Irms) of the output current ix is the polarity of the correction voltage Δvx from the first time 32 by using equation (5) instead of equation (2). The time change amount ta until reaching the time to switch to 38 is calculated in the form of gradually integrating the change of the value (effective value Irms) of the output current ix under a predetermined frequency f, and based on the calculation result The time 38 at which the polarity of the correction voltage Δvx is switched is corrected. However, time tx shown in Formula (5) is arbitrary time which exists between time t1 (1st time)-time t2 (correction voltage polarity switching time).

As described above, according to the second embodiment, before the polarity of the correction voltage is switched, the time at which the polarity of the correction voltage is switched in accordance with the value of the detected output current is gradually corrected, and the polarity of the correction voltage is switched accordingly. Therefore, even when the output current fluctuates due to load fluctuations or speed fluctuations, the polarity of the correction voltage can be switched more precisely, and the effect of further reducing the voltage error near zero cross of the output current can be obtained. Obtained.

(Embodiment 3)

Fig. 9 is a flowchart for explaining the corrective operation procedure of the voltage command correction means included in the power converter according to the third embodiment of the present invention. In addition, in FIG. 9, the same code | symbol is attached | subjected to the process sequence same or equivalent to the process sequence shown in FIG. Here, it demonstrates centering on the part which concerns on this Embodiment 3rd.

As shown in Fig. 9, the voltage command correcting means included in the power converter according to the third embodiment has a negative determination in ST14 in the process of transition from ST14 to ST17 in the processing procedure shown in Fig. 2. In this case, the process proceeds to ST17 after performing the process of ST40.

Referring to FIG. 7, in ST40, in the period 2 from the first time 32 to the correction voltage polarity switching time 38, the correction voltage polarity switching time 38 according to the frequency f for each sample timing. ) Correction processing. In addition, since the frequency f is the same as the frequency of the output current ix, the current command ix *, the output voltage vx, and the voltage command vx * in a steady state, you may correct | amend using either frequency.

If the frequency f is not constant and its value changes, the frequency f becomes a function of time t. That is, f = f (t). So, in this Embodiment 3, in this case, the value (effective value Irms) of the output current ix becomes the polarity of the correction voltage (DELTA) vx from the 1st time 32 using Formula (6) instead of Formula (2). The amount of time change ta until reaching the time to switch to 38 is calculated in such a manner that the change in the frequency f for each sample timing is gradually integrated with the value of the output current ix (effective value Irms) being constant. Based on the result, the time 38 at which the polarity of the correction voltage Δvx is switched is corrected. However, time tx shown in Formula (6) is arbitrary time which exists between time t1 (1st time)-time t2 (correction voltage polarity switching time).

As described above, according to the third embodiment, before the polarity of the correction voltage is switched, the time at which the polarity of the correction voltage is switched in accordance with the change of the frequency is gradually corrected, and the polarity of the correction voltage is switched based thereon. Even in the case where the frequency fluctuates due to the load fluctuation and the speed fluctuation, the polarity of the correction voltage can be switched more precisely, and the effect of further reducing the voltage error in the vicinity of zero cross of the output current is obtained.

(Embodiment 4)

10 is a block diagram showing a configuration of a power converter according to Embodiment 4 of the present invention. In addition, in FIG. 10, the same code | symbol is attached | subjected to the component same as or equivalent to the component shown in FIG. 1 (Embodiment 1). Here, it demonstrates centering on the part which concerns on this Embodiment 4.

As shown in FIG. 10, the power converter 1b which concerns on this Embodiment 4 is provided with the rotor 2b as an example instead of the AC load 2a in the structure shown in FIG. 1 (Embodiment 1). Moreover, the control apparatus 5b is provided instead of the control apparatus 5a. In the control apparatus 5b, the torque command setting means 8 which applies the torque command (tau) * to the voltage command calculation means 6b which changed the code, and the voltage command correction means 7b is provided.

