JP2012103240A - Random vibration test control apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a random vibration test control apparatus capable of performing control including control of a phase component, that is, significant information in vibration control.SOLUTION: A random vibration test control apparatus of the prevent invention comprises a target time series generating part 302, a target spectrum generating part 303, a response spectrum generating part 304, a drive spectrum generating part 305, a frequency series processing part 306, a time series decoding part 307, a DA converter part 308 and an AD converter part 309. A true random time series signal based on a PSD pattern to be tested in random vibration tests is generated, the true random time series signal is subjected to FFT processing for generating a target spectrum signal, and the random vibration test is controlled so that the target spectrum signal may accord with a response spectrum signal obtained through the FFT processing of a response signal supplied by a vibration generator, resulting in performing control in which the target and response time series signals accord with each other and instantaneous PSD and average PSD accord with each other.

Description

  The present invention relates to a random vibration test control device. More specifically, the earthquake resistance and durability of an industrial product are evaluated by applying the same random vibration to the industrial product in the laboratory.

In a conventional random vibration test control apparatus, only an average PSD (Power Spectrum Density) is used exclusively as a control target (Patent Document 1).
In order to perform control so that the target average PSD and the response average PSD coincide with each other by comparing the target average PSD of the PSD pattern to be tested to the test and the response average PSD obtained from the response signal from the vibration shaker device. It is not necessary to obtain a match between the target value and the response value in the series signal, instantaneous or instantaneous PSD, and no PSD probability density distribution function is defined for each frequency.
A conventional random vibration test control apparatus 100 will be described with reference to FIGS.
First, the instantaneous PSD and the average PSD will be described with reference to FIG.
A time-series signal cut out in frame units is subjected to FFT (Fast Fourier Transform) processing to calculate a Fourier spectrum in frame units, and an instantaneous PSD is obtained from the Fourier spectrum.
Each instantaneous PSD is expressed as instantaneous PSD 0 (f), instantaneous PSD 1 (f)... Instantaneous PSD i (f) 200, and the average PSD (f) 201 is obtained by averaging the instantaneous PSDs over time.

The configuration and operation of a conventional random vibration test control device 100 will be described with reference to FIG.
A response signal from the vibration exciter device (including the test body and sensor) 112 is converted into a digital signal by the AD converter 111.
The instantaneous PSD is calculated from the Fourier spectrum of the frame unit by processing the digital signal in the FFT conversion unit 110, and the instantaneous PSD is averaged for a certain period of time to generate an average PSD 108 of the response.
The target PSD 101 obtained from the PSD pattern under test is compared with the average PSD 108 from the response to generate a PSD manipulated variable 109.
A corrected PSD 102 is generated from the target PSD 101 and the PSD operation amount 109.
Using the transfer characteristics of the vibration exciter device 112, the reverse PSD multiplication 103 performs a transfer characteristic correction process on the modified PSD 102 to generate a drive PSD.
The white random generated by the random number generator 105 is normalized by the driving PSD, converted into a time series signal by the IFFT (Inverse Fast Fourier Transform) converter 104, and the time series signal is converted by the driving waveform synthesis 106. Overlay synthesis processing is performed, and a DA converter 107 generates a drive signal to the vibration exciter device 112.

Japanese Patent No. 4299953

In the conventional random vibration test apparatus 100, when a drive signal is generated to the vibration exciter apparatus 112, the control target correction PSD 102 obtained by comparing the target PSD 101 and the average response PSD 108 is subjected to the reverse transfer characteristic multiplication 103 and driven. The average PSD is calculated, but since the white random signal from the random number generator 105 is normalized by the driving average PSD to generate a true random signal, a driving time series signal is generated. Since the corrected phase component was lost, it took a long time to converge the random vibration test.
In addition, since the control target is control focusing only on average PSD, a probability density distribution function that PSD can take is set for each frequency, and random vibration test according to it cannot be performed.

