JP3194309B2 - Meshing gear rotation error detector - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention detects a change in the rotation angle of each of two gears that rotate while meshing with each other, and detects a rotation error of the gears based on the detected rotation angle change. BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an engagement type gear rotation error detecting device of a type for detecting.

[0002]

2. Description of the Related Art A meshing gear rotation error detecting device of this type generally comprises (a) a gear rotation signal generating device which discretely generates a rotation signal for each gear when each of two meshing gears rotates by a predetermined angle. Means, and (b) gear rotation error determination means for determining a rotation error of the gear based on the mutual relationship between the generation states of the rotation signals.

A conventional example of this type of meshing gear rotation error detecting device is disclosed in Japanese Patent Application Laid-Open No. 55-78229. This means that the gear rotation signal generating means outputs, for each gear, a rotation pulse signal in which a pulse is generated with an amplitude each time each of the two meshing gears rotates by a fixed angle,
In addition, the gear rotation error determining means determines that one pulse in the rotation pulse signal output for one gear and one in the rotation pulse signal output for the other gear.
The phase difference between the two pulses (that is, the difference between the absolute rotation angles of the respective gears from the reference rotation position and the two gears) is detected, and the rotation error (ie, the rotation error) is determined based on the detected phase difference. This is a meshing gear rotation error detection device that determines the amount of advance or delay of an angle.

[0004]

In this conventional meshing gear rotation error detecting device, the rotation error is specifically determined as follows. That is, in the rotation pulse signal output for one gear, one pulse of the rotation pulse signal output for the other gear is provided between the rising timing of a certain pulse and the rising timing of the next pulse. And the phase difference between the pulse of one gear and the pulse of the other gear on the assumption that the pulse that should be compared with the pulse of the one gear (hereinafter referred to as “comparative pulse”) always rises. Is determined based on the rotation error. That is, in the conventional meshing gear rotation error detecting device, a portion representing the rise of each rotation pulse in the rotation pulse signal is one mode of the “rotation signal”.

Therefore, as shown in the left part of the pulse waveform shown in the graph of FIG. 9, the rising timing of a certain pulse related to one gear (the driving gear in the figure) and the timing of the next pulse are described. If the pulse to be compared of the other gear (the driven gear in the figure) rises before the rising time of the pulse, the rotation error is correctly detected. However, due to a reason such as a large rotation error, as shown in the right part of the pulse waveform in the figure, the rising timing of one pulse related to one gear and the rising timing of the next pulse are related. If the comparison target pulse of the other gear does not rise during this time, a rotation error that is significantly different from the actual rotation error will be detected.

In short, in the conventional meshing gear rotation error detecting device, the comparison between the rotation signals is not always performed normally. For example, when the rotation error is considerably large, the rotation error can be accurately detected. There is a problem that it cannot be detected.

Against this background, the invention of claim 1 obtains a rotation signal of each gear in association with time,
The present invention has been made to solve the above-described problem by determining a rotation error using the rotation error.

According to a second aspect of the present invention, there is provided a mesh-type gear rotation error detecting device of the type in which a rotation error is determined based on a mutual relationship between absolute rotation angles of two gears at substantially the same time. An object of the present invention is to provide a preferred embodiment when applying the invention of Item 1.

[0009]

In order to solve the respective problems, the invention of claim 1 includes the gear rotation signal generating means 1 and the gear rotation error determining means 2 as shown in FIG. In the meshing type gear rotation error detecting device, the gear rotation error determining means 2 detects the generation time of each of the rotation signals for each of the two gears, thereby detecting the rotation of each gear from the reference rotation position. An absolute rotation angle, which is an angle, is discretely detected in association with time, and rotation errors of the gears are determined based on a mutual relationship between detection results of the respective gears.