In the fourth embodiment, the voltage command correction means 7b uses the first time 32, the frequency f, and the torque command as a process corresponding to ST16 in FIG. 2 as a process 38 for switching the polarity of the correction voltage Δvx. The time for switching the polarity of the correction voltage Δvx set by the method shown in Embodiment 1 (FIG. 2) as the processing (1) to be set based on τ * and the process corresponding to ST30 in FIG. 8 or ST40 in FIG. Both or either of the processes (2) for correcting (38) in accordance with the torque command τ * are performed. Here, a case of performing both will be described.

First, the processing (1) for setting the time 38 for switching the polarity of the correction voltage Δvx based on the first time 32, the frequency f, and the torque command τ * will be described. In Embodiment 1, although the relationship between the value (effective value Irms) of the output current ix, the exciting current id, and the torque current iq was shown by Formula (4), although the torque current iq and the load torque (tau) of the rotor 2b are shown, The relationship is represented by equation (7). However, Kt shown in Formula (7) is a torque constant inherent to the rotor 2b.

When the rotor 2b is an inductor, the excitation current id is set to a fixed value unique to each rotor as described in Embodiment 1, and when the rotor 2b is a permanent magnet synchronizer, the excitation current id Is set to 0 to control the excitation current id to a constant value by a known maximum efficiency control method or the like. And in the power converter 1b shown in FIG. 10, when load torque (tau) of the rotating machine 2b is controlled so that it may correspond to torque command (tau) *, following Formula (8) from Formula (4) and Formula (7) The relationship is established.

In Equation (8), the only variable is the torque command τ *, so that the value (effective value Irms) of the output current ix changes in accordance with the torque command τ *. For this reason, even if the time 38 for switching the polarity of the correction voltage Δvx is set in accordance with the torque command τ *, the zero cross of the output current is performed with the same precision as that set in the output current ix described in ST16 of FIG. Since the switching time 38 can be set by obtaining (31), the voltage error near the zero cross of the output current can be precisely reduced.

In addition, although the process (1) showed the method of setting the time 38 of switching the polarity of the correction voltage (DELTA) vx according to torque command (tau) *, besides, load torque (tau) of the rotor (2b), for example, by some method. The detection and adjustment may be performed according to the load torque τ, or the torque current iq may be obtained based on the equation (4) or (7) and set according to the torque current iq.

Next, the process (2) of correcting the time 38 for switching the polarity of the correction voltage Δvx set in the method shown in Embodiment 1 (FIG. 2) in accordance with the torque command τ * will be described. Since the value (effective value Irms) of the output current ix changes according to the torque command τ * in the relation shown in equation (8), the value set by the method according to the fourth embodiment described in the above process (1) is set. If the time at which the polarity of the correction voltage Δvx is switched is corrected according to the torque command τ * when the determination in ST14 of FIG. 2 is negative (NO), the processing proceeds to ST17.

According to this embodiment, since the time for switching the polarity of the correction voltage Δvx is gradually corrected for each sample timing before switching the polarity of the correction voltage, the polarity of the correction voltage is switched accordingly according to the second embodiment. Similarly to 3, even when the output current fluctuates due to load fluctuations or speed fluctuations, the polarity of the correction voltage can be switched more precisely, so that the voltage error near the zero cross of the output current can be further reduced. Effect is obtained.

In addition, although the process (2) showed the method of correct | amending the time which switches the polarity of the correction voltage (DELTA) vx according to torque command (tau *), in addition, the load torque (tau) of the rotor (2b) is detected by some method, for example. The correction may be performed according to the load torque τ, or the torque current iq may be obtained based on the equation (4) or (7) and corrected according to the torque current iq.

FIG. 11 is a flowchart for explaining the corrective operation procedure of the voltage command correction means 7b according to the third embodiment in which both of the above-described processes (1) and (2) are included in FIG. 2. In FIG. 11, in ST50 instead of ST16 shown in FIG. 2, the time 38 at which the polarity of the correction voltage Δvx is switched in accordance with the torque command τ * is set by the method described in the process (1). In addition, in ST51 provided between ST14 and ST17, the time 38 at which the polarity of the correction voltage Δvx is switched is corrected according to the torque command τ * by the method described in the process (2).

