JP2017078591A - Vibration control device and vibration control method, and vibration control program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a vibration control device, a vibration control method, and a vibration control program which are high in convergence performance of an excitation signal to be outputted, strictly following the specified energy of a measurement reference and also following the performance variation of a vibration exciter device during vibration testing and capable of adjusting the variation range within a set range.SOLUTION: A vibration control device is inputted with a PSD profile to be tested which is a spectrum distribution of the vibration generated in a vibration testing device, and generates a target time series signal from the PSD profile to be tested. The vibration control device includes: target signal generation means for generating a target frequency series signal Tf from the target time series signal; drive signal generation means for generating a drive time series signal from the target frequency series signal Tf; and drive signal correction means for correcting a drive frequency series signal Df which is a frequency spectrum of the drive time series signal so that the target frequency series signal Tf and a response frequency series signal Rf of the frequency spectrum of the response time series signal of the vibration testing device are matched.SELECTED DRAWING: Figure 8

Description

  The present invention relates to a vibration control device, a vibration control method, and a vibration for performing a random vibration test for evaluating the vibration resistance and durability of an industrial product by giving the industrial product a random vibration similar to that in an operating environment. It relates to the control program.

In the conventional random vibration control device, the acceleration average power spectrum density (PSD) and the root mean square (RMS) are exclusively used as control targets.
Specifically, control is performed so that the target average PSD and the response average PSD coincide with each other by comparing the target average PSD of the PSD profile to be tested to be tested and the response average PSD obtained from the response signal of the vibration test apparatus. .

Based on FIG. 12, the instantaneous PSD and the average PSD will be described.
First, a Fourier spectrum in a frame unit is calculated by Fourier transforming a time series signal cut out from a time series signal over a long period of time into a frame unit of a fixed time interval corresponding to a Fourier transform processing unit. In general, Fourier transform is used for the conversion from the time series signal to the frequency series signal, but here, it will be described as using fast Fourier transform (FFT transform) that performs Fourier transform at high speed.

  Next, an instantaneous PSD is obtained from the Fourier coefficient calculated by the Fourier transform using the following equation (4), and a root mean square (RMS) is obtained from the following equation (5).

なお、上記式（４）において、Ｎは所要周波数範囲の分割個数である。

In the above formula (4), N is the number of divisions in the required frequency range.

このとき、各瞬時ＰＳＤを瞬時ＰＳＤ＿０（ｆ），瞬時ＰＳＤ＿１（ｆ），・・・，瞬時ＰＳＤ＿ｉ（ｆ）と表し、これら瞬時ＰＳＤ＿０（ｆ）〜瞬時ＰＳＤ＿ｉ（ｆ）を時間平均したものが平均ＰＳＤ（ｆ）である。

At this time, each instantaneous PSD is expressed as an instantaneous PSD_0 (f), an instantaneous PSD_1 (f),..., An instantaneous PSD_i (f). PSD (f).

Based on FIG. 13, the structure and operation | movement of the conventional vibration control apparatus 100 are demonstrated.
First, the response time series signal Rt from the vibration test apparatus 150 is converted into a digital signal by the AD conversion unit 107. The vibration test apparatus 150 includes a response detector (not shown) that applies vibration to the device under test and measures a response signal from the device under test.

  The response time series signal Rt converted to the digital signal is Fourier-transformed by the FFT conversion unit 108 to calculate an instantaneous PSD, and the instantaneous PSD is time-averaged at a predetermined time to generate an average PSD.

Then, the PSD comparison unit 101 generates a PSD correction manipulated variable by comparing the target PSD obtained from the PSD profile under test with the average PSD.
Next, a modified PSD is generated from the target PSD and the PSD correction manipulated variable, and a reverse amplitude characteristic multiplication process is performed in the reverse amplitude characteristic multiplication processing unit 102 based on the amplitude characteristic of the transfer characteristic of the vibration test apparatus to generate a drive PSD.

  Further, in the Fourier inverse transform unit 103, the white random signal generated by the random number generation unit 104 is normalized by the driving PSD, converted into a time series waveform by the Fourier inverse transformation process, and developed, and the drive waveform synthesizing unit 105 converts the time. A series waveform overlapping process is performed, and the DA converter 106 generates a drive time series signal Dt that is an analog signal input to the vibration test apparatus 150.

  Generally, a vibration control device automatically stops when the displacement applied to the vibration test device by the drive time series signal exceeds the excitation limit region to prevent damage and ensure safety of the vibration test device. Therefore, it is necessary to set the test conditions with a profile within safety tolerance as much as possible.

