JPH0521166B2 - - Google Patents

Description

Detailed Description of the Invention [Field of Industrial Application] The present invention relates to a method and device for detecting the amount of rotation of a rotating body in absolute value, and in particular to a method and apparatus for detecting the rotation amount of a rotating body in absolute value. The present invention relates to a new method and apparatus for synthesizing the amount of rotation of a rotating body as an absolute value with high precision.

[Background Art and Problems] Conventionally proposed methods for detecting absolute values, for example, when used in machine tools, are By extracting the amount of rotation from the drive system in the direction, applying this amount of rotation to a deceleration mechanism consisting of two or three stages, and combining the values within one rotation of the rotation detector attached to each deceleration rotation shaft. This method was used to calculate absolute values.

However, in this conventional example, when trying to expand the effective measurement range, the deceleration mechanism becomes larger and its moment of inertia also increases.Since the values on each rotation axis have different weights, in other words, the values on each rotation axis have different weights. Since the amount of movement of the member to be moved (for example, the table) per rotation is different, there is a problem that errors are likely to occur. For this reason, the mechanical accuracy of each element of the reduction mechanism, mainly gears and bearing members, must always be maintained at a high level of accuracy, despite the vibrations and wear that accompany machine tool operation. .

Therefore, the present applicant previously proposed Japanese Patent Application No. 199882/1982 as a solution to such problems. This method detects the amount of mechanical change from a predetermined reference state position regarding the mechanical motion of the member to be measured as an absolute value. A detection means consisting of a plurality of detectors that emit signals is prepared, an electric signal corresponding to each cycle is extracted from the detection means and stored and held, and then one of the detection means according to the amount of mechanical change is prepared. The relative mechanical change amount between the (first detector) and the member to be measured is determined by detecting a value that is an integral multiple of the period (first period) corresponding to the first detector and a value of the same period. The absolute position is attempted to be detected by using a procedure of determination using the period corresponding to the other detector in the means and the stored value obtained from the other detector.

As a result, since this method calculates the absolute position by combining data from detectors with multiple periods, there is no weight between the measurement data from each detector as in the conventional method, and measurement errors are reduced. In addition to having almost no constraints, the moment of inertia of the gear train can be reduced, and there is almost no need to take measures against errors caused by gear friction.

By the way, it is conceivable to use this method even when detecting the amount of rotation of a rotating body such as a rotary table as an absolute value. However, even in this case, there is one drawback in terms of measurement accuracy, and when dividing a rotating body such as a rotary table into equal parts,
When dividing 360 degrees into equal parts, there is a drawback that the number of rotations of the main shaft is restricted depending on the number of divisions.

[Object of the Invention] The object of the present invention is to meet these demands and provide a method and device for highly accurately detecting the amount of rotation of a rotating body as an absolute value without restricting the number of rotations of the main shaft. It is about providing.

[Means and effects for solving the problem] Therefore, the method of the present invention is such that the rotational motion of the rotating body is connected to each other via the motion transmission mechanism, and at least one has a higher resolution than the other. , and a detection means consisting of a plurality of detectors that generate electrical signals with periods or amplitudes corresponding to different predetermined mechanical changes in the rotating body, is used to detect a predetermined mechanical rotational movement of the rotating body. A method for detecting the amount of rotation from a reference state position determined by an absolute value, wherein one of the detection means is used as a first detector, and the motion transmission means is the effective rotation of the rotating body. a first step of setting the relationship: number ∝ effective rotational speed of the first detector/mechanical transmission coefficient; and when a mechanical change occurs between the detection means and the rotating body, each of the detectors a second step of reading and storing electrical signals corresponding to each period from , one detector (first detector) of the detecting means, and one detector (first detector) of the detecting means associated with the amount of mechanical change in the second step; a third step of roughly specifying the relative positional relationship with the rotating body using a value that is an integral number (N) times the period corresponding to the first detector and a value that is less than one period of the same period; Integer value N in the third step
is determined using the period corresponding to the other detector among the detectors and the stored value obtained from the other detector, and the value obtained in this fourth step. Accurately calculate the absolute value using the obtained integer value N, the period corresponding to the high-resolution detector in the detection means, and the values obtained from the other detectors and stored in the second step. and a fifth step of determining.

