JP4700485B2 - Arithmetic apparatus and test apparatus - Google Patents

Description

  The present invention relates to an arithmetic device that calculates a phase difference and an amplitude of a response wave and an input wave when a frequency sweep wave (sweep wave) is applied to an object, and a test device using the arithmetic device.

このｆsを１／２周期（０〜πまで）積分した値Ｆsは、次の式（２）のようになる。

2. Description of the Related Art Conventionally, it is known to calculate a phase difference and an amplitude of a response signal when an input signal having a constant frequency is applied to an object by calculation.
That is, when the input signal (input wave) is a sine wave sin x having an amplitude of 1, the response signal (response wave) f (x) is generally represented by Asin (x + δ). A is the amplitude of the response signal, and δ is the phase difference with respect to the input wave.
Here, a value fs obtained by multiplying the response wave f (x) by the same waveform sinx as that of the input wave is expressed by the following equation (1).

A value Fs obtained by integrating fs by 1/2 period (from 0 to π) is expressed by the following equation (2).

このｆcを１／２周期（０〜πまで）積分したＦcは、次の式（４）となる。

Similarly, a value fc obtained by multiplying the response wave f (x) by a waveform cos x obtained by shifting the phase of the input wave sinx by 90 ° is expressed by the following equation (3).

Fc obtained by integrating this fc by 1/2 period (from 0 to π) is expressed by the following equation (4).

なお、上記においては積分期間を１／２周期（０→π）としたが、積分期間を１／２周期の正の整数倍の期間としたときも、同様にして振幅Ａと位相差δを求めることができる。 From the above equations (2) and (4), the amplitude A of the response wave and the phase difference δ between the input wave and the input wave are obtained as follows.

In the above description, the integration period is ½ cycle (0 → π). However, when the integration period is a positive integer multiple of ½ cycle, the amplitude A and the phase difference δ are similarly set. Can be sought.

Thus, the response wave f (x) = Asin (x + δ) when the input wave sinx having a constant frequency is applied.
Of the response wave f (x) multiplied by the same waveform sinx as the input wave, Fs obtained by integrating the response wave f (x) with a positive integer multiple of ½ period, and the response wave f (x). Fc obtained by integrating fc obtained by multiplying the input wave by a waveform cosx whose phase is 90 ° different from that of the input wave by a positive integer multiple of ½ cycle can be obtained by calculation from the Fs and Fc.
When digital processing is performed, sample data corresponding to fs and fc using sample data of an input wave at each sampling timing, sample data of a signal whose phase differs from the input wave by 90 °, and sample data of a response wave Is calculated, and the values corresponding to Fs and Fc are obtained by accumulating the sample values of fs and fc in the period corresponding to 1/2 period (or a positive integer multiple thereof), and the response is obtained by the above equation (5). The wave amplitude A and the phase difference δ may be calculated.

  Such arithmetic processing is also used, for example, when controlling the amplitude and phase difference of each axis when performing a tensile / compression test by repeatedly applying a load to the specimen in a multi-axis testing machine. It has been. Since the amplitude and phase difference of the response wave can be quickly obtained by calculation, feedback control can be performed by correcting the amplitude and phase of the input wave to each axis.

In a material testing machine, a peak detection test for obtaining a resonance point by applying a sweep wave (a signal whose amplitude is constant and a frequency continuously changes) to a specimen is performed.
Also, in a multi-axis testing machine, when a tension / compression test is performed by repeatedly applying a load to the specimen while controlling the amplitude and phase difference of each axis (for example, 6 axes), loading a sweep wave Is required.
As described above, when an input wave having a constant frequency is used, the amplitude (A) and the phase difference (δ) of the response wave can be obtained by calculation.
However, when the input wave is a sweep wave whose frequency changes continuously with time and the general formula described above is used in digital processing, the sampling interval (Δt) is adjusted in accordance with the change of the frequency of the sweep wave. It is necessary to keep the amount of change in the phase angle at each sampling timing constant, but it is difficult to process this with software, and special hardware is required.
Therefore, in the past, when detecting the phase of a sweep wave whose frequency changes, a special method that can output and measure a fixed number of sample data at a fixed phase angle in one cycle is a method for guaranteeing phase detection accuracy due to changes in frequency components. It was done with a configuration with simple hardware.
However, in this case, since the sampling interval (Δt) is not constant, there is a practical problem such that the collected test data cannot be used for other analysis (for example, FFT analysis).

