JP2006085383A - Control parameter setting method for control circuit in measurement control system, and measuring instrument - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a gain setting method for a control circuit by which a gain for accurately preventing the occurrence of hunting in a measurement control system can be set highly precisely and measured stably and rapidly. <P>SOLUTION: N pieces of gain G<SB>i</SB>are successively temporarily set in a control circuit 23, and a stylus 131 is brought into contact with an object W to be measured so as to be temporarily measured. A sensor detection signal to be outputted by a sensor detection circuit 21 is filtered by a filter 31, and only frequency components corresponding to the frequency of hunting to be generated in a closed loop L including the control circuit 23 are extracted. A magnitude S<SB>i</SB>of the frequency components is compared with a reference value S<SB>0</SB>, and the gains G<SB>i</SB>for establishing S<SB>i</SB><S<SB>0</SB>, and for preventing hunting from being generated in the closed loop L is extracted, and the maximum G<SB>i</SB>among those respective Gi is set in the control circuit 23 from the point of view of the improvement of the quick responsiveness of measurement or the like. Therefore, it is possible to stably and quickly perform measurement based on this set gain. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a control parameter setting method and a measuring apparatus for a control circuit in a measurement control system. More specifically, the present invention relates to a method and a measuring apparatus for setting optimal control parameters of a control circuit for stable and quick measurement.

The control parameters of the control circuit include a gain and a phase compensation frequency. Conventionally, as a gain setting method of the control circuit in the measurement control system, a data acquisition process for acquiring data necessary for gain setting, and based on this data. In addition, there is known a method including a determination step for determining an optimum gain for measurement and a setting step for setting a gain of a control circuit to a gain determined to be optimal (see, for example, Patent Document 1). Here, the optimal gain of the control circuit refers to a gain that can prevent a measurement control system from entering a self-oscillation state to realize a stable measurement state and can perform a quick measurement. And
Various methods can be used to determine the optimum gain, and Patent Literature 1 discloses the following methods (a) and (b).

(A) A measurer manually inputs a set of data necessary for gain setting (probe material, sample material, etc.), and compares it with a data table prepared in advance. In the data table, a plurality of sets of reference input data and values of the respective reference gains (such as proportional operation parameters PP) obtained empirically as optimum gains for the respective sets of reference input data are represented. . The measurer compares the actual input data set manually input with a plurality of reference input data sets, selects the nearest reference input data set, and selects the reference input data set. Is set in the control circuit as an optimum gain.
(B) The probe is fixed at one point on the object to be measured, temporary measurement is performed while changing the gain of the control circuit (proportional operation parameter PP or the like), and a current error amount ΔE is calculated for each gain. ΔE represents the degree of variation of the tunnel current value flowing between the probe and the sample with respect to the target current value. Since the tunnel current value changes so as to oscillate around the target current value in the measurement control state, when ΔE is large, the amplitude around the target current value is large, and the measurement control system is in a self-oscillation state (so-called hunting occurs). When ΔE is small, the amplitude around the target current value is small, the tunnel current value is substantially coincident with the target current value, and the measurement control system becomes stable. It is suggested that The determination of the optimum gain of the control circuit is performed from the viewpoint of adjusting the measurement control system to a stable state, and a gain value that realizes a small value of ΔE is determined as the optimum gain. Specifically, the value of ΔE for each gain is compared with a predetermined reference value, and a gain such that ΔE is equal to or less than this reference value is determined as the optimum gain.

JP-A-5-223519 (3rd to 6th pages, FIGS. 1 to 4, 7, and 8)

The method (a) has a problem that the burden on the measurer becomes heavy because it is necessary to precisely perform preliminary measurement for creating the data table. In addition, since the reference gain corresponding to the reference input data that most closely approximates the actual input data set is set as the optimum gain, this optimum gain is only an approximate value. For this reason, this method cannot be used when high accuracy is required in gain setting, such as during ultra-precise measurement.
In the method (b), the optimum gain of the control circuit is determined using the value of ΔE as an index from the viewpoint of preventing hunting from occurring in the measurement control system. However, the value of ΔE is merely a numerical value indicating the degree of variation in the tunnel current value, and is not a value directly related to the presence or absence of hunting. Therefore, just because ΔE is small, it cannot always be said that hunting has not occurred, and hunting may not be sufficiently prevented even if the optimum gain is set by the method (b).

  An object of the present invention is to provide a control parameter setting method for a control circuit in a measurement control system, which can set a control parameter for accurately preventing occurrence of hunting in a measurement control system with high accuracy and can perform measurement stably and quickly. And providing a measuring device.

A control parameter setting method for a control circuit in a measurement control system according to the present invention is based on measurement means for outputting a measurement signal in connection with measurement of an object to be measured, and a deviation between a predetermined target value and an output value of the measurement signal. A control circuit in a measurement control system comprising: a control circuit that outputs a control signal; and a control unit that controls the measurement unit so that an output value of the measurement signal matches the target value based on the control signal In the control parameter setting method, the control parameter of the control circuit is temporarily set to Q _i (i = 1, 2,..., N), and temporary measurement is performed. provisional measure temporarily measurement data acquisition step of acquiring for a predetermined time data S _i, different N types of control parameters to each other _{(Q 1, Q 2, ···} , Q N) After sequentially performed on, it is acquired based And performs frequency analysis on each of said temporary measurement data S _i, extraction step size of the frequency component corresponding to a predetermined vibration frequency in the measurement control system to extract the temporary measurement data S _i which is less than a predetermined reference value And a control parameter setting step of setting, as the control parameter of the control circuit, Q _i having the maximum frequency characteristic among the control parameters Q _i corresponding to the extracted temporary measurement data S _i. It is characterized by.

  In the measurement control system of the present invention, a closed loop is constituted by the measurement means, the control circuit, and the control means, and the measurement state is controlled by feedback control. The control parameters of the control circuit are variable, and the control parameters of the entire closed loop can be adjusted thereby. Hereinafter, control parameter setting of the control circuit will be described.