Further, the device of the present invention is a device that detects the amount of rotation from a predetermined reference state position with an absolute value with respect to the mechanical rotational movement of the rotating body, and A plurality of detections each connected via a transmission mechanism, at least one of which generates an electrical signal with a higher resolution than the others and with a period or amplitude each corresponding to a different predetermined amount of mechanical change with respect to the rotating body. a detection means consisting of a detector, and an electric signal corresponding to less than one cycle of each cycle extracted from each detector of the detection means in a state where the mechanical rotational movement of the rotary body is stopped, as a digital quantity. A memory means for storing memory, a relative positional relationship between one detector (first detector) of the detecting means and the rotating body is determined by an integer (N) of the period corresponding to the first detector. ) times the value and a value less than one period of the same period, the integer value N is set to the period corresponding to the other detector in the sensing means and the digital value stored in the memory means. an integer value determination means determined using a quantity, an integer value N obtained by the integer value determination means, a period corresponding to a high-resolution detector in the detection means, and a period obtained from the same other detector and and means for accurately determining an absolute value using a digital quantity stored in a memory means, the motion transmission mechanism being controlled by: effective rotational speed of the rotating body∝effective rotational speed of the first detector/mechanical transmission. It is characterized by being structured into a coefficient relationship.

[Description of Embodiment] FIG. 1 shows the overall configuration of a detection device of this embodiment. In the figure, one output shaft 72 of the motor 50 is connected to, for example, a shaft 71 of a rotary table 51 of a machine tool via a gear G(1) and a gear G(N). Further, the other output shaft 73 of the motor 50 is integrally connected with a shaft 77A of the gear G(29) ^* and a rotor shaft 77 of the resolver 56 in this order. The gear G(29) ^* has a gear G(29),
G (30) are meshed with each other, and further gear G (3
Gear G (23) provided on coaxial 75A with 0)
A gear G (24) is meshed with. The rotor shaft 74 of the resolver 53 is connected to the shaft 74A of the gear G (29), and the shaft 74 of the gears G (30) and G (23) is connected to the shaft 74A of the gear G (29).
A rotor shaft 75 of the resolver 54 is connected to 5A, and a rotor shaft 76 of the resolver 55 is connected to the shaft 76A of the gear G (24).

In each resolver 56, 53, 54, 55, electric wires 79, 78 are connected to the excitation circuit 52 for supplying excitation signals sine waves and cosine waves to the primary side excitation winding.
connected from. The excitation circuit 52 has an excitation signal for the resolvers 56, 53, 54, 55.
A selection switching circuit (not shown) is connected to switch the supply of sine waves and cosine waves. On the other hand, the secondary side output from each resolver 56, 53, 54, 55
P _W , P _V , P _U , P _T are electric wires 62, 59, 60, 6
1 to the relay circuit 63, where one is selected, and then sent to the isolator 64 via the electric wire 82. The output signal ground-separated by the isolator 64 is converted into a rectangular wave by a filter/comparator section 65, and then inputted to the register 58 by a flip-flop 66 and a NAND gate 67 as an enable signal EN. Then, the value of the 2000-decimal counter 57 is set in the register 58 via the line 80 in synchronization with the clock CK immediately after the enable signal EN is input. The value set in register 58 is sent to the central processing unit (CPU) via line 81.
68, processed in accordance with the procedure described later, and stored in memory 69.

Note that the gears G(29) ^* , G(29), G(3
0), G(23), and G(24) are respectively
At 29:29:30:23:24, the gears G(N) and G(1)
The ratio of the number of teeth is set to N (N is an arbitrary integer):1. In addition, resolvers 56, 53, 54,
55 uses resolvers with a 10-pole (10X), 1-pole (1X), 1-pole (1X), and 1-pole (1X) structure, respectively. In addition, the primary side terminal AG of the isolator 64
and secondary side terminal LG indicate analog ground and logic ground, respectively.

Absolute 0 point acquisition In Figure 2, the waveform diagram A in the diagram shows the excitation signal Sin
When applying wave, Cos wave, each resolver 56,5
Output signal output from the secondary side of 3, 54, 55
The waveforms of P _W , P _V , P _U , and _PT are shown, respectively.
By adjusting the mechanical angles of the resolvers 56, 53, 54, and 55, the phase difference between the four signals is eliminated and they are made to coincide with each other as shown in FIG. 2B.

Next, in Figure 3A, the top waveform is
2000 decimal counter 57 that performs 5KHz counting operation
When the counter 57 is run as a verification position counter, the state in which the counter 57 repeats counting from 0 to 1999, the waveform of the next stage is the basic clock, and the waveform of the next stage is the enable signal EN of the register 58 (filter/comparator section). 65 and flipprop 6
6 and is given as an output by a NAND gate 67. ), the bottom waveforms show the secondary output signals P _W , P _V , P _U , P _T of each resolver and how they form the enable signal EN when they cut off the zero voltage level, respectively. There is. In the examples of these waveform diagrams, it can be seen that the count value of the counter 57, which is set at the rising edge of the clock when the enable signal EN=0, is not zero. Therefore,
As shown in FIG. 3B, the counter 57 is set at the rising edge of the clock when the enable signal EN=0.
Adjust so that the count value becomes 0. In other words, the output signal of each resolver adjusted with phase difference = 0
P _W , P _V , P _U , and _PT are further adjusted so that the count value of the counter 57 becomes zero. The position obtained by the above operation is set to absolute 0 point (absolute 0 point).
point).

