JP2669540B2 - Rotary compressor controller - Google Patents

Description

TECHNICAL FIELD The present invention relates to a torque control device for a rotary compressor.

[Conventional technology]

As an example of a rotary machine driven by an electric element,
The hermetic rotary compressor will be described. FIG. 22 and FIG. 23 are a longitudinal sectional view and an AA transverse section showing a conventional structure of the hermetic rotary compressor. In these figures,
Reference numeral 1 denotes a container, in which an electric element and a compression element are housed. Reference numeral 2 denotes an electric element stator (stator).
And is fixed to the inner periphery of the container 1. This stator 2
A rotor (rotor) 3 of an electric element, which is fitted to the main shaft 4 and rotates integrally with the main shaft 4, is arranged inside the. The main shaft 4 is supported by a main bearing 5 and end bearings 6. These bearings 5 and 6 are fastened to a cylinder block 7. The cylinder block 7 is fixed to the container 1. A compression chamber 12 is formed in the cylinder block 7. A roller 8 eccentrically integrated with the main shaft 4 is provided in the compression chamber 12. Disposed vanes 9 are arranged, which form the compression element. When the main shaft 4 is rotationally driven by the electric element under such a configuration, the refrigerant gas is sucked from a suction accumulator 11 provided outside the container 1 and reaches a predetermined pressure by the rollers 8 in the compression chamber 12. It is pressurized and discharged outside the case along the direction of the arrow.

A rotary compressor having such a structure is disclosed in JP-A-58-187635.
No. 6,086,045.

By the way, in the compressor having such a structure, the electric element outputs an almost constant torque with respect to time, whereas the gas absorption torque in the compression element is extremely large during one rotation of the main shaft 4. There are fluctuations. Then, the difference between the gas absorption torque and the electromagnetic torque, that is, the residual torque acts as an exciting torque on the container 1, and a problem arises that a large rotational vibration is induced in the rotational direction of the entire compressor. .

FIG. 24 is a schematic diagram showing the cause of vibration in the rotary compressor as an example. For each rotary system (rotor 3, roller 8, main shaft 4) and fixed system (stator 2, block 7, container 1) of compression element and electric element,
The gas compression torque _TG and the electromagnetic torque T _M act as shown in the figure. In the figure, clockwise is shown as positive and counterclockwise is shown as negative.

The equations of motion of the rotating system and the fixed system at this time are and Indicated by However, Are the inertia torques of the rotary system and the fixed system, respectively,
J _R and J _S are the inertial moments of the rotating system and the fixed system, respectively, and T _G −T _M corresponds to the residual torque (= excitation torque), and K
Is the spring constant of the compressor support panel, φ _R is the rotation angle of the rotation system, and the rotation angular velocity of the rotation system There is a relationship.

Therefore, if the rotational speed ω _R of the rotary system is detected and feed back control is performed so that this is always equal to a constant rotational speed command value, Then, from equation (1), T _M −T _G = 0, and the electromagnetic torque T _M
And the load torque T _G are balanced with each other, the vibration torque in the equation (2) is eliminated, and vibration of the rotating machine and rotation direction vibration of the container, which is a cause of noise, can be suppressed.

Further, as shown in Japanese Patent Laid-Open No. 60-223181, the load torque T _G of the compressor fluctuates greatly during one rotation, but it is a cyclic pulsating load that is repeated with one rotation as a cycle, and attention should be paid to a certain rotation angle. There is also a conventional example in which torque control is performed by a method in which the rotation speed detected at each rotation angle is fed back at the same angle after one rotation, paying attention to the fact that the load torque value does not change much.

[Problems to be solved by the invention]

However, the rotation curve and the load torque _TG generally have different phases. FIG. 25 shows a load torque curve in the rotary compressor and a rotational speed fluctuation curve of the rotor together. As you can see from the figure,
The fluctuation of the rotation speed has a delay phase with respect to the torque fluctuation,
The point at which the rotation speed becomes the minimum is tens of degrees behind the peak value of the load torque.

