JP2008086083A - Pwm inverter control device, pwm inverter control method, and refrigeration air conditioner - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To conduct the drive of a three-phase voltage type PWM inverter which can be reduced in noise, irrespective of a PWM frequency without having to add a new device, is suppressed in the restriction of application with respect to an operation frequency and a motor to be driven, can generate a dead-time compensation voltage using an easy method, is reduced in current ripples, and is high in efficiency. <P>SOLUTION: A dead-time compensation computing part 70 calculates an output current of an inverter main circuit 2 detected by a phase-current detection means 10, the carrier frequency of a switching means 4, and a specified order component of a dead-time error voltage on the basis of a DC voltage and a preset dead time. A voltage command computing means 40 calculates the voltage command, on the basis of a frequency command, the output of the DC voltage detection means, and the output of the current detection means. A PWM signal generating means 80 generates a PWM signal, on the basis of an adding result of an output of the voltage command computing means 40 and the output of the dead-time compensation part 70. A PWM signal generating means 81 generates the PWM signal generated by the PWM signal generating means 80. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a PWM inverter control device and an inverter control method related to an inverter used for driving a motor for a refrigeration air conditioner and the like, and a refrigeration air conditioner using these.

  Conventionally, a short-circuit prevention time (dead time) including a turn-off time of the switching element is inserted in order to prevent an arm short circuit of the switching element of the three-phase voltage type PWM inverter used for driving the motor of the refrigeration air conditioner. In order to compensate for the error voltage between the PWM voltage command of the inverter and the output voltage due to this short-circuit prevention time, the polarity of the square-wave compensation voltage added to the PWM voltage command is set to the output current of the inverter. In general, a method of changing in synchronization with the time point of passing through the zero cross point is performed. However, according to the conventional dead time compensation method, since the polarity of the compensation voltage is changed stepwise at the zero cross point of the output current of the inverter, the ripple of the output current near the zero cross point increases. There has been a problem that uneven rotation due to ripples becomes large. In order to solve this problem, for example, in Patent Document 1, the inverter output current is compared with a reference value for each phase using a comparator, and the time according to the magnitude of the comparison result in the vicinity of the zero cross point of the output current. A pulse voltage of width is generated, and the compensation amount calculation unit calculates the dead time compensation amount (compensation amount amplitude) based on the PWM frequency, dead time, and DC power supply voltage, and outputs the comparator output and the compensation amount calculation unit. Multiply by the multiplier, integrate the multiplier output by the integrator to generate a trapezoidal waveform compensation amount signal, add this compensation amount signal to the command value from the PWM voltage command calculator, and PWS pulse calculator Thus, there is disclosed a dead time compensation method for adding a dead time after PWM calculation to the addition result and outputting a drive signal to the inverter main circuit. With this configuration, the polarity change of the compensation voltage that is performed as the inverter output current passes through the zero cross point can be performed with a predetermined gradient from the change start time before the zero cross point of the output current. Thus, the problem of increased ripple of the output current at the zero cross point is solved.

Japanese Patent No. 3738883 (FIG. 1, FIG. 2, paragraphs 0011 to 0017)

However, since the conventional dead time compensation device disclosed in Patent Document 1 is configured as described above, harmonics remain in the trapezoidal waveform dead time compensation voltage, and the specific order harmonics become a problem as noise. was there.
In addition, when low noise driving is required, the noise level near the PWM frequency is high, and there are significant restrictions in use particularly when the PWM frequency must be set in the audible range.
In addition, since the gradient of the compensation voltage near the zero cross changes depending on the magnitude of the output current, the ratio of the specific order harmonics changes, and the driving range is limited for motor driving where noise is a problem. In addition, since a comparator is used, the resistance to external noise is weak.
Further, since the integrator is included in the calculation of the compensation voltage, there is a problem such as overflow, and there is a problem that it is difficult to realize with a simple system.

  The present invention has been made to solve the above-described problems, and does not add a new device, and can reduce noise without depending on the PWM frequency. An object of the present invention is to obtain an inverter control inverter control device, an inverter control method, and a refrigeration air conditioner capable of generating a dead time compensation voltage by a simple method, reducing current ripple and realizing highly efficient operation.

  In order to solve the above problems, a three-phase voltage type PWM inverter control method and apparatus according to the present invention rotate at the same frequency as the phase current or the biaxial component of the phase current on the stator coordinate system or the electrical angular frequency of the rotor. A coordinate system {hereinafter referred to as a rotating coordinate system (dq coordinate system)} or a coordinate system rotating at the same frequency as the two-axis component of the phase current or the output voltage of the inverter {hereinafter referred to as a general coordinate system (γδ coordinate system)] A specific order component of the dead time error voltage is obtained from the biaxial component of the phase current, the DC voltage, the dead time, and the carrier frequency, and dead time compensation is performed.

  According to the present invention, noise can be reduced without adding a new device and without depending on the PWM frequency, and there are few application restrictions on the operating frequency and the motor to be driven, and the dead time compensation voltage is generated in a simple manner. In addition, it is possible to obtain an inverter control device, an inverter control method, and a refrigeration air conditioner that can realize a highly efficient operation with less current ripple.