The "gear rotation error determining means 2" here is based on, for example, the relationship between the two absolute rotation angles of the two gears taken at substantially the same time, and the advance / lag of the rotation angle of the gear. The amount may be determined as a rotation error. Further, it is also possible to adopt a mode in which the amount of advance / lag of the rotational angular velocity of the gears is determined as a rotational error based on the mutual relationship between the two relative rotational angles that the two gears take at substantially the same time. Here, the “relative rotation angle” is, for example,
It means the angle at which the gear rotates between the time when a rotation signal is generated one time and the time when the next rotation signal is generated.

According to a second aspect of the present invention, the "gear rotation error determining means 2" in the first aspect of the present invention includes substantially the same rotational angle among a plurality of absolute rotation angles detected for the two gears. For a time when two absolute rotation angles detected at the time exist, a rotation error is determined based on a relationship between the two detected absolute rotation angles, and at a time when the two absolute rotation angles do not exist, at least the rotation of the gears is determined. Estimating the absolute rotation angle that one is expected to take at that time based on the plurality of absolute rotation angles detected for the gear,
The rotation error is determined by using the estimated absolute rotation angle.

Here, the "gear rotation error determining means 2" can be, for example, in the following mode. That is, for each of the plurality of absolute rotation angles detected for one of the two gears, it is determined whether or not there is an absolute rotation angle detected for the other gear at substantially the same time, If present, the rotation error at that time is determined based on the relationship between the two existing absolute rotation angles, while if not present, the plurality of absolute rotations detected for the other gear are determined. Based on the angle, the absolute rotation angle expected to be taken by the other gear at that time is estimated, and based on the relationship between the estimated absolute rotation angle and the absolute rotation angle detected for the one gear, The rotation error at the timing can be determined.

The invention of claim 2 is merely an example of the embodiment of the invention of claim 1, and the invention of claim 1 can be embodied in other modes. For example, the "gear rotation error determination means 2" may be used to determine, for a plurality of timings detected for two gears, a timing at which substantially the same absolute rotation angle is detected for those gears.
The rotation error is determined based on the relationship between the existing timings, and for the absolute rotation angle that does not exist, the timing at which at least one of the gears is expected to take the absolute rotation angle is detected for the plurality of gears. , And the rotation error can be determined by using the estimated time.

[0014]

In the meshing type gear rotation error detecting device according to the first and second aspects of the present invention, each of the two meshing gears is rotated by a fixed angle by the gear rotation signal generating means. A rotation signal is generated discretely for each gear. Further, each time a rotation signal is generated by the gear rotation error determining means 2, the absolute rotation angle of each gear is detected discretely in association with time by detecting the generation timing of each rotation signal. As a result, the rotation error of the gears is determined based on the mutual relationship.

As described above, in the meshing type gear rotation error detecting device according to each of the first and second aspects of the present invention, the absolute rotation angle of each gear is discretely acquired in relation to time,
As a result, the absolute rotation angle of one gear and the absolute rotation angle of the other gear are obtained independently of each other via a common time axis. Therefore, the association of the absolute rotation angles between the two gears is always performed normally irrespective of the magnitude of the phase difference of the change situation of the absolute rotation angles, and as a result, the rotation error is always correctly determined. Becomes

In particular, in the meshing type gear rotation error detecting device according to the second aspect of the present invention, the gear rotation error detecting means 2 substantially includes the plurality of absolute rotation angles detected for the two gears. In the case where the absolute rotation angle detected at the same time exists, the rotation error is determined based on the relationship between the two detected absolute rotation angles. An absolute rotation angle that one is expected to take at that time is estimated based on the absolute rotation angle detected for the gear, and a rotation error is determined by using the estimated absolute rotation angle. Since the absolute rotation angle of each gear originally changes continuously with time,
Based on a plurality of actually detected absolute rotation angles, an absolute rotation angle that has not been actually detected can be estimated.