The random vibration test control apparatus according to claim 1 of the present invention for solving the above-mentioned problems is a random vibration test control apparatus (300) that applies a random vibration signal to the vibration exciter apparatus (310) to perform a vibration test of the specimen. ), A target time-series signal generating means (302) for generating a true random time-series signal as a target time-series signal Tt from a PSD pattern (301) under random vibration test, and a control target from the target time-series signal Tt. A target spectrum generating means (303) for converting the target spectrum Tf to a target spectrum Tf which is a frequency series signal, and a response time series signal obtained from a sensor for detecting the acceleration of the specimen, a response spectrum Rf which is a frequency series signal A response spectrum generating means (304) for converting to a vibration time and a drive time series signal Dt for driving the vibration exciter device (310). Drive spectrum generation means (305) for converting into a drive spectrum Df, which is a signal of several series, transfer function generation means (1104) for calculating a transfer function Gf from the response spectrum Rf and the drive spectrum Df, and the transfer function Gf The inverse transfer function generating means (1105) for calculating the inverse transfer function IGf which is the reciprocal of the above and the target spectrum Tf is multiplied by the inverse transfer function IGf to generate the drive Fourier spectrum Dr in which the amplitude and phase in the frequency domain are corrected. Frequency sequence processing means (306) including drive Fourier spectrum generation means (1103) for performing drive time series signal generation means (306) for generating a drive time series signal Dt from the drive Fourier spectrum Dr by inverse Fourier transform and overlap synthesis method ( 307), and the target time-series signal generating means (302) Is a means (501) for generating a Fourier spectrum with a random phase having a spectrum distribution of the PSD pattern (301) to be tested, an IFFT transform (inverse Fourier transform) of the Fourier spectrum with a random phase, and is called pseudo-random A pseudo-random time-series signal generating means (502) for generating a time-series signal, a process (503) for multiplying the pseudo-random with a Hanning Window and a process for superposing and synthesizing every half frame (504) A true random time series signal generating means (505) for generating a random time series signal is provided, and the target random time series signal and the response time series signal are controlled so as to match both amplitude and phase.
The random vibration test control device according to a second aspect of the present invention for solving the above-described problem is the random vibration test control device according to the first aspect, wherein the frequency sequence processing means (306) A corrected target spectrum generating means for comparing the average PSD and the average PSD of the response spectrum Rf by a comparator (1101) and generating a target spectrum (hereinafter referred to as “corrected target spectrum”) Tc corrected for the difference of the average PSD. 1102).

Transfer function correction of true random time series signal based on PSD pattern numerically defined by PSD probability density distribution function that distributes mean PSD and mean PSD as mean value for each frequency component that is the target of random vibration test Generated before, this true random time-series signal is subjected to FFT processing to obtain a target spectrum signal in frequency space.
Since the target spectrum signal and the response spectrum signal can be controlled by the transfer function correction calculation in the frequency space, the random vibration control in which the target time series signal and the response time series signal match can be performed.
As a result, the convergence time of the random vibration test control device is improved and a random test with a probability density distribution function of PSD for each frequency is realized.

1 is a block diagram of a random vibration test control device according to an embodiment of the present invention. It is a graph which shows an average PSD pattern. It is a flowchart which shows the flow of the process in a target time series production | generation part. It is a graph which shows the relationship between the Fourier spectrum component and the deflection angle θ of the complex plane. It is a flowchart which shows the flow of the process in a target spectrum production | generation part. It is a flowchart which shows the flow of a process in a response spectrum production | generation part. It is a flowchart which shows the flow of a process in a drive spectrum production | generation part. It is a block diagram which shows the detail of a frequency series process part. It is a matrix showing a general transfer function for correction. It is a block diagram which shows the detail of a time series decoding process part. It is a graph which shows the time series waveform of the control result of the random vibration test by the random vibration test control apparatus of this invention. It is a graph which shows the probability density distribution function in the arbitrary frequency in the case of having a separate PSD frequency distribution for each frequency. It is a graph which shows the probability density distribution function which added RandPSD and RandC. A PSD probability density distribution function at each frequency is generated from the average value, standard deviation (σ), and skewness (SK) of the PSD for each frequency obtained from actual field data, and this is used as a control target PSD pattern. It is a graph which shows the control result obtained by vibrating to. It is a graph which shows frequency distribution of PSD in frequency 5.00Hz of Drawing 12C. It is a graph which shows frequency distribution of PSD in frequency 40.00Hz of Drawing 12C. It is a block diagram which shows the hardware constitutions of the random vibration test apparatus by this invention. It is a block diagram which shows the conventional random vibration test control apparatus. It is a graph which shows the relationship between instantaneous PSD and average PSD.

  Hereinafter, embodiments of the present invention will be described with reference to examples shown in the drawings.