Hereinafter, specific embodiments of the present invention will be described.
Embodiment 1 FIG.
The PWM inverter control method and apparatus according to the first embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram illustrating a control device for a PWM inverter according to a first embodiment of the present invention. In the following text, when using a sensor that detects the rotor position, such as a pulse encoder, the electrical angular frequency of the rotor and the rotational frequency of the inverter are substantially the same, so the inverter rotates at the same frequency as the electrical angular frequency of the rotor. The coordinate system to be defined is defined as a rotating coordinate system (dq coordinate system). In addition, when a sensor for detecting the rotor position such as a pulse encoder is not used, the dq axis coordinates cannot be accurately captured by the inverter control unit 100, and in actuality, the phase difference between the rotating coordinate system (dq coordinate system) The inverter is rotating with a deviation of Δθ. Assuming such a case, a coordinate system that rotates at the same frequency as the output voltage of the inverter is referred to as a general coordinate system (γδ coordinate system), and is handled separately from the rotating coordinate system. The d-axis and q-axis in the text indicate the following meanings. That is, the N pole side on the rotor of the electric motor 6 is set as the d axis, and the phase advanced by 90 degrees in the rotation direction is set as the q axis.

  The apparatus shown in FIG. 1 performs a calculation using the current detection means 3a to 3b, the DC voltage detection means 90, the phase current value obtained by the current detection means, and the DC voltage value obtained by the DC voltage detection means, and outputs a PWM signal. And an inverter main circuit 2 that drives the motor 6 by turning on and off the switching elements 4a to 4f by a PWM signal.

  The current detection means 3a to 3b are constituted by a detection element / circuit for detecting a phase current value, an A / D converter capable of taking the output into a CPU or the like and converting the current, an amplifier, and the like. The phase current is detected using at least one detection element. FIG. 1 shows an example using a current transformer. Although control may be performed using three current transformers, in many cases, motor control is generally performed using two current transformers using the characteristics of a three-phase balanced inverter. An example of using two is shown. Moreover, the both-ends voltage which generate | occur | produces in the resistance inserted between the lower side switching elements 4d-4f of each phase and DC power source negative side is detected, and it detects according to the timing to which an electric current flows into a lower side switching element, Phase current detection may be performed by converting the current. In this case as well, control may be performed using three resistors, but motor control is performed using at least one detection element. Alternatively, the phase current may be detected by detecting the voltage across the resistor inserted in the path of the DC bus, and utilizing the relationship in which the switching pattern of the switching element and the flowing phase current are uniquely determined.

  The DC voltage detection means 90 includes a detection element / circuit for detecting a DC voltage value, an A / D converter capable of taking the output into a CPU or the like and converting the voltage, an amplifier, and the like.

  The inverter main circuit 2 includes switching elements 4a to 4f and diodes 5a to 5f.

Further, the inverter control unit 100 uses the DC voltage detection means 90 for obtaining the DC voltage value and the current values obtained by the current detection means 3a to 3b, as necessary, for a three-phase equilibrium condition (the sum of the three-phase current is 0). Phase current detection means 10 for obtaining the three-phase current value by calculating the remaining phase current value, and converting the three-phase phase current value into a biaxial component on the stator coordinate system (αβ coordinate system) Three-phase two-phase conversion means 20 for rotating, and rotating coordinate conversion means for converting the α-axis current Iα and β-axis current Iβ that are the outputs of the three-phase two-phase conversion means into a two-axis component on the rotating coordinate system or the general coordinate system 30 and the d-axis current Id · q-axis current Iq which is a biaxial component on the rotational coordinate system obtained by the rotational coordinate conversion means, or the γ-axis current Iγ and the δ-axis current Iδ which are biaxial components on the general coordinate system. , Α-axis current Iα, β-axis current Iβ, phase current values Iu to Iw, and DC voltage value V Speed control from a dead time compensation amount calculation unit 70 for calculating a dead time compensation amount from dc, a d-axis current Id / q-axis current Iq or a γ-axis current Iγ / δ-axis current Iδ, a frequency command ωe ^*, and a DC voltage value Vdc Voltage command calculating means 40 for performing various vector controls including the rotational coordinate system and calculating the biaxial voltage command on the general coordinate system, and the biaxial voltage command (Vα ^* , Vβ ^* ) on the stator coordinate system. Rotational coordinate reverse conversion means 50 for calculating each phase voltage command Vu ^{* to} Vw ^* , two phase three phase conversion means 60 for calculating each phase voltage command, adders 74a to 74c for adding a dead time compensation amount to each phase voltage command, The PWM signal generating unit 80 generates a PWM signal from each phase voltage command to which the dead time compensation amount is added, and the PWM signal generating unit 81 generates the obtained PWM signal.

  The dead time compensation amount calculation unit 70 calculates a specific order component of the dead time error voltage from the DC voltage value Vdc and calculates a dead time compensation amount amplitude, and a phase current value or stator. From the two-axis component of the phase current on the coordinate system or the two-axis component of the phase current on the rotating coordinate system or the general coordinate system, the zero phase of the phase voltage (hereinafter sometimes referred to as voltage zero phase) and the zero of the phase current The dead time compensation amount is calculated from the phase difference detecting means 72 for detecting the phase difference of the phase (hereinafter sometimes referred to as zero current phase), the dead time compensation amount amplitude, and the phase difference between the zero voltage phase and the zero current phase. Compensation amount calculation means 73 for calculating is configured. Here, the current zero phase is a phase of a voltage command when the current is zero.