As described above, in the meshing type gear rotation error detecting device according to the second aspect of the present invention, although the absolute rotation angle of each gear is discretely detected with respect to time, calculation of each gear is performed. It is possible to obtain the absolute rotation angle substantially continuously with respect to time. Therefore, the absolute rotation angle between the two gears is always associated with high accuracy, and as a result, the rotation error is determined with high accuracy.

[0018]

As is apparent from the above description, according to the first and second aspects of the present invention, the relationship between the changes in the absolute rotation angle between the two gears is determined by the magnitude of the rotation error and the like. Irrespective of the fact that it is always obtained correctly, the rotation error of the gear can be always accurately detected.

In particular, according to the invention of claim 2, there is an effect that the rotation error of the gear can be accurately detected even though the absolute rotation angle of each gear is discretely detected with respect to time. can get.

[0020]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A gear rotation error detecting device of a single-toothed mesh type, which is an embodiment common to the first and second aspects of the present invention, will be described in detail with reference to the drawings.

This gear rotation error detection device detects the rotation state of a gear pair rotating while meshing, and detects a rotation error of the gear pair based on the detection result. Here "gear pair"
As shown in FIG. 2, the rotation axes are the gear pairs 10 and 12 that are parallel to each other. However, the present invention is not limited to this, and for example, the rotation axes may cross each other or may be different.

Although not shown, the gear rotation error detecting device includes a drive motor for rotating the gear 10 as a drive gear, and a braking motor for applying a braking torque to the gear 12 as a driven gear. The brake motor cooperates with the drive motor to apply a load (for example, the number of revolutions and the rotational force assumed in the gear use state) to the gear pairs 10 and 12. Incidentally, the reduction ratio of the gear pair 10, 12, i.e., the ratio of number of teeth Z ₁ of the drive wheel 10 of the number of teeth Z ₂ of the driven gear 12, i.e., the Z ₂ / Z ₁ and be represented by "I".

The gear rotation error detecting device also includes rotary encoders 20 and 22 for the driving gear 10 and the driven gear 12, respectively, as shown in FIG. As is well known, each of the rotary encoders 20 and 22 is a disk that rotates together with the rotating shaft of each gear 10 and has a plurality of fan-shaped slits formed at a constant angular pitch in the circumferential direction, and a fixing member. A light-emitting / light-receiving element for detecting the passage of each slit of the disc, and a rotation pulse signal that repeats that the amplitude rises from a steady state and becomes a transition state each time each slit passes, that is, the amplitude Output a signal (see FIG. 3) described by a train of pulses. That is, in the present embodiment, these two rotary encoders 20, 22 are one aspect of the "gear rotation signal generating means 1" in the first and second aspects of the present invention.

The gear rotation error detecting device further includes frequency dividing circuits 30, 32 for each of the rotary encoders 20, 22.
And clock pulse counters 40 and 42 and divided pulse counters 44 and 46, and further, a clock pulse generating circuit 50 and a computer 60 are commonly provided for both.

Each of the frequency dividing circuits 30 and 32 is connected to each of the rotary encoders 20 and 22 and divides the period of each of the rotation pulses input from each of the rotary encoders 20 and 22. It is converted into each divided pulse signal. On the other hand, each clock pulse counter 40,
Reference numeral 42 denotes a clock which is connected to the frequency dividing circuits 30 and 32 and the clock pulse generating circuit 50, and which is inputted from the clock pulse generating circuit 50 from the start of rotation of the gear pair 10, 12 to the start of each divided pulse. The number _NCK of pulses (see FIG. 3) is counted, and the result is input to the computer 60. Further, each divided pulse counter 44,
Reference numeral 46 denotes a gear pair 10, 12 from each of the frequency dividing circuits 30, 32.
The number N _DV of divided pulses (not shown) input from the rotation start timing of the above to the rising timing of each divided pulse is counted, and the result is input to the computer 60.