<Temporary measurement data acquisition process>
First, the control parameters of the control circuit are temporarily set to N types of control parameters Q _{1 to} Q _N in order, and temporary measurement data S _i is obtained for each control parameter Q _i . Here, the provisional measurement data S _i is obtained by recording a predetermined measurement amount f _i based on a measurement signal from the measurement means at each time t, and is constituted by a plurality of sets of (f _i , t). Is done. If this is plotted on a coordinate plane and each point is smoothly connected, a function f _i (t) for the measurement time t is constructed. Incidentally, if to continuously record the measured amount f _i for a predetermined time in a temporary measure, f _i constitutes directly a function f _i a _(t).
By the way, since the feedback control is performed so that the output value of the measurement signal from the measurement unit coincides with the target value even in the temporary measurement, the value of the function f _i (t) based on the measurement signal is the target value. Usually, it changes with time while vibrating around the target position corresponding to the value. Here, if the measurement control system (closed-loop) has not occurred stable hunting, amplitude around the target position of the f i _(t) is attenuated with time t, f i _(t) is asymptotic to a target position . On the other hand, if the measurement control system if occurs is unstable and the hunting, f i _(t) without amplitude is attenuated over time t around the target position of, f i _(t) is above a certain Continue to vibrate while maintaining the amplitude.

Here, the relationship between hunting in the measurement control system and the control parameters of the measurement control system will be outlined. In the present invention, the control parameter of the control circuit is variable. Here, gain will be described as an example of a control parameter. The gain of the measurement control system is uniquely determined by the gain of the control circuit. Therefore, hereinafter, the gain will be described as the gain of the control circuit. FIG. 1 is a Bode diagram for the frequency transfer function of the control circuit. 1A is a gain diagram, and FIG. 1B is a phase diagram.
FIG. 1A shows gain characteristics when different gains G ₁ , G ₂ , and G ₃ (G ₁ <G ₂ <G ₃ ) are set in the control circuit. In this description, it is assumed that only changes in gain are considered, and changes in phase are not taken into consideration. For this reason, the phase diagram of FIG. 1B does not change. That is, one phase line in FIG. 1B is a common phase line for the _three gains G _{1 to} G ₃ . In FIG. 1A, the gain lines corresponding to the gains G _{1 to} G ₃ are drawn substantially parallel to each other, but this is for simplification of explanation, and the actual gain is determined for each frequency. It can be changed with different rate of change and can be set flexibly.

For determination of the stability of the measurement control system, that is, whether hunting occurs or not, as is generally known, the gain crossover frequency ω _{c at} which the gain value is 1 (0 on the gain diagram). And a phase crossover frequency ω ₀ at which the phase φ becomes −180 ° is used.
When the gain is G ₁ , ω _c1 <ω ₀ as shown in FIG. Therefore, when the phase is −180 ° or less, the gain is always 1 or less (minus on the gain diagram), hunting does not occur, and the measurement control system is stable.
When the gain is increased from G ₁ to G ₂ , ω _c2 = ω ₀ . This is a so-called stability limit state, and hunting occurs immediately when the gain is further increased.
That represents a state in which hunting occurs a gain curve for the gain G _3. At this time, ω _c3 > ω ₀ . In the frequency band of ω ₀ ≦ ω ≦ ω _c3 , the phase is −180 ° or less and the gain is 1 or more (plus on the gain diagram). Therefore, hunting of frequencies belonging to this frequency band (hereinafter referred to as self-oscillation frequency band) occurs, and the measurement control system becomes unstable.

From the above, it can be seen that the hunting frequency in the measurement control system is a frequency belonging to the self-oscillation frequency band ω ₀ ≦ ω ≦ ω _c3 . Here, since the lower limit ω ₀ is unchanged, the self-oscillation frequency band is uniquely determined by setting the gain G ₃ (> G ₂ ) and determining the upper limit ω _c3 . The vibration frequency in the present invention mainly means a frequency belonging to this self-oscillation frequency band.

Next, a case where the control circuit includes a phase compensation circuit and the characteristics of the phase diagram can be changed will be described.
In FIG. 1C, the phase compensation circuit is a phase lead compensation circuit, and the phase crossover frequency is ω ₀₁ (phase characteristic PH ₁ ), ω ₀₂ (phase characteristic PH ₂ ), ω ₀₃ (phase characteristic) by the phase lead compensation. An example that can be changed is shown as PH ₃ ). Also in this case, the generation principle of hunting is the same, and ω _C ≦ ω ₀ is a stable condition in the relationship between the gain crossing frequency ω _C and the phase crossing frequency ω ₀ . The combination of these gain characteristics and phase characteristics may be variable independently of each other, or may be variable while maintaining the correlation. For example, relationships such as control parameters Q ₁ (G ₁ , PH ₁ ), Q ₂ (G ₂ , PH ₂ ), Q ₃ (G ₃ , PH ₃ ), Q ₄ (G ₄ , PH ₄ ),. Or control parameters Q ₁ (G ₁ , PH ₁ ), Q ₂ (G ₁ , PH ₂ ), Q ₃ (G ₂ , PH ₁ ), Q ₄ (G ₂ , PH ₂ ),... Such a relationship may be used. In any case, the control parameter that satisfies the stability condition and maximizes the gain crossover frequency ω _C gives the maximum frequency characteristic of the measurement control system. Although ω _C ≦ ω ₀ is shown as the stability condition, it is preferable that the phase delay is about −150 deg at ω _C = 0 in consideration of stability. Furthermore, in this description, the case where the phase compensation circuit is a phase lead compensation circuit is shown, but the present invention is not limited to this, and a phase lag compensation circuit may be used, or a phase lead compensation circuit and a phase lag compensation circuit may be used in combination. The principle is the same. In addition, although the example in which the phase compensation circuit is included in the control circuit has been shown, the present invention is not limited thereto, and the phase compensation circuit may be inserted into the feedback circuit of the sensor detection signal.