In a normal acceleration vibration test, a PSD profile as a profile to be tested in the excitation frequency range and an RMS represented by the above equation (5) are given, and the probability density distribution of the acceleration time series signal is the square of the RMS. A normal distribution with variance σ ² is required.

  A random signal is a signal generated stochastically. Therefore, it is predicted that the acceleration time series signal having such a probability density distribution includes a signal that exceeds the allowable range of the vibration test apparatus.

  For example, even if this phenomenon is rare in terms of probability statistics, if a signal exceeding the allowable range of the vibration test equipment is generated, the vibration test equipment may be damaged. Since the function of automatically detecting the abnormality when the displacement of the allowable range is exceeded and stopping the vibration test automatically works, the test purpose is not completed. For this reason, a technique called clipping is used to remove acceleration signals outside the allowable range.

  Clipping is a process of controlling an acceleration random time series with a value obtained by multiplying σ, which is the square root of variance, by an appropriate magnification (clipping magnification) so as not to generate an excessive acceleration signal. If the clipping magnification is set appropriately, only excessive acceleration that rarely occurs in terms of probability statistics is suppressed, so that the entire acceleration average spectrum is not greatly affected. In many cases, a magnification value of 3.0 to 5.0 is used.

  The clipping process is used for the purpose of directly suppressing the generation of an excessive acceleration signal in a random acceleration vibration test, and is used in anticipation of the effect of suppressing the vibration physical range (displacement) of the vibration signal.

  FIG. 14 is a graph showing the difference in probability distribution of acceleration amplitude depending on the presence or absence of σ clipping, and FIG. 15 is a graph showing the difference in probability distribution of displacement amplitude depending on the presence or absence of σ clipping. Even if the acceleration is suppressed by the σ clipping shown in FIG. 14, the displacement is hardly suppressed as shown in FIG. That is, no significant displacement suppression effect due to clipping is observed.

  As described above, σ clipping with respect to the acceleration signal is effective as a technique for suppressing the excitation force, but has a very limited effect as a technique for suppressing the displacement (vibration physical range).

  Furthermore, in the conventional vibration control apparatus, the response time series signal is converted into a Fourier spectrum by FFT processing, but immediately after that, it is converted into an instantaneous PSD, and a response average PSD as a time average of the instantaneous PSD is calculated. Will be. At this time, vibration control is performed in comparison with the target average PSD, but phase components that are important information in vibration control are ignored. Furthermore, the vibration physical range (displacement) of the vibration signal resulting from the acceleration signal is outside the scope of control despite being an important limit physical quantity related to the vibration test apparatus.

  In addition, the random signal generated directly in the random vibration test is an acceleration random signal, and the displacement signal based on it is an infinite second-order integral of random acceleration. there were.

  Further, when a mechanism for suppressing and preventing displacement drift is provided in the vibration test apparatus, the displacement generated in the vibration test apparatus is also affected by the mechanism.

  If the random acceleration signal generated based on the PSD profile can suppress the displacement while maintaining the predetermined PSD profile and the RMS value, the application range of the vibration test apparatus can be expanded.

  As a method for suppressing displacement in this way, conventionally, a precautionary measure such as estimating the maximum displacement from the PSD profile under test in advance and correcting the PSD profile under test when the allowable performance range is exceeded, The over-acceleration signal is cut by clipping as described above.

  However, the clipping method for acceleration exhibits probabilistic regulation for random acceleration, and does not regulate displacement itself. As shown in FIG. 15, the effect of suppressing the displacement signal is limited. Met.

  In the present invention, in view of such a current situation, the output excitation signal has high convergence, strictly follows the prescribed energy of the measurement standard, and changes in characteristics of the excitation apparatus and the DUT during the vibration test. An object of the present invention is to provide a vibration control device, a vibration control method, and a vibration control program that can follow and keep a displacement range within a set range.

The present invention was invented in order to solve the problems in the prior art relating to the acceleration random vibration test as described above.
A vibration control device for controlling the vibration of a vibration test device that applies vibration to a device under test,
A target signal generating means for receiving a PSD (Power Spectrum Density) profile, which is a spectrum distribution of vibration generated in the vibration test apparatus, and generating a target frequency series signal Tf from the PSD profile under test;
Drive signal generating means for generating a drive time series signal Dt from the target frequency series signal Tf;
The drive frequency series signal that is the frequency spectrum of the drive time series signal Dt so that the target frequency series signal Tf matches the response frequency series signal Rf that is the frequency spectrum of the response time series signal Rt of the vibration test apparatus. Drive signal correction means for correcting Df;
With
The target signal generating means is
A random signal generator for generating a random time series signal having the PSD profile under test;
A target time series signal generating unit for generating a target time series signal Tt from the random time series signal;
A target frequency sequence signal generator for generating a target frequency sequence signal Tf from the target time series signal Tt;
It is characterized by including.