When the special hardware as described above is not used, the processing is performed by the step sweep method in which the frequency is changed stepwise.
In the step sweep method, a frequency component is given for each constant frequency, and thus no frequency component is given during the frequency interval.
FIG. 4 is a diagram illustrating an example of a result when a peak detection test is performed by the step sweep method, where the horizontal axis indicates the frequency and the vertical axis indicates the gain. As shown in this figure, in the case of the step sweep method, the frequency at which the gain reaches a peak may not be directly detected.
For this reason, in the case of the step sweep method, it is necessary to shorten the frequency interval so that an important frequency is not lost. Also, there is no guarantee that all frequencies will be given even if the frequency range is narrowed.
In addition, if you want to control the phase between multi-axis shakers with the step sweep method, it is necessary to synchronize the timing of measurement / control and frequency switching. You have to shake it. When controlling in such a manner, if a condition for sweeping within a certain time is given as a test condition, the frequency interval cannot be made more than a certain value in order to finish the sweep within this time. There is. For the above reasons, it is necessary to perform a continuous sweep in a test under such conditions.

  Therefore, the present invention uses an arithmetic device capable of detecting a phase difference and an amplitude of a response wave with respect to an input wave by calculation when the sweep wave whose frequency continuously changes with time is used as the input wave, and the arithmetic device. The purpose is to provide a test apparatus.

In order to achieve the above object, the arithmetic device of the present invention calculates a phase difference between a response signal and an input signal when an input signal, which is a sine wave signal having a constant amplitude and a swept frequency, is applied to an object. An arithmetic device for calculating, wherein the sampling data fsin of the input signal, the sampling data fcos of a signal whose phase differs from the input signal by 90 °, and the sampling data fo of the response signal at each sampling timing set at a fixed time interval And the phase difference change data Δθ of the input signal are inputted, and the product of the response signal sample data fo, the input signal sample data fsin, and the phase angle change amount data Δθ of the input signal is calculated. A first multiplying unit, a first adding unit for accumulating the calculation result (fo × fsin × Δθ) of the first multiplying unit over a period of a positive integer multiple of ½ cycle, the response Second multiplying means for calculating the product of the sample data fo of the signal and the sample data fcos of the signal whose phase is 90 ° different from the input signal and the data Δθ of the change amount of the phase angle of the input signal, the second multiplying means Of the calculation result (fo × fcos × Δθ) over a period of a positive integer multiple of ½ cycle, and the calculation result of the second addition unit and the first addition unit Computation means for calculating the phase difference δ between the input signal and the response signal based on the ratio of the computation results is provided.
And a means for calculating the amplitude A of the response signal by calculating the square root of the sum of the square of the calculation result of the first addition means and the square of the calculation result of the second addition means. It is.

Further, the test apparatus of the present invention is a test apparatus using the above-mentioned arithmetic unit, which performs a test by loading sweep waves having different phases on a plurality of loading points of a specimen, and each of the test apparatuses has a constant time interval. Sweep wave generating means for outputting sample data of the sweep wave, sample data of a signal that is 90 ° out of phase with the sweep wave, and change amount data of the phase angle of the sweep wave at a sampling timing, and the plurality of loading points A plurality of phase control means for controlling the phase of the sweep wave output from the sweep wave generation means, and a plurality of phase control means provided corresponding to the plurality of phase control means, respectively. A plurality of loading means for loading the sweep wave output from the corresponding loading point, and a plurality of loading points corresponding to the plurality of loading points, respectively. Wherein a plurality of detecting means for outputting a sample data of the response wave indicating the motion of the specimen, a plurality of operation means provided corresponding to said plurality of detection means that, each operation means, the corresponding Sample data of the response wave output from the detection means, sample data of the sweep wave from the sweep wave generation means, sample data of a signal that is 90 ° out of phase with the sweep wave, and amount of change in the phase angle of the sweep wave And a plurality of calculation means that are the calculation devices for calculating the phase difference of the response wave output from the detection means corresponding to the sweep wave, and the plurality of phase control means, respectively, Based on the phase difference calculated by the corresponding computing means, the phase of the sweep wave supplied to the corresponding loading means is controlled.