Next, the operation will be described with reference to the drawings.
FIG. 2A shows an x-phase voltage command where an arbitrary one of the three phases obtained by the PWM signal creation means 80 for creating the PWM signal is an x-phase. The state of the phase current that flows when this voltage is applied to the x phase is also shown. Here, the error voltage due to the dead time can be expressed by a square wave as shown in FIG. From the DC voltage Vdc obtained from the DC voltage detecting means 90, the carrier frequency fc of the inverter, and the dead time Td, the amplitude of the error voltage is obtained by the following equation (1).

  However, if the square wave or trapezoidal wave of FIG. 2B is used for dead time compensation, noise becomes a problem in inverter control using a carrier frequency (about 20 to 20 k [Hz]) in the audible range. There is a case. Therefore, a method has been devised in which a specific order component (especially a low-order component) of the dead time error voltage is used as the dead time compensation voltage without using a conventional square wave or trapezoidal wave for the compensation pattern.

  For example, Fourier series expansion is used to extract the specific order component. The dead time compensation amount calculation unit 70 uses the Fourier series to calculate the dead time compensation amount. A method for calculating the specific order component of the dead time error voltage (square wave) by the Fourier series expansion in the dead time compensation amount specific order calculating means 71 from the DC voltage Vdc will be described below with respect to the square wave as shown in FIG. . Equation 2 shows a functional expression related to the square wave from 0 to T / 2 when the square wave period is T. T is an arbitrary value determined by the electrical angular velocity of the inverter. T represents an arbitrary time.

  Here, the Fourier series can be expressed by Equation 3. Here, n represents the order of the nth harmonic of the fundamental wave (n is a positive integer). Further, t represents an arbitrary time, and ω represents a rotating electrical angular velocity (= 2π / T) (T represents the above-described square wave period).

Since Equation 2 is an even function, it is clear that a _n = 0. Further, Equation 2 is a symmetric wave, and the half-cycle integrated value may be doubled, and Equation 4 may be obtained.

b ₀ and b _n are obtained as in Expression 5 and Expression 6.

As described above, Expression 7 is obtained as a result of Fourier series expansion of the even function as shown in Expression 2.

  Although the dead time compensation voltage is obtained as an even function here, it may be obtained as an odd function.

  Since Equation 7 is a symmetric wave with respect to the y-axis as apparent from FIG. 3, it includes only odd-order harmonics. If noise due to odd-order harmonics of the third or higher order becomes a problem depending on the specifications of the PWM inverter and the motor device, only the primary component in Equation 7 may be used.

  Further, the phase difference detection means 72 detects the phase difference αx (x = U, V, W) between the zero voltage phase and the zero current phase for each phase from the phase currents Iu to Iw. Specifically, the phase difference αx (x = U, V, W) between each voltage zero phase and each current zero phase is determined based on the switching timing of the polarity of the phase current. That is, the phase difference αx between the point at which the polarity of the zero cross of the current changes from − to + and the current phase of the voltage command at this time is obtained. Note that a point where the polarity of the zero crossing of the current changes from + to-may be used. FIG. 2A shows an example in which αx is obtained from the point where the polarity of the zero cross of the x-phase current changes from + to − and the phase difference between the voltage commands with reference to the x-phase voltage command. Alternatively, assuming that αu, αv, and αw are equal, the dead time compensation voltage for each phase may be obtained from the phase difference information of the phase of the voltage command for one phase and the current phase. When noise or the like becomes a problem, the phase difference αx between the voltage command of each phase and the phase current may be obtained by time integration of the phase current value. The phase difference αx can be obtained directly from the phase current as described above, but can also be obtained from biaxial current information in the stator coordinate system (αβ coordinate system). Next, a method for detecting the phase difference between the voltage command of each phase and the phase current from the biaxial current information in the stator coordinate system will be described below. The phase current values of the respective phases are represented by Iu, Iv, and Iw, and are expressed as in Expression 8. Here, θ is the phase of the voltage command, Ia is the current amplitude of each phase (same for simplification), and αx is the phase difference between the voltage command phase of each phase and the phase current phase.

  If the two-axis components on the stator coordinate system after the three-phase to two-phase conversion are set as α-axis current Iα and β-axis current Iβ, Iα and Iβ are expressed by the following equation (9). However, although the rotation direction of the inverter is assumed to be counterclockwise, the same method can be used for clockwise rotation.

  For simplification, assuming that a three-phase equilibrium state is maintained, αu, αv, αw are set as αx, and when Equation 8 is substituted into Equation 9, the phase difference αx between the voltage command and the phase current of each phase is Can be derived as follows.

That is, the phase difference αx between each phase voltage command and phase current in FIG. 2 (a) can be calculated from the biaxial components of the stator coordinate system, and even when inverter control is performed based on the voltage phase, The phase difference αx between the voltage command and phase current can be calculated, and dead time compensation can be performed easily and accurately. Also, when the phase current is in the vicinity of 0 [A], the detected current polarity swings in the positive direction or in the negative direction and is unstable, so the section where the detected current value is less than the predetermined value (hereinafter this section is It is possible to realize highly accurate dead time compensation without requiring special processing such as not performing dead time compensation in a forbidden band).
Furthermore, when the noise in the actual environment is large, a filter may be used in the input / output of several tens.