In this case, the drive signal generating means is
A frequency sequence signal processing unit that generates a drive frequency sequence signal Df from the target frequency sequence signal Tf;
A re-conversion processing unit from the driving frequency series signal Df to a time series signal for generating the driving time series signal Dt;
It is preferable to contain.

Further, the drive signal correction means includes the frequency series signal processing unit,
The frequency series signal processing unit,
The drive time series signal so that the target frequency series signal Tf and the response frequency series signal Rf, which is a response spectrum generated from the response time series signal Rt detected by the response detector of the vibration test apparatus, coincide with each other. It is preferable that the driving frequency series signal Df which is a driving spectrum generated from Dt is corrected.

Further, the vibration control method of the present invention includes:
A vibration control method for controlling vibration of a vibration test apparatus that applies vibration to a device under test,
A target time series signal Tt is generated from a PSD (Power Spectrum Density) profile that is a spectrum distribution of vibrations generated in the vibration test apparatus, and a target frequency series signal Tf is generated from the target time series signal Tt.
Generating a driving time series signal Dt from the target frequency series signal Tf;
The drive frequency series signal that is the frequency spectrum of the drive time series signal Dt so that the target frequency series signal Tf matches the response frequency series signal Rf that is the frequency spectrum of the response time series signal Rt of the vibration test apparatus. While correcting Df,
A displacement signal for the target time series signal Tt is always generated, the set limit value of the displacement region is always determined, and the random vibration, the target PSD profile, and the RMS value are maintained constant without exceeding the displacement limit value. A random acceleration signal that continues to be generated is generated.

In this case, while generating the drive frequency sequence signal Df from the target frequency sequence signal Tf,
The drive time series signal Dt is preferably generated from the drive frequency series signal Df.

  The target frequency series signal Tf and the response frequency series signal Rf, which is a response spectrum generated from the response time series signal Rt detected by the response detector of the vibration test apparatus, coincide with each other during the driving. It is preferable to correct the drive frequency series signal Df which is a drive spectrum generated from the series signal Dt.

The vibration control program of the present invention is
A vibration control program for controlling vibration of a vibration test apparatus that applies vibration to a device under test,
A target time series signal Tt is generated from a PSD (Power Spectrum Density) profile that is a spectrum distribution of vibrations generated in the vibration test apparatus, and a target frequency series signal Tf is generated from the target time series signal Tt. Further, a displacement signal for the target time series signal Tt is always generated, the set limit value of the displacement region is always determined, and the random vibration, the target PSD profile, and the RMS value are kept constant without exceeding the displacement limit value. A target signal generation procedure for generating an acceleration random signal that continues to be maintained,
A drive signal generation procedure for generating a drive time series signal Dt from the target frequency series signal Tf;
The drive frequency series signal that is the frequency spectrum of the drive time series signal Dt so that the target frequency series signal Tf matches the response frequency series signal Rf that is the frequency spectrum of the response time series signal Rt of the vibration test apparatus. A drive signal correction procedure for correcting Df;
Is executed by a computer.

In this case, the drive signal generation procedure is:
A frequency sequence signal processing procedure for generating a drive frequency sequence signal Df from the target frequency sequence signal Tf;
A time series signal decoding processing procedure for generating the drive time series signal Dt from the drive frequency series signal Df;
It is preferable to contain.

The drive signal correction procedure includes the frequency sequence signal processing procedure,
The frequency sequence signal processing procedure includes:
The drive time series signal so that the target frequency series signal Tf and the response frequency series signal Rf, which is a response spectrum generated from the response time series signal Rt detected by the response detector of the vibration test apparatus, coincide with each other. It is preferable to correct the drive frequency series signal Df which is a drive spectrum generated from Dt.

  According to the present invention, by correcting the drive frequency sequence signal so that the target frequency sequence signal matches the response frequency sequence signal, only the displacement frequency distribution can be obtained without suppressing the acceleration frequency distribution of the vibration test apparatus. Stable and continuous vibration test can be performed without exceeding the vibration physical range (displacement) of the vibration test apparatus.