According to the arithmetic device and the test device of the present invention, the phase difference and amplitude of the response wave when sampling is performed at a constant time interval can be detected, so that the sampling period is changed in synchronization with the sweep. Therefore, the hardware can be configured only by the D / A converter and the A / D converter, and all can be processed only by the software.
In addition, since the phase detection accuracy can be guaranteed without using special hardware for guaranteeing the phase detection accuracy, no special hardware is required and the cost can be reduced.
Further, since the controller can perform arithmetic processing simultaneously with sampling, the phase and amplitude can be detected at high speed, phase control can be performed at high speed, and frequency characteristic analysis can be performed at high speed.
Furthermore, since the waveform data is collected at a constant sampling interval, the data can be used for other analysis as it is.

FIG. 1A is a functional block diagram showing a system using the arithmetic device of the present invention, and FIG. 1B is a functional block diagram showing a configuration of the arithmetic device of the present invention. The arithmetic device of the present invention can actually be realized by software by a central processing unit (CPU).
In FIG. 1A, 1 is a function generator (sweep wave generator) that generates a sine wave (sweep wave) whose amplitude is constant and the frequency continuously changes with time, and 2 is the application of the sweep wave. For example, a specimen (test piece) loaded with the sweep wave. Reference numeral 3 denotes an arithmetic unit according to the present invention, which includes sample data of a sweep wave (input wave) output from the sweep wave generator 1 and sample data of a response wave output from a sensor attached to the object 2. To calculate the amplitude of the response wave and the phase difference with respect to the input wave.
The sweep wave generator 1 includes, for example, a waveform memory that stores waveform sample data of a sine wave sin θ, and reads and outputs a waveform sample corresponding to the phase at each sampling timing from the waveform memory. In addition to the sample data fsin of the sweep wave (input wave) to be applied to the object 2, the sample data fcos of a signal whose phase is 90 ° different from the input wave and the change amount Δθ of the phase angle of the input wave at each sampling timing Is configured to output. Here, the time interval (sampling interval) between each sampling timing is fixed.
The input wave sample data fsin output from the sweep wave generator 1 is supplied to the object 2 via a D / A converter, an actuator, or the like (not shown). Further, the sample data fsin of the input wave, the sample data fcos of a signal whose phase is different by 90 °, and each data of the phase angle variation Δθ at each sampling timing are supplied to the arithmetic unit 3. On the other hand, sample data fo of the response wave to the input wave is input from the object 2 to the arithmetic unit 3. That is, the output of the sensor attached to the object 2 is supplied to the arithmetic unit 3 via an A / D converter (not shown).

FIG. 1B is a functional block diagram showing the internal configuration of the arithmetic device 3. The input wave sample data fsin, the response wave sample data fo, and the phase angle variation Δθ are multiplied by the first multiplier 4, and the multiplication result (fsin × fo × Δθ) The calculator 5 accumulates for a predetermined period. This integration period may be a period that is a positive integer multiple of a half cycle of the input wave (between the zero cross point and the next zero cross point). For example, one cycle (from the zero cross point to the next zero cross point) Period).
The sample data fcos of the signal whose phase is 90 ° different from that of the input wave, the sample data fo of the response wave, and the change amount Δθ of the phase angle are multiplied by the second multiplier 6, and the multiplication result (fcos × fo × Δθ) is accumulated by the second adder 7 for a predetermined period (for example, one cycle) similar to the above.
The integration result Fs (= Σ (fo × fsin × Δθ)) in the first adder 5 and the integration result Fc (= Σ (fo × fcos × Δθ)) in the second adder 7 are arithmetic units. 8, the computing unit 8 calculates the amplitude A of the response wave fo and the phase difference δ of the response wave fo based on the following equations (6) and (7).