Absolute position (angle) calculation (1) Now, as shown in FIG. It is assumed that is located at the absolute 0 point. From this state, as shown in FIG. 4, the rotary table 51 is rotated to θ ₀ , and each resolver 56, ₅₃ , 54,
The output signals P _W (w), P _V (v), P _U (u), P _T (t) from 55 are
Assuming that w ₀ , v ₀ , u ₀ , and t ₀ , the absolute position θ ₀ is: θ ₀ = R _v・e+v ₀・e/2000 (1) R _v = Number of revolutions of resolver 53 (positive integer) (2) can be expressed as This is resolver 5
3 and resolver 56 are gears G(29), G(29) ^*
This is because they are connected by G(29):G(29) ^*
=1:1. Also, here motor 5
The amount of movement (angle) of the rotary table 51 per rotation of 0 is set to e=360/N [°/rev].

On the other hand, in the resolver 54, there is a relationship of G(29) ^* :: G(30) = 29:30, so u ₀ is u ₀ · e / 2000 = 29/30 (R _v · e + v ₀ · e /2000)-A
(3) A=iFiX[29/30(R _v・e+v ₀・e/2000)/e]e (4) It can be expressed as follows. However, iFiX[α] is α
It means converting to an integer.

In addition, in the resolver 55, since the relationship is G(29) ^* :G(24)=29:24, t ₀ is t ₀ · e / 2000 = 667 / 720 (R _v · e + v ₀ · e / 2000-C(5
) C=iFiX[66/720(R _v・e+v ₀・e/2000)/e]e (6).

Then, in equations (3) and (4), R _v ={0, 1,
2, ..., (720)} to find R _v = {i, j, k, ..., q} that satisfies equations (3) and (4), and from among them, equation (5) , find R _v (only one exists) that satisfies (6) as R _v0 . And this R _v0
By substituting R _v in Equation (1) into Equation (7) with R _v0 , θ ₀ = R _v0・e+v ₀・e/2000 (7) From Equation (7), the absolute position (angle) θ ₀ can be found.

However, in this equation (7), the accuracy is e/2000 [°/pulse], which is a fairly 'coarse' resolution, so the output signal from the resolver 56 is also required.
Determining the absolute position (angle) θ ₀ with an accuracy of 20000 [°/pulse], θ ₀ =R _v0・e+ (8) ψ=iFiX[v ₀ /200]・200・e/2000+w ₀ e/20000(9 ) becomes.

The flowchart in Figure 5 is divided into two parts: flow() and flow().
In flow (), R _v that simultaneously satisfies equations (3) and (4)
The process of selecting multiple values of , in flow (), the process of selecting only 1 for R _v0 from among them,
are shown respectively.

Effective Detection Range The range that can be detected by the detection device shown in FIG. 1 is determined by the following formula.

Now, if any absolute position (angle) is θ, then θ=R _v・e+v・e/2000 R _v = Number of revolutions of resolver 53 (positive integer) (10) becomes. However, t, u, v are arbitrary positions (corners)
This is measured position data at θ. Here, data t, u, v are t=u=v=0 (including absolute 0 point)
Then, from equations (11) and (12), becomes. Then, becomes. Now, we introduce the functions f and g to this equation (14), and get f(R _v )=29/30R _v −iFiX[29/30 R _v ](15) g(R _v )=667/ 720R _v −iFiX[667/720R _v ] (16) Then, f(R _v and g(R _v ) are shown in FIG.

Therefore, the solutions of the equations f(R _v )=0, g(R _v )=0 are as follows from FIG. 6: R ^f _v =30α (17) R ^g _v =720β (18). However, α and β are 0 or positive integers.

From the above, R ^f _v = R ^g _v , 30α = 720β ∴α = 720/30β = 24β. Substituting this α into equation (17) yields R ^f _v =30×24β=720β (=R ^g _v ). Therefore, the electrical position data becomes zero.
R _v is R _v = 0 (absolute 0 point) when β = 0, and R _v when β = 1.
= 720, R _v = 1440 when β = 2, R _v = 2160 when β = 3,
...becomes.

Therefore, when β=1, the effective detection range θmax is expressed with rough e/2000 [°/pulse] accuracy as follows: θmax=R _v max・e+vmax・e/2000=(720−1)・e+1999・e /2000 = 719.9995·e [°], and 2e circumferences of the rotary table 51 are the rotation angle range that can be detected absolutely.