FIG. 1 is a control process diagram for explaining a control system of a vibration control apparatus according to an embodiment of the present invention. FIG. 2 is a graph showing an example of the PSD profile under test. FIG. 3 is a flowchart for explaining a flow of generating a pure random signal from the PSD profile under test in the random signal generation unit and the target time series signal generation unit. FIG. 4 is a graph showing the relationship between the Fourier spectrum component and the deviation angle θ of the complex plane. FIG. 5 is a graph showing an example of a pure random time series signal. FIG. 6 is a flowchart for explaining a flow of generating an acceleration random signal whose displacement is suppressed. FIG. 7 is a graph showing the PSD profile when the acceleration random test is executed at RMS = 12.5 m / sec ² and the average PSD of the response signal. FIG. 8 is a processing block diagram for explaining the configuration of the frequency sequence signal processing unit. FIG. 9 is a processing block diagram for explaining the configuration of the time-series signal reconversion processing unit. FIG. 10 is a graph showing a probability distribution of acceleration amplitude when a vibration test is performed using the vibration control apparatus of the present embodiment. FIG. 11 is a graph showing a probability distribution of displacement amplitude when a vibration test is performed using the vibration control apparatus of the present embodiment. As a result, a displacement suppression effect of 35% was confirmed by the present embodiment. FIG. 12 is a conceptual diagram for explaining the instantaneous PSD and the average PSD. FIG. 13 is a control process diagram for explaining the configuration and operation of a conventional vibration control apparatus. FIG. 14 is a graph showing a difference in probability distribution of acceleration amplitude depending on the presence or absence of clipping. FIG. 15 is a graph showing the difference in probability distribution of displacement amplitude depending on the presence or absence of clipping.

Hereinafter, embodiments (examples) of the present invention will be described in more detail with reference to the drawings.
FIG. 1 is a control process diagram for explaining a control system of the vibration control device 10 in the present embodiment.

  As shown in FIG. 1, the vibration control device 10 includes a random signal generation unit 12, a target time series signal generation unit 14, a target spectrum generation unit 16, a response spectrum generation unit 18, a drive spectrum generation unit 20, A frequency series signal processing unit 22, a time series signal re-conversion processing unit 24, a DA conversion unit 26, and an AD conversion unit 28 are provided.

  Further, the vibration control device 10 is connected to a vibration test device 50 that applies vibration to the device under test via the DA conversion unit 26 and the AD conversion unit 28. The vibration test apparatus 50 includes a power amplifier and a response detector (not shown), and the vibration signal output from the DA converter 26 is input to the vibration test apparatus via the power amplifier to generate vibration. The test apparatus generates vibration based on the excitation signal. Then, the response vibration from the DUT obtained according to the vibration is detected by the response detector and input to the AD converter 28 as a response analog signal.

Hereinafter, each process will be described.
(1) Random signal generator 12 to target time series signal generator 14
For example, a random signal generator 12 receives a PSD profile to be tested, which is a spectral distribution of vibrations generated in the vibration test apparatus 50 as shown in FIG. The vibration control apparatus 10 according to the present embodiment generates an excitation signal based on the input profile under test.

  The random signal generator 12 generates a random signal having a PSD profile under test for each frame, and sequentially repeats this to generate a random time-series signal that maintains the PSD profile under test.

  The necessary condition for the random time-series signal to be completely white noise is that the frequency of appearance in the time domain is uniformly random, the amplitude distribution in the frequency domain is random, and the frequency domain The phase distribution is uniformly random.

  As in this example, in the random vibration test defined by the PSD profile under test shown in FIG. 2, the amplitude distribution in the frequency domain is replaced with the PSD profile under test, the frequency of appearance in the time domain is random, and A random signal having a random phase distribution in the frequency domain is generated.

FIG. 3 is a flowchart for explaining a flow of generating a pure random signal from the PSD profile under test in the random signal generation unit 12 and the target time series signal generation unit 14.
First, a Fourier spectrum having a random phase is generated from the PSD profile under test.

  The time series signal f (t) and its Fourier transform function F (ω) have a relationship represented by the following formula (1).

上記式（１）において、ｎ＝・・・，−３，−２，−１，０，１，２，３，・・・である。

In the above formula (1), n =..., −3, −2, −1, 0, 1, 2, 3,.

  Moreover, the real component a and the imaginary component b in the function after the FFT process, which is a discrete Fourier transform, have a relationship represented by the following equations (2) and (3).

とすれば、

given that,

  Further, the absolute value of the Fourier spectrum can be expressed by the following equation (6) using equations (3) and (4).