ここで、ω₀は初期角速度、ω₁は単位時間当たりの角速度増加量を示す。
ある時間での角度（位相角）θ(t)は角速度の積分で与えられるので、次式のようになる。

ここで、積分定数ｃは、ｔ＝０のときθ(t)＝０とすると、ｃ＝０となる。
すなわち、

したがって、スイープ波である入力波sinθ(t)は次式で表される。

この入力波に対する応答波ｆ(x)は次式で表される。

Ａは応答波の振幅、δは入力波に対する位相差である。 Hereinafter, the calculation executed in the calculation device 3 will be described in detail.
First, the case of an analog signal will be described.
When the frequency changes in proportion to time (in the case of linear sweep), the angular frequency (angular velocity) ω (t) is expressed by the following equation.

Here, ω ₀ represents the initial angular velocity, and ω ₁ represents the amount of increase in angular velocity per unit time.
Since the angle (phase angle) θ (t) at a certain time is given by the integration of the angular velocity, the following equation is obtained.

Here, the integral constant c is c = 0 when θ (t) = 0 when t = 0.
That is,

Therefore, the input wave sinθ (t) that is a sweep wave is expressed by the following equation.

The response wave f (x) with respect to this input wave is expressed by the following equation.

A is the amplitude of the response wave, and δ is the phase difference with respect to the input wave.

ここで、

とすると、置換積分により、

となる。
ここで、積分区間をｘ＝０→π（１／２周期）とすると、前記式（２）と同様に、

となる。 When the response wave f (x) is multiplied by the input wave sin θ (t) and the angular frequency ω (t) and integrated, the following equation is obtained.

here,

Then, by substitution integration,

It becomes.
Here, if the integration interval is x = 0 → π (1/2 period), as in the above equation (2),

It becomes.

上と同様の置換積分により、

となる。
積分区間をｘ＝０→π（１／２周期）とすると、前記式（４）と同様に、

となる。 Further, when the response wave f (x) is multiplied by the waveform cos θ (t) and the angular frequency ω (t) that are 90 ° different in phase from the input wave and integrated, the following equation is obtained.

By permutation integration similar to the above,

It becomes.
Assuming that the integration interval is x = 0 → π (1/2 period), as in the equation (4),

It becomes.

すなわち、応答波ｆ(x)と入力波sinθ(t)と角周波数ω(t)（＝ω₀＋ω₁ｔ）の積をｘ（＝θ(t)＝ω₀ｔ＋(1/2)ω₁ｔ²）について０→πまで積分した値Ｆsと、応答波ｆ(x)と入力波と９０°位相が異なる波形cosθ(t)と角周波数ω(t)の積をｘについて０→πまで積分した値Ｆcとを用いて、上記式（１６）からスイープ波に対する応答波ｆ(x)の振幅Ａと位相差δを計算により求めることができる。 Therefore, the amplitude A and the phase difference δ of the response wave can be obtained by Expression (16) similar to Expression (5) described above.

That is, the product of the response wave f (x), the input wave sin θ (t), and the angular frequency ω (t) (= ω ₀ + ω ₁ t) is expressed as x (= θ (t) = ω ₀ t + (1/2) ω ₁ t ² ), the value Fs integrated from 0 to π, and the product of the response wave f (x), the input wave and the waveform cosθ (t) that is 90 ° out of phase with the angular frequency ω (t) for x 0 → π The amplitude A and the phase difference δ of the response wave f (x) with respect to the sweep wave can be calculated from the above equation (16) using the value Fc integrated up to.