A random vibration test control apparatus according to an embodiment of the present invention is shown in FIGS.
First, the overall control process of the random vibration test control apparatus will be described with reference to FIG.
As shown in FIG. 1, the random vibration test control apparatus 300 includes a target time series generation unit 302, a target spectrum generation unit 303, a response spectrum generation unit 304, a drive spectrum generation unit 305, a frequency sequence processing unit 306, and a time series decoding processing unit. 307, a DA conversion unit 308, and an AD conversion unit 309, and a vibration vibration exciter device (including a power amplifier and a vibration response detector) 310 is connected to the random vibration test control device 300.
The PSD pattern 301 to be tested is given as a test condition for a random vibration test.

<Target time series generation unit 302>
The details of the target time series generation unit 302 will be described in the case where the PSD pattern 401 in FIG. 2 is given as the PSD pattern 301 under test.
The random vibration test must be performed with a random signal having a spectral distribution defined by the PSD pattern 401.
In the generation of the random signal, a random signal having the PSD pattern 401 is generated for each frame, and a random time series signal in which the PSD pattern 401 is sustained is generated by sequentially repeating this.

The requirements for complete white random are uniform random appearance frequency in time series space, uniform random amplitude distribution in frequency space, and uniform random phase distribution in frequency space.
In the vibration test defined by the PSD pattern 401, the amplitude distribution in the frequency space is replaced with the PSD pattern 401 instead of uniform random, and the appearance frequency in the time series space is uniform random and the phase distribution in the frequency space is random. A random signal that meets the requirements will be generated.

A case where the target time series generation unit 302 generates a true random time series signal from the PSD pattern 401 will be described with reference to FIG.
First, a Fourier spectrum having a random phase is generated from the PSD pattern 401 (step 501).
Assuming that the fundamental frequency f ₀ = 1 / T and the fundamental angular frequency ω ₀ = 2πf ₀ , the time series signal x (t) and its Fourier transform X _n

There is the following relationship between the real component a and the imaginary component b of the output of the FFT, which is a discrete Fourier transform.

Between the real component a _n and the imaginary component b _n and PSD (n),

However,
N: Number of sampling points of time series data during Fourier transform
N = 2 ⁱ i = 1, 2, 3, 4, 5...
n: 1, 2, 3, ..., (N / 2-1)
Δf: Frequency resolution at the time of Fourier transform
Δf = 1 / (dt · N)
dt: There is a relationship of sampling time intervals of time series data, and Absolute (n) of the Fourier spectrum uses PSD (n),

It is expressed. An asterisk (*) indicates a product operator. Hereinafter, the same applies to the drawings. The Fourier spectrum component indicating PSD (n), where Absolute is the base and the declination of the Fourier spectrum is θ,

Can be written.
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 (PSD value) 601 shown in FIG.
Absolute (PSD value) 601 has a degree of freedom in the length direction, and can generate an arbitrary distribution according to a predetermined PSD density distribution.
On the other hand, since the phase value θ has 360 degrees of freedom, Absolute (PSD value) 601 can be generated by a probability density distribution function of PSDs having different lengths for each frequency, and θ can be changed randomly. For example, a random phase Fourier spectrum satisfying an arbitrary PSD probability density distribution function can be generated.
If a Fourier spectrum having a random phase is subjected to IFFT processing (inverse FFT transform), an average PSD pattern 401 in which sine waves having different phases are overlapped, and a spectrum of probability density distribution functions of PSDs different for each frequency as shown in FIG. 12A. A normally distributed random time series signal having a distribution is generated (step 502).
A process of generating Absolute (PSD value) 601 from the probability density distribution function of PSD will be described with reference to FIGS. 12A and 12B.
An example of a PSD probability density distribution function distributed from PSD _Min to PSD _Max for each frequency is shown in FIG. 12A.
PSD (i) in FIG. 12A is a value calculated for each PSD resolution interval (ΔPSD), and Pr (i) is a probability density at which PSD (i) appears.
As shown in FIG. 12B, white (uniform) random numbers RandPSD = PSD (i) appearing evenly at resolution intervals (ΔPSD) from PSD _Min to PSD _Max are generated based on the Rand function, and the probability density distribution function The probability density Probab in RandPSD is obtained from the above, and another white random number RandC uniformly distributed from 0.0 to 1.0 is generated at the same time.
As shown in FIG. 12B, when the value of RandC ≦ Probab, RandPSD is adopted as the output value, and otherwise it is discarded.
Here, RandC is used as a sieve for adopting / destroying PSD (i) as a PSD generation value in order to appear at the probability density of Probab.