なお、上記式（４）において、Ｎは所要周波数範囲の分割個数である。

In the above formula (4), N is the number of divisions in the required frequency range.

  With this absolute value as a reference, when the declination of the Fourier spectrum is θ, the Fourier spectrum component indicating PSD [n] can be expressed by the following formula (7).

That is, the relationship between the Fourier spectrum component and the deviation angle θ of the complex plane is as shown in FIG.
The PSD is determined by the absolute value (PSD value) shown in FIG. 4, but since there is a degree of freedom in the phase value θ, the PSD distribution is not randomized by randomly varying the phase value θ. A phase Fourier spectrum can be generated.

  By subjecting the random phase Fourier spectrum to inverse FFT transformation, a normally distributed random time series signal having a spectrum distribution of the PSD profile under test in which sine waves with different phases are overlapped is generated.

  However, since the inverse FFT transform used in the digital signal processing is a discrete process, it does not include all frequency components without omission, and the missing frequency components are compensated for by irregularly changing the frame phase. effective.

  Nevertheless, since a certain amount of frequency components are lost, the normally distributed random time series signal is called a pseudo-random signal. Therefore, the following processing is performed in order to generate a pure random signal from the pseudo random signal.

First, a Hann window (Hanning Window) is applied to the time-series signal of each pseudo-random signal.
The random time series signal to which the Hann window is applied is overlap-combined by 1/2 frame.

  The time series is mixed by the process of applying the Hann window and the overlap synthesis process by 1/2 frame, and a pure random time series signal maintaining the PSD profile under test as shown in FIG. 5 is generated. .

  The target time series signal generation unit 14 sends the pure random signal as a target time series signal to the target spectrum generation unit 16.

(2) Target spectrum generator 16
In the target spectrum generation unit, a target instantaneous Fourier spectrum, a target instantaneous PSD, and a target average PSD are obtained by performing FFT processing on the target time series signal, and a target frequency series signal having all these information is generated.

Since a _n + ib _n includes a phase component and is obtained for each frame, this is called an instantaneous Fourier spectrum. Also, an instantaneous PSD is calculated from the instantaneous Fourier spectrum based on the above equation (4), and an average PSD is calculated by taking a time average with respect to the instantaneous PSD.

  That is, the target spectrum generation unit 16 calculates the target instantaneous Fourier spectrum, the target instantaneous PSD, and the target average PSD based on the equations (1) to (7) by inputting the target time series signal Tt. A target frequency sequence signal Tf having the following information is generated.

  The displacement of the target acceleration random generated by the above method rarely occurs in terms of probability statistics. However, in the vibration test using an actual apparatus, if the displacement is exceeded even once, the machine may be damaged, and the test must be stopped at that time. For this reason, the target signal generation means for continuously generating the acceleration random target time-series signal Tt that does not cause displacement excess can further include an acceleration random signal generation unit that performs the following processing.

FIG. 6 is a flowchart for explaining a flow of generating a random acceleration target signal in which displacement is suppressed.
The acceleration random signal is first generated for each frame, and is sequentially connected to be a continuous signal without periodic repetition.

Here, generation of an acceleration random signal of frame i + 1 in which displacement over does not occur will be described as the next stage after frame i is generated.
The acceleration random signal of the frame i + 1 is generated in the random signal generation unit 12 to the target time series signal generation unit 14 described above.

  At the same time, accurate displacement prediction is performed by performing second-order numerical integration using the final displacement data of frame i as an initial value. If the displacement prediction under this assumption exceeds the preset allowable displacement range, the acceleration random signal of frame i + 1 is discarded, and the acceleration random signal is generated again by the same procedure, and the same displacement prediction is performed. .

  If the predicted displacement value falls within the preset allowable range, the acceleration random frame data generated here is adopted as the frame i + 1. This process generates an acceleration random signal in which a desired displacement is suppressed.

  If the required displacement suppression amount exceeds 40%, the displacement may not be suppressed even if the above-described method is tried several times in succession. In this case, displacement can be suppressed by appropriately reducing the PSD value in the low frequency region of 5 Hz or less. In reality, such a case rarely occurs when a large amount of deterrence is set, and does not greatly affect the average PSD.

  By sequentially processing the acceleration random signal thus obtained as the next original data, it is possible to continuously generate the acceleration random signal in which the displacement is suppressed.

  With the above displacement suppression algorithm, it is possible to suppress the actual displacement range while satisfying the following conditions for the random acceleration of acceleration defined by the PSD profile.

