JPH0519396B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION <Technical Field of the Invention> The present invention relates to a vector control device for an induction motor, and particularly to a vector control device using a pulse width modulation (PWM) type transistor inverter.

<Prior art and problems> In recent years, as a control method to improve the quick response of induction motors, the primary current of the motor is divided into exciting current and secondary current, and the secondary magnetic flux and secondary current vector are always orthogonal. A vector control method has been proposed that attempts to obtain responsiveness equivalent to that of a DC machine by

As such a vector control method, we use a voltage-based vector control method that uses a PWM inverter as a power converter that supplies AC power to the motor.
The applicant of the present application has already proposed a non-interference control method that can cancel mutual interference between the secondary magnetic flux component and the secondary current component (Japanese Patent Laid-Open No. 165982/1982). An outline of this will be explained below with reference to FIG.

In controlling the motor 1 by supplying a voltage-controlled primary voltage from the PWM transistor inverter 2 so that the magnetic flux and the secondary current are orthogonal to each other, the direction of the magnetic flux is set as the α axis, and the direction of the secondary current is set as the α axis. The two-phase α-phase primary voltage e ₁ 〓 and β-phase primary voltage e ₁ 〓 - are obtained from the α-phase primary current i 1 〓 ^* and β-phase primary current i ^* ₁ 〓 as the command values on the β _- axis perpendicular ^to the α-axis. To obtain the voltage signal, the motor 1 is
Interference with the magnetic flux caused by the β-phase primary current i ₁ 〓 and interference with the secondary current caused by the α-phase primary current i ₁ 〓 are removed. By this correction calculation circuit 3, the α-phase primary voltage e ₁ 〓 and the β-phase primary voltage e ₁ 〓 become magnetic flux and secondary current command signals that do not interfere with each other, and these signals can be processed using the polar coordinate method or the two-axis method. The three-phase voltage command signal e _a ^* of the inverter 2 is calculated by the phase voltage calculation circuit 7 according to
Converted to e _b ^* , e _c ^* .

The β-phase primary current command i ₁ 〓 ^* is taken out as the output of the speed regulator 5 by matching the speed setting value v _s ^* with the detected value ω _r of the speed detector 4 of the motor, and the power supply angular frequency ₀ is obtained by angular frequency calculation obtained by circuit 6. In addition, the sine wave/cosine wave signals sINω ₀ t and cosω ₀ t necessary for two-phase/three-phase conversion in the phase voltage calculation circuit 7 are obtained from the trigonometric function generation circuit 8 using the power supply angular frequency ω ₀ , and the inverter 2 A triangular wave signal T _ri as a carrier wave necessary for pulse width modulation in is obtained from the triangular wave generation circuit 9 using ω ₀ . 10 is a rectifier that supplies DC power to the inverter 2.

In this way, the method of vector-controlling the primary voltage of a motor using a PWM inverter performs feed-forward control of the primary voltage, unlike conventional current-controlled vector control, by performing correction calculations for non-interference control. As a result, it has excellent responsiveness, and has been confirmed to have response characteristics that are better than those of DC machines.

However, since this system controls the primary voltage in an open loop, voltage reduction due to dead time between the transistors of the transistor inverter 2 may appear as a control error.

<Objective of the Invention> An object of the present invention is to provide a vector control device that improves control performance by compensating for control errors caused by dead time of a transistor inverter.

<Summary of the invention> The present invention calculates the voltage drop due to dead time in phase voltage calculation using the polar coordinate method, which converts the two-axis control voltage signals e ₁ 〓, e ₁ 〓 to polar coordinates and further converts them to three phases. It is characterized by performing compensation by adding the result to the polar coordinate one-three phase transformation.

<Principle description of the invention> The inverter 2 in FIG. 1 is composed of transistors Tr ₁ to Tr ₆ and a feedback diode D ₁ as shown in FIG. 2.
In an inverter main circuit 2A in which ~ _D6 parallel circuits are bridge-connected, for example, when the upper and lower arms of transistors _Tr1 and _Tr2 are commutated, if there is a period in which both transistors are in the firing state at the same time, turn-off loss will occur. Because of this, control is performed to slightly delay the transistors that turn on compared to the transistors that turn off.

