JP2015010992A - Rotation position error correction device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress the load.SOLUTION: A micro computer calculates, as a correction value Δθ, the difference between the rotation position on a straight line L2 and the rotation position θby a resolver signal, the correction value Δθis added to (or subtracted from) the rotation position θby the resolver signal, thereby correcting the error of the rotation position θ. The straight line L2 is a straight line interconnecting the point A as the rotation position θat the start time of one period T in an NM signal S4, the point B as the rotation position θat the final time of the one period T, and the point E as the average value of the rotation position θat the time of a half period T/2 determined by averaging processing.

Description

  The present invention relates to a rotational position error correction device that corrects an error in the rotational position of a rotor in a rotating electrical machine.

  In a rotating electrical machine, the rotational position of a rotor is detected by a sensor (for example, a resolver). An error is included in the rotational position of the rotor detected by the sensor. For this reason, in the control device for a rotating machine described in Patent Document 1, in order to correct this error, a correction value is calculated, and the detection value of the rotation angle is corrected based on this correction value.

  First, the control device for a rotating machine described in Patent Document 1 starts and ends an error calculation period, which is a period from the timing at which the detected value of the rotation angle becomes 0 ° to the timing at which one cycle of a sinusoidal error elapses. A reference line, which is a straight line drawn on the time axis, is calculated by connecting the detected values (0 ° and 360 °) of the rotation angle at the timing. Then, the control device calculates a difference Δ between the reference line and the rotation angle detection value θ in the error calculation period, and based on the calculated difference Δ, an average value ΔAVE of the differences Δ in the error calculation period. Is calculated. Next, the control device calculates a correction value Δθ of the rotation angle detection value θ with reference to the rotation angle detection value at the timing when the difference Δ becomes the average value ΔAVE. The correction value Δθ is a value obtained by subtracting the average value ΔAVE from the difference Δ every time. Then, the control device reads the correction value Δθ corresponding to the rotation angle detection value θ from the memory, and adds the read correction value Δθ to the rotation angle detection value θ, thereby obtaining the rotation angle detection value θ each time. to correct.

JP 2012-105515 A

Incidentally, in the rotational position error correction device, it is desired to suppress the load.
An object of the present invention is to provide a rotational position error correction device capable of suppressing a load.

  A rotational position error correction device that solves the above problem is based on a rotational position signal that represents a rotational position of a rotor in a rotating electrical machine and a reference position signal that represents the start and end of one cycle of the rotor. A rotational position error correction device for correcting an error of the rotational position of the at least two rotational positions based on the rotational position signal at the same time before and after the same period before and after the half cycle by the reference position signal. Mean value calculating means for calculating an average value, the rotation position by the rotation position signal at the start time of one cycle in the reference position signal, and the rotation position signal by the rotation position signal at the end time of one cycle in the reference position signal. The rotation position on the straight line connecting the rotation position and the average value of the rotation position in a half cycle by the average value calculating means, and the rotation position signal Correction value calculating means for calculating a difference between the rotational position as a correction value, by comprising the gist.

  According to this, the difference between the rotation position based on the rotation position signal at the start point and the end point of one cycle and the rotation position on the straight line connecting the average values and the rotation position based on the rotation position signal becomes the correction value. For this reason, the correction value can be easily calculated, and the load on the rotational position error correction device can be suppressed.

  With respect to the rotational position error correction device, it is preferable that the average value calculating means calculates the average value from the rotational positions at more points as the number of rotations per unit time of the rotor in the rotating electrical machine increases.

  According to this, when the number of rotations per unit time increases, the correction accuracy may be reduced. However, it is possible to suppress a decrease in correction accuracy by calculating an average value from rotation positions at many points in time.

  According to the present invention, the load can be suppressed.

Schematic which shows the rotational position detection apparatus of embodiment. (A)-(d) is a time chart for demonstrating the process which a rotation position error correction apparatus performs. The schematic diagram which shows the resolver signal containing an error. The flowchart for demonstrating an averaging process. Explanatory drawing which shows the calculation method of a correction value.

Hereinafter, an embodiment of the rotational position error correction apparatus will be described.
As shown in FIG. 1, the rotational position detection device 1 includes a resolver 10 as a rotational position sensor. The resolver 10 has a rotor 11 that is attached to a rotor or a rotating shaft portion of a rotating electrical machine (motor) and rotates integrally with the rotor. Around the rotor 11, a sin voltage output coil 13 and a cos voltage output coil 14 are disposed 90 degrees apart from each other. An excitation coil 12 is disposed around the rotor 11 so as to be 180 degrees apart (opposite) from the cos voltage output coil 14.

  An excitation signal generator 15 is connected to the excitation coil 12. As a result, when an excitation signal is input to the excitation coil 12, the resolver 10 generates a dielectric voltage corresponding to the rotational position (rotation angle) of the rotor 11 (rotor of the rotating electrical machine) between both ends of the sin voltage output coil 13 and cos. It occurs at both ends of the voltage output coil 14.