The acceleration random target signal for displacement suppression generated in this way is as shown in FIG.
(A) The average PSD of the measured response signal matches the designated PSD profile.
(B) The RMS value matches the specified value.
(C) The probability density distribution of the acceleration time series is a normal distribution.
(D) There is no repetition of the same signal.
The basic requirements for an acceleration random vibration signal are completely satisfied.
As shown in FIG. 11, as a result of an acceleration vibration test using the vibration control apparatus of this example, a displacement suppression effect of 35% was confirmed.

(3) Frequency sequence signal processing unit 22
FIG. 8 is a processing block diagram for explaining the configuration of the frequency sequence signal processing unit 22.
As shown in FIG. 8, the frequency sequence signal processing unit 22 includes an average PSD comparison unit 221 that compares the target average PSD and the response average PSD, and a spectrum that corrects the target instantaneous PSD to generate a corrected target frequency sequence signal Tc. The correction control unit 222, the inverse transfer function multiplier 223 that generates the drive signal Dr by multiplying the correction target frequency sequence signal Tc by the inverse transfer function IGf, and the ratio of the response frequency sequence signal Rf and the drive frequency sequence signal Df A transfer function calculation unit 224 that calculates a certain transfer function Gf and an inverse transfer function calculation unit 225 that calculates an inverse transfer function IGf that is an inverse characteristic of the transfer function Gf are provided.

(4) Drive frequency signal generation unit The average PSD comparison unit 221 receives the target frequency sequence signal Tf and the response frequency sequence signal Rf, and compares the target average PSD with the response average PSD. An average PSD correction vector TCv is calculated for each frame. Then, it is sent to the spectrum correction control unit 222 as a correction amount for the target instantaneous PSD of the target frequency series signal Tf.

  The spectrum correction control unit 222 adds the average PSD correction vector TCv to the target instantaneous PSD of the target frequency series signal Tf, thereby generating a corrected target frequency series signal Tc in which the target instantaneous PSD is corrected for each frame.

  By configuring in this way, the spectrum correction control unit 222 does not correct the phase component, but performs an agile control operation with respect to the characteristic variation of the vibration test apparatus as the amplitude correction of the target instantaneous PSD. Become.

  In addition, the transfer function calculation unit 224 constantly calculates a transfer function Gf that is a ratio of the response frequency sequence signal Rf and the drive frequency sequence signal Df and an average value thereof. When the change in the transfer function Gf exceeds a predetermined amount, the latest transfer function Gf is sent to the inverse transfer function calculation unit 225.

  In the inverse transfer function calculation unit 225, the inverse transfer function IGf is calculated based on the input transfer function Gf, and is sent to the inverse transfer function multiplier 223. As described above, the inverse transfer function IGf is applied to the corrected target frequency sequence signal Tc. The inverse transfer function IGf is multiplied to generate a drive signal Dr that is a Fourier spectrum signal for driving.

(5) Re-conversion processing unit 24 for driving time series signals
FIG. 9 is a processing block diagram for explaining the configuration of the time-series signal reconversion processing unit 24.
In the time series signal reconversion processing unit 24, the IFFT conversion unit 241 decodes the drive signal Dr from the frequency series signal to the time series signal by IFFT conversion processing. Since the decoded time series signal is discontinuous between frames, the overlap processing unit 242 performs overlap synthesis processing to generate a continuous primary drive time series signal Dtc.

The control cycle of the re-conversion processing unit 24 to the time series signal is synchronized with a control sampling clock (for example, 5.12 × 10 ³ sample / s when the control band is 2 kHz) except for the IFFT conversion process and the overlap synthesis process. Works.

The operations of the IFFT conversion processing and the overlap synthesis processing are processed in the background of the operation of the time-series signal reconversion processing unit 24, and this calculation is completed within the processing time of one frame.
The generated drive time series signal Dt is accumulated in a buffer and used in the next control cycle.

  In the signal calculation in the frequency series, the FFT calculation process is always performed on the time series signal of the appropriate length (one frame) that is logically obtained. For this reason, in a time-series signal obtained by performing IFFT conversion processing on the result calculated in the frequency series, the frames are discontinuous.

  Since time-series signals are usually controlled on the assumption that they are continuous, discontinuity between frames is eliminated, and the time-series signals obtained by signal operations in the frequency sequence are kept continuous. The overlap synthesis process is performed. In this embodiment, the overlap synthesis process is performed every 1/2 frame.

  The target / response time series signal comparison unit 244 compares the target time series signal Tt and the response time series signal Rt, and generates a time series signal correction amount Dc based on, for example, PID control.

