JPH0222638B2 - - Google Patents

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention provides a method in which, when applying vector control to an AC motor by applying a microprocessor function, primary current control is executed after sampling the rotor speed. The present invention relates to a speed calculation device that eliminates the response delay caused by the time required for calculations.

[Technical background of the invention]

Vector control is performed on induction motors or synchronous motors, and here we will discuss its application to synchronous motors.

First, FIG. 1 shows a basic control system when vector control is applied to a synchronous motor.

In FIG. 1, the rotation speed signal N _FB of the speed detector 2 directly connected to the rotor of the synchronous motor 1 and the speed reference value N _ref are added by the adder 3 with the polarity shown, and this speed error is amplified. The speed amplifier 4 obtains the amplified value |i _a | of the primary current command. A vector control circuit 5 connects the output |i _a | of the speed amplifier 4 and a magnetic flux position detector 6 directly connected to the rotor of the synchronous motor 1.
The primary current command value i _a is output based on the magnetic flux position signal φ output from the magnetic flux position signal φ. A current amplifier 7 detects the primary current of the synchronous motor 1 with a current detector 8 and supplies the synchronous motor 1 with a primary current corresponding to the output signal i _a of the vector control circuit 5 .

As described above, this device obtains the rotational speed of the synchronous motor 1 corresponding to the speed reference value _Nref .

On the other hand, devices that apply microprocessors to these controls are emerging. As this is a digital process, there is no problem of information drift due to temperature, etc., the excellent functions of a microprocessor can be easily added, new control theories can be easily added, and it is programmable, so it has a wide range of applications. Benefits include a wider space.

On the other hand, since digital processing or processor processing is performed, the magnetic flux or speed becomes time division sampling with a certain period, and the primary current command value is calculated from this information, so it has the following drawbacks.

Because the primary current is controlled discontinuously, during high-speed operation, current distortion causes torque pulsation, torque decrease, large loss, power factor drop, and vibration noise.

Since there is a lag between the time when the speed data is sampled and the time when the primary current is controlled based on this, there is a delay in response during speed transient response. (There is a similar delay in detecting the magnetic flux position, but the torque is cos(φ−Δφ), so φ≒
The effect is small near 90°. ) [Object of the Invention] The present invention aims to eliminate these drawbacks, and maintains the characteristic of smooth control of the conventional analog control method, while also solving the drawbacks of the analog control method. A microprocessor that does not have temperature drift or troublesome adjustment is used, and the excellent functions inherent in the microprocessor are fully utilized.

[Structure of the invention]

The present invention detects the speed and magnetic flux position as digital quantities using a detector directly connected to the rotor of the AC motor, and detects the primary order of the AC motor according to the deviation between these detected values and the speed command and the position command. In a speed calculation device for AC motor control that calculates a current command value and applies this primary current command value to a current amplifier to control the AC motor, the speed calculation device corresponds to the rotor speed of the AC motor detected by the detector. I did 2
A one-stage shift register that inputs a base digital quantity at every fixed sampling period τ and always holds and outputs the speed data of one sample before, and the sampled digital quantity and the digital quantity of one stage of the shift register. A subtraction circuit that calculates a digital quantity corresponding to the linear approximation of acceleration by subtraction, and a time delay by multiplying the digital quantity calculated by this subtraction circuit by a constant K determined from the sampling period τ, calculation delay time _τc , etc. An AC motor comprising: a multiplication circuit that calculates a correction amount; and an addition circuit that adds the sampled digital amount and the digital amount of the multiplication circuit and outputs corrected speed information _NFB . This relates to a speed calculation device for control.

[Specific description of the invention]

FIG. 2 is a block diagram showing the configuration of an embodiment of the present invention. Since a resolver is used to detect the magnetic flux position and velocity, the principle of detecting the magnetic flux position and velocity by the resolver will first be explained.

Figure 3 shows the configuration of the resolver, where the two-phase excitation windings 200 and 201 have cosω ₀ t and sinω ₀ t, respectively.
If the synchronous motor is excited with a two-phase current, the detection winding 202 directly connected to the rotor of the synchronous motor will have
A resolver signal of V〓 is obtained.

V = Esin (ω ₀ t + φ _e (t)) ......Equation 1 However, φ _e (t) = Pφ _n (t)/2 P: Number of resolver poles φ _e (t): Rotor electrical angle φ _n (t): Rotor mechanical angle According to the first equation, the magnetic pole position (magnetic flux position) can be detected from the phase difference between the resolver signal and the excitation signal.

Since t ₁ or t ₂ in Figure 4 is information that corresponds to the magnetic flux position, if t ₁ and t ₂ (time) are digitally counted using the reference clock, information expressed as a digital quantity of the magnetic flux position can be obtained. It will be done.

Next, assuming that the frequencies of the resolver signal and the excitation signal are f and f ₀ , respectively, this deviation Δf is expressed by the second equation.