As described above, according to the fourth embodiment, since the zero cross 31 of the output current can be obtained with the same precision as in the first embodiment according to the torque command, the voltage error in the vicinity of the zero cross of the output current is precisely determined. Can be reduced. In addition, similarly to Embodiments 2 and 3, since the polarity of the correction voltage can be switched more precisely, an effect of further reducing the voltage error near the zero cross of the output current is obtained.

(Embodiment 5)

It is a flowchart explaining the correction operation procedure of the voltage command correction means with which the power converter which concerns on Embodiment 5 of this invention is equipped. In addition, in FIG. 12, the same code | symbol is attached | subjected to the process sequence same or equivalent to the process sequence shown in FIG. Here, it demonstrates centering on the part which concerns on this Embodiment 5.

In the power converter according to the fifth embodiment, the components are the same as those of the power converter 1a shown in Fig. 1 (Embodiment 1) or the power converter 1b shown in Fig. 10 (Embodiment 4). The voltage command correction means included in the power converter according to 5 performs a slightly different operation from the voltage command correction means 7a and 7b. That is, in FIG. 12, the voltage command correcting means included in the power converter according to the fifth embodiment executes the processing of ST60 in place of the processing of ST16 (ST50) in the processing procedure shown in FIG. 2 (FIG. 11). Then, the ST61 processing is executed in place of ST17, the ST62 processing is executed in place of ST18, and the ST63 processing is executed in place of ST19.

In ST60, the second time is set using the same method as the setting method of the time for switching the polarity of the correction voltage Δvx for correcting the AC voltage command vx * described in the first embodiment. In ST61, it is determined whether or not the second time has passed. In ST62, the polarity of the AC voltage command vx * is recorded in the polarity setting signal signx. In ST63, the second time is reset.

That is, the voltage command correction means according to the fifth embodiment determines that the value of the output current ix detected at the sample timing has entered the range from outside the range of the current range 30 for the first time at the sample timing (ST14). (Example) executes the respective processes of ST15 and ST60, and proceeds to ST21. Although the processing order of ST15 and ST60 may be reversed, in ST15, the polarity of the detected value ix [N-1] of the output current ix before one sampling is recorded in the polarity setting signal signx, and in ST60, the range of the current range 30 At the first time (symbols 32 and 34 shown in FIG. 3), which is the time from the outside into the range, the time of switching the polarity of the correction voltage Δvx for correcting the AC voltage command vx * described in the first embodiment (FIG. 7). The 2nd time is set by the method similar to the setting method of (38) shown by the following.

In addition, in ST14, the voltage command correcting means according to the fifth embodiment determines that the value of the output current ix detected at the sample timing is kept within the range of the current range 30 at the sample timing. In the case (ST14: NO), since the second time has already been set as described above, it is determined whether the second time has passed (ST61).

As a result, when the second time has not passed (ST61: NO), since the polarity of the correction voltage Δvx is not switched, the polarity of the correction voltage Δvx is the same as that of the correction voltage Δvx before the first time 32. Fit to That is, the polarity recorded in the polarity setting signal signx at ST15 matches the polarity of the correction voltage Δvx (ST20).

Then, if it is determined that the second time has elapsed in ST61 after the first time 32 and the period just before the second time has elapsed (ST61: YES), the voltage command correction according to the fifth embodiment is performed. The means writes the polarity of the AC voltage command vx * in the polarity setting signal signx as processing in the period until the value of the output current ix (effective value Irms) goes out of the current range 30 (ST62). That is, the polarity of the correction voltage Δvx is set to the polarity of the AC voltage command vx *. Then, the previously set second time is reset (ST63).