When digital processing is performed, the amplitude A and the phase difference δ of the response wave can be similarly obtained by calculation using the sample data at each sampling timing at a constant time interval. That is, values corresponding to the Fs and Fc are calculated, and the amplitude A and the phase difference δ of the response wave are calculated using the equation (16).
Here, among the three terms in the equation (12) for obtaining the Fs, sample data of the response wave as a value corresponding to Asin (ω ₀ t + (1/2) ω ₁ t ² + δ) = f (x) fo is used, and the sample data fsin of the input wave sinθ (t) is used as a value corresponding to sin (ω ₀ t + (1/2) ω ₁ t ² ) = sinθ (t). Further, (ω ₀ + ω ₁ t) = ω (t) = 2πf (t) (angular frequency) is a derivative of the phase angle θ (t), and the change amount (difference) of the phase angle θ at each sampling timing (θ _n− θ _n−1 = Δθ) is used.
Therefore, the product of the response wave sample data fo, the input wave sample data fsin at each sampling timing, and the amount of change (Δθ) of the phase angle θ at each sampling timing is x corresponding to ½ period of the input wave. A value (Σ (fo × fsin × Δθ)) integrated in the range of π (or a positive integer multiple thereof) is obtained as a value corresponding to Fs in the equation (13).
The product (fo × fsin × Δθ) of the response wave sample data fo, the input wave sample data fsin at each sampling timing, and the change amount (Δθ) of the phase angle θ at each sampling timing by the multiplier 4 in FIG. ) And the output of the multiplier 4 corresponding to the range where x is 0 to π (or a range of a positive integer multiple thereof) is integrated by the adder 5, and Σ (fo × fsin × Δθ) = Fs Ask for.

In addition, the sample data fcos is used as a value corresponding to the waveform cos (ω ₀ t + (1/2) ω ₁ t ² ) that is 90 ° different in phase from the input wave in the equation (14) for obtaining Fc.
Therefore, the product of the sample data fo of the response wave at each sampling timing, the sample data fcos of a signal whose phase differs from the input wave by 90 °, and the amount of change (Δθ) of the phase angle θ at each sampling timing is 1/2 cycle of the input wave A value (Σ (fo × fcos × Δθ)) obtained by integrating x corresponding to x in a range of 0 to π (or a positive integer multiple thereof) is obtained as a value corresponding to Fc in the equation (15). It is done.
That is, the product of the sample data fo of the response wave at each sampling timing and the sample data fcos of the signal whose phase is 90 ° different from the input wave by the multiplier 6 and the change amount (Δθ) of the phase angle θ at each sampling timing (fo × fcos × Δθ) is obtained, and the output of the multiplier 6 corresponding to the range where x is 0 to π (or a positive integer multiple thereof) is integrated by the adder 7, and Σ (fo × fcos × Δθ ) = Fc.

すなわち、前記演算器８において、前記足算器５から出力されるＦsと前記足算器７から出力されるＦcを用い、応答波の振幅Ａと位相差δを算出する。 Using the thus-obtained Fs and Fc, the amplitude A and the phase difference δ of the response wave can be obtained using the above-described equations (6) and (7).

That is, the computing unit 8 calculates the response wave amplitude A and phase difference δ using Fs output from the adder 5 and Fc output from the adder 7.

ここで、ω₀は初期角速度、ω₁は単位時間当たりの角速度増加倍率を示す。
ある時間での角度θ(t)は、次式のようになる。

ここで、積分定数ｃは、ｔ＝０のときθ(t)＝０とすると、

となり、角度θ(t)は、次式のようになる。

したがって、入力波sinθ(t)は次式で表される。

この入力波に対する応答波ｆ(x)は、次式で表される。

Ａは応答波の振幅、δは入力波に対する位相差である。 The above is the case of the linear sweep in which the frequency changes linearly in proportion to the time. Similarly, in the case of the log sweep in which the frequency changes exponentially with respect to the time, similarly, the amplitude A of the response wave And the phase difference δ can be calculated.
In the case of log sweep, the angular frequency ω (t) is expressed by the following equation:

Here, ω ₀ represents an initial angular velocity, and ω ₁ represents an angular velocity increase magnification per unit time.
The angle θ (t) at a certain time is expressed by the following equation.

Here, if the integral constant c is θ (t) = 0 when t = 0,

Thus, the angle θ (t) is as follows.