Measurement Errors The previous discussion of FIG. 1 concerned the method process without taking into account errors in the measured data t ₀ , u ₀ , v ₀ . However, in reality, these measured data t ₀ , u ₀ , v ₀
includes quantization errors affected by electrical resolution and mechanical errors of the detector and gear train.
Produces results different from theoretical values. Therefore, next we check the variation range of such errors.

Now, the position data t, u, v measured corresponding to the absolute position (angle) θ from each resolver 53, 54, 55 in Fig. 1 are expressed as t=t _T +Δt (19) u=u _T +Δu (20) Expressed as v=v _T +Δv (21). Here, t _T , u _T , v _T are true values,
Δt, Δu, and Δv are errors.

Therefore, equations (3) and (4) are (u _T +Δu)・e/200029/30{R _v・e+(v _T +Δv
)×e/2000}-A(22) A=iFiX[29/30{R _v・e+(v _T +Δv)e/2000}
/e]e(23). Here, U _T・e/2000 is u _T・e/2000=29/30(R _v ×e+v _T・e/2000)−iF
iX[29/30(R _v・e+v _T・e _/ 2000)/e]e(24) In order to express it as
The error Δu _obtained by substituting _into e]×e−iFiX [29/30(R _v・e+v _T・e/2000)e]e(26) Therefore, |γ|=0 From e, Δu29/30×ΔV〓−0 2000 (27) Similarly, from equations (5) and (6), Δt・667/720Δv〓 (28). Now, let the quantization error be ε and the mechanical error be δ, and set these on the detector as ε=360/2000 [°/pulse] and δ=40 [′/pulse], respectively, and let Δv have ε and δ. , substituting into equations (27) and (28), [Δu] ^* 29/30 (360/2000+40/60)・[Δv] 0.818・[Δv] [Δt] ^* 667/720 (360/2000+40/60)・[Δv] 0.784・[Δv] If [Δv] = ±3, then [Δu] ^* ±2.454[°] (29) [Δt] ^* ±2.352[°] (30) However, [ΔQ] is defined as the actual error pulse number of ΔQ. Furthermore, [ΔQ] ^* is defined as the actual error amount of ΔQ.

Now, in R _v0 determined by a series of steps in the flowchart shown in Figure 5, when [Δv] = ±3, the possible range of u ₀ is as follows when u ₀ = (1999, 0) 〈u _T −2.454/360・2000〉≦u ₀ ≦1999 (31) Or, 〈u _T +2.454/360・2000〉≧u ₀ ≧0 (32) When u ₀ ≠ (1999, 0) vicinity , u _T −2.454/360・2000≦u ₀ ≦u _T +2.454/360・2000(33) The possible range of t ₀ is 〈t _T −2.352/ when t ₀ = (1999, 0) vicinity. 360・2000〉≦t ₀ ≦1999 (34 ₎ Or, when t _{0 ≠} (1999, 0), t _T −2.352/ It is necessary to judge within the range of 360・2000≦t ₀ ≦t _T +2.352/360・2000(36). Furthermore, here,
<D> is D-iFiX [D/2000] if D≧0.
2000, and if D<0, it is defined as 2000+D.

When [Δv]=± ₄ , the possible range of u ₀ is 〈u _T −3.272/360・2000〉≦u≦1999 (37) or 〈u _T +3.272/360・2000〉≧u ₀ ≧0 (38) When u ₀ ≠ (1999, 0), u _T −3.272/360・2000≦u ₀ ≦u _T +3.272/360・2000( 39) When t ₀ is near (1999, 0), the possible range of t ₀ is 〈t _T −3.136/360・2000〉≦t ₀ ≦1999 (40) or 〈t _T +3.136/360・2000〉≧t ₀ ≧0 (41) When t ₀ ≠ (1999, 0), judge in the range of t _T −3.136/360・2000≦t ₀ ≦t _T +3.136/360・2000(42) There is a need to.

Note that the value u _T in the inequalities of equations (31) to (33) and (37) to (39) is required, but this u _T is calculated as v _T →v ₀ in equation (24). use The same goes for _tT .

From the above, when a correction formula is determined by setting [Δv]=±4, since the measurable position data u ₀ and t ₀ are all integer values, the determination is actually made using the inequality equation below.