Δf=f−f ₀ ... Second equation From this second equation, the rotation speed N _FB of the synchronous motor can be obtained as the third equation.

N _FB = 120・Δf/P _M = 120 (f - f ₀ )/P _M = 120 (T ^-1 - T ₀ ^-1 )/P _M ......3rd formula, where T = f ^-1 , T ₀ = f ₀ ^-1 , P _M : Number of motor poles Therefore, since the excitation frequency of the rotation speed N _FB is known in advance, if the period T of the resolver signal is detected, it can be calculated from the third equation. can. In other words, the rotation speed can be determined from digital information obtained by digitally counting T using the reference clock.

Next, the configuration and operation of FIG. 2 will be explained.

10 is a magnetic flux position detecting resolver directly connected to the rotor of the synchronous motor. 22 is a counter that uses the zero crossing of the excitation signal (sinω ₀ t) of the resolver 10 as a start and reset signal, and digitally counts clock pulses. 23 is resolver 1
This counter uses the zero cross of the 0 resolver signal 24 as a latch timing signal, latches the digital amount of the counter 22, and holds digital information representing the amount corresponding to the magnetic flux position in time (t ₁ ).
The zero cross signal of the resolver signal 24 is also a timing indicating that the latch circuit 23 has held new digital information.

11 is a speed detection resolver directly connected to the rotor of the synchronous motor. 12 starts the resolver signal (zero cross signal) 14 of the resolver 11,
A counter 13 for digitally counting clock pulses as a reset signal is a latch circuit that uses the resolver signal as a latch timing to latch the digital count amount of the counter 12 and holds digital information representing an amount corresponding to the speed as a time T.
The zero cross signal of the resolver signal 14 is also a timing indicating that the latch circuit 13 has held new digital information.

The T/φ conversion circuit 25 in the microprocessor processing block 102 converts the magnetic flux position expressed in time to the magnetic flux expressed in electrical angle based on the first equation. This timing is resolver signal 2
This is done at the timing of 4, and the magnetic flux position is sampled in a time-divided manner. T/N _FBo conversion circuit 15
converts the digital information expressed by the period T of the resolver signal of the latch circuit 13 into digital information of the speed based on the third equation. This is resolver signal 1
4, resulting in time-divided speed sampling.

16 is a one-state shift register,
The T/N _FBo conversion circuit 15 always outputs speed information (N _FBo-1 ) of one cycle before (n-1) sampled at time n.

17 is a subtraction circuit that subtracts (N _FBo -N _FBo-1 ), and 18 is a multiplication circuit that multiplies the result of the subtraction circuit 17 by a constant K. Reference numeral 19 denotes an addition circuit that adds the calculation result of the multiplication circuit 18 and the result converted by the T/N _FBo conversion circuit 15 with the polarity shown. The output _NFB of this adder circuit 19 is expressed by the fourth equation: _NFB = _NFBo +K( _NFBo -N _FBo-1 ).

The adder circuit 20 receives the speed command value N _ref and the speed N _FB
are added with the polarities shown, and the difference is amplified by the speed amplification circuit 21 with proportional and integral constants.
Obtain the amplitude value of the next current |i _a | (torque command).

The pulse generation circuit 27 divides the clock of the reference clock generation circuit 26 according to the rotational speed _NFB ,
A pulse train with a frequency proportional to the digital amount of _NFB is generated, and the pulse train is counted by a counter 28.

The digital addition circuit 29 is a T/φ conversion circuit 25
Digital information (base portion) of the magnetic flux position (base portion) φ and the amount of change in magnetic flux position (offset portion) during the sampling period are obtained from the counter 28, and the two are added to obtain interpolated magnetic flux position information. Thereby, the primary current command value is also interpolated, and smooth primary current control can be performed.

Next, a method of generating this primary current command will be explained.
The memory table 30 includes a digital adder circuit 29
The interpolated magnetic flux position information obtained is used as address information, and two-phase (cos, sin) digital quantities are output. The outputs are input to multiplier circuits 31 and 32, and multiplied by the amplitude value |i _a | of the primary current command obtained by the speed amplifier circuit 21, respectively, to obtain the two-phase primary current command value (digital quantity). is output. The output is converted into an analog quantity by the D/A conversion circuits 33 and 34, and the analog quantity of the two-phase primary current command value is output to the current amplifier 7 in FIG.

Next, the function of the speed calculation block 101 will be explained with reference to FIG. 5, taking as an example a transient state in which the speed changes in proportion to time.

The sampling period of speed information by T/N _FBo conversion circuit 15 is τ, resolver 11, counter 1
2. Speed detection delay due to latch circuit 13 is τ ₁ /2
(τ ₁ : resolver signal period), and if there are no other delays, there will already be a time delay between the actual rotor speed A and the detected speed B, as shown in Figure 5a. ing.