As another example of the vibration frequency in the measurement control system, there is a resonance frequency ω _r of an electric circuit constituting the measurement control system. When a measurement signal containing a resonance frequency component of ω _r or more is input to this electric circuit, the resonance frequency component is hardly attenuated, so that the measurement signal continues to be oscillated at the resonance frequency ω _r. . In particular, when the gain of the measurement control system at the resonant frequency omega _r is 1 or more, the resonance frequency component from being amplified amplitude is increased, to destabilize the measurement control system. This is exactly the same as the situation for the self-oscillation frequency band. That is, even if the measurement signal approaches the target value by the control by the measurement control system, the measurement signal continues to vibrate around the target value because the vibration of ω _r remains.

<Extraction process>
Now, following the provisional measurement data acquisition process, an extraction process is performed. The extraction step performs frequency analysis for each temporarily measurement data S _i. That is, each function f _i (t) in each temporary measurement data S _i is Fourier-transformed and displayed as a frequency, and the magnitude of the frequency component corresponding to the vibration frequency is compared with a predetermined reference value. Since the vibration frequency is a frequency at which hunting occurs, if the frequency component corresponding to this is large, hunting has occurred. Conversely, if the frequency component is small, the occurrence of hunting is appropriately prevented. The reference value is set in advance as an optimum value for determining the occurrence / non-occurrence of hunting. The reference value may be set as a different value for each vibration frequency in the self-oscillation frequency band.
Through the above frequency analysis, M (≦ N) pieces of provisional measurement data S _j (j = 1, 2,..., M) are extracted so that the magnitude of the vibration frequency component is less than the reference value.

<Control parameter setting process>
The extracted temporary measurement data S _j is a data group that ensures that hunting does not occur and the measurement control system is in a stable state. Therefore, if the control parameters Q _j corresponding to these S _j are set in the control circuit, the stability of the measurement control system is ensured.
As long as the stability can be secured, for example, the gain may be set as large as possible in order to increase the steady-state characteristics, the quick response (responsiveness), and the like. In the control parameter setting step, among the control parameters Q _j that guarantee the stability of the measurement control system, the parameter having the maximum frequency characteristic is set as the control parameter of the control circuit.

As described above, by using the control parameter setting method of the present invention, it is possible to set control parameters in the control circuit that can realize high steady-state characteristics, rapid response, etc. while ensuring stability in the measurement control system. Stable, highly accurate and quick.
In the present invention, the current error amount ΔE in the method (b) of Patent Document 1 does not focus on the amount that is only indirectly related to the occurrence of hunting, but directly to the occurrence of hunting. Since the optimum control parameters are set by paying attention to the magnitude of the related vibration frequency component of the measurement signal, the occurrence of hunting can be prevented directly and accurately, and a high degree of stability can be realized.

  As described above, in the present invention, it is preferable that the control parameter of the control circuit is at least one of a gain and a phase compensation frequency.

In the present invention, in the provisional measurement data acquisition step for each control parameter Q _i , the provisional measurement data is obtained by filtering the measurement signal with a frequency filter that passes only a frequency component corresponding to the vibration frequency. are S _i, it is preferable.

According to this configuration, the function f _i (t) in each S _i has only the vibration frequency component that has passed through the frequency filter. Therefore, since it is not necessary to perform the Fourier transform of f _i (t), the optimum gain setting for bringing the measurement control system into a stable state can be performed more quickly and easily.

In the present invention, the storage step of storing the control parameter Q of the entire measurement control system determined by the control parameter Q _i set in the control circuit in the control parameter setting step, the measuring means, and the measured device The control parameter of the control circuit is corrected so as to cancel the apparent change in the control parameter of the measuring means caused by replacing at least one of the objects with a different property, and the control parameter of the entire measurement control system It is preferable to carry out a control parameter correction step for holding Q at the stored Q.

Since the measurement signal output from the measurement means is also changed by the properties of the measurement means and the object to be measured, such as hardness, elasticity, etc., when at least one of the measurement means and the object to be measured is replaced with one having a different property, The control parameter of the measuring means is apparently changed. Then, since the control parameter of the entire measurement control system deviates from the optimum control parameter Q set in the control parameter setting step of the control circuit, the measurement stability may be hindered.
In the present invention, first, the optimum control parameter Q of the entire measurement control system is stored in the storing step. Thereafter, when the control parameter of the measuring means is apparently changed by exchanging at least one of the measuring means and the object to be measured, a control parameter correcting step is performed, and the control parameter of the control circuit is corrected to correct the entire measurement control system. The control parameter is held in the stored optimum control parameter Q. Therefore, according to the present invention, the control parameters of the entire measurement control system can be held at the optimum control parameters regardless of the measurement means and the object to be measured, and the measurement can be performed stably and with high accuracy.

  Further, in the present invention, the measuring means is a contact measuring element that performs measurement while being in contact with the object to be measured, and the control parameter correction step is performed after replacement of at least one of the contact measuring element and the object to be measured. The contact measuring element is pressed into the object to be measured by a predetermined amount to perform contact measurement, and at the same time, a preliminary measurement step for measuring the indentation amount, an output value of the measurement signal in the preliminary measurement step, and a measurement A calculation step of calculating an apparent change of the control parameter in the contact probe from the relationship between the pressed amount and the control amount of the measurement control system as a whole based on the calculation result in the calculation step It is preferable to correct the control parameter of the control circuit so that the parameter is held at Q.

In this invention, since the measurement load received by the contact probe differs depending on the difference in properties such as hardness and elasticity of the contact probe and the object to be measured, at least one of the contact probe and the object to be measured has a different property. When they are exchanged, a change occurs in the output value of the measurement signal output from the contact probe, and an apparent change occurs in the control parameter of the contact probe.
In the preliminary measurement step of the present invention, the contact measuring element is pushed into the object to be measured by a predetermined amount to perform contact measurement, and the output value of the measurement signal and the push-in amount at this time are detected. In the subsequent calculation step, an apparent change in the control parameter of the contact probe is calculated using the correlation between the output value of the measurement signal and the push amount. In the final control parameter correction step, the control parameter of the control circuit is corrected so that the apparent change of the control parameter of the contact probe is canceled and the control parameter of the entire measurement control system is held in the optimal control parameter Q. The
As described above, according to the present invention, the entire measurement control system is held at the optimum control parameter Q regardless of the contact probe and the object to be measured, so that the measurement can be performed stably and accurately.