That is, for equation (37), iFiX [〈u _T −3.272/360・2000〉]≦u ₀ ≦1999 (37A) For equation (38), iFiX [〈u _T +3.272/360・2000〉]≧u ₀ ≧0 (38A) For equation (39), iFiX[u _T −3.272/360・2000]≦u ₀ ≦iFiX[u _T +3.272/360・2000] (39A ) For equation (40), iFiX[〈t _T −3.136/360・2000〉]≦t ₀ ≦1999 (40A) For equation (41), iFiX[〈t _T +3.136/360・2000〉】≧t ₀ ≧0 (41A) For equation (42), iFiX[t _T −3.136/360・2000]≦t ₀ ≦iFiX[t _T +3.136/360・2000] (42A ) is as follows.

Next, let us consider the possibility that the correct absolute position (angle) may be incorrectly calculated due to the presence of the measurement error described above. As shown in Fig. 7, the tooth ratio of each gear is G(29):G(30):G(2
3):G(24)=29:30:23:24, and the resolvers 53, 54, and 55 are all 1× resolvers. Therefore, if gear G (29) rotates once from the absolute 0 point position and then continues to rotate in the same direction, gear G (30)
Suppose that it rotates once for the first time, during which time G(3
The number of pulses du generated by the rotation of 0) is du=30−29/30·200066.7. Also, gear G(29) is
Similarly, the number of pulses in 4) is dt=720−667/7202000147.2. Therefore, 2.[Δu]<du(2.3.272/360.2000<66.7) (43) 2.[Δt]<dt(2.3.136360.2000<147.2) (44) and the measurement data u ₀ , Error [Δu] in t ₀ ,
[Δt], even if the data u _{0 and} t ₀ containing the error are used when deriving R _v0 in the flowchart of Fig. 5, equations (43) and (44) are satisfied. As long as R _v0 is true, the correct value will be obtained. In other words, it is not affected by errors. However, since the data v ₀ includes an error, if we express the absolute position (angle) θ ₀ with an accuracy of e/2000 [°/pulse] as θ ₀ = R _v0・e+v ₀・e/2000, the error will remain as θ It will join ₀ . Therefore, we will discuss the correction method in the next section.

Correction of absolute position error Design accuracy of resolvers 53 and 56 and gear G
(29), G(29) ^* When performing ``absolute 0 point acquisition'' with the combination accuracy, two methods can be considered: absolute 0 point acquisition and absolute 0 point acquisition, as shown in FIG.

Absolute 0 point acquisition occurs when the P _V signal has a phase lag than the P _W signal (P _W (w) is within 9 pulses at maximum), and absolute 0 point acquisition occurs when the P _V signal has a phase lead than the P _W signal. Now, when the motor 50 rotates from the absolute zero point and the rotary table 51 stops at an arbitrary position θ, data v is measured with ±4 pulses due to quantization error and mechanical error, and data w is measured with ±1 pulse due to quantization error. Assume that errors are included. Therefore, both data v and w are (1999, 0)
When the neighborhood is shown, as shown in the examples of FIGS. 9 and 10, v≪w or v≫w, and e/
When calculating the absolute position (angle) with an accuracy of 20000 [°/pulse], the general formula of equations (8) and (9) θ=R _v0・e+ (45) = iFiX [v/200]・200・e /2000+w・e/2000
0(46) requires correction.

From this, the absolute position (corner) is
If the accurate absolute position (angle) θ ^* is calculated based on the resolver 56 that derives the e/20000 [/pulse] accuracy, the flowchart shown in FIG. 11 will be obtained. In this flowchart, the resolvers 53 and 56 use data v ₀ and w ₀ in STP1.
Calculate a value that does not include an error within one rotation of , and set this as the value. In STP2, it is checked whether the data w ₀ is smaller than 80 (immediately after one cycle of P _W (w) has passed). If YES in STP2,
Moving to STP3, the context of data v ₀ with respect to data w ₀ is checked. Also, if NO, STP5
Move to. In STP3, if v ₀ is before w ₀ and when the value of v ₀ is divided into 10 equal parts and it is at the rear of each division (YES), the process moves to STP4. Otherwise (NO), the process moves to STP8 and no correction is made. In STP4, the correction value e/10 is added to the value obtained in STP1, and the process moves to STP8.

Also, in STP5, it is checked whether the data w ₀ is greater than 1920 (just before P _W (w) enters the next cycle). If YES in STP5,
Proceeding to STP6, the context of data v ₀ with respect to data w ₀ is checked. Also, if NO, STP8
, and no correction is made. In STP6, if v ₀ is after w ₀ and when the value of v ₀ is divided into 10 equal parts and it is in the front of each division (YES), the process moves to STP7. In other cases (NO), the process moves to STP8 and no correction is made. In STP7, the correction value -e/10 is added to the value obtained in STP1, and the process moves to STP8. As described above, in STP8, using the rotation speed R _v 0 and the corrected
Calculate the accurate absolute position (angle) θ ₀ ^* .