However, IFFT processing used for digital signal processing is a discrete process, so it does not include all frequency components without omission, and there is an effect that missing components are made up by mixing the phases, but there are still some missing components. Since this occurs, this normally distributed random time series signal is called pseudo-random. In order to generate true random from pseudo-random, the following processing is performed.
A process of multiplying a time series signal for each pseudo-random frame by a Hanning Window is performed (step 503).
A process of overlapping the random time series signal multiplied by the Hanning window every 1/2 frame is performed (step 504).
The time series are combined by the process of multiplying the Hanning window and the overlap synthesis process every 1/2 frame to generate a true random time series signal maintaining the PSD pattern 401 defined as the pattern under test (step 505). ).
The true random time series signal is sent to the target spectrum generation unit 303 as the target time series signal Tt.

<Target spectrum generation unit 303>
Details of the target spectrum generation unit 303 of FIG. 1 will be described with reference to FIG.
In order to complete the control of the random vibration test, the FFT is performed from the target time series Tt generated from the PSD pattern, the response time series Rt generated from the response signal, and the drive time series Dt generated from the drive signal. By processing, it is necessary to obtain the amplitude and phase for each of the divided frequencies (Δf) in the frequency band under test, and to control the amplitude and phase.

Assuming that the fundamental frequency f ₀ = 1 / T and the fundamental angular frequency ω ₀ = 2πf ₀ , the time series signal x (t) and its Fourier transform X _n

Between the real component a and the imaginary component b of the FFT output, which is a discrete Fourier transform

There is a relationship.
(A _n −ib _n ) is called a Fourier spectrum, includes a phase component, and is obtained for each frame. As shown in FIG. 5, the Fourier spectrum obtained by subjecting the target time series Tt to FFT conversion is called a target instantaneous Fourier spectrum 1011, the target instantaneous PSD 1012 is calculated from the target instantaneous Fourier spectrum 1011, and the time average with respect to the target instantaneous PSD 1012 is calculated. As a result, a target average PSD 1013 is obtained.
The target instantaneous Fourier spectrum 1011 includes amplitude and phase components, but the target instantaneous PSD 1012 and the target average PSD 1013 do not include phase components.
A target spectrum Tf as an aggregate of the target instantaneous Fourier spectrum 1011, the target instantaneous PSD 1012, and the target average PSD 1013 is generated.

<Response spectrum generation unit 304>
The operation of the response spectrum generation unit 304 in FIG. 1 will be described with reference to FIG.
The response analog signal from the vibration exciter apparatus 310 is converted into a response time series signal Rt by the AD conversion unit 309 and sent to the response spectrum generation unit 304.
The response time-series signal Rt is also subjected to FFT processing on the time-series signal similarly cut out for each frame at the same time as the target time-series signal Tt, and the amplitude and phase of each divided frequency (Δf) in the frequency band to be tested are obtained. Ask.
As shown in FIG. 6, the response instantaneous Fourier spectrum 1021 is calculated from the response time series signal Rt by FFT processing, the response instantaneous PSD 1022 is calculated from the response instantaneous Fourier spectrum 1021, and the time average with respect to the response instantaneous PSD 1022 is calculated. An average PSD 1023 is determined.
The response instantaneous Fourier spectrum 1021 includes amplitude and phase components, but the response instantaneous PSD 1022 and the response average PSD 1023 do not include phase components.
A response spectrum Rf as an aggregate of the response instantaneous Fourier spectrum 1021, the response instantaneous PSD 1022, and the response average PSD 1023 is generated.

<Drive spectrum generation unit 305>
The operation of the drive spectrum generation unit 305 in FIG. 1 will be described with reference to FIG.
A drive time series signal Dt generated as a drive signal of the vibration exciter apparatus 310 is branched at the previous stage where it is converted into a drive analog signal by the DA converter 308 and sent to the drive spectrum generator 305.
The drive time series signal Dt is also subjected to FFT processing on the time series signal cut out for each frame at the same time as the target time series signal Tt and the response time series signal Rt, and each frequency (Δf) divided in the frequency band to be tested Find the amplitude and phase of each.
As shown in FIG. 7, the driving instantaneous Fourier spectrum 1031 is calculated from the driving time series signal Dt by FFT processing, the driving instantaneous PSD 1032 is calculated from the driving instantaneous Fourier spectrum 1031, and the time average for the driving instantaneous PSD 1032 is calculated to calculate the driving average. PSD 1033 is required.
The drive instantaneous Fourier spectrum 1031 includes amplitude and phase components, but the drive instantaneous PSD 1032 and the drive average PSD 1033 do not include phase components.
A drive spectrum Df as an aggregate of the drive instantaneous Fourier spectrum 1031, the drive instantaneous PSD 1032, and the drive average PSD 1033 is generated.