Next, T/N _FBo in microprocessor processing
If it takes time τ _c for conversion, calculation of i _a , T/φ conversion, etc., there will be an additional time delay between rotor speed A and detected speed B, as shown in Figure 5b. ing. That is, the velocity detected by the resolver
When calculating N _FBo as speed information and obtaining the primary current command value, a control delay has already occurred. In such a case, by adding the speed calculation block 101, if the constant K of the multiplier circuit 18 is set to, for example, K=(τ ₁ /2+τ _c )/τ, the speed information N _{FB is calculated as N FB} from the fourth equation. _FB = N _FBo + {(τ ₁ /2 + τ _c /τ} × (N _FBo −N _FBo-1 ...... It is modified to the formula 5. This is calculated by changing the sampling rate N _FBo-1 before one cycle and the new velocity sampled at
This means that a linear approximation of speed change, that is, a first-order approximation of acceleration, is performed from N _FBo to predict the speed at the time when the primary current is controlled. Also, by setting K=0 to 1 no matter what point (time) of the primary current control cycle the speed sampling is at,
Speed information during primary current control can be predicted, and delays in primary current control can be reduced (FIG. 5c).

Furthermore, if we add 0.5 to K=(τ ₁ /2 + τ _c )/τ for the control delay due to speed change between one primary current control and the next primary current control, we get the result as shown in Figure 5d. Corrected control can be performed.

FIG. 6 is a graph showing how the transient response is improved by the present invention, where a shows the speed change when K=0 (no compensation) and b shows the speed change when K=3/2. .

Although the speed calculation circuit 101 has been described as processing outside the microprocessor processing block 102, it may be processed within the same microprocessor processing block or may be processed by a separate microprocessor.

Although the above description has been made regarding a synchronous motor, the same effect can be obtained by performing similar control on an induction motor. As for the detector, it goes without saying that something other than a resolver can be used.

〔Effect of the invention〕

As described above, the speed calculation device of the present invention has the effect that the speed response delay due to the difference between the time when the speed data is sampled and the time when the primary current is controlled is significantly improved, and the transient response can be improved. It is something that plays.

[Brief explanation of drawings]

Fig. 1 is a block diagram showing a basic control system when vector control is applied to a synchronous motor, Fig. 2 is a block diagram showing the configuration of an embodiment of the present invention, and Fig. 3 is a block diagram showing a basic control system when vector control is applied to a synchronous motor.
The figure is an explanatory diagram explaining the principle of position detection by a resolver, Figure 4 is a waveform diagram showing the relationship between the excitation signal and the resolver signal, Figure 5 is an explanatory diagram showing the relationship between the actual rotation speed and the detected speed, FIG. 6 is a waveform diagram comparing the speed response when the compensation according to the present invention is performed and the speed response when no compensation is performed. 1: Synchronous motor, 2: Speed detector, 3: Adder, 4: Speed amplifier, 5: Vector control circuit,
6: Magnetic flux position detector, 7: Current amplifier, 8: Current detector, 10, 11: Resolver, 12, 22: Counter, 13, 23: Latch circuit, 14, 24:
Resolver signal, 15: T/N _FBo conversion circuit, 1
6: Shift register, 17: Subtraction circuit, 18: Multiplication circuit, 19, 20: Addition circuit, 21: Speed amplification circuit, 25: T/φ conversion circuit, 26: Reference clock generation circuit, 27: Pulse generation circuit, 28 : Counter circuit, 29: Digital addition circuit, 30: Memory table, 31, 32: Multiplication circuit, 33, 3
4: D/A conversion circuit.

Claims (1)

[Claims] 1. Speed and magnetic flux position are detected as digital quantities by a detector directly connected to the rotor of an AC motor,
In AC motor control, a primary current command value of the AC motor is calculated according to the deviation between these detected values and a speed command and a position command, and this primary current command value is given to a current amplifier to control the AC motor. In the speed calculation device, a binary digital quantity corresponding to the rotor speed of the AC motor detected by the detector is inputted at every fixed sampling period τ, and the speed data of the previous sampling is always held and outputted. a stage shift register 16; a subtraction circuit 17 for subtracting the sampled digital quantity and the digital quantity of one stage of the shift register 16 to obtain a digital quantity corresponding to a linear approximation of acceleration; a multiplication circuit 18 that calculates a time delay correction amount by multiplying the calculated digital amount by a constant K determined from the sampling period τ, calculation delay time _τc , etc.; 1. A speed calculation device for controlling an AC motor, comprising an addition circuit 19 that adds the digital quantities and outputs corrected speed information _NFB . 2. The speed calculation device for AC motor control according to claim 1, wherein the constant K is K=(τ ₁ /2+τ _c )/τ, where τ ₁ is set according to the speed detector signal period. . 3. The constant K is K=(τ ₁ /2+τ _c )/τ+0.5, where τ ₁ is set by the speed detector signal period. Computing device.