Arbitrary 0 point acquisition The above-mentioned 'absolute 0 point acquisition' is Resolver 5
6, 53, 54, 55 and gears G (29), G
(29) ^* There was a restriction (matching of mechanical angles) in the configuration for G(30), G(23), and G(24). This means that even if you manually change the position (angle) when the power is turned off, the position data t, u, v, w obtained when the power is turned on again.
While this is advantageous in that it allows you to calculate the current position using only a single button, it does add a considerable burden to the design and configuration of the detector.

Therefore, the absolute (corner) is detected using the memory 69 shown in FIG. Now, resolvers 56, 53, 54, 55 and gears G (29), G
(29) ^* , G(30), G(23), and G(24) are optionally configured (mechanical angle matching in the resolver is not required). Without changing this configuration, the motor 50 is rotated to the required zero point position and then stopped, and the obtained data is P _T (t)=t ₀ P _U (u)=u ₁ P _V (v)=v ₀ P _W (w)=w ₀ , and write each value to the memory 69 via the CPU 68. FIG. 12 shows how the position obtained by this operation is set as an ``arbitrary 0 point.'' From now on, the absolute position (angle) is calculated using this arbitrary 0 point as 0 [°].

Absolute Position/(Angle) Calculation (2) Now, as shown in FIG. 12, it is assumed that the motor 50 is rotated from an arbitrary 0 point and stopped at an arbitrary position (angle) θ. At this time, the absolute position (angle) θ is based on equations (10), (11), and (12), θ=R _v・e+(v−v ₀ )・e/2000 R _v = pseudo rotation of resolver 53 Number (integer) (47) It can be expressed as However, when the pseudo rotation speed R _v of the resolver 53 is R _v = {0, 1, 2}, the equations (48) and (49) become the following equations (48A) and (4), respectively.
9A) needs to be considered.

u・e/2000=29/30 {R _v・e+(v−v ₁ )・e/20
00} +u ₁・e/2000 (48A) t・e/2000=667/720 {R _v・e+(v−v ₁ )・e/20
00}+t ₁・e/2000(49A) This means that the absolute position θ (angle) is the 13th
As shown in the figure, when θ≦30/29 (e−u ₁・e/2000) or θ≦720/667 (e−t ₀・e/2000), Equation (48), ( 49), it is necessary to treat the 'constant' e-u ₁ ·e/2000 e-t ₁ ·e/2000 as a 'variable'.

From the above, equations (47), (48), (49), (48A),
A flowchart for determining the absolute position (angle) θ using (49A) is shown in FIG. but,
Absolute position (corner) obtained with this flow
Since θ has e/2000 [°/pulse] accuracy and is a rough value, the absolute position (angle) can be calculated with e/2000 [°/pulse] accuracy as in the flowchart shown in Figure 11. is shown in FIG.

In the flowchart of FIG. 15, flow ( ) corresponds to STP1 in the flowchart of FIG. 11, and calculates the pseudo movement angle (less than one rotation) Σ using values within one rotation of the resolvers 53 and 56. It shows the situation. Also, flows () and (V) are the first
The flowchart in Figure 1 corresponds to the processing from STP2 to STP7, and shows how a correction value ±e/10 is added in consideration of the influence of data measurement errors. From the above, the rotation speed R _v ^* and the corrected Σ
Calculate the accurate absolute position θ using

The present invention has been explained above with reference to the drawings, in which R _v0 is expressed by equations (3) and (4) based on the period of the resolvers 54 and 55 and values u ₀ and t ₀ that are less than one period of the period.
Determined according to (5) and (6). But R _v0 in this case
may be determined using the ratio of the period of the resolver 56 as a high-resolution detector to the period of the resolver 53 and a value w ₀ less than one period of the resolver 56.

Furthermore, Figs. 5, 11, 14, and 15 merely show one method of processing measured data. It may be determined by substituting a regular and periodic number.

Also, in Figure 1, a resolver was used as the rotary position detector, but as a rotary position detector,
It is not limited to resolvers. Also, it does not have to be a rotating type, for example, as long as it has a constant period and the absolute quantities (such as w, v, u, t, etc.) can be measured within one period,
The detector is not limited to a rotary type, but may be a linear type detector, such as an inductosin or a magnescale. In this case, the detection means may be configured by a combination of a rotary position detector and a linear position detector.