The resolver 10 is connected to the RD converter 30. Specifically, the sin voltage output coil 13 and the cos voltage output coil 14 are connected to an RD converter 30 (resolver digital converter) that converts a resolver signal into a digital output angle. The RD converter 30 is connected to a microcomputer 40 having a CPU 41 and a memory 42. The memory 42, a program for correcting the error of the rotational position theta _d by the resolver signal detected by the resolver 10 are stored.

  2A and 2B, when an excitation signal S1 is applied to the excitation coil 12, a sin signal S2 (sin output voltage) is induced in the sin voltage output coil 13, and a cos voltage output coil is generated. 14, a cos signal S3 (cos output voltage) is induced. 2A to 2C, the horizontal axis represents time and the vertical axis represents voltage. In FIG. 2D, the horizontal axis represents time, and the vertical axis represents voltage and angle. The sin signal S2 and the cos signal S3 can be expressed by the following equations (1) and (2).

S2 = Refθ × sin θ (1)
S3 = Refθ × cosθ (2)
Here, Refθ is the excitation signal S1. That is, the sin signal S2 is obtained by multiplying the excitation signal S1 by the sine wave W1 shown in FIG. The cos signal S3 is obtained by multiplying the excitation signal S1 by the cosine wave W2 shown in FIG.

Then, the RD converter 30 calculates the rotational position θ _d based on the resolver signal of the rotor in the rotating electrical machine from the following equation (3).
θ _d = tan ⁻¹ (S2 / S3) (3)
Thus, it is possible to calculate the rotational position theta _d by the resolver signal. That is, in this embodiment, the resolver signal (sin signal S2 and the cos signal S3) becomes the rotational position signal representing the rotational position theta _d of the rotor in a rotating electrical machine.

As shown in FIG. 2 (d), (3) expression by that calculates the rotational position theta _d by the resolver signals, it is possible to obtain a waveform line L1 representing the rotational position theta _d by the resolver signal. However, as shown in FIG. 3, when an error is included in the resolver signal (sin signal S2 and the cos signal S3), the error in the rotational position theta _d to output the RD converter 30 by this error occurs. This error is caused by the mounting accuracy of the rotor 11 and external noise.

Then, in order to correct the error included in the rotational position θ _d , the microcomputer 40 calculates a correction value Δθ ₁ for correcting the rotational position θ _d based on the resolver signal. That is, the microcomputer 40 is a rotational position error correction device. The rotational position error correcting device of the present embodiment (the microcomputer 40), when the rotor of the rotating electrical machine is rotating at a constant speed, to correct the errors by calculating the correction value [Delta] [theta] _1.

As shown in FIG. 5, the CPU 41 calculates the time (period T) required for the rotor of the rotating electrical machine to make one rotation. RD converter 30, a pulse-like North marker signal S4 when the rotational position theta _d by the resolver signal reaches 0 degrees (360 degrees) (hereinafter referred to as NM signal) to the microcomputer 40. Then, the CPU 41 calculates the cycle T by regarding the output cycle of the NM signal S4 as one cycle of the rotor in the rotating electrical machine. Therefore, the NM signal S4 becomes the reference position signal.

Then, CPU 41 calculates the average value of the rotational position theta _d by the resolver signal at the time when the same time α spaced front and rear from the half period T / 2 time (E point) (average value calculating means). The number of calculation points (number of rotational positions θ _d ) used in the averaging process for calculating the average value may be any number as long as it is an even number. The number of rotations per unit time of the rotor in the rotating electrical machine (rotation of the rotor) By calculating the average value from the rotational positions θ _d at more points as the speed increases, it is possible to suppress a decrease in correction accuracy. Hereinafter, the averaging process will be specifically described.

  As shown in FIGS. 4 and 5, the CPU 41 calculates the rotational speed of the rotor (step S10). The rotation speed can be calculated from the period T (NM signal S4). Next, the number of calculation points used for the averaging process is read from the map (step S20). This map is stored in the memory 42, and the number of calculation points used for the averaging process is set in association with the rotation speed.

Next, CPU 41 reads the rotational position theta _d according to the calculated number of half-period T / 2 by the same time α away resolver signal at the time (step S30). In the present embodiment, the C point and the D point, which are the rotational positions θ _d by the two resolver signals separated from the half cycle T / 2 by the same time, are read. Then, CPU 41 calculates the average value of the rotational position theta _d read in step S30. When obtaining the average value of two points, the average value can be obtained by obtaining point E indicating the midpoint of a straight line drawn by connecting points C and D with a straight line (step S40).