Next, FIG. 13 is a time chart summarizing and explaining the operation | movement of the voltage error resulting from the dead time realized by the correction process shown in FIG. In addition, in FIG. 13, the content similar to what was shown in FIG. 7 is shown. The difference is that the correction voltage polarity switching time 38 shown in FIG. 7 is the second time 40 in FIG. 13, and that the polarity of the correction voltage Δvx in the period 3 is the period in FIG. 7. While the polarity set in (1) is the inverted polarity, the same point as that of the AC voltage command vx * is shown in FIG. 13.

In the processing procedure according to the fifth embodiment, when determining whether or not the value of the output current ix reaches the second time 40, the polarity and the AC voltage command of the correction voltage Δvx at the second time 40 are determined. Compare the polarity of vx *. As a result, when the polarity of the AC voltage command vx * is inverted, the polarity of the correction voltage Δvx is switched when the second time 40 elapses. On the other hand, in the case of the same polarity, the polarity of the correction voltage Δvx is switched when the polarity of the AC voltage command vx * is switched. Therefore, the operation of switching the polarity of the correction voltage Δvx with respect to the second time 40 is slow. As a result, when the phase difference between the output current ix and the AC voltage command vx * (output voltage vx) is large, the polarity of the correction voltage Δvx is switched in the vicinity of zero cross of the output current ix, similarly to the first and fourth embodiments. Can be.

Therefore, according to the fifth embodiment, when the phase difference between the output current and the AC voltage command (output voltage) is large, the voltage error in the vicinity of zero cross of the output current can be reduced. Rotation unevenness can be reduced, and the effect which drive performance improves is acquired.

Embodiment 6

Fig. 14 is a flowchart for explaining the corrective operation procedure of the voltage command correction means included in the power converter according to the sixth embodiment of the present invention. 14, the same code | symbol is attached | subjected to the process sequence same or equivalent to the process sequence shown in FIG. Here, it demonstrates centering on the part which concerns on this 14th Embodiment.

As shown in FIG. 14, in the voltage command correcting means included in the power converter according to the sixth embodiment, in the processing procedure shown in FIG. 12, the determination in ST14 is negative in the transition process from ST14 to ST61. In this case, the process proceeds to ST61 after performing the process of ST70.

Referring to FIG. 13, in ST70, in the period 2 from the first time 32 to the second time 40, the correction processing of the second time 40 is performed according to the value of the output current ix. Conduct. As the value of the output current ix, any one of an effective value, an amplitude value, and an instantaneous value may be used. In addition, the above-mentioned current command ix * may be used instead of the output current ix.

The correction of the second time 40 in accordance with the value of the output current ix can be performed in the same manner as the correction method of the correction voltage switching time described in ST30 in FIG. 8. That is, when the output current ix is not constant and its value (effective value Irms) changes, the value (effective value Irms) of the output current ix is similarly determined from the first time point 32 using Equation (5). The amount of time change ta until reaching the second time 40 is calculated in the form of gradually integrating the change in the value of the output current ix (effective value Irms) under a constant frequency f, and based on the result of the calculation, the second time Calibrate 40. However, time tx shown in Formula (5) is arbitrary time which exists between time t1 (1st time)-time t2 (2nd time).

As described above, according to the sixth embodiment, before the polarity of the correction voltage is switched, the second time is gradually corrected according to the detected value of the output current, and the polarity of the correction voltage is switched based on the load variation. Even when the output current fluctuates due to the speed variation, the polarity of the correction voltage can be switched more precisely, and the effect of further reducing the voltage error in the vicinity of zero cross of the output current is obtained.

(Embodiment 7)

Fig. 15 is a flowchart for explaining the corrective operation procedure of the voltage command correction means included in the power converter according to the seventh embodiment of the present invention. In addition, in FIG. 15, the same code | symbol is attached | subjected to the process sequence same or equivalent to the process sequence shown in FIG. Here, it demonstrates centering around the part which concerns on this Embodiment.

As shown in FIG. 15, the voltage command correcting means included in the power converter according to the seventh embodiment has a negative determination in ST14 in the process of transition from ST14 to ST61 in the processing procedure shown in FIG. 12. In this case, the process proceeds to ST61 after the process of ST80 is performed.