Further, in the above embodiment, the position detector uses a phase modulation method, but an amplitude modulation method position detector may be used. In this case, as shown in FIG.
2, the excitation signal ei=sinωt is supplied to each amplitude modulation type detector 153, 154, 155, 156,
The selector 201 selects the output signals e ₀ =K・sinθ・ei from each of the detectors 153, 154, 155, and 156.
After selecting, the detection circuit 202 and filter 2
03, the A/D hold 204 converts it into a digital quantity, and the CPU 205 inputs the converted data. However, the circuit configuration after the selector 201 is not particularly limited as long as it can detect the amplitude value of the feedback signal.

Furthermore, the motor 50 in FIG. 1 may be a pulse motor, a synchro motor, or the like. Further, the effective value of the rotational speed of the motor 50 is not specified, and can be expanded arbitrarily using a gear transmission mechanism. In other words, the gear transmission mechanism may be configured to have the following relationship: Effective rotation speed of table 51 ∝ Effective rotation speed of resolver 53 / Mechanical transmission coefficient. Further, the gear reduction mechanism is not limited to the above embodiment.

[Effects of the Invention] As described above, according to the present invention, it is possible to provide a method and a device that can detect the amount of rotation of a rotating body with high precision as an absolute value.

[Brief explanation of drawings]

FIG. 1 is an overall configuration diagram showing one embodiment of the detection device of the present invention, FIG. 2 is a diagram illustrating matching the phases of the secondary side outputs of each resolver in FIG. 1, and FIG.
The figure is a diagram explaining the counting operation of the collation position counter (2000 base counter) in Figure 1 and the zero cross adjustment (absolute zero point adjustment) that adjusts the phase of each resolver output. A waveform diagram serves as a reference for explaining the process of absolute position detection using the detection device shown in the figure, Figure 5 is a flowchart explaining the data processing process in the central processing unit of Figure 1, and Figure 6 is a flow chart of the process of detecting the absolute position using the detection device shown in Figure 1. A diagram explaining the process for calculating Ru at which the electrical position data becomes zero in the detection device, Figure 7 is a diagram to explain the influence of measurement error in the detection device in Figure 1, and Figure 8 FIG. 3 is a diagram explaining the phase relationship between the secondary side outputs of resolvers 56 and 53 that occurs during zero-cross adjustment, FIG. 9 is a diagram explaining possible measurement errors from FIG. 8, and FIG. Similar to Figure 9 _, Figure 11 is a diagram explaining another measurement error that may occur from Figure 8.
A flowchart explaining the data processing process for calculating the absolute position (corner) with high precision using FIG. 13 is a waveform diagram that can explain the processing process when R _v = {0, 1, 2} in the data processing process in the central processing unit in FIG. Figure 1 is a flowchart explaining the processing process for calculating the absolute position (corner) from arbitrary zero points using the central processing unit, and Figure 15 is a flowchart explaining the process of calculating the absolute position ( _corner ⁾ from arbitrary zero points using the central processing unit. FIG. 16 is a flowchart illustrating a processing process for calculating the angle) from an arbitrary zero point, and FIG. 16 is a circuit diagram showing a measurement system when an amplitude modulation type detector is used as a position detector. 50...Motor, 51...Rotary table as a rotating body, 53, 54, 55...1X resolver,
56...10X resolver, 52...excitation circuit, 57...
2000-decimal counter, 58...Register, 63...Relay circuit, 64...Isolator, 65...Filter/comparator section, 66...Flip-flop, 68...
central processing unit, 69...nonvolatile memory;

Claims (1)

