JP3220418B2 - Signal control system and signal control method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a signal control system and a signal control method, and more particularly to a technique for controlling a drive signal given to a controlled system driven using a plurality of drive signals.

[0002]

2. Description of the Related Art A device that controls the vibration of a specimen to give a target vibration has been developed (for example, Kazuyoshi Ueno: “Random Vibration Controller SX-2000”, Nihon Kogyo Shuppan, Measurement Technology, 15, No. 3, pp. 75-79). This is a so-called random vibration control device. It is used for controlling a vibration generator by a random vibration control device, giving a target vibration to a specimen, and testing the vibration resistance of the specimen.

[0003] Vibrations applied to equipment (test pieces) placed in an actual operating environment or transport environment generally have an irregular waveform. Therefore, in this random vibration control device,
A PSD (power spectrum density) of an irregular waveform actually applied is calculated, and a vibration having this PSD is applied to the specimen.

A point to be considered here is that the vibration generator itself also has a frequency response characteristic. Therefore, simply providing a drive waveform having a target PSD to the vibration generator cannot provide a device having the target PSD. In consideration of the frequency response characteristics of the vibration generator, the drive PSD
And a drive waveform having this drive PSD is generated and supplied to the vibration generator.

As shown in FIG. 30, a plurality of vibration systems including such a random vibration control device and a vibration generator (vibrator) are prepared, and vibration is simultaneously applied to the specimen from each vibration system. It is conceivable to perform a multidimensional vibration test on the test sample by providing a desired target PSD to each of the vibrating systems while configuring so as to obtain the vibration.

However, such a conventional multidimensional vibration test has the following problems. In the conventional multidimensional vibration test, each vibration system was regarded as a system independent of each other. However, in practice, since all the exciters constituting the exciter system are connected in one specimen, there is a limit in mechanically avoiding the occurrence of interference between the exciters.

In the so-called multi-point vibration system shown in FIG. 31, there is in principle interference between the vibrators.

Therefore, the distortion of the response based on the interference between the vibrators cannot be corrected by the conventional multidimensional vibration test in which the respective vibrating systems are regarded as independent systems. For this reason, it has not been possible to accurately reproduce each of the target PSDs or accurately apply a vibration having a desired relationship to the target PSD to the specimen.

The present invention solves such a problem,
It is an object of the present invention to provide a signal control system and a signal control method capable of accurately controlling a response of a controlled system driven using a plurality of drive signals so as to have a desired relationship with an original target. I do.

[0010]

According to a first aspect of the present invention, there is provided a signal control system comprising control characteristic storing means for storing control characteristics of a controlled system, wherein a response of the controlled system driven by using a plurality of drive signals is used. In consideration of the possibility of being affected by the plurality of drive signals, the plurality of drive signals and the control system for this drive signal are controlled so that the response is substantially equal to the control target converted into a signal on the time axis. The control characteristic of the controlled system stored in the control characteristic storage means is updated based on the plurality of main responses, and the signal on the frequency axis and the control characteristic obtained based on the control target converted into the signal on the time axis The transfer function matrix of the controlled system stored in a matrix
First control means for generating a drive signal on the time axis based on a signal on the frequency axis obtained as a product of the control characteristic of the controlled system represented by an inverse matrix of Second control means for generating a control target based on the response and the original target given by the power spectral density.

[0011] The signal control system according to claim 2 is based on claim 1.
In the above signal control system, the second control means includes:
The control target is generated such that the amount obtained based on the response is substantially equal to the original target.

According to a third aspect of the present invention, there is provided a signal control system.
In the above signal control system, the second control means includes:
The control target is generated such that the amount obtained based on the response does not exceed a given limit value.

According to a fourth aspect of the present invention, there is provided a signal control system.
In the above signal control system, the second control means includes:
Provisional control target generation means for generating a tentative control target based on the response and the original target so that the amount obtained based on the response does not exceed a given limit value; A control target generating means for generating a control target based on the response and the temporary control target so as to be substantially equal to the temporary control target is provided.