Referring to FIG. 13, the ST80 corrects the second time 40 in accordance with the frequency f for each sample timing in the period 2 from the first time 32 to the second time 40. Is carried out. In addition, since the frequency f is the same as the frequency of the output current ix, the current command ix *, the output voltage vx, and the voltage command vx * in a steady state, you may correct | amend using either frequency.

The correction of the second time 40 in accordance with the frequency f can be performed in the same manner as in the correction method of the correction voltage switching time described in ST40 in FIG. 9. In other words, when the frequency f is not constant and its value is changed, the value (effective value Irms) of the output current ix is similar to the second time 40 from the first time 32 using equation (6). The amount of time change ta to reach is calculated by gradually integrating the change in frequency f for each sample timing, provided that the value of output current ix (effective value Irms) is constant, and based on the result of the calculation Calibrate 40. However, time tx shown in Formula (6) is arbitrary time which exists between time t1 (1st time)-time t2 (2nd time).

As described above, according to the seventh embodiment, before switching the polarity of the correction voltage, the second time is gradually corrected in accordance with the change of the frequency, and the polarity of the correction voltage is switched based on the change in load, and thus the speed fluctuation. Even when the frequency fluctuates, the polarity of the correction voltage can be switched more precisely, and the effect of further reducing the voltage error near the zero cross of the output current is obtained.

Embodiment 8

In the eighth embodiment, a method of adjusting the amplitude of the correction voltage generated by the voltage command correction means described above with reference to Figs. 16 and 17 will be described. 16 is a view for explaining the necessity of adjusting the amplitude of the correction voltage as Embodiment 8 of the present invention. It is a figure which shows an example of the relationship characteristic table of the amplitude of a correction voltage and the value of an output current used for the amplitude adjustment of a correction voltage.

As described in the first embodiment, the magnitude (absolute value) of the error due to the dead time occurring between the AC voltage command vx * output by the voltage command calculating means 6a and the output voltage vx of the power conversion means 3 is Is theoretically obtained as the product | Verr | of the carrier frequency [Hz] and the DC link voltage [V] of the power conversion means 3 and the dead time [sec]. However, as a premise of the above theory, there is an assumption that the timing of actually on / off operation of the power device elements of the upper and lower arms constituting the power conversion means 3 coincides with the desired timing specified by the drive signal.

In FIG. 16, the relationship between the upper and lower arm ON signals which are drive signals, and the ON operation of the upper and lower arms actually performed, and the relationship between the set dead time 42 and the dead time 43 actually occurring are shown. As shown in Fig. 16, dead times 42 and 43 are defined as time intervals between the off operation of the upper arm and the on operation of the lower arm.

Since there is a switching delay in the power device elements of the upper and lower arms constituting the power conversion means 3, as shown in Fig. 16, the ON and OFF states designated by the upper and lower arm ON signals as the drive signals are designated. With respect to the operation timing, the operation timing of the on / off of the upper and lower arms actually performed is delayed by the switching time. The time of this switching delay is often different in the on operation and in the off operation, and also varies according to the magnitude of the output current ix. Therefore, an error occurs between the set dead time 42 and the dead time 43 actually occurring, as shown in FIG.

Therefore, setting the amplitude | Δvx | of the correction voltage? Vx to the same magnitude as the product | Verr | may result in excessive correction or insufficient correction for the voltage error actually occurring, resulting in an error due to voltage correction.

Therefore, in order to reduce the error due to voltage correction, for example, as shown in Fig. 17, the characteristic table of the amplitude | Δvx | of the correction voltage Δvx that is changed in advance according to the output current ix is transferred to the voltage command correction means 7a, 7b. The amplitude | Δvx | of the correction voltage [Delta] vx | area can be computed and adjusted gradually according to the output current ix for every sample timing.