<Frequency sequence processing unit 306>
Details of the frequency sequence processing unit 306 of FIG. 1 will be described with reference to FIG.
The basic function of the frequency sequence processing unit 306 is to multiply the target instantaneous Fourier spectrum 1011 by an inverse transfer function that is an inverse characteristic of the transfer characteristic (transfer function including amplitude and phase) of the vibration exciter apparatus 310 that performs a random test, By generating a Fourier spectrum signal for driving that has undergone transfer characteristic correction processing, the target time-series signal and the response time-series signal coincide with each other, and as a result, control that realizes coincidence of instantaneous PSD and average PSD is realized. .

During random vibration control, the average PSD comparator 1101 calculates the difference between the target average PSD 1013 of the target spectrum Tf and the response average PSD 1023 of the response spectrum Rf, and the average PSD correction vector TCv for each frame.
The spectrum correction PID controller 1102 adds / subtracts the average PSD correction vector TCv to / from the average PSD of the target spectrum Tf to generate a corrected target spectrum Tc in which the average PSD is corrected for each frame.
The spectrum correction PID controller 1102 does not function as phase component correction, but performs an agile control operation as PSD amplitude correction.

In the random vibration test, the initial transfer function of the vibration exciter apparatus (including the test body) 310 is identified before the test is performed, or in the process of increasing from the initial target level at the start of the test to the test target level. Identify the initial transfer function.
As the initial transfer function, the transfer function calculator 1104 in FIG. 8 identifies the instantaneous transfer function from the ratio between the response instantaneous Fourier spectrum 1021 of the response spectrum Rf and the drive instantaneous Fourier spectrum 1031 of the drive spectrum Df, and the time average of the instantaneous transfer function is obtained. Find and save the average transfer function.
During the random vibration test, if there is no characteristic change in the transfer characteristic of the vibration exciter device 310, the initial transfer function is continuously used as the correction transfer function Gf for the characteristic correction calculation.

The transfer function calculator 1104 has a function of continuously updating the average transfer function.
When the transfer characteristic of the vibration exciter apparatus 310 changes for some reason and exceeds a certain set value, the transfer function Gf for correction is changed / updated by the transfer function correction / change function of the transfer function calculator 1104 to transfer the transfer function. Corresponding to changes over time.

For the correction transfer function Gf, the inverse transfer function calculator 1105 generates an inverse transfer function IGf.
For the transfer function correction calculation of the corrected target spectrum Tc, the inverse transfer function IGf is used and the inverse transfer function multiplier 1103 performs the correction calculation to generate the drive Fourier spectrum Dr with the amplitude and phase corrected.

In the case of simultaneous random vibration test control using a plurality of shakers (multi-axis), the correction transfer function Gf is generally expressed in a matrix as shown in FIG.
9, components G ₁₁ , G ₂₂ , G ₃₃ ,..., G _nn are transfer functions of their own excitation axes, and components G ₁₁ , G ₂₂ , G ₃₃ ,. _The components other than _nn are called interference components, and the motions of the individual excitation shafts interfere with each other, affecting each other's transfer functions.

When the simultaneous random vibration test is controlled by a plurality of vibrators (multi-axis) including interference components in the control method according to the present invention, random time series signals for the PSD pattern under test for each axis are generated in advance, After the target random time-series signal is converted into a spectrum signal in the frequency space, the transfer function correction control calculation including an interference component is performed so that the response spectrum signal of each axis matches the target spectrum signal. Multiple shaker (multi-axis) simultaneous random vibration test is possible.
In the present invention, in particular, since an antiphase component with respect to the interference component can be positively added, stable control can be performed even when the interference component becomes larger than the excitation average PSD value of its own excitation axis.

<Time-series decoding processing unit 307>
Details of the time-series decoding processing unit 307 in FIG. 1 will be described with reference to FIG.
The basic function of the time series decoding processing unit 307 is to decode a Fourier spectrum signal for driving from a frequency series signal to a time series signal by IFFT transform (inverse fast Fourier transform), and to perform discontinuous decoding between frames. In this process, an overlap process is performed on the time-series signal thus generated to generate a continuous time-series signal.