Therefore, the torque control is performed by detecting the rotation speed of the rotor and increasing the motor output when the detected speed is lower than the command speed, and decreasing the motor output when the detected speed is higher than the command speed. However, there is a problem that the motor output torque cannot be perfectly synchronized with the load torque due to this phase delay, and some phase correction means is required.

However, this phase correction is difficult to realize for the following reason. That is, as shown in FIG. 26, the torque component of the rotary compressor includes high-frequency components such as secondary, tertiary ... In addition to the rotational primary component corresponding to the operating frequency of the compressor, They have a certain phase difference from each other. Referring to the equation (1), the rotation speed ω _R is the excitation torque Δ
It has an integral relationship with Tr. Here, focusing on the excitation torque component ΔTn of the nth rotation, by integrating ΔTn = ΔTn o sin nωt From this, it can be seen that the nth-order component of the rotation speed has a phase delay of π / 2n with respect to the nth-order component of the torque. The actual rotor rotation speed is a combination of all the nth-order components, and the ratio of these components is different depending on the operating conditions, so there is a problem that the optimum phase correction angle cannot be easily determined. .

The present invention has been made in view of the above problems, and an object thereof is to synchronize the motor output torque with the load torque without phase delay at each rotation angle of the rotor to eliminate the excitation torque, and to reliably and quickly reduce the torque. An object of the present invention is to provide a torque control device for a rotary electric machine that can achieve vibration.

[Means to solve the problem]

The above-mentioned objects are: an electric motor, a rotary compressor rotatably driven by the electric motor, a rotational acceleration detecting means for outputting the rotational speed of the rotary compressor, a means for issuing a rotational acceleration command, the rotational acceleration command and the output. In a control device for a rotary compressor, comprising: a current command generating unit that generates a current command to be given to the electric motor based on a deviation from the rotational acceleration, and a unit that drives the electric motor based on the current command. The rotation speed detecting means divides an angle of one rotation of the rotary compressor into a plurality of pieces, and outputs the rotational acceleration of the divided angle sheets, and the current command generating means has a plurality of angles. This is achieved by having a plurality of integrating means for integrating each time based on the deviation of the angle.

[Action]

Rotational acceleration obtained by differentiating the rotational speed Changes in the same phase as the torque fluctuation with a value that reflects the magnitude of the torque fluctuation. Since the present invention is used for torque control, the phase correction as described in the section of the problem is not required.

That is, according to the present invention, the control amount required to make the exciting torque zero is calculated from the rotational acceleration at each rotational angle position during one rotation of the rotating spindle, and the load torque is calculated at each rotational angle without phase delay. Since the output torque of the electric element can be made to follow the fluctuation, the vibration torque can be reduced reliably and promptly.

〔Example〕

Hereinafter, a case will be described in which a rotary compressor, which is a type of rotating machine driven by an electric element (DC motor), is driven by an inverter. Figure 2 shows Figures 22 and 23
Since the structure is almost the same as that of the conventional compressor shown in the figure, only different parts will be described and the same parts will be denoted by the same reference numerals and the description thereof will be omitted.

In FIG. 2, the rotor 2 and the compression element 21 of the electric element 20 are shown.
One end of the rotary main shaft 4 that is directly connected to the roller 8 is extended to fix a gear 22 that is a rotational position detecting member.
Is adapted to rotate integrally with the main shaft 4. Container 1
A position detecting sensor (for example, an electrode pick-up or a gear-up sensor) 23 is fixed to the position detecting sensor 23 and detects a gear train of the gear 22 to output a sinusoidal signal corresponding to the rotational angle position of the main shaft 4.