In synchronization with the start of rotation of the gear pairs 10 and 12, a trigger signal is input from a trigger signal generating circuit (not shown) to two clock pulse counters 40 and 42 and two divided pulse counters 44 and 46. Is done. Then, each of the clock pulse counters 40 and 42 receives the first clock pulse number _NCK from the clock pulse generation circuit 50 from the input timing of the trigger signal to the rising timing t _{(1) of the} first divided pulse. It is designed to count as the number of clock pulses. Each of the divided pulse counters 44 and 46 counts the first divided pulse number N _DV from the input timing of the trigger signal to the rising timing t _{(1) of the} first divided pulse.
It is designed to count as the number of divided pulses input from 0 and 32. Therefore, the gear pair 10,
12 rotation start timing, two rotary encoders 2
It is not essential to make an initial adjustment of the gear rotation error detection device so that the first rotation pulse rises simultaneously from 0 and 22.

As shown in FIG. 4, the computer 60 mainly includes a CPU 70, a ROM 72, a RAM 74, a clock 76, an input interface 78, and an output interface 80. The computer 60 uses the input interface 78 to
Is connected to the number of clock pulse counter 40, 42 and two divided pulse counter 44, based from which the clock pulse number N _CK and dividing the number of pulses N _DV is momentarily entered, CPU 70 is stored beforehand in ROM72 By running the program, the gear pair 1
The rotation errors ε of 0 and 12 are calculated and stored in the RAM 74.

The programs stored in the ROM 72 in advance include an absolute rotation angle calculation routine shown in the flowchart of FIG. 5 and a gear rotation error calculation routine shown in the flowchart of FIG. Hereinafter, the contents of these routines will be described individually.

First, the absolute rotation angle calculation routine shown in FIG. 5 will be described. First, a brief description will be given. In this routine, first, each clock pulse counter 40, 42
Every time the number of clock pulses N _{CK (k)} (k: the number of divided pulses) is input, the latest divided pulse output from each of the frequency dividing circuits 30 and 32 (hereinafter, “this divided pulse”)
₎ (See FIG. 7). Specifically, the rising timing t _(k) of the current divided pulse is calculated as the product of the number N _{CK (k)} of clock pulses and the period T _CK (fixed value) of the clock pulses. Next, in this routine, each of the gears 1 and 12 is rotated from the rotation start timing of each of the gears 10 and 12 (that is, the input timing of the trigger signal) to the rising timing t _{(k) of the} current divided pulse.
The absolute rotation angle θ _(k) at which 0 and 12 are rotated is calculated. Specifically, each divided pulse counter 44, 46
Is calculated as the product of the number of divided pulses N _{DV (k)} input from the above and the fixed minute angle Δφ of the rotary encoders 40 and 42.

That is, in the present embodiment, the portion representing the rise of each divided pulse in the divided pulse signal is:
This is one mode of the “rotation signal” in each of the first and second aspects of the invention.

It should be noted that, in this embodiment, the rotation start timing t of each pair of divided pulses and the absolute rotation angle θ at the rise timing t of each divided pulse are not limited. Is calculated on the basis of the number of clock pulses or divided pulses from the divided pulse.
When the number of clock pulses and the like generated between the time and the next divided pulse rising time t is sequentially acquired for each time, and the total value thereof is used to calculate the rising time t and the like of each divided pulse. In comparison, the accumulation of rounding errors due to the digital amount is avoided, and the unique effect of improving the accuracy of acquiring the rising timing t of each divided pulse is obtained. However, it is possible to implement the present invention in the latter mode.

Next, this routine will be described in detail with reference to FIG. This routine is executed for each of the gears 10 and 12. First, at step S,
1 (hereinafter referred to as “S1”; the same applies to other steps), the value of the divided pulse number k (hereinafter, “i” for the driving gear 10 and “j” for the driven gear 12) is changed. It is initialized to 1 and subsequently, in S2, it enters a state of waiting for an input request signal from each of the clock pulse counters 40 and 42. If so, the determination is YES, and in S3, the current number of clock pulses _{NCK (k)} is input from the clock pulse counters 40 and 42 that have requested input.