Therefore, the input wave sinθ (t) is expressed by the following equation.

The response wave f (x) with respect to this input wave is expressed by the following equation.

A is the amplitude of the response wave, and δ is the phase difference with respect to the input wave.

ここで、

とすると、置換積分により、

となる。
積分区間をｘ＝０→π（１／２周期）とすると、前記式（２）と同様に、

となる。 When the response wave f (x) is multiplied by the input wave sin θ (t) and the angular frequency ω (t) and integrated, the following equation is obtained.

here,

Then, by substitution integration,

It becomes.
Assuming that the integration interval is x = 0 → π (1/2 period), as in the equation (2),

It becomes.

上と同様の置換積分により、

となり、積分区間をｘ＝０→π（１／２周期）とすると、前記式（４）と同様に、

となる。
したがって、応答波の振幅Ａ及び位相差δを前述した式（１６）で求めることができる。

デジタル処理の場合にも、上述の場合と同様にして、上述した式（６）及び式（７）により応答波の振幅Ａ及び位相差δを算出することができる。
このように、ログスイープの場合にも、上述したリニアスイープの場合と同様にして、応答波の振幅Ａ及び位相差δを算出することができる。 Further, when the response wave f (x) is multiplied by the waveform cos θ (t) and the angular frequency ω (t) that are 90 ° different in phase from the input wave and integrated, the following equation is obtained.

By permutation integration similar to the above,

Assuming that the integration interval is x = 0 → π (1/2 period), as in the equation (4),

It becomes.
Therefore, the amplitude A and the phase difference δ of the response wave can be obtained by the above-described equation (16).

Also in the case of digital processing, similarly to the above-described case, the amplitude A and the phase difference δ of the response wave can be calculated by the above-described equations (6) and (7).
Thus, also in the case of log sweep, the amplitude A and the phase difference δ of the response wave can be calculated in the same manner as in the case of the linear sweep described above.

Thus, according to the present invention, the amplitude A and the phase difference δ of the response wave with respect to the sweep wave can be calculated by the arithmetic processing by the arithmetic device 3. At this time, as described above, since the integration is performed by multiplying the change amount Δθ of the phase angle θ at each sampling timing, the sampling period (time interval of each sampling timing) may be constant, as described above. No special hardware is required.
The sweep wave generator 1 can be realized by software using a waveform memory, and implements an apparatus without using special hardware other than the A / D and D / A converters. be able to.

FIG. 2 is a diagram for explaining the range of integration executed by the adders 5 and 7.
In this figure, the horizontal axis represents time, the vertical axis represents amplitude, the broken line represents an input wave, and the solid line represents an example of a response wave. Further, t _{0 to} t ₁₀ are sampling timings, and θ _{0 to} θ ₁₀ are phase angles of input waves at the respective sampling timings t _{0 to} t ₁₀ . Each sampling timing interval is a constant time interval Δt as shown in the figure.
As described above, in the adders 5 and 7, Fs and Fc output from the multipliers 4 and 6 are in the range of a multiple of 1/2 period of the input wave (here, 1 period). Accumulate. At this time, if one period which is an integration period is determined in accordance with the sampling timing, a period T ′ from time t ₁ to time t ₁₀ is obtained as shown in the figure. That is, the adders 5 and 7 integrate the outputs Fs and Fc of the multipliers 4 and 6 with respect to ten sample data from t _{1 to} t ₁₀ .

Here, when the number of sampling points is small, an accurate result may not be obtained by integration according to the sampling timing as described above. That is, in the figure, from point P to Q is an accurate ½ cycle, from Q to R is an accurate ½ cycle, and when performing integration for one cycle, integration from point P to point R It is desirable to do.
Therefore, the zero cross point (P, Q, R) of the input wave is detected, the value of ε ₁ is used for the section from θ ₀ to θ ₁ , and the value of ε ₂ is used for the section from θ ₉ to θ ₁₀ . Using the value of the portion, the outputs of the multipliers 4 and 6 are integrated. Thereby, it is possible to obtain an integration result for an accurate section of one cycle T, and to reduce errors. In addition, when calculating 1/2 cycle instead of 1 cycle, it may be calculated for each section of T1 and T2 shown in the figure.