FIGS. 3 to 7 show a series of processes for obtaining the rotational speed and rotational acceleration of the rotary spindle 4 from the sinusoidal signal detected by the position detection sensor 23. Now the third
As shown in FIG. 4, when the gear 22 rotates, the position detecting sensor 23 outputs a sinusoidal detection waveform 24 as shown in FIG. The rotation pulse signal 25 shown in FIG. 5 is obtained by shaping the waveform of the threshold voltage 27 based on the threshold voltage 27 and converting it into a pulse train. Each pulse value of this pulse time series 25 tT ₁ , T ₂ ,
... T _n is measured, and the reciprocal number N _i = 1 / T _i determines the rotation speed N _i at each time. Rotational speed at each rotation angle during one rotation
FIG. 6 shows 28. On the other hand, the rotational acceleration _i at each time, from the rotational speed at i pulses of the rotational pulse signal 25 N _i and (i-1) the rotational speed N _i-1 at the time of pulse, _i =
It is obtained as (N _i −N _i-1 ) / T _i . Alternatively, (i +
1) After obtaining the rotation speed W _{i + 1} at the pulse, _i = (N _{i + 1} −
N _i-1 ) / (T _i + T _{i + 1} ) or
It can be similarly obtained by three-point interpolation of N _i−1 , N _i , and N _{i + 1} . FIG. 7 shows the rotation acceleration 101 at each rotation angle during one rotation.

FIG. 8 shows another method of calculating the rotation speed and the rotation acceleration from the detection pulse. The pulse width of the detection pulse 25 at intervals of one cycle is measured, and the pulse widths T ₁ , T ₂ , ... T _i are measured while shifting the measurement position by half cycles. Its reciprocal
The rotation speed of each point is obtained from N _i = 1 / T _i, and _i = (N _{i + 1} −N _i-1 ) / is calculated from the speeds N _i−1 , N _{i + 1} and the pulse width T _i on both sides.
It can also be obtained as T _i .

FIG. 9 shows the overall configuration of the torque control device. That is, the detection signal 24 obtained by the position detection sensor 23 is converted into the rotation pulse signal 25 by the waveform shaping circuit 29 and sent to the CPU 31 of the microcomputer 38 via the interface 30. The rotation pulse signal 25 activates the interruption of the microcomputer 38 to execute a series of operations of detection-calculation-command. That is, the pulse interval of the rotation pulse signal 25 is clocked by the timer built into the microcomputer to obtain the time interval T _i , from which the rotation speed N _i and the rotation acceleration are calculated.
_i is calculated. Then, this rotational acceleration _i is 0, and it is necessary for the rotational speed N _i to be equal to the commanded speed N _c ,
A current command value given to the electric element 20 of the compressor is calculated and output. The control signal is sent to the control unit 34 to form a chopper signal for driving the base driver 35. By the current control operation of the base driver 35, the inverter 36 controls the winding current supplied to the electric element 20 to perform torque control, so that the compressor rotation speed is always equal to the command speed with a fluctuation of a certain allowable value or less. In addition, control is continuously performed so that the rotor rotational acceleration is minimized at each rotational angle. These series of control loops are written in the ROM 33.

A reference control flow of this torque control device will be described in detail with reference to the block diagram shown in FIG. Note that S shown in the drawing is a Laplace operator, and 1 / S is an integral element. The load torque T _G of the compressor acts as a disturbance against the electromagnetic torque T _M (105) of the electric element, and its residual torque
T _M −T _G = ΔT _r is the excitation torque, and the rotation speed ω _R is Fluctuates accordingly.