Subsequently, in S4, the product of the input clock pulse number N _{CK (k)} and the clock pulse period T _CK is calculated as the rising timing t _(k) of the current divided pulse, and is stored in the RAM 74. It is stored (see FIG. 8). Thereafter, in S5, each divided pulse counter 44, 46
, The number of division pulses N _{DV (k) of} this time is input, and
In S6, the current absolute rotation angle θ _{(k) of} each of the gears 10 and 12 is calculated based on the current divided pulse number N _{DV (k)} . The calculated absolute rotation angle θ _(k) of this time is associated with the rising time t _(k) of the current divided pulse,
It is stored in M74 (see FIG. 8). Subsequently, in S7, the value of the number k of the divided pulse is incremented by one. Thereafter, in S8, it is determined from the current value of the number k of the divided pulse whether or not the execution of this routine may be terminated. If the execution should be continued, the determination is NO, and the process returns to S2. If the process can be ended, the determination is YE
At S, the execution of this routine ends.

Therefore, when the execution of this routine is completed, the driving gear 10 and the driven gear 12
The data shown in FIG. 8 is stored in the RAM 74, and the absolute rotation angle θ is discretely detected for each of the gears 10 and 12 in association with the time t. Here, “time t” means a time on a time axis common to the gears 10 and 12, whereas “timing t _{1 (i)} ” in FIG. Angle θ _{1 (i)}
, And “timing t _{2 (j)} ” means the timing of detecting each absolute rotation angle θ _{2 (j)} of the driven gear 12. FIG. 3 is a graph showing the relationship between the time t and the absolute rotation angle θ for each of the gears 10 and 12.

Next, the gear rotation error calculation routine of FIG. 6 will be described. First, a brief description will be given. In this routine, based on the mutual relationship between the absolute rotation angles θ ₁ and θ ₂ obtained at substantially the same time (hereinafter simply referred to as “simultaneous period”) t for each gear 10, 12, the gear pair 10, 1
2, the rotation error ε at that time t is calculated.

If the gears 10 and 12 have the same diameter, that is, if the reduction ratio I is 1, the difference between the absolute rotation angles θ ₁ and θ ₂ obtained at the same time is immediately ε, that is, the amount of advance or delay of the rotation angle of the gear pair 10, 12. However, if the reduction ratio I is not 1,
The situation is different. Because the pitch circle radii are different between the two gears 10 and 12, even if the actual rotation error ε is 0, the difference between the absolute rotation angles θ ₁ and θ ₂ obtained at the same time is 0. Because it does not become.

Therefore, in the present embodiment, the rotation error ε is calculated as follows. That is, first, by multiplying the reduction ratio I in absolute rotational angle theta ₁ of the drive wheel 10, the absolute rotation angle theta ₁ is driven gear when the drive gear 10 is assumed to have been replaced in the same diameter as the driven gear 12 10
Is converted to the expected value. Then, by subtracting the converted value of the absolute rotation angle θ ₁ of the driving gear 10 (ie, I · θ ₁ ) from the absolute rotation angle θ ₂ of the driven gear 12, the rotation error ε is calculated, and as a result, the rotation error ε is calculated using the equation θ ₂ −I · θ ₁ .

That is, in the present embodiment, the relationship represented by this equation is one aspect of the "relationship between the detection results of the respective gears" in the first aspect of the present invention, and the second aspect of the present invention. Is an aspect of the “relationship between two absolute rotation angles”.

However, the absolute rotation angles θ ₁ and θ ₂ obtained for the gears 10 and 12 at the same time are not always present. Therefore, in this routine, the rotation error ε is determined as follows.