  The time series signal correction unit 243 performs correction processing based on the primary drive time series signal Dtc and the time series signal correction amount Dc, and the corrected primary drive time series signal Dtc is shown as a drive time series signal Dt in FIG. As shown in FIG. 6, the vibration is output to the vibration test apparatus 50 via the DA converter 26.

  As described above, the vibration control apparatus 10 of the present embodiment configured as described above includes the random signal generation unit 12, the target time series signal generation unit 14, and the target spectrum generation unit 16 when the PSD profile to be tested is input. A target frequency series signal Tf suitable for the purpose of displacement suppression is generated by the target signal generation means.

Next, a drive time series signal Dt, which is a random time series signal, is generated by a drive signal generation unit including a frequency series signal processing unit 22 and a time series signal reconversion processing unit 24.
In addition, a response frequency that is a response spectrum generated by the drive signal correction means including the frequency sequence signal processing unit 22 from the target frequency sequence signal Tf and the response time series signal Rt detected by the response detector of the vibration test apparatus 50. A new drive signal Dr is generated by correcting the drive frequency series signal Df, which is a drive spectrum generated from the drive time series signal Dt, so that the series signal Rf matches, and the drive signal Dr is driven from this drive signal Dr. In the series signal generation means, a new drive time series signal Dt is generated.

  By performing such feedback control, the displacement in the vibration test apparatus 50 can be suppressed.

(6) Response spectrum generator 18
The response analog signal input from the vibration test apparatus is converted into a response time series signal Rt by the AD conversion unit and sent to the response spectrum generation unit.
The response spectrum generation unit calculates the response instantaneous Fourier spectrum, response instantaneous PSD, and response average PSD based on the above formulas (1) to (6) by inputting the response time series signal Rt, and all the information Is generated as a response frequency sequence signal Rf.

(7) Drive spectrum generation unit 20
The drive time series signal Dt generated as the drive signal of the vibration test apparatus 50 is branched at the previous stage where it is converted into a drive analog signal by the DA converter and input to the drive spectrum generator 20.
That is, the drive spectrum generation unit 20 receives the drive time series signal Dt, and calculates the drive instantaneous Fourier spectrum, the drive instantaneous PSD, and the drive average PSD based on the above formulas (1) to (7). A drive frequency series signal Df having all information is generated.

  The preferred embodiments of the present invention have been described above, but the present invention is not limited to this. For example, each processing unit described above may be hardware or software depending on the device configuration. Various modifications can be made without departing from the object of the present invention, such as a program for executing the above-described processing contents in a computer.

  FIGS. 10 and 11 show the probability distribution of the acceleration amplitude and the probability distribution of the displacement amplitude when the vibration test is performed using the vibration control device 10 of the present embodiment and the conventional vibration control device.

  When a conventional vibration control apparatus is used, the maximum displacement of the vibration test apparatus 50 is ± 115 mm in measurement using a displacement meter provided in the vibration test apparatus 50.

  On the other hand, when the vibration control apparatus 10 of the present embodiment is used, the maximum displacement of the vibration test apparatus 50 can be suppressed to ± 75 mm in the measurement by the displacement meter provided in the vibration test apparatus 50, and it is continuously performed for 100 hours or more. In operation, stable displacement suppression of 35% was confirmed.

DESCRIPTION OF SYMBOLS 10 Vibration control apparatus 12 Random signal generation part 14 Target time series signal generation part 16 Target spectrum generation part 18 Response spectrum generation part 20 Drive spectrum generation part 22 Frequency series signal processing part 221 Average PSD comparison part 222 Spectral correction control part 223 Reverse transmission Function multiplication unit 224 Transfer function calculation unit 225 Inverse transfer function calculation unit 24 Re-conversion processing unit 241 IFFT conversion unit 242 Overlap processing unit 243 Time-series signal correction unit 244 Target / response time-series signal comparison unit 26 DA Conversion unit 28 AD conversion unit 50 Vibration test device 100 Vibration control device 101 PSD comparison unit 102 Inverse amplitude characteristic multiplication processing unit 103 Inverse Fourier transform unit 104 Random number generation unit 105 Drive waveform synthesis unit 106 DA conversion unit 107 AD conversion unit 108 FFT conversion 150 vibration test equipment

Claims (9)