Rotational angular position phi _R of the rotor based on the rotation speed omega _R
The time t _F ( ^k ) at the time of the k pulse of the rotation pulse signal 25 obtained by sampling
Measure with 38 built-in timers. The microcomputer 38 uses k pulse time t _F ( ^k ) and (k-1) pulse time t _F ( ^k-1 )
Request interval ^{_{ΔT F (k) = t F}} (k) -t F (k-1). Then, by the method described in FIG. 2, the rotation speed N _F ( ^k ) (28) and the rotation acceleration _F ( ^k ) (101) at the time of k pulses are ^obtained.
Ask for. In FIG. 1, the rotational acceleration at the time of the k pulse is shown.
_{As a} calculation for obtaining _F ( ^k ), the rotation speed _F at the time of k pulses
^(K) and (k-1) pulses when the rotational speed N _F ^(k-1) using the forward difference method ^{_{F (k) = (N F}} (k) -
N _F ( ^k-1 )) / ΔT _F ( ^k ) is used, but instead of this, as described above, the rotation speed N _F ((k + 1) pulse) N _F (
^{k + 1)} and (k-1) pulses when the rotational speed N _F ^(k-1) central difference using method ^{_{F (k) = (N F}} (k + 1) -N F (k-1))
^{_{/ (ΔT F (k) +}} ΔT F (k-1)) may be utilized,
Alternatively, a backward difference method or three-point interpolation may be used.

Alternatively, the detection method according to FIG.
+1) Pulse rotational speed N _F (k ^{+ 1)} and (k-1) the rotational speed at the time of pulse N _F ^(k-1) and K Pulse interval Δ
Using T _F ( ^k ), _F ( ^k ) = (N _F ( ^{k + 1} ) −
It may be calculated as N _F ( ^k−1 )) / ΔT _F ( ^k ). However, it should be noted that the rotational acceleration _F ( ^k ) at the time of the K pulse confirms that the rotational speed N _F ( ^{k + 1} ) at the time of the (k + 1) pulse.
Is determined, and it is possible only with a control system having a dead time for one rotation.

N _F ( ^k ) (28) obtained in this way, and
_F ( ^k ) (101) is feedback-backed, and command speed _Nc and command acceleration _c ( _Nc = _c when constant ₎
It is. ) To obtain the deviations ΔN ( ^k ) (102) and Δ ( ^k ) (103). The torque control system operates so that each deviation becomes equal to zero by a PI (proportional / integral) control operation for each deviation amount. That is, the current command value i _c ( ^k ) (104) at the time of the k pulse of the microcomputer that determines the electromagnetic output torque of the electric element is calculated according to the following equation.

Here, the first term is the operation amount for the speed deviation ΔN ( ^k ), and the second term is the operation amount for the acceleration deviation Δ ( ^k ). K _Pn , K _Pa , K _In , and K _Ia are the proportional gain and integral gain of velocity deviation and acceleration deviation, respectively, and K _nc and K _na are respectively
This is a conversion constant for converting a proportional + integral value of the speed deviation and the acceleration deviation into a current command value. Current command value obtained in this way
i _c ( ^k ) or electromagnetic torque T _M ( ^k ) of the electric element at the time of k pulses
(105) is determined as T _M ( ^k ) = k _DA · K _T · i _c ( ^k ). Here, K _DA is the D / A conversion constant of the digital amount and analog output in the microcomputer, and K _T is the current-torque conversion constant (torque constant).

Here, considering the physical meaning of the above equation, it is as follows. No.
13 figure shows the curve of the electromagnetic torque T _M in the case of 1 no curve and the torque control of the load torque T _G during rotation. The shaded area surrounded by this curve T _M and the horizontal axis is 1
It is the energy required to drive the compressor during rotation.
Now, in order to make the rotation speed of the compressor equal to the command speed _Nc on average, the area surrounded by this electromagnetic torque curve may be made equal to the area surrounded by the load torque curve. The first term in the above equation corresponds to this magnitude, so to speak, to the DC component of the current supplied to the electric element. On the other hand, in order to prevent the rotation pulsation at each rotation angle, it is necessary to make the electrode torque curve equal to the load torque curve, and the second term in the above equation is the magnitude for this (thus positive and negative at each rotation angle). , Which corresponds to the AC component of the current (see Fig. 14).