  Similarly, the voltage command of each phase and the phase difference αx between the phase currents can be directly obtained from the current information on the rotating coordinate system (dq coordinate system). For the sake of simplification, the rotation coordinate conversion means 30 changes the U-phase direction and the α-axis direction to 0 [deg] of the rotation angle with respect to the α-axis / β-axis current output from the three-phase two-phase conversion means 20. Perform rotation coordinate transformation. That is, a rotation matrix such as Equation 11 is multiplied.

  For simplicity, it is assumed that a three-phase equilibrium state is maintained. The phase difference αu, αv, αw between the voltage command phase and current phase of each phase is set as αx, and the U-phase direction is set to the reference phase (0 [deg] ) In the same manner as in the stator coordinate system, αx can be derived as shown in Equation 12 by expanding Equation 11. Although the rotational direction is taken counterclockwise in Equation 11, αx can be calculated by the same method even if it is developed clockwise in some cases.

  That is, the phase difference αx between each phase voltage command and phase current in FIG. 2 (a) can be calculated from the biaxial current component on the rotating coordinate system, and even when the inverter control is performed based on the voltage phase, each phase It can be seen that a phase difference αx between the voltage command and the phase current is obtained. Therefore, the phase difference αx between the voltage command of each phase and the phase current can be obtained regardless of the polarity of the phase current, and dead time compensation can be performed easily and accurately. Also, even when the current level is near zero, highly accurate dead time compensation can be realized without the need for special processing such as setting of a forbidden band. Further, when the motor is driven, current ripple can be reduced and high efficiency can be realized. Further, when the noise in the actual environment is large, a filter may be used in the input / output of Equation 12.

  Even in the case of handling in the general coordinate system, αx can be similarly obtained by considering the phase difference Δθ with respect to the rotating coordinate system. That is, θ in Equation 11 is set as θ + Δθ, and αx can be obtained as shown in Equation 13 from γ-axis current and δ-axis current corresponding to d-axis current and q-axis current.

  From the phase difference αx of each phase voltage command and phase current obtained in this way and the specific order component obtained in the dead time compensation amount specific order component calculation means 71, the dead time compensation amount ΔVx (x = U, V, By calculating W), the phase difference αx between the voltage command of each phase and the phase current can be obtained regardless of the polarity of the phase current, and dead time compensation can be performed easily and accurately. Further, even when the current level is near zero, it is possible to realize highly accurate dead time compensation without requiring special processing such as setting a prohibited band. Further, when the noise in the actual environment is large, a filter may be used in the input / output of Equation 13. In this way, in the phase difference detection means 72, the phase currents Iu to Iw, the α-axis current Iα / β-axis current Iβ on the stator coordinate system, or the d-axis current Id · q-axis current Iq on the rotor coordinate system, Alternatively, αx can be obtained with high accuracy by using either the γ-axis current Iγ or the δ-axis current Iδ in the general coordinate system.

  From the Fourier series expansion result f (t) of the square wave obtained as described above and the phase difference αx between the voltage command phase and the current phase of each phase, the dead time compensation amount calculating means 73 performs the dead time compensation amount ΔVx. (X = U, V, W) is calculated. Equation 14 shows the dead time compensation amount of each phase. However, Equation 14 is an example in which the U-phase voltage phase is used as a reference and the rotor rotation direction is counterclockwise. In addition, it is calculated | required by the same view also around the clock.

  Also, as an example, the primary component of the dead time compensation voltage is expressed by Equation 15.

FIG. 4 shows a first-order component waveform example of the dead time compensation voltage of Formula 15 corresponding to FIG. By adding the phase compensation components ΔVu to ΔVw obtained by Equations 14 and 15 to the respective phase voltage commands Vu ^{* to} Vw ^* before compensation as shown in Equation 16 by the adders 74a to 74c, Vu ^** to Obtain Vw ^** .

Based on the compensated phase voltage commands Vu ^{** to} Vw ^** , the PWM signal generating means 80 obtains energization times Tup to Twn in one carrier of the switching element. Thereafter, the PWM signal generating means 81 issues drive signals Up to Wn to the switching elements 4a to 4f from a microcomputer, a CPU or the like, and the electric motor 1 can be driven. Inverter control that compensates for the dead time can be realized in this way, ripple reduction near the phase current zero crossing becomes possible, and compensation performance can be improved by adding a forbidden band near the zero cross without adding a new device. Noise reduction and high efficiency motor drive that can accurately compensate for dead time without deterioration can be realized.