In this case, since the amplitude | Δvx | of the correction voltage Δvx can be adjusted to the optimum value for each sample timing, the amplitude error of the correction voltage resulting from the switching delay of the power device elements of the upper and lower arms constituting the power conversion means 3 is adjusted. It can reduce, and the correction accuracy of a voltage error improves further. It is also possible to create in the characteristic table as shown in Fig. 17 using the current command ix * in place of the output current ix, and adjust the amplitude | Δvx | of the correction voltage Δvx in accordance with the current command ix *.

Next, in the processing of ST11 in Figs. 2, 8, 9, 11, 12, 14, and 15, various forms can be adopted, and with reference to Fig. 3, these are summarized below. Explain.

(A) The upper limit value ΔI1 and the lower limit value -ΔI2 of the current range 30 set in ST11 may be adjusted and set according to the output current ix. In Embodiment 1, it was shown that what is necessary is just to adjust the time 38 of switching the polarity of correction voltage (DELTA) vx based on the 1st time 32 and either of Formula (1)-Formula (3). However, the period from the first time 32 obtained by applying the formula (1) or the formula (2) to the zero cross 31 in which the polarity of the output current ix is switched is sufficiently compared with the period of alternating current as described above. This is a theoretical time calculated assuming short.

Therefore, when the output current ix is small, the time 38 at which the polarity of the correction voltage Δvx is switched from the first time 32 is adjusted and set so as to make the upper limit value ΔI1 and the lower limit value −ΔI2 of the current range 30 smaller. It is desirable that the period up to not be longer than the period of the exchange.

Therefore, if the upper limit value ΔI1 and the lower limit value -ΔI2 of the current range 30 are gradually calculated and adjusted according to the output current ix for each sample timing, and the current range 30 is set, regardless of the value of the output current ix, Since the period from the time 32 to the zero cross 31 which is the time when the polarity of the output current ix is switched is not too long compared with the period of alternating current, the time 38 for switching the polarity of the correction voltage Δvx is precisely zero. As can be set with the cross 31, the current range 30 having the optimum upper limit value ΔI1 and the lower limit value -ΔI2 can be adjusted and set.

In addition, the value of the output current ix at the time of adjusting the upper limit (DELTA) I1 and the lower limit-(DELTA) I2 of the current range 30 may be any of an effective value, an amplitude value, and an instantaneous value. Alternatively, the current command ix * may be used.

(B) The upper limit value ΔI1 and the lower limit value -ΔI2 of the current range 30 set in ST11 may be adjusted and set in accordance with the frequency f. In this case, since the frequency f is the same as the frequency of the output current ix, the current command ix *, the output voltage vx, and the voltage command vx * in a steady state, either frequency may be used.

For example, if the voltage command calculating means 6a, 6b are controlled so that the ratio of the AC voltage command vx * and the frequency f becomes constant, the output voltage vx also increases as the frequency f increases. On the other hand, since the correction voltage (DELTA) vx becomes almost constant regardless of frequency f, the correction voltage (DELTA) vx with respect to the output voltage vx becomes small relatively. Therefore, at the time 38 when the polarity of the correction voltage Δvx is switched, the change amount of the correction voltage Δvx is small and the change of the output current ix is small, so that a current chattering phenomenon occurs near the zero cross 31 of the current. Tends to be easy.

Therefore, the frequency f increases when the upper limit value ΔI1 and the lower limit value -ΔI2 of the current range 30 are gradually calculated and adjusted according to the frequency f of the AC voltage command vx * for each sample timing, and the current range 30 is set. Since the current range 30 can be adjusted accordingly, the chattering phenomenon occurring at the zero cross 31 of the current can be suppressed.

(C) The upper limit value ΔI1 and the lower limit value -ΔI2 of the current range 30 set in ST11 may be set to a constant value. In this case, the values of the upper limit value ΔI1 and the lower limit value -ΔI2 may be set to the same constant value, and the offset may be added to the output current ix depending on the performance of the current detecting means 4, and the upper limit value ΔI1 may be detected. Each value of the lower limit -ΔI2 may be set to a different value.