That is, for each of the plurality of absolute rotation angles θ ₁ detected for the driving gear 10, the absolute rotation angle θ ₂ detected for the driven gear 12 at the same time as the timing t _{1 (i)} when it is detected. Is determined, and if so, the timing t _{1 (i)} is determined based on the mutual relationship between the two existing absolute rotation angles θ _{1 (i)} and θ ₂ as described above. rotation error epsilon _(i) is determined in. On the other hand, when the driven gear 1 does not exist, the driven gear _{1 is set at} the time t _{1 (i)} based on the plurality of absolute rotation angles θ ₂ detected for the driven gear 12.
2 'is estimated, the estimated absolute rotation angle theta _2' absolute rotational angle theta ₂ which it is expected to take on the basis of the absolute rotational angle theta _{1 (i)} mutual relationship between the drive gear 10, the rotational error ε _(i)
Is determined.

In FIG. 8, each data is a time t common to the gears 10 and 12 from the upper side to the lower side of the figure.
When the time t _{1 (i)} and the time t _{2 (j)} coincide with each other, the time t _{1 (i)}
And the timing t2 _(j) are described on the same horizontal line, while when they do not match, they are respectively described on different horizontal lines.

The "absolute rotation angle theta ₂ of the estimated" here, in the present embodiment, specifically, is carried out as follows.
That is, first, in the portion corresponding to the driven gear 12 in the storage area of the RAM 74, immediately before the position corresponding to the time t1 _(i) in the absence of the above ₍ time t1 _(i)).
Retrieved timing t _2F is positioned in the traveling direction), and will be loaded absolute rotation angle theta _2F corresponding thereto, further, the traveling direction immediately rearward (time t positions should correspond to the time t _(i) is The timing t _2R located in the opposite direction) is searched, and the corresponding absolute rotation angle θ _2R is read. Moreover, by their timing t _2F and timing t _2R is internally divides the absolute rotation angle theta _2F and theta _2R at the same ratio as that is internally divided by the time t _{1 (i),} at that time t _{1 (i)} An absolute rotation angle θ ₂ ′ expected to be taken by the driven gear 12 is estimated. The estimated absolute rotation angle θ ₂ ′ is also stored in the RAM 74.

This estimation will be described more specifically based on the example of FIG. In the example of the drawing, there is no absolute rotation angle theta ₂ of the driven gear 12 corresponding to the timing t _{1 (2).} Therefore, first, the search time t _{2 (3)} as the time t _2F, the absolute rotation angle theta _{2 (3)} is read as an absolute rotational angle theta _2F, further search period t _{2 (2)} is a timing t _2R Then, the absolute rotation angle θ _{2 (2)} is read as the absolute rotation angle θ _2R . further,
The absolute rotation angles θ _{2 (2)} and θ _{2 (3)} are internally divided at the same ratio as that at which the timings t _{2 (2)} and t _{2 (3)} are internally divided by the timing t _{1 (2)} . As a result, the driven gear 1 at time t _{1 (2)}
Absolute rotational angle theta ₂ that 2 takes the expected ₍₂₎ 'is estimated.

[0044] In this routine, by substituting the equation an absolute rotation angle theta ₂ 'estimated in the above manner each time obtained by t _{1 (i),} the timing t _{1 (i)}
Rotation error epsilon _(i) is calculated in.

In this embodiment, the absolute rotation angle θ not actually detected is estimated by performing linear interpolation on the actually detected absolute rotation angle θ. For example, it can be estimated by a technique such as quadratic interpolation.

Further, in the present routine, since the plurality of rotation errors ε _(i) thus obtained are considered to be a combination of a static error and a dynamic error, the rotation errors ε _(i) By performing a low-pass filter process on, a static error is extracted, and by performing a high-pass filter process, a dynamic error is also extracted. Here, “static error” represents the absolute value of the average elastic strain of each gear 10, 12 (that is, the amount of elastic deformation of each gear 10, 12 due to the load), while “dynamic error” represents Each gear 1
It is believed to represent a mesh transmission error of 0,12.

Next, this routine will be described in detail with reference to FIG. This routine is started after the execution of the absolute rotation angle calculation routine is completed.