If necessary, compensation may be performed by adjusting the dead time compensation voltage amplitude of the primary component so that the time integral value of the dead time error voltage is equal to the time integral value of the dead time compensation voltage of the primary component. good. FIG. 2C shows the state of the dead time compensation voltage when only the primary component is used. Now, the dead time compensation voltage corresponding to the dead time error voltage in FIG. 2B is compared with the dead time compensation voltage in FIG. The time integral value S1 of the positive-side half cycle of the dead time compensation voltage corresponding to the dead time error voltage in FIG. 2B is obtained as πfcTdVdc / ω by time-integrating Equation 2.
Further, the time integration value S ₂ of the positive-side half cycle of the dead time compensation voltage in FIG. 2C is also 8fcTdVdc / πω by time-integrating Equation 15.
Therefore, in the compensation of only the first-order component by the Fourier series expansion of FIG. 2C, S2 / S1 = 8 / π ² (about 81 [ %]). Accordingly, the compensation amount calculation means 73, so that the time integral value shown in FIG. 2 (b) and FIG. 2 (c) are matched, the amplitudes and times [pi ^2/8 primary component amplitude shown in FIG. 2 (c) By setting, the error of the time integral value is small, the influence on noise by specific order harmonics is reduced, and at the same time, the startability, controllability improvement and high efficiency can be improved. The compensation amount calculation means 73 calculates the sum of the time integration value of the dead time error voltage calculated by the compensation amount specific order component calculation means 71 and the time integration value of at least one specific order component of the dead time error voltage. By creating the dead time compensation voltage so as to be equal, it is possible to reduce the noise caused by the contained harmonics of the dead time compensation voltage.

In addition, depending on the motor device and application, the order of the harmonics that generate noise is determined, and it may be desired to reduce current ripple due to the effect of dead time as much as possible. In this case, since the dead time compensation component other than the first order is about 19%, the dead time compensation may be performed by combining the remaining orders excluding the order in which noise is a problem. With respect to Equation 14, low-noise and high-efficiency dead time compensation can be performed by performing compensation by combining the order components.
Further, the 3n-order component does not contribute to actual dead time compensation. Equation 17 shows the dead time compensation voltage in the third-order case.

Here, when A = −4 · fc · Td · Vdc and x = θ−αx, and the three-phase two-phase conversion shown in Equation 18 is performed, the α · β-axis voltage component of Equation 19 is obtained. .

Next, when the rotational coordinate conversion shown in Equation 20 or 21 is performed, the d · q axis voltage component of Equation 22 is obtained, and as a result, both components become zero.

  It can be seen that the third order multiple can be derived in the same manner and does not contribute to the dead time compensation. As described above, the compensation amount calculation means 73 obtains at least one odd-order component other than the 3n-order component among the specific order components of the dead time error voltage, and performs dead time compensation, thereby reducing the calculation load. Time compensation can be performed. In addition, the compensation amount calculation means 73, if necessary, out of the specific order component of the dead time error voltage, the magnitude of the operating frequency or the load torque, the magnitude of the output voltage, or the ratio of the output voltage to the bus voltage (hereinafter, Depending on the modulation rate), at least one order component to be used is selected from at least one specific order component of the dead time error voltage to compensate for the dead time. It can be driven with low noise against the noise of the other.

FIG. 5 is a flowchart showing the dead time compensation operation of the control unit 100 according to Embodiment 1 of the present invention. Next, the operation of the control unit 100 will be described with reference to FIG.
First, the control unit 100 acquires the current value of each phase using the phase current detection means 10. In addition, the control unit 100 acquires the DC voltage of the DC power source 1 using the DC voltage detection unit 90. Furthermore, a frequency command value is acquired from the outside. If there is a rotor sensor, the position of the rotor is detected. (Step 501). Next, the control unit 100 calculates the phase difference αx between the phase of the phase current and the phase voltage from the acquired phase current, DC voltage, frequency command value, and rotor position (step 502). Next, the control unit 100 calculates a compensation amount specific order component from the acquired DC voltage, a preset dead time, and frequency (step 503). Next, the control unit 100 calculates a dead time compensation amount using the phase difference αx between the phase current and the phase voltage obtained in step 502 and the compensation amount specific order component obtained in step 503 (step 504). . However, regarding the dead time compensation amount calculation in step 504, the amplitude of the dead time compensation voltage specific order component may be adjusted so that the dead time error voltage and the time integral value of the dead time compensation voltage specific order component to be compensated match. . Next, the control unit 100 adds the dead time compensation amount to the command voltage calculated based on the frequency command, and creates a PWM signal based on the added value (step 505). Next, the control unit 100 outputs a PWM signal to the inverter main circuit 2 (step 506).

  As described above, according to the first embodiment, the three-phase voltage type PWM inverter control method and apparatus are the two-axis component on the phase coordinate or the stator coordinate system of the phase current or the two-axis component on the rotational coordinate system of the phase current. Since the specific order component of the dead time error voltage is obtained from the two axis components on the general coordinate system of the axis component or phase current, the DC voltage, the dead time, and the carrier frequency, dead time compensation is performed, so that a new device is not added. Even in the vicinity of the zero cross of the phase current, dead time compensation can be performed with high accuracy without hunting at the time of detection and without deterioration in compensation performance due to the provision of a forbidden band near the zero cross. Inverter that can reduce noise without depending on PWM frequency, has few restrictions on operation frequency and drive target motor, generates dead time compensation voltage in a simple way, has less current ripple, and realizes high-efficiency operation A control method and apparatus can be obtained.