The control cycle of the time series decoding processing unit 307 operates in synchronization with the control sampling clock (5.12 K samples / second in the case of the control band of 2 KHz) except for the processing of the IFFT conversion 1205 and the overlap synthesis 1206.
The operations of IFFT transformation 1205 and overlap synthesis 1206 operate as background processing of the time-series decoding processing unit 307, and the calculation is completed within the processing time of one frame.
The generated drive time series signal Dt is stored in the loop buffer and used in the next control cycle.

In the signal calculation in the frequency series, since the calculation is always performed on the time series signal of the appropriate length (frame) determined by the FFT calculation logic, when the IFFT transform (inverse fast Fourier transform) is obtained from the signal calculation result The sequence signal is discontinuous between frames.
Since the time series signal is originally subjected to control calculation on the premise that it is piecewise continuous, overlap processing is performed in the overlap synthesis 1206 in order to remove the discontinuity of the time series signal.
In the present invention, overlap processing is performed every 1/2 frame.

In order to control the time series signal at a time interval shorter than the control for each frame, the target / response time series comparison 1201 is provided, and the target time series Tt and the response time series Rt are compared to generate the driving time series correction amount Dc. To do.
The driving primary time series Dtc generated by the overlap synthesis 1206 is PID-corrected by the time series signal PID correction 1208 by the driving time series correction amount Dc to generate the driving time series Dt.
The signal of the drive time series Dt is converted into an analog signal by the DA converter 308 and output as a drive signal of the vibration exciter device 310.

FIG. 11 is a time-series waveform of the control result of the random vibration test by the random vibration test control apparatus of the present invention.
As shown in FIG. 11, the target time series waveform and the response time series waveform match. The drive signal is generated so that the target time-series waveform and the response time-series waveform match.

FIG. 12C generates a probability density distribution function of the PSD at each frequency from the average value, standard deviation (σ), and skewness (SK) of the PSD for each frequency obtained from actual field data, and this is used as the control target PSD. This is control result data actually obtained as a pattern.
Even when different probability density distribution functions of PSD are given for each frequency, random control is performed in which the target average PSD and the response PSD match.
FIG. 12D is a measurement of the PSD frequency distribution at a frequency of 5.00 Hz in FIG. 12C, which is in good agreement with the control target value, and random control is performed according to the probability density distribution function.
FIG. 12E is a measurement of the PSD frequency distribution at a frequency of 40.00 Hz in FIG. 12C. The PSD frequency distribution matches well with the control target value, and random control is performed according to the probability density distribution function.

<Hardware configuration and operation>
The hardware configuration and operation of the random vibration test apparatus according to the present invention will be described with reference to FIG.
The basic hardware configuration is completely synchronized with the basic sampling clock of 51.2K samples / second, independent of the operation of the small computer 2101 and the small computer 2101 equipped with general commercial operating software. It comprises a control function that operates and a sub-function device 2102 that performs input / output conversion of analog signals.

The small computer 2101 and the sub-function device 2102 exchange information with each other via a DMA (Direct Memory Access) circuit 2105 capable of high-speed data transfer.
The DMA circuit 2105 performs a mutual transfer of a large amount of signal data by sewing a gap between hardware operations of the small computer 1201 without performing a read / write operation for each unit data by the small computer 2101.

  The small computer 2101 is provided with an arithmetic control unit 2103 and a multi-channel FIFO (first-in first-out) 2104, and the sub-function device 2102 is provided with a loop buffer 2106, an arithmetic control unit 2107, and an AD / DA converter 2108.

Since the commercial operating software of the small computer 2101 does not have a real time control function, the small computer 2101 cannot perform the real time operation required for the random vibration test control device, but the operation speed is a sub-function device. It is much faster than 2102.
The sub-function device 2102 includes 51.2 Ksamples / second in which all the functional blocks including the AD / DA converter 2108 oscillate from the inside of the sub-function device 2102 independently of the operation of the small computer 2101. It operates in complete synchronization with the basic sampling clock.
The small computer 2101 and the sub-function device 2102 execute asynchronous operations that are independent in time.