[Scope of Claims] 1. Predetermined mechanical changes that are respectively connected to the rotational movement of the rotating body via a motion transmission mechanism, at least one of which has a higher resolution than the others, and that are different from each other with respect to the rotating body. The amount of rotation from a predetermined reference state position regarding the mechanical rotational movement of a rotating body can be determined as an absolute value using a detection means consisting of multiple detectors that generate electrical signals with a period or amplitude corresponding to the amount of rotation. A method of detecting, wherein when one of the detectors among the detection means is a first detector, the motion transmission means is set to the effective rotation number of the rotating body∝effective rotation number of the first detector/mechanical a first step of setting a relationship called a transmission coefficient; and when a mechanical change occurs between the detection means and the rotating body, reading and storing electrical signals corresponding to each cycle from each of the detectors; a second step in which the relative positional relationship between one of the detectors (first detector) of the detecting means and the rotating body is determined according to the amount of mechanical change in the second step; a third step in which the integer value N in this third step is roughly specified by a value that is an integer (N) times the period corresponding to one detector and a value that is less than one period of the same period; a fourth step of determining the period corresponding to the other detector in the detector and the stored value obtained from the other detector; and an integer value N obtained in this fourth step. and a period corresponding to the high-resolution detector in the detection means and the values obtained from the other detectors and stored in the second step to accurately determine the absolute value. A method for detecting the amount of rotation of a rotating body using an absolute value, the method comprising the following steps: 2. In claim 1, in the fourth step, the integer value N in the third step is
determined using the ratio of the period corresponding to the high-resolution detector in the detection means to the period of the first detector and the values obtained from the other detectors and stored in the second step. A method for detecting the amount of rotation of a rotating body using an absolute value. 3. In claim 1 or 2, the relative positional relationship between the rotating body and the first detector is θ, and the period of the first detector is d ₁ , which corresponds to the same period. Let Δd ₁ be the measured value stored in memory, and let N be the integer value, and let these values θ, d ₁ ,
Specify △d ₁ , N using the relational expression θ=N・d ₁ + △d ₁ to obtain an integer value N. Furthermore, in order to obtain θ with higher precision, for the measured value Δd ₁ ,
Using the measured value Δd ₁ ^* stored in the high-resolution detector, calculate θ from θ=N・d ₁ +f(Δd ₁ , Δd ₁ ^* ),
A method for detecting the amount of rotation of a rotating body using an absolute value, which is characterized by using this as an absolute position. 4 A device that detects the amount of rotation from a predetermined reference state position with an absolute value regarding the mechanical rotational movement of a rotating body, which is connected to the rotational movement of the rotating body through a motion transmission mechanism. detection means comprising a plurality of detectors, at least one of which has a higher resolution than the others and generates an electrical signal with a period or amplitude corresponding to a different predetermined mechanical change amount with respect to the rotating body; a memory means for storing and holding as a digital quantity an electric signal corresponding to less than one cycle of each cycle, which is extracted from each detector of the detection means when the mechanical rotational movement of the rotating body is stopped; One detector among the detection means (first detector)
and the rotating body relative to the first
The integer value N corresponds to other detectors in the detection means, so as to roughly define a value that is an integer (N) times the period corresponding to the detector and a value that is less than one period of the same period. an integer value determination means determined using a period and a digital quantity stored in said memory means; an integer value N obtained by said integer value determination means and corresponding to a high-resolution detector in said detection means; means for accurately determining an absolute value using digital quantities obtained from periodic and other detectors and stored in memory means; 1. A device for detecting the amount of rotation of a rotating body using an absolute value, characterized in that the relationship is: effective rotation speed of a detector/mechanical transmission coefficient. 5. A device for detecting the amount of rotation of a rotating body in absolute values, wherein the detection means is comprised of a plurality of rotational detectors. 6 In claim 5, the plurality of rotary detectors include a plurality of detectors whose period is one rotation, and a high-resolution detector whose period is a rotation angle range obtained by equally dividing one rotation. A device for detecting the amount of rotation of a rotating body using an absolute value. 7. A device for detecting the amount of rotation of a rotating body using an absolute value, as set forth in claim 4, wherein the detection means is comprised of a linear position detector. 8. According to claim 4, the detection means detects the amount of rotation of the rotating body using an absolute value, characterized in that the detection means is configured by a combination of a rotary position detector and a linear position detector. device to do. 9. A device for detecting the amount of rotation of a rotating body as an absolute value according to claim 4, wherein the detection means is configured to provide a phase shift (modulation) signal. 10 Claim 5 or 6, wherein the motion transmission mechanism is configured by a rotation transmission mechanism consisting of a shaft and a gear train, and detects the amount of rotation of the rotating body using an absolute value. Device. 11. A device for detecting the amount of rotation of a rotating body using an absolute value, as set forth in claim 10, wherein the rotation transmission mechanism is constituted by a plurality of gears in which the number of teeth is mutually prime. . 12. A device for detecting the amount of rotation of a rotating body using an absolute value, as set forth in claim 10 or 11, wherein the motion transmission mechanism includes an electric rotation drive means. 13. The apparatus for detecting the amount of rotation of a rotating body using an absolute value according to claim 12, characterized in that a pulse motor is used as the electric rotation driving means. 14. The apparatus for detecting the amount of rotation of a rotary body using an absolute value according to claim 12, characterized in that a synchro motor is used as the electric rotation drive means. 15. In any one of claims 4 to 14, the memory means for storing and holding the electrical signal as a digital quantity has a counter that periodically calculates a certain number at a certain time period. ,
A device for detecting the amount of rotation of a rotating body using an absolute value, further comprising a register that stores the count contents of the counter corresponding to the extracted electric signal as measurement data at that time. 16. In any one of claims 4 to 15, the integer value determining means is configured to combine a measured digital value with an arbitrary integer value (negative or negative value) and a cycle corresponding to each measured value. 1. A device for detecting the amount of rotation of a rotating body using an absolute value, comprising a computer that determines the amount of rotation by sequentially changing the integer value that satisfies the relationship.