FIG. 10 shows a second reference example of the torque control device. Regarding the control of the rotation speed, it is only necessary that the average speed of one rotation becomes equal to the command speed. Therefore, in FIG. 6, the rotation speed is detected and controlled once per rotation. That is, regarding the rotation speed, the rotation pulse signal 25 generated during one rotation of the rotor is divided by the counter 39 to be input as one rotation and one pulse, and when this one rotation and one pulse is input n pulses (that is, the rotor is rotated). N times) time t _n
And a detection speed N _Fn corresponding to the average speed of one rotation is calculated from the difference ΔT _n between (n-1) pulse time (that is, (n-1) rotation time) time t _{n-1. Fn} (100) is fed back and compared with the command speed N _c, and one rotation speed deviation ΔN _n
(108) is obtained, and speed control is performed by this. The acceleration detection-control system is the same as in the reference example shown in FIG.

FIG. 11 shows the first embodiment of the present invention. This is because when the number of pulses of the rotation pulse signal 25 during one rotation is n pulses, n integrators (109) for integral control with respect to acceleration deviation are prepared and switched for each pulse input of the rotation pulse signal 25. use. As a result, each integrator (109) corresponds to each rotation angle, and operates to integrate the rotational acceleration deviation input at the same angle in the past independently for each angle. This is an effective method when the fluctuation of the load torque has a periodic property.

FIG. 12 shows a second embodiment, which replaces the feed back of the detected rotation speed N _F ( ^k ) (28) in FIG. 11 with the feed back of the one rotation average detection rotation speed (100) in FIG. Is used. Other parts are first
This is similar to the embodiment shown in FIG.

In each of the above-described reference examples and examples, the rotational acceleration at each rotation angle during one rotation is feedback-backed at each rotation angle during the same one rotation, in other words, at each rotation angle during the current one rotation. The rotational acceleration is fed back at each of the same rotational angles in the current one rotation, so to speak, the feedback is performed in real time. On the other hand, in the case of a load torque that repeats periodic fluctuations, the torque control device described below feeds back the rotational acceleration at each rotation angle during one rotation at the same rotation angle after one rotation.

The overall configuration of this torque control device will be described with reference to FIG. The detection signal 24 obtained by the position detection sensor 23 is converted into a rotation pulse signal 25 by a waveform shaping circuit 29 and is converted into an interface signal.
It is sent to the CPU 31 of the microcomputer 38 via 30. This rotation pulse signal 25 causes the microcomputer 38 to start an interrupt and execute a series of operations of detection-calculation-command. That is, the rotation pulse signal
The 25 pulse intervals are clocked by the timer built in the microcomputer to obtain the time interval T _{i in} FIG. 2, and the rotational speed N _i and rotational acceleration _i are calculated from this. Then, the current command value given to the electric element 20 of the compressor necessary for the rotational acceleration _{i to} be 0 and the rotational speed N _i to be equal to the commanded speed N _c is calculated, and the data is stored in the RAM 32 until the time when the data is needed. Remember. Next, the data stored in RAM32 one rotation before is C
It is read by the PU 31 and sent to the control unit 34 as a control signal to form a chopper signal for driving the base driver 35.
By the current control operation of the base driver 35, the inverter 36 controls the amount of winding current supplied to the electric element 20 to perform torque control, and the compressor rotation speed is always equal to the command speed within a fluctuation below a certain allowable value. And continuously control so that the rotor rotation acceleration is minimized. These series of control loops are written in the ROM 33.