Now, considering the period when the a-phase output current ia is in the direction shown in the figure, the transistor
In order to commutate current from Tr ₁ to transistor Tr ₂ , diode D ₂ conducts when transistor Tr ₁ is turned off and continues to flow current ia, and even if the firing of transistor Tr ₂ is delayed, there is no effect. On the contrary, transistor
Commutation from Tr ₂ to Tr ₁ occurs because transistor Tr ₂ is originally non-conducting and diode D ₂ is conducting.
Until the transistor Tr ₁ is turned on, a current ia flows through the motor 1 from the negative potential through the diode D ₂ . This operation is the same even during the period in which the current ia is in the opposite direction to that illustrated by replacing Tr ₁ and Tr ₂ and D ₂ and D ₁ .

These relationships will be explained with reference to FIG. Although the phase control angle current ia turns on and off transistors Tr ₁ and Tr ₂ according to the PWM waveform obtained by comparing the control voltage signal e _a and the triangular wave Tr ₁ during the positive period Tp in the direction of the arrow in FIG. When the potential at the connection point between Tr ₁ and Tr ₂ changes to positive polarity, there is a delay by the firing delay (dead time T _d ) of the transistor Tr ₁ . Conversely, when the current ia is in the negative period _TN , the change in potential to negative polarity is delayed by the dead time _Td set in the transistor _Tr2 . As shown in Figure d, this delay is equivalent to a pulsed voltage E _d of width T _d added to the opposite polarity, and the fundamental wave component obtained by Fourier expansion of this voltage is the voltage ea ^* that was originally intended to be output. Since the polarity is opposite to , it acts to lower the fundamental wave output voltage. In this way, the control voltage signals e _a ^* , e _b ^* , e _c ^*
In contrast, the control output decreases due to the dead time set in the transistor, causing an error in the intended control output.

The average value _DB of the reverse voltage caused by this dead time T _d is expressed by the following equations (1), (2), and (3).

_DB = E _d 1/πQ _d (P-1)/2 ...(1) Q _d = 2π・T _d・f ...(2) _DB = E _d・T _d・(P-1)・f ... ...(3) Here, E _d is the DC voltage of inverter 2, and P is the triangular wave for one period of the control voltage signal (e _a ^* , etc.)
The number of pulses of Tr ₁ , f is the frequency of the control voltage signal.

Next, the fundamental wave voltage e ₁ obtained by ideal control without dead time T _d is expressed by the following equation (4).

e ₁ =Ed/2μsinωt (4) Here, μ is the control rate that is the ratio of the control voltage signal (e _a ^* , etc.) amplitude to the triangular wave amplitude.

For this fundamental wave voltage e ₁ , the dead time voltage
e _DB takes into account the phase angle (power factor angle) and becomes the following equation (5).

e _DB = π/2・E _d・T _d・(P−1)・f・sin(ωt−ω
) ...(5) In this formula, if we set π・T _d・(P-1)・f=μ _d , the inverter output voltage e (fundamental wave) is as follows (7), (8)
It becomes a ceremony.

e=e ₁ +e _DB ......(6) =E _d /2 {μ・sinωt+μ _d・sin (ωt−)} ...(7) =E _d /2√ ² + ² _d +2・_d・sin (ωt
+ α) ...(8) However, α = tan ^-1 μ _d・sinμ−μ _d・cos From the above, the present invention
Considering in advance the voltage e _DB (formula (5) above) that decreases with T _d , the control voltage signal (e _a ^* , e ₁ 〓,
Obtain accurate output voltage by compensating for e ₁ 〓).

<Embodiment> FIG. 4 is a main circuit diagram showing an embodiment of the present invention. The phase voltage calculation circuit 7A using the polar coordinate method is
A polar coordinate conversion circuit 11 that converts the control voltage signals e ₁ 〓, e ₁ 〓 to polar coordinates, and a voltage generated by this conversion circuit 11.
Each phase control voltage signal e _a ^* , e _b ^* , e _c from E ₁ and phase angle δ
^* A polar coordinate one-three phase conversion circuit 12 is provided. Here, the conversion circuit 12 corrects the polar coordinate one-three phase conversion using the dead time voltage _EDB .
For this purpose, a dead time voltage calculator 13 is provided.