Embodiment 2. FIG.
In FIG. 1, after the dead time compensation components ΔVu to ΔVw are added to the respective phases via the adders 74 a to 74 c with respect to the respective phase voltage command values Vu ^{* to} Vw ^* obtained by the two-phase / three-phase conversion means 60. In the example, a PWM signal is generated by the PWM signal generating means 80 and a PWM signal is generated by the PWM signal generating means 81.
The calculation of the dead time compensation amount may be performed from the stator coordinate system.
In the second embodiment, a case where the calculation of the dead time compensation amount is performed from the stator coordinate system (αβ coordinate system) will be described.
FIG. 6 is a diagram illustrating a PWM inverter control device according to the second embodiment of the present invention.
FIG. 6 differs from FIG. 1 in that the dead time compensation is performed. Hereinafter, the difference in the dead time compensation method and apparatus for the inverter according to the second embodiment of the present invention from the first embodiment will be described with reference to the drawings. While explaining.
In FIG. 6, dead time compensation is performed by adding a compensation voltage component to the biaxial component of the voltage command on the stator coordinate system. That is, FIG. 6 shows that the α-axis voltage command Vα ^* and β-axis voltage command Vβ ^* obtained by the rotating coordinate inverse conversion means 50 are compared with the α-axis voltage command compensation components ΔVα ^* and β calculated by the compensation amount calculating means 71. The dead time compensation is performed by adding the shaft voltage command compensation component ΔVβ ^* to each component of the αβ axis via the adders 76a to 76b. The voltage components Vα ^** and Vβ ^{** on} the αβ axis after the dead time compensation can be expressed by Equation 23.

  The operation will be described below. Based on Equation 7, the dead time compensation amount of each phase obtained as shown in Equation 14 is subjected to the three-phase two-phase transformation of Equation 18 in the compensation amount calculation means 73, and the biaxial component on the stator coordinate system is converted into a biaxial component. Convert.

Next, the compensation amount components are added through the adders 76a to 76b to the α-axis voltage command Vα ^* and β-axis voltage command Vβ ^* obtained by the rotational coordinate inverse transform 50, and Vα ^* ′ (= Vα ^* + ΔVα ), Vβ ^* ′ (= Vβ ^* + ΔVβ) is obtained. Vα ^* ′ and Vβ ^* ′ are used for the two-phase / three-phase conversion means to calculate each phase voltage command Vu ^{* to} Vw ^* . Subsequent operations are the same as those in the first embodiment.
Thus, dead time compensation can be performed even on the stator coordinate system. In addition, although the example of the primary component is shown in Equation 15, the specific order component shown in Equation 14 can be performed in the same way.

  According to the second embodiment, in addition to the same effects as those of the first embodiment, the number of adders is one less than that of the first embodiment.

Embodiment 3 FIG.
The calculation of the dead time compensation amount may be performed from the rotating coordinate system (dq coordinate system).
In the third embodiment, a case where the calculation of the dead time compensation amount is performed from the rotating coordinate system (dq coordinate system) will be described.
FIG. 7 is a diagram illustrating a PWM inverter control device according to the third embodiment of the present invention.
FIG. 7 differs from FIG. 1 in the point where dead time compensation is performed.
Hereinafter, the inverter dead time compensation method and apparatus according to the third embodiment of the present invention will be described with reference to the drawings, with respect to differences from the first embodiment.
In FIG. 7, dead time compensation is performed by adding a compensation voltage component to the biaxial component of the voltage command on the rotating coordinate system or the general coordinate system. That is, FIG. 7 shows compensation for the d-axis voltage command Vd ^* (or γ-axis voltage command Vγ ^* ) and q-axis voltage command Vq ^* (or δ-axis voltage command Vδ ^* ) obtained by the voltage command calculation means 40. The adder 75a adds the d-axis voltage command compensation component ΔVd ^* (or γ-axis voltage command compensation component ΔVγ ^* ) and q-axis voltage command compensation component ΔVq ^* (or δ-axis voltage command compensation component ΔVδ ^* ) calculated by the quantity calculation unit 70. The dead time compensation is performed by adding to each component of d-axis voltage and q-axis voltage (or γ-axis voltage and δ-axis voltage) through ˜75b. The voltage components Vd ^** (Vγ ^** ) and Vq ^** (Vδ ^** ) on the dq axis (γδ axis) after the dead time compensation can be expressed by Expression 24 and Expression 25.

  The operation will be described below. Based on Equation 7, the dead time compensation amount of each phase obtained as shown in Equation 14 is subjected to the three-phase two-phase transformation of Equation 18 in the compensation amount calculation means 73, and the biaxial component on the stator coordinate system is converted into a biaxial component. Convert.

Next, the rotational coordinate transformation of Equation 20 is applied to the α-axis voltage command Vα ^* and β-axis voltage command Vβ ^* obtained by Equation 18, and the compensation amount component is transformed into a biaxial component on the rotor coordinate system. By adding each component of the compensation amount via the adders 75a to 75b to the d-axis voltage command Vd ^* and the q-axis voltage command Vq ^* obtained by the voltage command calculation means 40 as shown in Equation 24, Dead time compensation can be performed even on a rotating coordinate system.