As shown in FIG. 5, CPU 41 calculates the point A is the rotational position theta _d at the start of one period T in the NM signal S4. Similarly, the CPU 41 calculates a point B that is the rotational position θ _d at the end of one cycle T. Then, CPU 41 is, A point, calculates the difference between the rotational position theta _d by the rotation position and the resolver signal on a line L2 connecting the point E and point B as the correction value [Delta] [theta] ₁ (correction value calculating means). When the points A, E, and B are located on the same straight line, the straight line L2 is a single line segment. On the other hand, when the points A, E, and B are not located on the same straight line, the straight line L2 includes two lines, a line segment L3 connecting the points A and E, and a line segment L4 connecting the points E and B. Minutes. In this embodiment, since the point A, the point E, and the point B are not on the same straight line, the straight line L2 has two line segments. For this reason, the straight line L2 and the straight line L5 connecting the points A and B do not overlap.

The CPU 41 stores the correction value Δθ ₁ in the memory 42, and adds (or subtracts) the correction value Δθ ₁ to the rotational position θ _d based on the resolver signal from the period following the period T in which the correction value Δθ ₁ is calculated. and it outputs the rotational position theta _t after.

Next, the operation of the rotational position error correction device of this embodiment will be described.
The microcomputer 40 calculates the difference between the rotational position θ _d (detected value) based on the resolver signal and the rotational position on the straight line L2 connecting the points A, E, and B as the correction value Δθ ₁ . For this reason, the correction value Δθ ₁ can be easily calculated.

Therefore, according to the above embodiment, the following effects can be obtained.
(1) The microcomputer 40, the difference between the rotational position theta _d by the rotation position and the resolver signal on the line L2, i.e., it calculates the difference between the waveform line L1 and the straight line L2 as the correction value [Delta] [theta] _1, the correction value [Delta] [theta] ₁ By adding (or subtracting) the rotational position θ _d based on the resolver signal, the error of the rotational position θ _d is corrected. By the difference of the waveform line L1 and the straight line L2 and the correction value [Delta] [theta] _1, it is possible to easily calculate the correction value [Delta] [theta] _1, it is possible to reduce the load of the microcomputer 40. In particular, in Patent Document 1, it is necessary to calculate the average value ΔAVE based on the difference Δ between the reference line and the detected value θ of the rotation angle, or to calculate the correction value by subtracting the average value ΔAVE from the difference Δ. is there. On the other hand, the rotational position error correction device (microcomputer 40) of this embodiment does not require these processes, and the load is reduced.

(2) When the number of rotations per unit time of the rotor in the rotating electrical machine increases, the correction accuracy may decrease (the error of the correction value Δθ ₁ may increase). However, the more number of rotations per unit of rotor time, it is possible to suppress a decrease in correction accuracy by calculating the average value from the rotational position theta _d many times (calculated points).

(3) it is also conceivable to correct the rotational position theta _d by the resolver signal the difference between the rotational position theta _d by the rotation position and the resolver signal on a line L5 connecting the points A and B as the correction value. However, when the resolver signal includes an error having a period, the rotational position θ _d by the resolver signal at the points A and B may include an error, and the waveform of the straight line L5 connecting the point A and the point B and the waveform Even if the difference from the line L1 is used as a correction value, the error may not be reduced. In the present embodiment, by calculating the point E, which is the average value at the time of the half cycle T / 2, even if the resolver signal includes an error having a cycle, the error can be reduced.

In addition, you may change embodiment as follows.
In the embodiment, the number of calculation points is increased according to the number of rotations of the rotor per unit time, but the number of calculation points may be changed depending on the characteristics of the resolver 10.

As the rotational position sensor, for example, an encoder such as an optical rotary encoder may be used instead of the resolver 10.
Next, the technical idea that can be grasped from the embodiment and other examples will be described below.

  (A) A rotational position detection device for detecting a rotational position of a rotor in a rotating electrical machine, comprising the rotational position error correction device according to claim 1 or 2.

  L1 ... waveform line, L2 ... straight line, S2 ... sin signal, S3 ... cos signal, S4 ... north marker signal, 10 ... resolver, 40 ... microcomputer.

Claims (2)

A rotational position error that corrects an error in the rotational position of the rotor based on a rotational position signal that represents the rotational position of the rotor in the rotating electrical machine and a reference position signal that represents the start and end of one cycle of the rotor. A correction device,
An average value calculating means for calculating an average value for at least two rotational positions according to the rotational position signal at the same time before and after the half period by the reference position signal;
The rotation position by the rotation position signal at the start of one cycle in the reference position signal, the rotation position by the rotation position signal at the end of one cycle in the reference position signal, and a half cycle by the average value calculation means Correction value calculation means for calculating, as a correction value, a difference between the rotation position on a straight line connecting the average values of the rotation positions at the rotation position and the rotation position based on the rotation position signal. Rotational position error correction device.   2. The rotation according to claim 1, wherein the average value calculation unit calculates the average value from the rotation position at a larger time as the number of rotations per unit time of the rotor in the rotating electrical machine is larger. Position error correction device.