The control flow of this control circuit system will be described in detail with reference to the block diagram shown in FIG. Note that S shown in the figure is a positive operator, 1 / S is an integral element, and e- ^ST is a time element that is only the time T. The load torque T _G of the compressor acts as a disturbance against the electromagnetic torque T _M of the electric element, and its residual torque T _M −T _G = ΔT _r becomes the excitation torque.
The rotation speed ω _R is Fluctuates accordingly. The rotation speed position φ _R of the rotor based on the rotation speed ω _R is sampled by the gear 22 or the like and the time t _F ( ^k ) at the time of the k pulse of the rotation pulse signal 25 is measured. Microcomputer 38 has time t _F at k pulse
Interval ΔT _F between ( ^k ) and (k-1) pulse time t _F ( ^k-1 )
^(K) = Request _{^{t F (k) -t F (}} k-1). And the third
The rotational speed N _F ( ^k ) at the time of k pulse is obtained by the method described with reference to FIG. 7 and the rotational acceleration _F ( ^k ) at the time of K pulse is obtained using this. In FIGS. 13 and 14, in order to obtain the acceleration _F ( ^k ) at the time of K pulse, (K +
1) Central difference expression _F ( ^k ) = using the pulse rotation speed N _F ( ^{k + 1} ) and the (K-1) pulse rotation speed N _F ( ^k-1 )
^{_{(N F (k + 1)}} -N F (k-1)) / (ΔT F (k) + Δ
Although T _F ( ^k−1 )) is used, as described above, the forward difference expression _F ( ^k ) = (N _F ( ^k ) −N _F ( ^k−1 )) / ΔT _F ( ^k )
Alternatively, a backward difference equation, three-point interpolation, or the like may be used.
Alternatively, the k pulse rotational acceleration _F ( ^k ) may be used by using the method described in FIG. On the other hand, in the same manner as described in Example of FIG. 10, the number of pulses n in one rotation of the rotor and divided by the counter 39 performs a sampling of one rotation pulse, and the pulse counting time t _n at the n-rotating (N-1)
ΔT _n = t _n −t _n-1 is calculated from the time t _n-1 during rotation, and the rotational speed N _Fn = 60 / ΔT _Fn (100) corresponding to the average speed of one rotation is calculated.

The rotational speed N _Fn (100) and the rotational acceleration _F ( ^k ) (101) thus obtained are feedback-backed, and the command speed N _c and the command acceleration _c (N _c = _c = 0 during constant steady operation, respectively). ), And the speed deviation ΔN _n
(108), the acceleration deviation Δ _F ( ^k ) (103) is obtained. However, in this embodiment, the acceleration _F ( ^k ) at the k-th pulse in the n-th rotation is 1 by utilizing the periodicity of the fluctuation.
In order to feed back at the same angle after rotation (the k-th pulse in the (n + 1) th rotation), n shown in FIG.
It is fed back through the number of dead time element e ^-ST (106). Further, n number of integrators as shown in FIG. 15 relative to the acceleration deviation Δ _{F ^(k)} Are prepared and operated by switching them for each pulse input similarly to the embodiment of FIG. 11, each integrator can be operated independently corresponding to each rotation angle of one rotation. As a result, the current command value i _c ( ^k ) (104) at the time of the k-th pulse in the n-th rotation is calculated according to the following equation.

Here, the first term is a manipulated variable with respect to the speed deviation ΔN _n and is controlled by PI (proportional / integral) control,
The second term is the acceleration deviation at the k-th pulse during the (n-1) th rotation. Is controlled by I (integral) control. In FIG. 15, a term of proportional control may be further added to the acceleration control system. In the figure, K _Pn , K _In , K _nc
Are respectively a proportional gain of the speed feedback system, an integral gain and a proportional + integral value of the speed deviation and a conversion constant of the current command value. Similarly, K _Ia1 ... K _Ian is an integral gain of the first pulse to the nth pulse, and K _ac is a conversion constant of the integrated value of the acceleration deviation and the current command value.

Next, FIG. 16 shows an example in which a control system is configured by using the acceleration detecting method shown in FIG. K as described above
The rotational acceleration _F ( ^k ) at the time of pulse is determined by (k +
1) after the determination of the rotational speed at pulse ^{_{N F (k + 1) =}} 1 / t 1 = (k + 1), the computed actual microcomputer so further requires computation time, FIG. 16 In this case, the feedback of _F ( ^k ) occurs at the time of sampling the (K + 2) pulse. Then, in FIG. 16, at the time when _F ( ^k ) is fed back, the (k + 2) th integrator output of the n-division integrator is made, and this is added to the average speed loop output, and the k pulse current command value i _c ( ^K ) is obtained.