Next, a test apparatus using the above-described arithmetic device will be described.
FIG. 3 is a functional block diagram showing the configuration of a two-axis tester that is an embodiment of the test apparatus of the present invention. The same sweep wave is added to the two actuators while shifting the phase, the phase of the movement of the specimen at that time is detected, and the two actuators are controlled so as to have the target phase difference.
In FIG. 3, 10 is a specimen, 11 is a sweep wave generator, 12 is a first phase detector, 13 is a first phase controller, 14 is a first actuator, and 15 is a first actuator 14. Servo valve, 16 is a sensor for detecting a load or displacement applied by the first actuator 15, 17 is a second phase detector, 18 is a second phase controller, 19 is a second actuator, 20 is A servo valve for driving the second actuator 19 and a sensor 21 for detecting a load or displacement applied by the second actuator 19.
The first phase detector 12, the first phase controller 13, the first actuator 14, the servo valve 15 and the sensor 16 constitute a feedback control system related to the first actuator 14, and the second The phase detector 17, the second phase controller 18, the second actuator 19, the servo valve 20 and the sensor 21 constitute a feedback control system related to the second actuator 19. Then, the first actuator 14 and the second actuator 19 are driven by the same sweep wave whose phase is different by a predetermined amount (for example, 90 °).

The sweep wave generator 11 is configured in the same manner as the sweep wave generator 1 shown in FIG. 1 and continuously changes the frequency with a constant amplitude at each sampling timing having a constant time interval. Sample data fsin of a sine wave (sweep wave), sample data fcos having a waveform that is 90 ° different in phase from the sweep wave, and data Δθ of a change amount of the phase of the sweep wave at each sampling timing are output.
The first phase detection unit 12 and the second phase detection unit 17 are both configured in the same manner as the arithmetic device 3 shown in FIG.
Further, the first phase control unit 13 and the second phase control unit 18 perform a phase shift on the sweep wave output from the sweep generator 11 to the servo valve 15 or 20 of the corresponding actuator. Output. This can be configured, for example, by a memory that reads sample data of a sweep wave after a lapse of time corresponding to the write phase shift amount.
In practice, the functions of the sweep wave generator 11, the phase detection units 12 and 17, and the phase control units 13 and 18 are all realized by software by a controller provided in the tester. be able to.

Then, the first phase detector 12 uses an A / D converter (not shown) to convert the output signal of the sensor 16 that detects the movement of the specimen 10 at the first loading point by the first actuator 14 into digital data. Sampled data fo ₁ of the response wave converted to, sample data fsin of the sweep wave from the sweep oscillator 11, sample data fcos having a waveform that is 90 ° different in phase from the sweep wave, and the amount of change in the phase of the sweep wave at each sampling timing On the basis of the data Δθ, the amplitude A ₁ of the response wave fo ₁ at the first loading point and the phase difference δ ₁ with respect to the sweep wave f sin are calculated based on the aforementioned equations (6) and (7). Then, a control signal for controlling the phase shift amount in the first phase controller 13 is output so that the phase difference δ ₁ becomes a set phase difference (for example, 0 °). Thereby, feedback control is performed so that the movement of the specimen 10 at the first loading point has a phase (for example, the same phase) set in advance with respect to the sweep wave fsin from the sweep wave generator 11.

In addition, the second phase detector 17 includes sample data fo ₂ of the response wave from the sensor 21 that detects the movement of the specimen 10 at the second loading point by the second actuator 19, and the sweep oscillator 11. Based on the respective signals from, the amplitude A ₂ of the response wave fo ₂ at the second loading point and the phase difference δ ₂ between the sweep wave f sin are calculated. Then, a control signal for controlling the phase shift amount in the second phase controller 18 is output so that the phase difference δ ₂ becomes a preset phase difference (for example, 90 °). Thereby, feedback control is performed so that the movement of the specimen 10 at the second loading point has a phase (for example, 90 °) set in advance with respect to the sweep wave fsin from the sweep wave generator 11.
In this way, it is possible to load with the same sweep wave whose phase is shifted from a plurality of axes by a predetermined amount.