Further, according to the present invention, it is possible to configure a measuring apparatus including a measurement control system including a control circuit in which a control parameter is set by the control parameter setting method of the control circuit.
According to this measuring apparatus, measurement can be performed stably, with high accuracy, and quickly.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 2 shows a measurement apparatus according to the present embodiment, and FIG. 3 shows a block diagram schematically showing feedback control in the measurement apparatus.
In FIG. 2, a sensor support 12 is attached to the main body 1 of the measurement apparatus via an actuator 11, and a sensor 13 as a measurement means involved in the measurement of the workpiece W is attached to the sensor support 12.
The main body 1 can be moved in a three-dimensional direction with respect to the workpiece W by the driving device 1A. Thereby, it is possible to perform measurement while moving the sensor 13 along the surface of the workpiece W. Here, the displacement sensor 14 attached to the main body 1 is configured such that when the main body 1 and the object to be measured W are relatively moved, the vertical distance between them (Z-axis direction) is determined in a non-contact manner, for example, optical, electrostatic capacitance. It is a detector that detects by the formula.
The actuator 11 serving as a control means is a braking element that can be expanded and contracted based on a drive signal from a drive circuit 24 to be described later, whereby the sensor 13 approaches / separates along the Z direction with respect to the object W to be measured. Positioning control is performed.
That is, the sensor 13 positioned by driving the main body 1 is further finely positioned by the actuator 11, so that the measurement accuracy is improved.
The sensor 13 as a contact measuring element includes a stylus 131 that is constantly vibrated in the axial direction. When the stylus 131 is in contact with the workpiece W, the amplitude of the stylus 131 is reduced according to the contact load. The sensor 13 outputs a sensor signal as a measurement signal in accordance with the amount of change in the amplitude of the stylus 131, and the measurement state is controlled based on this sensor signal. It is assumed that the amplitude of the stylus 131 and the output value of the sensor signal are linearly changed with respect to the measurement load received by the stylus 131.

The sensor signal from the sensor 13 is subjected to signal processing in the sensor detection circuit 21 and output as a sensor detection signal. This sensor detection signal also constitutes a measurement signal in the present invention. As shown in FIG. 3, the sensor detection signal is branched, and one of the sensor detection signals is input as a feedback signal to a closed loop L for performing feedback control, and the other is a gain (control parameter Q) of the control circuit 23 described later. Is input to the control circuit gain adjustment unit 3 for performing the adjustment.
The sensor detection signal as the measurement signal input to the closed loop L is compared with the target value 22, and the deviation between the two is input to the control circuit 23. The control circuit 23 amplifies the deviation based on a gain G _C (described later) adjusted by the control circuit gain adjustment unit 3 and inputs the amplified deviation to the drive circuit 24 as a control signal. The drive circuit 24 outputs a drive control signal based on the input control signal toward at least one of the actuator 11 and the drive device 1A, and the actuator 11 is expanded or contracted accordingly, or the main body 1 is driven in the Z-axis direction. Is done. Then, the sensor 13 is displaced in the Z-axis direction with respect to the workpiece W and the measurement load between them is changed, so that the sensor signal from the sensor 13 is changed. Thus, the actuator 11 and the drive device 1A constitute the control means of the present invention that controls the sensor 13 based on the drive control signal.
As described above, the closed loop L for controlling the measurement state is configured in the order of the sensor 13 -the sensor detection circuit 21 -the control circuit 23 -the drive circuit 24 -the actuator 11 and the drive device 1A -the sensor 13. The measurement is performed under feedback control so that the sensor detection signal from the sensor detection circuit 21 matches the target value 22.

Subsequently, a measurement method using the measurement apparatus of the present embodiment will be described.
First, by driving the main body 1 in the −Z-axis direction (downward in FIG. 2) by the driving device 1A, the stylus 131 of the sensor 13 is pushed into the object W to be measured by a predetermined amount, and then the main body 1 is stopped. . Thereafter, the movement of the main body 1 in the Z-axis direction is limited, and the vertical distance (Z-axis direction) between the main body 1 and the object W to be measured is kept constant.
When the stylus 131 is pushed into the workpiece W, the amplitude (Z-axis direction) of the stylus 131 is changed according to the measurement load. For example, since the measurement load received by the stylus 131 increases as the push amount increases, the amplitude of the stylus 131 decreases.

Information regarding the amplitude of the stylus 131 is output from the sensor 13 as a sensor signal, and then processed by the sensor detection circuit 21 to be a sensor detection signal and compared with the target value 22. That is, the target value 22 defines a target value of the amplitude of the stylus 131 (hereinafter referred to as target amplitude), and control is performed so that the amplitude of the stylus 131 becomes equal to the target amplitude during measurement.
More specifically, when the amplitude of the stylus 131 becomes smaller than the target amplitude, the actuator 13 is contracted to drive the sensor 13 in the + Z-axis direction, thereby reducing the measurement load applied to the stylus 131 and reducing the amplitude. Increase to target amplitude. On the contrary, when the amplitude of the stylus 131 becomes larger than the target amplitude, the sensor 13 is driven in the −Z-axis direction by extending the actuator 11 to increase the measurement load applied to the stylus 131, and the amplitude is set to the target amplitude. Reduce to.

  Thus, under feedback control by the closed loop L, the amplitude of the stylus 131 is matched with the target amplitude. That the amplitude of the stylus 131 is constant means that the vertical distance (Z-axis direction) between the sensor 13 and the workpiece W is constant. Therefore, when the main body 1 is scanned in the XY directions in this state, the sensor 13 is displaced in the Z-axis direction following the unevenness of the surface of the workpiece W by driving the actuator 11. If this Z-axis displacement is detected by an appropriate displacement detecting means, the unevenness on the surface of the object to be measured W can be detected, so that the measurement of the surface of the object to be measured W can be performed.