Here, the rotational acceleration _F ( ^k ) and the current command value i _c ( ^k )
FIG. 17 shows the positional relationship of FIG. Rotational speed
While the N _F ( ^k ) and the rotational acceleration _F ( ^k ) are the measured values at the time of detecting the k pulse from the relation of the difference method, the current command value
i _c ( ^k ) is the section output value from the (k-1) pulse detection to the k pulse detection. Therefore, for example, the detected acceleration reflected in the current command value i _c ( ^k ) is _F ( ^k-1 )
Two types of _F ( ^k ) can be considered.

As a first method, the acceleration reflected in the current command value _ic ( ^k ) is calculated as an average value at (k-1) hour and k hour. It is a method of determining.

As the second method, the acceleration _F ( ^k ) at the time of k pulse is used.
(When the acceleration deviation Δ ( ^k ) = − _F ( ^k ) is focused on, a condition is attached to this and the determination is made so that it is reflected in either i _c ( ^k ) or i _c ( ^{k + 1} ).

For example, _F ( ^k ) 0 and _F ( ^k ) <0 are divided into two types, and _F ( ^k )>0; i _c ( ^k ) / i _c ( ^{k + 1} ) is the larger one (Δ
( ^K ) <0) Add Δ ( ^k ). ^{_{F (k) <0; i}} c (k) / i c smaller towards the ^{(k + 1)} (Δ
( ^K )> 0) Add Δ ( ^k ).

It is possible to use This method is advantageous in reducing a detection error particularly when the detection pulse includes a detection error.

FIG. 18 and FIG. 19 show these two methods.

The current command value i _c at the time of n rotation k pulses thus obtained
From ( ^k ) (104), the electromagnetic torque T _M ( ^k ) (10
5) is determined by T _M ( ^k ) = K _{D / A} · K _T · i _c ( ^k ).
Here, K _{D / A} is a digital amount in the microcomputer and a D / A conversion constant of analog output, and K _T is a current-torque conversion constant (torque constant).

Here, considering the physical meaning of the above equation, it is as follows. 1
In Figure 13, which shows a curve of the electromagnetic torque T _M when no curve torque control of the load torque T _G during rotation, the curve T _M
The area enclosed by the horizontal axis and the horizontal axis is the energy required to drive the compressor during one rotation. Now, in order to make the rotation speed of the compressor equal to the command speed _Nc on average, the area surrounded by this electromagnetic torque curve may be made equal to the area surrounded by the load torque curve. The first term in the above equation corresponds to this magnitude, so to speak, to the DC component of the current supplied to the electric element. On the other hand, in order to prevent the rotation pulsation from occurring at each rotation angle, it is necessary to make the electromagnetic torque curves equal in the load torque curve, and the second term in the above equation is the magnitude for this (thus, at each rotation angle (Changes positively or negatively), which is equivalent to the AC component of the current (see FIG. 14). The integral term in the second term indicates the current output for the torque pulsation corresponding to each pulse detection position, and the second term performs the repeated learning control for the periodic load torque fluctuation, and the current command value gradually increases. It gradually approaches the pulsating load pattern.

FIG. 20 shows another embodiment of the present invention,
In FIG. 15, instead of feeding back one rotation average rotation speed N _Fn (100), the rotation speed N _F ( ^k ) (28) at each rotation angle position of one rotation is obtained and this is fed back. The acceleration detection-control system is the same as that in FIG.

FIG. 21 shows a method of determining the acceleration command _c when the compressor is in a speed increasing / decelerating state. The speed command N _c ( ^k ) at the time of k pulse is the speed command at the time of (k-1) pulse
If it is different from _Nc ( ^k-1 ), it means the speed-up or deceleration mode. In this case, the difference ΔN _c ( ^k ) = N _c ( ^k ) −N _c (
^k-1 ), on the other hand, the period T _c ( ^k ) = 60 / N _c ( ^k ) is calculated from the speed command N _c ( ^k ) at the time of k pulse, and _c ( ^k )
= ΔN _c ( ^k ) / T _c ( ^k ), an acceleration command _c ( ^k ) is obtained.