The polarity of the output current ix is not detected incorrectly by the phenomenon that the offset is overlapped with the output current ix by the current detection means 4 or by the current chattering phenomenon near the zero cross 31 of the current. It is preferable to set the respective values of the upper limit value ΔI1 and the lower limit value −ΔI2 by obtaining a current range 30 which is not obtained by measurement or the like.

If the relationship between the current range 30 without misdetection of the polarity of the output current ix and the rating of the power converters 1a and 1b or the rating of the AC loads 2a and 2b is known, these ratings are known. The respective values of the upper limit value ΔI1 and the lower limit value -ΔI2 may be set on the basis of.

In this way, when the upper limit value ΔI1 and the lower limit value -ΔI2 of the current range 30 are constant values, the upper limit value ΔI1 and the lower limit value -ΔI2 are stored in the voltage command correction means 7a, 7b once, so that the operation of the power converters 1a, 1b. In the meantime, it is not necessary to always calculate the upper limit value ΔI1 and the lower limit value -ΔI2 of the current range 30. Therefore, the amount of calculation performed by the control devices 5a and 5b of the power converters 1a and 1b can be reduced, and the load of a calculator including a microcomputer stored in the control devices 5a and 5b can be reduced.

(D) In Embodiment 4, Formula (8) showed that the value (effective value Irms) of the output current ix changes with torque command (tau) *. Therefore, the upper limit value ΔI1 and the lower limit value -ΔI2 of the current range 30 set in ST11 in FIG. 11 (Embodiment 4) may be adjusted and set by calculating the difference in accordance with the torque command τ * for each sample timing.

Also in this method, the period from the first time 32 to the zero cross 31, which is the time when the polarity of the output current ix is switched, is not too long compared to the period of alternating current, and thus the time when the polarity of the correction voltage Δvx is switched. It is possible to adjust and set the current range 30 having the optimum upper limit value ΔI1 and the lower limit value −ΔI2, which can precisely set 38 to the current zero cross 31.

In addition, similarly to ST50 and ST51 in FIG. 11 (Embodiment 4), in place of torque command τ *, the load torque τ of the rotor 2b is detected by some method, and depending on the load torque τ or, The torque current iq may be obtained based on the equation (4) or (7), and the upper limit value ΔI1 and the lower limit value −ΔI2 of the current range 30 may be adjusted according to the torque current iq.

As described above, the power converter according to the present invention is useful as a power converter in which the power conversion means outputs AC power like an AC voltage command by reducing the voltage error caused by the value dead time near the zero cross of the output current. Do.

Claims (16)