The scanning measurement has been described above. However, according to the measurement apparatus of the present embodiment, it is possible to perform the touch measurement of the measurement object.
For example, after the expansion / contraction state of the actuator 11 is fixed to the maximum extension state, the main body 1 is driven in the −Z-axis direction until the amplitude of the stylus 131 is reduced to the target amplitude by the measurement load from the object W to be measured. Then, the main body 1 is brought closer to the object to be measured W. When the amplitude of the stylus 131 matches the target amplitude, the -Z axis drive of the main body 1 is stopped, and the three-dimensional coordinates (X, Y, Z) of the main body 1 at that time are represented by a three-dimensional displacement sensor (not shown). To detect (touch measurement). By performing this at each measurement point on the workpiece W, the shape of the workpiece W can be measured.

  The touch measurement can also be performed by extending and contracting the actuator 11 after fixing the Z-axis position of the main body 1. That is, the contact state between the stylus 131 and the object to be measured W is adjusted by controlling the expansion and contraction of the actuator 11. The expansion / contraction is stopped when the amplitude of the stylus 131 matches the target amplitude, and the expansion / contraction amount (Z-axis direction) of the actuator 11 and the XY coordinates of the main body 1 at this time are detected (touch measurement). By performing this at each measurement point on the workpiece W, the shape of the workpiece W can be measured.

  As described above, the scanning measurement and the touch measurement have been described as examples of the measurement using the measurement apparatus of the present embodiment. However, as a premise for appropriately performing these measurements, it is necessary to appropriately perform feedback control using the closed loop L. When the control state becomes unstable, hunting occurs in the closed loop L, and the output value of the sensor detection signal from the sensor detection circuit 21 vibrates greatly around the target value 22. This means that the amplitude value of the stylus 131 vibrates around the target amplitude value with an amplitude that cannot be ignored. Since both the scanning measurement and the touch measurement described above are based on the premise that the amplitude of the stylus 131 can match the target amplitude, if hunting occurs, the measurement cannot be performed properly, and the measurement accuracy may be significantly deteriorated. .

In order to suppress the occurrence of hunting and appropriately perform feedback control by the closed loop L, it is necessary to adjust the gain G of the closed loop L to an optimum value. As shown in FIG. 3, the gains of the control circuit 23, the drive circuit 24, the actuator 11 or the drive device 1A, the sensor 13, and the sensor detection circuit 21 are set to G _C , G _AD , G _A , G _S , G _{SD, respectively.} Then, the gain G of the entire closed loop L is expressed by the following formula 1.

In Equation 1, only the gain G _C of the control circuit 23 is adjustable by the control circuit gain adjuster 3. Therefore, in order to stably and accurately perform measurement optimally adjust the gain G of the closed loop L, it is necessary to set the gain G _C optimally. Hereinafter, detailed setting of the _{G C.}

First, after fixing the XY position of the main body 1, the main body 1 is driven in the Z-axis direction and stopped when the stylus 131 is pushed into the workpiece W by a predetermined amount. Thereafter, in this step, the XYZ position of the main body 1 is fixed at this stop position.
At this time, the feedback control by the closed loop L is not started, and the actuator 11 is fixed in a fixed expansion / contraction state, for example, the maximum expansion state. Since both the main body 1 and the actuator 11 are fixed, the contact load between the stylus 131 and the workpiece W is constant, so the amplitude of the stylus 131 is also constant. Now, the description will be continued assuming that this constant amplitude (hereinafter referred to as initial amplitude) is smaller than the target amplitude defined by the target value 22. The following description is almost the same even when the initial amplitude is larger than the target amplitude.

From this state, feedback control by the closed loop L is started, and temporary measurement is performed. Temporarily measurement, the control circuit gain adjuster each N different gains which are sequentially provisionally set in the control circuit 23 by the _{3 G i (i = 1,2,3,} ···, N) is performed for each. That is, first, after temporarily setting to one gain, the control is started, and the sensor detection signal output from the sensor detection circuit 21 at that time is taken into the control circuit gain adjustment unit 3 for a predetermined time. Thereafter, the control is stopped, and after changing to another temporarily set gain, the control is resumed, and the sensor detection signal is taken into the control circuit gain adjusting unit 3. Hereinafter, for N provisionally set the gain G _i everything incorporation of the sensor detection signal is completed, repeat the same operation.

FIG. 4 is a graph showing the relationship between the sensor detection signal output value f taken into the control circuit gain adjustment unit 3 and the temporary measurement time t. In this figure, C _j generally indicates a curve when the temporarily set gain of the control circuit 23 is appropriate, and C _k generally indicates a curve when the temporarily set gain is inappropriate. Here, t = 0, the provisional measurement start point, i.e., represents the time when the feedback control by the closed loop L for each temporary setting gains G _i is started. f ₀ indicates the output value of the sensor detection signal when the stylus 131 is at the initial amplitude before the feedback control is started, and f _t indicates the target value 22 of the sensor detection signal. At the time of t = 0, the amplitude of the stylus 131 is constant regardless of the temporary setting the gain G _i of the control circuit 23 (the initial amplitude), the output value of the sensor detection signal also is constant (f _0), The starting points of C _j and C _k are both (0, f ₀ ).

In C _j , since the temporarily set gain (hereinafter referred to as G _j ) is appropriate, the sensor detection signal f gradually approaches the target value f _t . In this remains a predetermined time has elapsed, since substantially coincides with f and f _t as shown in FIG. 4, according to the provisional setting gain G _j, can stably hold the control state of the closed loop L, stable and accurate measurement Can be done.
On the other hand, at C _k , since the temporarily set gain (hereinafter referred to as G _k ) is inappropriate, hunting occurs, and the sensor detection signal f continues to vibrate around the target value f _t . Even after the elapse of this state a predetermined time, do not amplitude of hunting is hardly attenuated as shown in FIG. 4, there is no possibility that the f and f _t are constantly coincide. Thus, according to the provisional setting gain G _k, becomes unstable control state of the closed loop L, it is impossible to properly perform the measurement.
Here, according to the present embodiment, the provisional measurement data acquisition step and the extraction step does not cause hunting among the N different gains G _i (C _j see) the gain G _{j (j} = 1,2, · .., M) and a gain G _k (k = 1, 2,..., NM) that causes hunting (see C _k ) can be distinguished. Hereinafter, both these steps will be described.