As described above, the embodiment in which the torque control is performed when the hermetic rotary compressor is driven by driving the electric element by the inverter has been described, but the present invention is not limited to this.
It can be effective torque control means for all rotary electric machines driven by electric elements.

〔The invention's effect〕

As described above, according to the present invention, for a rotary electric machine having a load element driven by an electric element, the rotational acceleration changes the torque fluctuation in phase with a value that reflects the magnitude of the torque fluctuation. Mismatches of electromagnetic torque and load torque fluctuations of electric elements can be eliminated, and a vibrating torque can be greatly reduced, and a vibration in a rotating direction of a machine can be greatly reduced. Furthermore, since the output torque of the electric element can be made to follow the variation of the load torque at each rotation speed without phase delay, the response of the control system can be made faster, and the torque control execution area can be expanded to a higher speed area The rotation speed region can be reduced in vibration. Therefore, it is possible to contribute to downsizing of the entire system, power saving by wide-range variable speed operation, and improvement of comfort.

[Brief description of the drawings]

FIG. 1 is a control block diagram of the torque control device, FIG. 2 is a longitudinal sectional view of a compressor to which the torque control device is applied, FIG.
4, FIG. 5, FIG. 6, FIG. 7, FIG. 8 and FIG. 8 are diagrams showing a method for obtaining the rotation speed and the rotation acceleration from the rotation detection signal, and FIG. 9 is a schematic configuration diagram of the torque control device. FIG. 11 is a control block diagram of the torque control device, FIGS. 11 and 12 are control block diagrams of the first and second embodiments of the present invention, respectively.
13 and 14 are diagrams for explaining the operation of the present invention.
FIG. 15 is a control block diagram of another embodiment of the present invention, FIG.
FIG. 20 is a control block diagram showing still another embodiment of the present invention, FIG. 17, FIG. 18, and FIG. 19 are diagrams showing a method for determining the current command value, and FIG. 22 and 23 are a longitudinal sectional view and an AA transverse sectional view of a conventional compressor, FIG. 24 is a diagram for explaining a torque generated in the compressor, and FIG. FIG. 26 is a diagram showing fluctuations in torque and rotation speed during one rotation, and FIG. 26 is a diagram showing rotation order components of a torque waveform. 4 ... rotating main shaft, 20 ... electric element, 21 ... compression element, 22
……gear.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Tsunehiro Endo 4026 Kujimachi, Hitachi City, Ibaraki Prefecture Hitachi Research Laboratory, Hitachi, Ltd. (72) Kenichi Iizuka, 800, Tomita, Ohira-machi, Shimotsuga-gun, Tochigi Hitachi, Ltd. In Tochigi Factory (72) Inventor Hiroaki Hatake 800 Tomita, Ohira-cho, Shimotsuga-gun, Tochigi Hitachi Co., Ltd. Tochigi Factory (56) Reference JP 62-71483 (JP, A) JP 61-164494 (JP, A) JP-A-56-86083 (JP, A)

Claims (1)

(57) [Claims] 1. An electric motor, a rotary compressor rotationally driven by the electric motor, a rotational acceleration detecting means for outputting rotational acceleration of the rotary compressor, a means for generating a rotational acceleration command, and the rotational acceleration command. A rotary compressor control device comprising: a current command generation unit that generates an electric command to be given to the electric motor based on a deviation from the output rotational acceleration; and a unit that drives the electric motor based on the electric current command. , The rotational acceleration detecting means,
The rotary compressor divides one rotation angle into a plurality of parts and outputs a rotational acceleration for each of the divided angles.
The control apparatus for a rotary compressor, wherein the current command generating means has a plurality of integrating means for integrating for each of the plurality of divided angles based on the deviation of the angle.