Power conversion means for generating AC power supplied to an AC load in accordance with an input AC voltage command, voltage command calculation means for calculating an AC voltage command of a frequency to be applied to the power conversion means, and the power conversion means being used for the AC conversion. A power converter comprising current detecting means for obtaining an output current which is a current component in AC power supplied to a load, Between the power conversion means and the voltage command calculation means, a correction voltage for correcting an AC voltage command obtained by the voltage command calculation means is based on a voltage error due to a dead time in the power conversion means. A voltage command correcting means for calculating and applying the corrected voltage to the AC voltage command determined by the voltage command calculating means and applying it to the power converting means; The voltage command correction means provides a predetermined current range including the zero level with respect to the output current, and the first time at a first time when the value of the output current enters from inside of the outside of the current range. And calculating the zero cross timing of the output current using the frequency and the frequency, and setting the obtained zero cross timing to a time for switching the polarity of the correction voltage. Power converter characterized in that. The method of claim 1, And the voltage command correction means obtains a zero cross timing of the output current using the first time, the frequency, and the output current. The method of claim 1, And the voltage command correction means adjusts and sets an upper limit value on the anode side and a lower limit value on the cathode side with respect to the zero level of the current range in accordance with the output current. The method of claim 1, And the voltage command correction means adjusts and sets an upper limit value on the anode side and a lower limit value on the cathode side with respect to the zero level of the current range according to the frequency. The method of claim 1, And the voltage command correction means adjusts and sets the upper limit value on the anode side and the lower limit value on the cathode side to a constant value with respect to the zero level of the current range. The method of claim 1, The voltage command correction means corrects the time for switching the polarity of the correction voltage according to the output current in a period from the first time until the time for reaching the time for switching the polarity of the correction voltage. Power converter. The method of claim 1, The voltage command correction means corrects the time for switching the polarity of the correction voltage according to the frequency in the period from the first time until the time for reaching the time for switching the polarity of the correction voltage. Power converter. The method of claim 1, And the voltage command correction means adjusts an amplitude of the correction voltage in accordance with the output current when calculating the correction voltage for correcting an AC voltage command obtained by the voltage command calculation means. A power conversion means for generating AC power supplied to a rotor, which is an AC load, in accordance with an input AC voltage command, and an AC voltage command of a frequency to be applied to the power conversion means according to a torque command for applying a control parameter of the rotor. And a voltage command calculating means, a torque command setting means for generating the torque command, and a current detecting means for obtaining an output current which is a current component in AC power supplied to the rotor by the power converting means. In Between the power conversion means and the voltage command calculation means, a correction voltage for correcting the AC voltage command obtained by the voltage command calculation means is calculated based on the voltage error caused by the dead time in the power conversion means. Providing a voltage command correction means for adding a correction voltage to the AC voltage command obtained by the voltage command calculation means and applying it to the power conversion means; The voltage command correction means provides a predetermined current range including the zero level with respect to the output current, and the first time at a first time when the value of the output current enters from inside of the outside of the current range. And calculating the zero cross timing of the output current using the frequency and the torque command, and setting the obtained zero cross timing to a time for switching the polarity of the correction voltage. Power converter characterized in that. Power conversion means for generating AC power supplied to an AC load in accordance with an input AC voltage command, voltage command calculation means for calculating an AC voltage command of a frequency to be applied to the power conversion means, and the power conversion means being used for the AC conversion. A power converter comprising current detecting means for obtaining an output current which is a current component in AC power supplied to a load, Between the power conversion means and the voltage command calculation means, a correction voltage for correcting the AC voltage command obtained by the voltage command calculation means is calculated based on the voltage error caused by the dead time in the power conversion means. Providing a voltage command correction means for adding a correction voltage to the AC voltage command obtained by the voltage command calculation means and applying it to the power conversion means; The voltage command correction means provides a predetermined current range including the zero level with respect to the output current, and the first time at a first time when the value of the output current enters from inside of the outside of the current range. And the zero cross timing of the output current using the frequency and the frequency, the zero cross timing obtained is set to a second time, and the polarity of the correction voltage at the second time and the AC voltage obtained by the voltage command calculating means. When the polarity of the command is different, the polarity of the correction voltage is switched at the second time, while the polarity of the correction voltage at the second time and the polarity of the AC voltage command determined by the voltage command calculating means are the same. The polarity of the correction voltage is switched when the polarity of the AC voltage command determined by the voltage command calculating means is switched. Will Power converter characterized in that. 11. The method of claim 10, And the voltage command correction means obtains a zero cross timing of the output current using the first time, the frequency, and the output current. 11. The method of claim 10, And the voltage command correction means adjusts and sets an upper limit value on the anode side and a lower limit value on the cathode side with respect to the zero level of the current range in accordance with the output current. 11. The method of claim 10, And the voltage command correction means adjusts and sets an upper limit value on the anode side and a lower limit value on the cathode side with respect to the zero level of the current range according to the frequency. 11. The method of claim 10, And the voltage command correction means adjusts and sets the upper limit value on the anode side and the lower limit value on the cathode side to a constant value with respect to the zero level of the current range. 11. The method of claim 10, And the voltage command correction means corrects the time for switching the polarity of the correction voltage according to the output current in a period from the first time until the second time is reached. 11. The method of claim 10, And the voltage command correction means corrects the time for switching the polarity of the correction voltage according to the frequency in the period from the first time until the second time is reached.

Images