The conversion operation of the polar coordinate conversion circuit 11 is E ₁ =√ ² ₁ 〓 + ² ₁ 〓 ……(9) δ=t−1an ⁻¹ e ₁ 〓/e ₁ 〓 ……(10). In addition, the dead time voltage calculator 13 calculates the following equation (11) from the angular frequency ω, the number of triangular wave pulses P, and the set dead time T _d E _DB = π/2・E _d・T _d・(P−1)・F... Find the magnitude of the dead time voltage e _DB from (11).

Then, the polar coordinate one-to-three phase conversion circuit 12 performs calculations for each phase by adding the dead time voltage _eDB .
For example, the a-phase control voltage signal e _a ^* is determined by the following calculation.

e _a ^* =e ₁ +e _DB =E ₁ sin(ωt+δ) +E _DB sin(ωt+δ−) =√ ₁ ² + _DB ² +
2E ₁ E _DB cos ・sin(ωt+δ−α) ……(12) However, α＝tan ⁻¹ E _DB sin／E ₁ −E _DB cos ……(13) Figure 5 shows the experimental results based on the present invention. FIG. For torque control current setting value i ₁ 〓,
Torque characteristic with dead time compensation compared to torque characteristic T ₁ (shown by solid line) before dead time compensation
T ₂ (shown by the broken line) has excellent linearity in the area where the torque is large. With this torque characteristic T ₂ , the rated torque T ₀
It increases linearly with respect to the torque command i ₁ 〓 to about three times as much. The characteristic f indicates the output frequency, the characteristic Va indicates the a-phase output voltage before compensation, and Va' indicates the output voltage after compensation. Note that the dead time T _d is 35 μs.
And so.

In the embodiment, if the power factor is good, it may be determined by calculating e=E _d /2(μ+μ _d ) sinωt with cos=1 in the above equation (8) or (12).

<Effects of the Invention> According to the present invention, in vector control using a PWM type transistor inverter, voltage drop due to transistor dead time can be compensated for and control performance can be improved. In particular, the present invention allows accurate compensation even when the load power factor changes due to the induction motor acting as the load, and furthermore, compensation focusing on the fundamental wave component where the influence of the dead band is large allows compensation on polar coordinates. This has the effect of simplifying the compensation circuit.

[Brief explanation of the drawing]

Fig. 1 is a block diagram of a vector control system for non-interference control, Fig. 2 is a main circuit diagram of a transistor inverter, Fig. 3 is a waveform diagram for explaining errors due to dead time, and Fig. 4 is an embodiment of the present invention. FIG. 5 is a circuit diagram of a main part showing an example, and a characteristic diagram showing experimental results based on the present invention. 2...Inverter, 3...Compensation circuit, 7,7A
... Phase voltage calculation circuit, 8 ... Trigonometric function generation circuit,
9...Triangular wave generation circuit, 11...Polar coordinate conversion circuit, 12...Polar coordinate one-three phase conversion circuit, 13...Dead time voltage calculator.

Claims (1)

[Claims] 1. Convert the control voltage signals e ₁ 〓, e ₁ 〓 of the secondary magnetic flux and secondary current of the induction motor into polar coordinates of voltage E ₁ and phase angle δ, and from these polar coordinates, each phase control voltage signal is obtained. e _a ^* , e _b ^* ,
In a vector control device for an induction motor, the control voltage signal of each phase is converted into a control voltage signal of a pulse width modulation type transistor inverter by performing polar coordinate-three phase conversion to e _c ^* , and control by the dead time T _d set in the transistor inverter. Voltage drop e _DB magnitude E _DB is calculated by the following formula E _DB = π/2E _d・T _d・(P-1)・f E _d ; Inverter DC voltage P; Triangular wave pulse number f; Control voltage signal frequency A dead time voltage calculator is provided to obtain the dead time _voltage according to the above equation ^.
The following equation is the addition of the control voltage drop e _DB to e _b ^* , e _c ^* : ea=√ ₁ ² + _DB ² +2 _{1 DB}・sin (ωt+δ−α) α=tan ^-1 E _DB sinφ/E ₁ −E _DB cosφ φ: A vector control device for induction motors that is characterized by calculating from the power factor angle.