First, in S101, the number i of the divided pulse of the driving gear 10 is initialized to 1, and subsequently, in S10
2, the RAM 74 is associated with the current time t _{1 (i).}
The absolute rotation angle θ _{1 (i) of the} drive gear 10 stored in the storage device is read. Thereafter, at S103, the time t
_{1 (i)} , the absolute rotation angle θ ₂ detected for the driven gear 12
Is stored in the RAM 74. If it is stored, the determination becomes YES and S104
In, read the corresponding absolute rotation angle theta ₂ is from RAM 74, subsequently, in S105, therewith by substituting the absolute rotation angle theta _{1 (i)} and the said formula, timing t
The rotation error ε _{(i) at 1 (i)} is calculated. The calculated rotation error ε _(i) is stored in the RAM 74 in association with the timing t _{1 (i)} . Thereafter, in S106, the value of the divided pulse number i is incremented by one, and in S107, it is determined from the current value of the divided pulse number i whether or not the calculation of all the plurality of rotation errors ε to be obtained has been completed. You. Assuming that it has not been completed yet,
And returns to S102.

On the other hand, at time t _{1 (i)} , the driven gear 12
When the absolute detected rotation angle theta ₂ is not stored in the RAM74 For, S103 the determination is NO,
The process proceeds to S108 and subsequent steps. In S108, the absolute rotation angles θ _2F and θ _{2R of the} driven gear 12 stored in the RAM 74 in association with the times t _2F and t _2R before and after the time t _{1 (i)} are read. Then, S10
9, the timings t _2F , t _2R and their existing 2
Based on the absolute rotation angles θ _2F and θ _2R , as described above, the absolute rotation angle θ ₂ that does not exist, that is, the timing t _{1 (i)}
, An absolute rotation angle θ ₂ ′ expected to be taken by the driven gear 12 is estimated. Subsequently, in S110, the estimated absolute rotation angle θ ₂ ′ is set to the absolute rotation angle θ ₂ for convenience, and then the process proceeds to the steps after S105.

As a result of performing the above process many times, if all the plurality of rotation errors ε to be obtained have been calculated,
The determination in S107 is YES, and in S111, R
The high-pass filter processing and the low-pass filter processing are respectively performed on the plurality of rotation errors ε stored in the AM 74, whereby the mesh transmission error and the rigidity (the amount of elastic deformation) of the gear pairs 10 and 12 are calculated, respectively. It is stored in the RAM 74. This is the end of the execution of this routine.

When the execution of this routine is completed, the relationship between the rotation error ε and the time t is obtained as shown in the graph of FIG. 3, and the relationship between the meshing transmission error and the time t is obtained. Is obtained as represented by a graph in FIG.

[0052] As apparent from the above description, in this embodiment, independently of one another absolute rotation angle theta ₁ and the absolute rotation angle theta ₂ of the driven gear 12 via a common time axis to those of the drive gear 10 Thus, the rotation error ε can be detected between the driving gear 10 and the driven gear 12 without directly comparing the pulses. Therefore, it is possible to avoid the situation where the detection accuracy of the rotation error ε is reduced due to the pulse condition.

Further, in this embodiment, the driven gear 1
Absolute rotational angle theta ₂ of 2 Despite being discretely detected with respect to time t, calculated on, as required absolute rotational angle theta ₂ of the driven gear 12, substantially continuously obtained for time t Therefore, it is possible to compare the absolute rotation angles θ to be truly compared between the gears 10 and 12, and it is possible to detect the rotation error ε with sufficiently high accuracy. Can be

Further, in this embodiment, since the fluctuation component and the absolute component of the rotation error ε of each of the gears 10 and 12 can be individually extracted, the extraction result can be used. The mesh transmission error and the stiffness (elastic strain) of the pairs 10 and 12 can be individually evaluated, and the characteristic effect that the characteristics of the gear pairs 10 and 12 can be evaluated over a wider range can be obtained. Can be

As is clear from the above description, in the present embodiment, the frequency dividing circuits 30, 32, the clock pulse counters 40, 42, the divided pulse counters 44, 46, the clock pulse generating circuit 50 and the computer 60 cooperate with each other. Thus, it constitutes one aspect of the "gear rotation error determining means 2" in each of the first and second aspects of the present invention.