The calculation of the dead time compensation amount may be performed from a general coordinate system (γδ coordinate system). Even in the case of position sensorless control in which the inverter rotates with a phase difference of Δθ with respect to the actual rotor position, the control position deviation Δθ corresponding to the rotor position is taken into consideration as shown in Equation 21. This can be done in the same way. By adding each component of the compensation amount to the γ-axis voltage command Vγ ^* and the δ-axis voltage command Vδ ^* obtained by the voltage command calculation unit 40 via the adders 75a to 75b as shown in Equation 25, Dead time compensation can also be performed on a general coordinate system. Further, by taking Δθ into consideration, dead time compensation can be performed with high accuracy even during acceleration / deceleration. Further, depending on the application and operating state, Δθ may be ignored.

  According to the third embodiment, in addition to the same effects as those of the first embodiment, the number of adders is one less than that of the first embodiment.

Embodiment 4 FIG.
The calculation of the dead time compensation amount may be performed from a voltage command.
In the fourth embodiment, a case where the calculation of the dead time compensation amount is performed from a voltage command will be described.
FIG. 8 is a diagram illustrating a PWM inverter control device according to the fourth embodiment of the present invention.
FIG. 8 differs from FIG. 1 in the point where dead time compensation is performed.
Hereinafter, the inverter dead time compensation method and apparatus according to the fourth embodiment of the present invention will be described with reference to the drawings, with respect to differences from the first embodiment.
In FIG. 8, the voltage command V ^** is created in consideration of the dead time compensation component. The voltage command V ^* before dead time compensation can be expressed as shown in Equation 26.

The operation will be described below.
FIG. 8 shows various vector controls in the voltage command calculation means 40 from information on the d-axis current Id (or γ-axis current Iγ), the q-axis current Iq (or δ-axis current Iδ), the frequency command ωe ^*, and the DC voltage value Vdc. Voltage command compensation calculated by the compensation amount calculation unit 70 using Equation 27 for the voltage command V ^* obtained by calculating Equation 26 from the voltage component of either the dq axis or the γδ axis obtained by performing A voltage command value V ^** is created using the component ΔV ^* . The voltage command value V ^** after the dead time compensation can be expressed by Equation 28. A PWM is created and output using each voltage component or voltage command compensated as described above.

In a PWM method such as space vector modulation often used in inverter control, a PWM signal is created based on the voltage command V * obtained by Equation 26. Therefore, it is controlled to add a compensation amount to the voltage command V *. It may be convenient for the algorithm. Therefore, according to the fourth embodiment, the biaxial compensation amount component obtained on the rotating coordinate system or the general coordinate system is calculated as the compensation amount component ΔV ^* of the voltage command as shown in Equation 27, and By calculating the command voltage, various modulation methods can be handled more flexibly.

  Further, according to the fourth embodiment, in addition to the same effects as those of the first embodiment, an adder is unnecessary.

It is a figure which shows the control apparatus of the PWM inverter concerning Embodiment 1 of this invention. Compensation voltage for compensating the primary component of the error voltage obtained by the positional relationship between the phase difference between the arbitrary phase voltage command and the phase current according to the first and second embodiments of the present invention, an example of the error voltage due to the dead time, and Fourier series expansion It is a figure which shows the example of a pattern. It is a figure which shows the example of the periodic function (dead time error voltage) which performs the Fourier series expansion by Embodiment 1-2 of this invention. It is a figure which shows the example of the U-phase voltage component of the dead time compensation voltage by Embodiment 1-2 of this invention. It is a flowchart which shows operation | movement of the control part of the PWM inverter concerning Embodiment 1 of this invention. It is a figure which shows the control apparatus of the PWM inverter concerning Embodiment 2 of this invention. It is a figure which shows the control apparatus of the PWM inverter concerning Embodiment 3 of this invention. It is a figure which shows the control apparatus of the PWM inverter concerning Embodiment 4 of this invention.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 DC power supply, 2 Inverter main circuit, 3a-3c Current detection means, 4a-4f Switching element, 5a-5f Diode, 6 Electric motor, 10 phase current detection means, 20 3 phase 2 phase conversion means, 30 Rotation coordinate conversion means, 40 voltage command calculation means, 50 rotational coordinate reverse conversion means, 60 two-phase three-phase conversion means, 70 dead time compensation section, 71 compensation amount specific order component calculation means, 72 phase difference detection means, 73 compensation amount calculation means, 74a- 74c adder, 75a to 75b adder, 76a to 76b adder, 80 PWM signal generation means, 81 PWM signal generation means, 90 DC voltage detection means, 100 inverter control unit, 113 to 115 adder, 116 PWM pulse calculator 117 Voltage command calculator, 121/131/141 comparator, 122/132/142 The amount calculator, 123, 133, 143 multipliers, 124, 134, 144 integrator 151 frequency setting device.