<Temporary measurement data acquisition process>
Sensor detection signals for each temporary setting gain G _i for the control circuit gain adjuster 3 taken over a predetermined time is filtered by the frequency filter 31.
The filter 31 is configured to pass only the frequency component corresponding to the frequency at which hunting occurs (vibration frequency) in the sensor detection signal (see FIG. 4) whose waveform is displayed over time. Therefore, the larger the component that has passed through the filter 31, the larger the hunting that occurs in the closed loop L, and the more unstable the control state is. As will be described later, the magnitude of the component passing through the filter 31 is compared with a predetermined reference value, and if it is greater than or equal to the reference value, it is determined that hunting has occurred, and if it is less than the reference value, hunting has occurred. It is determined that it is not. Then, one gain satisfying a predetermined requirement is set in the control circuit 23 as the optimum gain among the temporarily set gains G _j giving data determined that hunting has not occurred.
As described above, the magnitude of the frequency component that passes through the filter 31 is an important index for setting an optimum gain (maximum frequency characteristic) that does not cause hunting. It is necessary to set a frequency band strictly and pass only a desired frequency component. Hereinafter, this setting will be described.

As described above with reference to FIG. 1, the self-oscillation frequency band in which hunting mainly occurs is obtained by using the phase crossover frequency ω ₀ and the gain crossover frequency ω _{C of the} entire closed loop L, and ω ₀ ≦ ω ≦ ω _C. (However, only when ω ₀ ≦ ω _C ). In the present embodiment, the phase is fixed in consideration of the focus on the gain adjustment of the closed loop L. Therefore, ω ₀ as the lower limit of the self-oscillation frequency band is a common value for each temporarily set gain G _i . On the other hand, omega _C as the upper limit of the self-oscillating frequency band is different value for each temporary setting gain G _i. As shown in FIG. 1, the larger the gain value, the larger the value of ω _{C and} the wider the self-oscillation frequency band.
Incidentally, the hunting amplitude of a frequency belonging to the self-oscillation frequency band has a positive correlation with the magnitude of the gain (amplification factor) at that frequency. As shown in FIG. 1, the gain of the control system is generally configured to monotonously decrease with respect to the frequency. Therefore, in the self-oscillation frequency band, near the lower limit (ω ₀ ). Is greater than the gain near the upper limit (ω _C ). Therefore, the hunting amplitude of the frequency belonging to the self-oscillation frequency band is large in a frequency band near the lower limit ω ₀ (hereinafter referred to as a main vibration frequency band), ω ₀ ≦ ω ≦ ω ₀ + Δω (<ω _C ). Thus, it can be said that this large-amplitude hunting is the main factor that makes the measurement control state unstable.
Therefore, the passable frequency band of the filter 31 is set to coincide with the main vibration frequency band. According to the filter 31 set in this way, only the frequency component corresponding to the main hunting can be passed and analyzed, so that the optimum gain that does not cause hunting can be easily determined. It is assumed that an optimum value is set in advance as appropriate for the width Δω of the passable frequency band of the filter 31 set as described above.

The output value S _i of the sensor detection signal for each temporary set gain G _i filtered by the filter 31 as described above is stored in the memory 32 as temporary measurement data of the present invention based on the sensor detection signal as the measurement signal. Is done.

<Extraction process>
Subsequently, each S _i is compared with a predetermined reference value S ₀ set in advance. The reference value S ₀ is set in advance as an optimum value for determining the presence or absence of hunting in the closed loop L, and hunting occurs for the temporarily set gain G _{k where} S _k ≧ S _0, and S _j <S ₀ That is, no hunting occurs for G _j . In the extraction process, only the temporarily set gain G _j (S _j <S ₀ is satisfied) that does not cause hunting is extracted and stored in the memory 33.

<Gain setting process>
In the subsequent gain setting step, the extracted G _j having the maximum value is selected and set to the gain (G _C ) of the control circuit 23. Here, the maximum G _j is used in order to increase the steady-state characteristics, speed response, etc. of the closed loop L, that is, to maximize the frequency characteristics of the measurement control system.

As described above, according to the gain setting method of the present embodiment, the stability in the closed loop L is obtained by extracting the gain G _j (j = 1, 2,..., M (≦ N)) that does not cause hunting. By selecting the gain G _j having the maximum value while ensuring the above, high steady-state characteristics, rapid response, etc. can be realized. Therefore, measurement can be performed stably, with high accuracy, and quickly.

<Storage process>
Hereinafter, in the gain setting step, a gain set in the control circuit 23 as the optimum gain and G _C. The gain of the entire closed loop L at this time is G. G is stored in the memory 41 of the control circuit gain correction unit 4 as the optimum gain of the entire closed loop L.

<Gain correction process>
Next, gain correction will be described.
Replacing at least one of the stylus 131 and the workpiece W in a different one of characteristics, the gain G _S of the sensor 13 is changed apparently in FIG.
For example, when the object to be measured W is changed from a soft material to a hard material, the measurement load that the stylus 131 receives from the object W increases. As a result, input from the actuator 11 or the driving device 1A (extension / contraction of the actuator 11). The rate of change of the amplitude of the stylus 131 with respect to the amount or the amount of drive in the Z-axis direction of the drive device 1A) increases, and as a result, the rate of change of the sensor signal from the sensor 13 increases. This means that the gain G _S of the sensor 13 (= output value of the sensor signal / input from the actuator 11 or the driving device 1A) is apparently changed.
Here, it is assumed that the apparently changed gain of the sensor 13 is G _S ′, and the gain of the entire closed loop L calculated by substituting this G _S ′ into Equation 1 is G ′. Since G ′ ≠ G, the gain of the closed loop L at this time is deviated from the optimum gain, and if measurement is performed as it is, problems may occur in terms of stability, steady-state characteristics, quick response, and the like. Therefore, it is necessary to correct the gain of the closed loop L from G ′ to the optimum gain G so as to cancel the apparent change in the gain of the sensor 13. In this embodiment, the control circuit gain correction unit 4 adjusts the gain of the control circuit 23 to adjust to the optimum gain. Details will be described below.