As described above, one embodiment common to the first and second aspects of the present invention has been described in detail with reference to the drawings. However, each of these aspects can be embodied in other modes.

For example, in the above embodiment, it is determined whether or not the absolute rotation angle θ ₂ for the driven gear 12 was not actually detected at the same time as the absolute rotation angle θ ₁ actually detected for the drive gear 10. If it is determined and not detected, the absolute rotation angle θ ₂ of the driven gear 12 is estimated. For example, each of the absolute rotation angles θ actually detected for the driven gear 12 _It is also determined whether or not the absolute rotation angle θ ₁ of the drive gear 10 was actually detected at the same time as the period _2, and if not detected, the absolute rotation angle θ ₁ of the driven gear 10 is also estimated. That can be done. That is, the inventions of claims 1 and 2 can be carried out such that the estimation of the absolute rotation angle θ is performed not only for the driven gear 12 but also for the driving gear 10.

In the above embodiment, the gear pair 1
The rotation error is detected by rotating the 0 and 12 in a load state, and the rotation error is detected. However, the present invention is an embodiment in which the rotation error is detected by rotating in a quasi-static state (no load and low speed state). Can also be implemented.

In addition to these, without departing from the scope of the claims, various inventions can be carried out in various modifications and improvements based on the knowledge of those skilled in the art.

[Brief description of the drawings]

FIG. 1 is a diagram conceptually showing a configuration of each of the first and second aspects of the present invention.

FIG. 2 is a system diagram showing a gear rotation error detecting device of a single-tooth-surface meshing type which is an embodiment common to the first and second aspects of the present invention.

FIG. 3 is a diagram for explaining divided pulses output from a frequency dividing circuit in FIG. 2;

FIG. 4 is a diagram conceptually showing a configuration of a computer in FIG.

FIG. 5 is a flowchart illustrating an absolute rotation angle calculation routine stored in a ROM of the computer in FIG. 4;

FIG. 6 is a flowchart showing a gear rotation error calculation routine stored in the ROM.

FIG. 7 is a diagram conceptually showing a configuration of a part of a RAM in FIG. 4;

FIG. 8 is a graph conceptually showing a flow of signal processing in the embodiment.

FIG. 9 is a diagram for explaining how pulse signals are compared in a conventional meshing gear rotation error detecting device.

[Explanation of symbols]

 Reference Signs List 10 driving gear 12 driven gear 20, 22 rotary encoder 30, 32 frequency dividing circuit 40, 42 clock pulse counter 50 clock pulse generating circuit 60 computer

Claims (2)

(57) [Claims] 1. A gear rotation signal generating means for discretely generating a rotation signal for each gear when each of two meshing gears rotates by a predetermined angle, and each time a rotation signal is generated individually for each of the gears, By detecting the occurrence timing, the absolute rotation angle, which is the rotation angle of each gear from the reference rotation position, is discretely detected in relation to time, and the rotation of each gear is determined based on the mutual relationship between the detection results of each gear. And a gear rotation error determining means for determining an error. 2. The gear rotation error determination means according to claim 2, wherein, among the plurality of absolute rotation angles detected for the two gears, two absolute rotation angles detected for the gears at substantially the same time are present. For the period of time, the rotation error is determined based on the relationship between the two detected absolute rotation angles, and for the period of non-existence, at least one of the gears is expected to take that period. 2. The meshing gear rotation error detection according to claim 1, wherein the rotation angle is estimated based on a plurality of absolute rotation angles detected for the gear, and the rotation error is determined by using the estimated absolute rotation angle. apparatus.