  Although FIG. 3 shows the case of a two-axis tester, when there are more actuators (for example, in the case of six axes), a sensor, a phase detector and a phase controller are provided for each actuator. What should I do?

It is a figure which shows the structure of the arithmetic device of this invention, (a) is a block diagram which shows the system which uses the arithmetic device of this invention, (b) is a functional block diagram which shows the structure of the arithmetic device of this invention. It is a figure for demonstrating the range of integration. It is a functional block diagram which shows the structure of one Embodiment of the testing apparatus using the arithmetic unit of this invention. It is a figure for demonstrating the peak detection by the conventional step sweep method.

Explanation of symbols

  1: Sweep wave generator, 2: Object, 3: Computing device, 4: Multiplier, 5-adder, 6: Multiplier, 7: Adder, 8: Calculator, 10: Specimen, 11: Sweep wave generator, 12, 17: Phase detector, 13, 18: Phase controller, 14, 19: Actuator, 15, 20: Servo valve, 16, 21: Sensor

Claims (3)

An arithmetic device that calculates a phase difference between a response signal and an input signal, which is a sinusoidal signal whose amplitude is constant and frequency is swept, to an object,
The sample data fsin of the input signal, the sample data fcos of a signal that is 90 ° out of phase with the input signal, the sample data fo of the response signal, and the phase angle of the input signal at each sampling timing set at a fixed time interval Change data Δθ is input,
First multiplication means for calculating a product of the response signal sample data fo, the input signal sample data fsin, and a phase angle change amount data Δθ of the input signal;
First addition means for integrating the calculation results (fo × fsin × Δθ) of the first multiplication means over a period of a positive integer multiple of ½ cycle;
Second multiplication means for calculating a product of sample data fo of the response signal, sample data fcos of a signal whose phase is 90 ° different from the input signal, and data Δθ of the change amount of the phase angle of the input signal;
Second addition means for integrating the calculation result (fo × fcos × Δθ) of the second multiplication means over a period of a positive integer multiple of ½ cycle; and
An arithmetic device comprising: arithmetic means for calculating a phase difference δ between the input signal and the response signal based on a ratio of a calculation result of the second addition means and a calculation result of the first addition means.   And a means for calculating an amplitude A of the response signal by calculating a square root of a sum of a square of a calculation result of the first addition means and a square of a calculation result of the second addition means. The arithmetic unit according to claim 1. A test apparatus using the arithmetic unit according to claim 1, wherein the test is performed by loading sweep waves having different phases on a plurality of loading points of the specimen.
Sweep wave generation that outputs sample data of the sweep wave, sample data of a signal that is 90 ° out of phase with the sweep wave, and data on the amount of change in the phase angle of the sweep wave at each sampling timing set at a fixed time interval Means,
A plurality of phase control means that are provided corresponding to the plurality of loading points, respectively, and that control the phase of the sweep wave output from the sweep wave generation means;
A plurality of loading means provided corresponding to each of the plurality of phase control means, and loading a sweep wave output from each phase control means at a corresponding loading point;
A plurality of detecting means provided corresponding to the plurality of loading points, respectively, for outputting sample data of response waves indicating the movement of the specimen at each loading point;
A plurality of calculating means provided corresponding to the plurality of detecting means , each calculating means including response wave sample data output from the corresponding detecting means and the sweep wave from the sweep wave generating means; Based on the sample data of the wave, the sample data of the signal whose phase is 90 ° different from that of the sweep wave, and the data of the change amount of the phase angle of the sweep wave, the response wave output from the detection means corresponding to the sweep wave A plurality of calculation means which are calculation devices according to claim 1 for calculating a phase difference;
The plurality of phase control means controls the phase of the sweep wave supplied to the corresponding loading means based on the phase difference calculated by the corresponding calculation means.

Images