<Preliminary measurement process>
After exchanging at least one of the stylus 131 and the workpiece W, the XY coordinates of the main body 1 are fixed, the main body 1 is driven in the Z-axis direction, and the stylus 131 is pushed into a point on the workpiece W. Perform contact measurements. At this time, the vertical distance (Z-axis direction) between the displacement sensor 14 and the workpiece W is changed, and a displacement signal corresponding to the amount of change (push-in amount) is output from the displacement sensor 14. This displacement signal is input to the arithmetic circuit 42 in the control circuit gain correction unit 4 together with the sensor signal from the sensor 13.

<Calculation process>
The arithmetic circuit 42 calculates the gain G _S ′ of the sensor 13 that is apparently changed based on these two signals.
The calculation of G _S ′ is performed as shown in FIG. 5, for example. 5A and 5B, (i) the sensor 13 (main body 1) is brought close to the workpiece (workpiece) W at a constant speed, and (ii) the stylus 131 is moved to the workpiece W. And (iii) a graph representing the relationship between time and displacement signal, and a graph representing the relationship between time and sensor signal when the sensor 13 is moved away from the workpiece W at the same speed. FIG. 5C is a graph showing the relationship between the displacement signal and the sensor signal based on (A) and (B).
5A and 5B, the stylus 131 and the workpiece W are in contact with each other in the time region of t1 ≦ t ≦ t2, and the sensor signal is changed according to the amount of pressing of the stylus 131. However, this is represented as a straight line portion having a constant inclination in FIG. By appropriately taking the coordinate origin in FIG. 5C, the extension line of this straight line portion can pass through the origin, so that the inclination is expressed by a sensor signal / displacement signal. Here, since the displacement signal corresponds to an input from the driving device 1A to the sensor 13 (drive of the main body 1 in the Z-axis direction), the inclination = an output from the sensor 13 / an input to the sensor 13, Is equal to the gain G _S ′ of the sensor 13.

Further, when the arithmetic circuit 42 substitutes the calculated G _S ′ for the right side of the equation 1, the left side is held in the optimum gain G (stored in the memory 41) of the closed loop L. The value of the gain G _C ′ is calculated. Specifically, the correction gain G _C ′ of the control circuit 23 is calculated as a unique value that satisfies the following formula 2 including G _S ′ and the optimum gain G.

The control circuit gain correction unit 4 sets G _C ′ calculated as described above in the control circuit 23. Accordingly, even when at least one of the stylus 131 and the workpiece W is replaced with one having a different property, the gain of the entire closed loop L can be held at the optimum gain G, so that the measurement stability and steady state characteristics can be maintained. High speed response can be maintained.

It should be noted that the present invention is not limited to the above-described embodiments, and modifications, improvements, and the like within the scope that can achieve the object of the present invention are included in the present invention.
For example, in the embodiment, as passable frequency band of the filter 31, but sets the common principal vibration frequency band _{(ω 0 ≦ ω ≦ ω 0} + Δω) in the temporary setting gain G _i, in the present invention, passable frequency band of the N different for each temporary setting gains G _i may be set. For example, the self-oscillation frequency band determined for each temporary setting gains G _i a _{(ω 0 ≦ ω ≦ ω C} ), if passable frequency band for each temporary setting gain G _i, to accurately prevent the occurrence of hunting The gain can be set in the control circuit 23. The switching of the N passable frequency band cross may be performed by adjusting the passable frequency band of one filter, also self each passable frequency band of each temporarily set the gain G _i N filters that are equal to the oscillation frequency band are prepared, and the filter to be used may be switched according to the temporarily set gain.

In the above embodiment, the passable frequency band (ω ₀ ≦ ω ≦ ω ₀ + Δω) of the filter 31 is set by paying attention only to the self-oscillation frequency band (ω ₀ ≦ ω ≦ ω _C ) defined previously. However, if there is a frequency (or frequency band) that causes hunting in the closed loop L in addition to the self-oscillation frequency band, the passable frequency band of the filter 31 is set so as to be included. Can do.
For example, in addition to the frequency belonging to the self-oscillation frequency band, the filter 31 may be used so that the resonance frequency ω _r of the electric circuit constituting the closed loop L is also a passable frequency. When ω _r is different for each temporarily set gain G _i , N filters with a passable frequency set to match each ω _r are adjusted by adjusting the passable frequency of one filter or N filters. prepared, by switching them in response to the G _i, it can be extracted by filtering the frequency component corresponding to the omega _r.

In the above embodiment, the gain correction when the sensor 13 having the stylus 131 is used as the measurement unit has been described. However, the gain correction method according to the above embodiment can be used even when the measurement unit is a non-contact probe. Applicable.
For example, when an optical probe that receives and measures reflected light from an object irradiated with light is used as a non-contact type probe, a difference in optical properties such as reflectance of the object to be measured Accordingly, the amount of light received by the optical probe changes, and the measurement signal output from the optical probe changes. For this reason, when the object to be measured is replaced with one having a different optical property, the gain of the optical probe may change apparently, and the closed loop gain of feedback control may deviate from the optimum gain. According to the present invention, the apparent change in the gain of the optical probe can be calculated from the change in the amount of light received by the optical probe accompanying the exchange of the object to be measured, and the gain of the entire closed loop can be maintained at the optimum gain. The gain of the control circuit can be calculated and set. Therefore, it is possible to maintain high stability of measurement, steady characteristics, quick response, etc. regardless of the object to be measured.

  In the above embodiment, the sensor detection signal at the time of temporary measurement is filtered by the filter 31. However, in the present invention, the filter is not always necessary, and it is also optimal by performing frequency analysis directly on the sensor detection signal. Gain can be determined. Specifically, the sensor detection signal is Fourier-transformed and displayed as a frequency, and only the frequency component corresponding to the vibration frequency is extracted from this, and this is compared with a predetermined reference value. When this vibration frequency component is smaller than the reference value, it is determined that hunting has not occurred in the closed loop, and if the temporarily set gain at this time is set in the control circuit, the measurement can be performed stably.