Claims (16)

DC voltage detection means for detecting the voltage of the DC power supply of a voltage source PWM inverter having a DC power supply and switching means;
Current detection means for detecting an output current of each phase of the voltage-type PWM inverter;
Based on the output of the DC voltage detection means and the output of the current detection means, an error voltage between the voltage command of the voltage-type PWM inverter and the output voltage caused by the dead time for preventing arm short circuit of the switching means is obtained. A dead time compensation unit for generating a compensation amount for compensation;
Voltage command calculation means for calculating a voltage command based on a frequency command, an output of the DC voltage detection means, and an output of the current detection means;
PWM signal creation means for creating a PWM signal based on the output of the voltage command calculation means and the output of the dead time compensation unit;
PWM signal generating means for generating a PWM signal obtained by the PWM signal generating means,
The dead time compensation unit includes a phase difference detection unit that calculates a phase difference between a phase voltage and a phase current based on an output of the current detection unit;
A compensation amount specific order component calculating means for calculating a specific order component of a dead time error voltage based on a carrier frequency of the switching means, an output of the DC voltage detecting means and a predetermined dead time;
A PWM inverter control apparatus comprising: a compensation amount calculation unit that generates the compensation amount based on an output of the phase difference detection unit and an output of the compensation amount specific order component calculation unit. Comprising three-phase two-phase conversion means for converting the phase current detected by the current detection means into two-axis components on a stator coordinate system;
2. The PWM inverter according to claim 1, wherein the phase difference detection unit calculates a phase difference between the phase voltage and the phase current based on an output of the three-phase / two-phase conversion unit and an output of the DC voltage detection unit. Control device. Rotational coordinate conversion means for converting the phase current detected by the current detection means into a biaxial component of the phase current on a coordinate system that rotates at the same frequency as the electrical angular frequency of the rotor;
2. The PWM inverter control device according to claim 1, wherein the phase difference detection unit calculates a phase difference between a phase voltage and a phase current based on an output of the rotating coordinate conversion unit and an output of the DC voltage detection unit. . A rotation coordinate conversion means for converting the phase current detected by the current detection means into a two-axis component of the phase current on a coordinate system rotating at the same frequency as the output voltage of the inverter;
2. The PWM inverter control device according to claim 1, wherein the phase difference detection unit calculates a phase difference between a phase voltage and a phase current based on an output of the rotating coordinate conversion unit and an output of the DC voltage detection unit. .   The compensation amount calculating means creates a dead time compensation voltage using at least one specific order component of the dead time error voltage calculated by the compensation amount specific order component calculating means. 4. The PWM inverter control device according to claim 4.   The compensation amount calculation means is configured such that the sum of the time integration value of the dead time error voltage calculated by the compensation amount specific order component calculation means and the time integration value of at least one specific order component of the dead time error voltage is equal. The PWM inverter control device according to claim 1, wherein each order component amplitude of a dead time error voltage used in the step is set.   The compensation amount calculation means obtains at least one odd-order component other than the 3n (n is a positive integer) order component among the specific order components of the dead time error voltage calculated by the compensation amount specific order component calculation means. The PWM inverter control device according to claim 1, wherein dead time compensation is performed.   The compensation amount calculating means, among the specific order components of the dead time error voltage calculated by the compensation amount specific order component calculating means, according to the operating frequency or the magnitude of the load torque, the magnitude of the output voltage or the modulation rate, The dead time compensation is performed by selecting at least one order component to be used from among at least one specific order component of at least one dead time error voltage. PWM inverter control device.   9. An adding means for adding each phase component of a dead time compensation amount of the compensation amount calculating means to each phase voltage command from the voltage command calculating means. The PWM inverter control device described.   The said dead time compensation part adds each axis | shaft component of a dead time compensation amount with respect to the biaxial voltage command on the coordinate system which rotates with the same frequency as the output voltage of an inverter. PWM inverter control device.   A refrigeration air conditioner comprising the PWM inverter control device according to any one of claims 1 to 10. The voltage of the voltage source PWM inverter caused by the dead time for preventing arm short circuit of the switching means of the voltage source PWM inverter based on the DC voltage and the AC output current obtained from the DC voltage by the voltage source PWM inverter. A dead time compensation step for generating a compensation amount for compensating an error voltage between the command and the output voltage;
A voltage command calculation step for calculating a voltage command based on the frequency command, the DC voltage, and the AC output current;
A PWM signal creating step for creating a PWM signal based on the voltage command obtained by the voltage command calculation step and the compensation amount obtained by the dead time compensation step;
A PWM signal generating step for generating the PWM signal generated by the PWM signal generating step;
With
The dead time compensation step calculates a phase difference between a phase voltage and a phase current based on the DC voltage and an output current of the voltage source PWM inverter, and
A compensation amount specific order component calculating step for calculating a specific order component of a dead time error voltage based on a carrier frequency of the switching means, the DC voltage, and a predetermined dead time;
A PWM inverter control method comprising: a compensation amount calculating step for generating the compensation amount based on an output of the phase difference calculating step and an output of the compensation amount specific order component calculating step.   13. The PWM inverter control according to claim 12, wherein in the phase difference detection step, a phase difference between the phase voltage and the phase current is calculated based on the two-axis component on the stator coordinate system and the DC voltage in the phase difference detection step. Method.   In the phase difference detection step, the phase difference between the phase voltage and the phase current is calculated based on the two-axis component of the phase current on the coordinate system that rotates the phase current at the same frequency as the electrical angular frequency of the rotor and the DC voltage. The PWM inverter control method according to claim 12.   In the phase difference detection step, calculating the phase difference between the phase voltage and the phase current based on the two-axis component of the phase current on the coordinate system rotating the phase current at the same frequency as the output voltage of the inverter and the DC voltage. The PWM inverter control method according to claim 12, wherein:   16. The PWM inverter control method according to claim 12, wherein in the compensation amount calculation step, a dead time compensation voltage is created using at least one specific order component of the dead time error voltage. .

Images