A vibration control device for controlling the vibration of a vibration test device that applies vibration to a device under test,
A PSD (Power Spectrum Density) profile, which is a spectrum distribution of vibrations generated in the vibration test apparatus, is input, a target time series signal Tt is generated from the PSD profile under test, and a target is obtained from the target time series signal Tt. Target signal generating means for generating a frequency series signal Tf;
Drive signal generating means for generating a drive time series signal Dt from the target frequency series signal Tf;
The drive frequency series signal that is the frequency spectrum of the drive time series signal Dt so that the target frequency series signal Tf matches the response frequency series signal Rf that is the frequency spectrum of the response time series signal Rt of the vibration test apparatus. Drive signal correction means for correcting Df;
With
The target signal generating means is
A displacement signal for the target time series signal Tt is always generated, a set limit value of the displacement region is always determined, and the normal distribution of the time series probability density without exceeding the displacement limit value, the target PSD profile The vibration control device further includes an acceleration random signal generation unit that generates an acceleration random signal that keeps the RMS value constant. The drive signal generating means is
A frequency sequence signal processing unit that generates a drive frequency sequence signal Df from the target frequency sequence signal Tf;
A re-conversion processing unit from the driving frequency series signal Df to a time series signal for generating the driving time series signal Dt;
The vibration control device according to claim 1, comprising: The drive signal correction means includes the frequency series signal processing unit,
The frequency series signal processing unit,
The drive time series signal so that the target frequency series signal Tf and the response frequency series signal Rf, which is a response spectrum generated from the response time series signal Rt detected by the response detector of the vibration test apparatus, coincide with each other. The vibration control apparatus according to claim 2, wherein the vibration control apparatus is configured to correct the drive frequency series signal Df that is a drive spectrum generated from Dt. A vibration control method for controlling vibration of a vibration test apparatus that applies vibration to a device under test,
A target time series signal Tt is generated from a PSD (Power Spectrum Density) profile that is a spectrum distribution of vibrations generated in the vibration test apparatus, and a target frequency series signal Tf is generated from the target time series signal Tt.
Generating a driving time series signal Dt from the target frequency series signal Tf;
The drive frequency series signal that is the frequency spectrum of the drive time series signal Dt so that the target frequency series signal Tf matches the response frequency series signal Rf that is the frequency spectrum of the response time series signal Rt of the vibration test apparatus. While correcting Df,
A displacement signal for the target time series signal Tt is always generated, the set limit value of the displacement region is always determined, and the random vibration, the target PSD profile, and the RMS value are maintained constant without exceeding the displacement limit value. A vibration control method characterized by generating an acceleration random signal that continues. A drive frequency series signal Df is generated from the target frequency series signal Tf,
The vibration control method according to claim 4, wherein the drive time series signal Dt is generated from the drive frequency series signal Df.   The drive time series signal so that the target frequency series signal Tf and the response frequency series signal Rf, which is a response spectrum generated from the response time series signal Rt detected by the response detector of the vibration test apparatus, coincide with each other. The vibration control method according to claim 5, wherein the drive frequency series signal Df, which is a drive spectrum generated from Dt, is corrected. A vibration control program for controlling vibration of a vibration test apparatus that applies vibration to a device under test,
A target time series signal Tt is generated from a PSD (Power Spectrum Density) profile that is a spectrum distribution of vibrations generated in the vibration test apparatus, and a target frequency series signal Tf is generated from the target time series signal Tt. Further, a displacement signal for the target time series signal Tt is always generated, the set limit value of the displacement region is always determined, and the random vibration, the target PSD profile, and the RMS value are kept constant without exceeding the displacement limit value. A target signal generation procedure for generating an acceleration random signal that continues to be maintained,
A drive signal generation procedure for generating a drive time series signal Dt from the target frequency series signal Tf;
The drive frequency series signal that is the frequency spectrum of the drive time series signal Dt so that the target frequency series signal Tf matches the response frequency series signal Rf that is the frequency spectrum of the response time series signal Rt of the vibration test apparatus. A drive signal correction procedure for correcting Df;
Vibration control program for causing a computer to execute. The drive signal generation procedure includes:
A frequency sequence signal processing procedure for generating a drive frequency sequence signal Df from the target frequency sequence signal Tf;
A time series signal decoding processing procedure for generating the drive time series signal Dt from the drive frequency series signal Df;
The vibration control program according to claim 7, comprising: The drive signal correction procedure includes the frequency sequence signal processing procedure,
The frequency sequence signal processing procedure includes:
The drive time series signal so that the target frequency series signal Tf and the response frequency series signal Rf, which is a response spectrum generated from the response time series signal Rt detected by the response detector of the vibration test apparatus, coincide with each other. 9. The vibration control program according to claim 8, wherein the drive frequency series signal Df, which is a drive spectrum generated from Dt, is corrected.

Images