  Further, in the above embodiment, a so-called vibration probe having a stylus 131 that is constantly vibrated is used as a measurement means. Can be used. For example, an optical sensor using a CCD, an interferometer, a capacitance sensor, an electromagnetic induction sensor, an acoustic sensor using ultrasonic waves, a hinge-type probe having a strain gauge, or the like can be employed.

In the above embodiment, in the preliminary measurement process and the calculation process, the main body 1 is driven in the Z-axis direction, and the gain G _S ′ of the sensor 13 is calculated using the displacement signal. the body 1 by stretching the actuator 11 instead of driving the Z-axis direction by driving the sensor 13 in the Z-axis direction, G _S by utilizing the expansion and contraction amount of the actuator 11 at that time, and a sensor signal from the sensor 13 'May be calculated.

  In the above embodiment, the gain setting by the control circuit gain adjustment unit 3 and the gain correction by the control circuit gain correction unit 4 are combined. However, in the present invention, it is not always necessary to combine them. May be used. In particular, when the control circuit gain correction unit 4 is used alone, the optimum gain of the measurement control system (closed loop) is calculated in advance by some method, and at least one of the measurement means and the device under test is exchanged. When the gain of the measuring means changes, the control circuit gain correction unit 4 is used to correct the gain of the control circuit 23 so that the gain of the entire measurement control system is held at the optimum gain.

Furthermore, in the above-described embodiment, an example in which only the gain G of the control circuit 23 is made variable as the control parameter Q has been shown. However, the present invention is not limited to this, and the control circuit 23 is provided with a phase compensation circuit. The phase characteristic PH may be made variable so that the phase crossover frequency ω ₀ of the closed loop L can be changed. For example, if a phase advance compensation circuit is provided as a phase compensation circuit, the frequency characteristics of the measurement control system can be further improved. Also, by providing a phase lag compensation circuit as the phase compensation circuit, for example, harmful phase fluctuations near the phase crossover frequency ω ₀ can be reduced, and the stability of the measurement control system can be improved.

  The present invention can be used for various measuring devices, for example, a three-dimensional measuring machine, an atomic force microscope, and a focused displacement measuring device.

The Bode diagram about the frequency transfer function of the control circuit in the present invention. The figure which shows the measuring apparatus concerning embodiment of this invention. The figure which shows the closed loop which performs feedback control in the measuring apparatus concerning the said embodiment. The graph showing the relationship between the output value of the sensor detection signal taken into the control circuit gain adjustment part, and temporary measurement time in the case of temporary measurement using the measuring apparatus concerning the said embodiment. The figure which shows the method of calculating the apparent change of the gain in a sensor using a displacement signal and a sensor signal in the said embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Main body 1A ... Drive device 3 ... Control circuit gain adjustment part 4 ... Control circuit gain correction part 11 ... Actuator 13 ... Sensor 14 ... Displacement sensor 21 ... Sensor detection circuit 22 ... Target value 23 ... Control circuit 31 ... Filter 131 ... Stylus L ... Closed loop W ... DUT

Claims (6)

A measuring means for outputting a measurement signal in connection with the measurement of the object to be measured;
A control circuit that outputs a control signal based on a deviation between a predetermined target value and an output value of the measurement signal;
Control means for controlling the measurement means based on the control signal so that the output value of the measurement signal matches the target value;
In the control parameter setting method of the control circuit in the measurement control system comprising:
Temporary measurement is performed after temporarily setting the control parameter of the control circuit to Q _i (i = 1, 2,..., N), and temporary measurement data S _i based on the measurement signal output at that time is obtained. After performing the temporary measurement data acquisition process acquired over a predetermined time sequentially for N different control parameters (Q ₁ , Q ₂ ,..., Q _N ),
Perform frequency analysis obtained the each temporary measurement data S _i, the magnitude of the frequency component corresponding to a predetermined vibration frequency in the measurement control system to extract the temporary measurement data S _i which is less than a predetermined reference value An extraction process;
A control parameter setting step of setting, as a control parameter of the control circuit, Q _i having the maximum frequency characteristic among the control parameters Q _i corresponding to the extracted temporary measurement data S _i ;
A control parameter setting method for a control circuit. In the control parameter setting method of the control circuit according to claim 1,
The control parameter of the control circuit is at least one of a gain and a phase compensation frequency.
A control parameter setting method for a control circuit. In the control parameter setting method of the control circuit according to claim 1 or 2,
In the provisional measurement data acquisition step for each control parameter Q _i , the provisional measurement data S _i is obtained by filtering the measurement signal with a frequency filter that passes only a frequency component corresponding to the vibration frequency.
A control parameter setting method for a control circuit. In the control parameter setting method of the control circuit according to any one of claims 1 to 3,
A storage step of storing the control parameter Q of the entire measurement control system determined by the control parameter Q _i set in the control circuit in the control parameter setting step;
Correcting the control parameter of the control circuit so as to cancel the apparent change in the control parameter of the measuring means caused by replacing at least one of the measuring means and the object to be measured with a different property; A control parameter correction step of holding the control parameters of the entire measurement control system in the stored Q;
A control parameter setting method for a control circuit. In the control parameter setting method of the control circuit according to claim 4,
The measuring means is a contact measuring element that performs measurement by contacting the object to be measured,
The control parameter correction step includes
After exchanging at least one of the contact measuring element and the object to be measured, a preliminary measurement step of performing a contact measurement by pressing the contact measuring element into the object to be measured by a predetermined amount and simultaneously measuring the indentation amount. When,
From the relationship between the output value of the measurement signal in the preliminary measurement step and the measured push amount, an arithmetic step of calculating an apparent change of the control parameter in the contact measuring element,
Based on the calculation result in this calculation step, the control parameter of the control circuit is corrected so that the control parameter of the entire measurement control system is held in the Q.
A control parameter setting method for a control circuit.   6. A measuring apparatus including a measurement control system including a control circuit in which control parameters are set by the control parameter setting method for a control circuit according to claim 1.

Images