  In the present invention, in order to utilize the high speed of the small computer 2101, the process of random vibration control in real time is divided into a process that depends on the sampling clock and a process that does not depend on the sampling clock, and the process that depends on the sampling clock is performed. By implementing the sub-function device 2102 and the processing independent of the sampling clock in the small computer 2101, the random vibration test device as a whole can achieve both high-speed processing and real-time processing.

The small computer 2101 takes advantage of its high speed, generates a target signal and a part of the drive signal for random vibration control, and performs a prior write to the multi-channel FIFO 2104 regardless of the control progress of the sub-function device 2101.
In the multi-channel FIFO 2104, one stage of the target signal and drive signal for a maximum of 60 seconds is stored in advance.

The sub-function device 2102 has a loop buffer 2106 having the same capacity as the multi-channel FIFO 2104 mounted on the small computer 2101.
The sub-function device 2102 fetches one signal from the loop buffer 2106 in synchronization with the sampling clock and executes control. When the control for one frame is completed, the small computer 2101 indicates that the loop buffer 2106 is empty. Notify

  When one frame of free space occurs in the loop buffer 2106 of the sub-function device 2102, the signal accumulated in advance in the small computer 2101 is automatically used by the DMA circuit 2105 to automatically open the loop in the sub-function device 2102. Send to buffer 2106 area.

  As described above, even if the small computer 2101 that operates at a unique timing speed and the sub-function device 2102 that executes AD / DA processing in complete synchronization with the sampling clock operate asynchronously with each other, the entire random vibration testing device is obtained. Real-time control is achieved that is fully synchronized with the sampling clock.

  INDUSTRIAL APPLICABILITY The present invention can be widely used industrially as a random vibration test control device for evaluating the seismic resistance and durability of an industrial product by giving the industrial product the same random vibration as in the use environment in a laboratory. .

DESCRIPTION OF SYMBOLS 300 Random vibration test control apparatus 301 PSD pattern to be tested 302 Target time series generation unit 303 Target spectrum generation unit 304 Response spectrum generation unit 305 Drive spectrum generation unit 306 Frequency sequence processing unit 307 Time series decoding processing unit 308 DA conversion unit 309 AD conversion 310 vibration exciter device

Claims (2)

In the random vibration test control device (300) for applying a random vibration signal to the vibration exciter device (310) and performing a vibration test of the test body,
Target time series signal generating means (302) for generating a true random time series signal as a target time series signal Tt from a PSD pattern (301) to be tested in a random vibration test;
Target spectrum generating means (303) for converting the target time series signal Tt into a target spectrum Tf which is a frequency series signal to be controlled;
Response spectrum generating means (304) for converting a response time series signal obtained from a sensor for detecting the acceleration of the test body into a response spectrum Rf which is a frequency series signal;
Drive spectrum generation means (305) for converting a drive time series signal Dt for driving the vibration exciter device (310) into a drive spectrum Df which is a frequency series signal;
Transfer function generating means (1104) for calculating a transfer function Gf from the response spectrum Rf and the drive spectrum Df, inverse transfer function generating means (1105) for calculating an inverse transfer function IGf which is the reciprocal of the transfer function Gf, and Frequency sequence processing means (306) comprising drive Fourier spectrum generation means (1103) for multiplying the target spectrum Tf by the inverse transfer function IGf to generate a drive Fourier spectrum Dr in which the amplitude and phase in the frequency domain are corrected;
Driving time series signal generating means (307) for generating a driving time series signal Dt from the driving Fourier spectrum Dr by an inverse Fourier transform and an overlap synthesis method;
With
The target time series signal generating means (302)
Means (501) for generating a phase-random Fourier spectrum having a spectral distribution of the PSD pattern under test (301);
The phase-random Fourier spectrum is subjected to IFFT transform (inverse Fourier transform) to generate a pseudo-random time-series signal generating means (502) for generating a time-series signal called pseudo-random, and the pseudo-random is multiplied by a Hanning Window. A true random time series signal generating means (505) for generating a true random time series signal by the process (503) and the process of superposing and synthesizing every half frame (504),
A random vibration test control apparatus that performs control such that a target random time-series signal and a response time-series signal are matched in amplitude and phase.   The frequency sequence processing means (306) compares the average PSD of the target spectrum Tf before the inverse transfer function IGf is multiplied with the average PSD of the response spectrum Rf by a comparator (1101), and the difference between the average PSDs The random vibration test control apparatus according to claim 1, further comprising a corrected target spectrum generation unit (1102) that generates a target spectrum Tc that is corrected.

Images