According to a fifth aspect of the present invention, there is provided a signal control system.
In the signal control system according to any one of claims 4 to 6, one or more control groups are set, and an original target, a second control means, a control target, a drive signal, and a response belonging to each control group are defined. One or more predetermined responses among the responses belonging to the control group are defined as a main response of the control group, and the first control means includes control characteristic storage means for storing control characteristics of the controlled system. Considering that the main response belonging to the control group may be affected by the drive signals belonging to the control group other than the corresponding control group, a plurality of drive signals and a plurality of main responses from the controlled system to the drive signal are provided. The control characteristic of the controlled system stored in the control characteristic storage means is updated based on the above, and the signal on the frequency axis obtained based on the control target converted into the signal on the time axis and the matrix in the control characteristic storage means Remembered in format The transfer function matrix of the controlled system were
Drive signal on the time axis based on the signal on the frequency axis obtained as the product of the control characteristics of the controlled system expressed by the inverse matrix of the drive within the time required to guarantee real-time performance. By performing the calculation and giving it to the controlled system, the main response belonging to each control group is controlled so as to be substantially equal to the control target converted into a signal on the time axis belonging to the corresponding control group. Wherein the second control means generates a control target belonging to the corresponding control group based on the response belonging to the corresponding control group and the original target.

According to a sixth aspect of the present invention, there is provided a signal control system.
In the signal control system of (1), the second control means belonging to each control group performs processing across control groups by also referring to responses belonging to control groups other than the corresponding control group .
That is, a control target belonging to a corresponding control group is generated so as to be an appropriate control target .

A signal control system according to claim 7 is a signal control system according to claim 5.
7. The signal control system according to claim 6, wherein at least some of the responses including the main response and the drive signal are substantially represented as signals on a time axis, and the original target and the control target. It is expressed as an amount corresponding to the amplitude spectrum on the frequency axis, the control target
The converted into signals on the time axis a first conversion means for providing the first control means, converts the response expressed as a signal on a time axis in an amount corresponding to the frequency characteristic second control means And a second conversion means for providing

The signal control system according to claim 8 is a signal control system according to claim 7.
In the signal control system, the first control means includes:
Control characteristic storage means for storing control characteristics of the controlled system,
Based on the control target converted to a signal on the time axis by the conversion means and the control characteristics stored in the control characteristic storage means, the drive signal is calculated within the time required to guarantee real-time performance. And a control signal stored in the control characteristic storage means based on the drive signal calculation means provided to the controlled system, the drive signal output from the drive signal calculation means, and a main response to the drive signal from the controlled system. And control characteristic updating means for updating the control characteristic.

According to a ninth aspect of the present invention, there is provided a signal control system according to the eighth aspect.
Wherein the control characteristic is expressed by a predetermined inverse matrix of a transfer function matrix of the controlled system having a plurality of transfer functions of the controlled system representing a relationship between a drive signal and a main response. It is characterized by.

According to a tenth aspect of the present invention, in the signal control system of the ninth aspect, the drive signal calculating means performs a Fourier transform on a control target signal vector having a control target converted into a signal on a time axis as an element. Fourier transform means for obtaining a control target spectrum vector having a complex Fourier spectrum as an element, multiplication means for obtaining a drive spectrum vector which is a product of a predetermined inverse matrix of the transfer function matrix and the control target spectrum vector, a drive spectrum vector Is inverse Fourier-transformed to output a drive signal vector having a drive signal as an element.

According to an eleventh aspect of the present invention, there is provided the signal control system according to any one of the first to tenth aspects, wherein the controlled system receives the drive signal from the test object and drives the test object by a plurality of vibrators. When a vibration generator that gives vibration and a vibration detector that detects vibration of the specimen at one or more response points are provided, the purpose is to provide the specimen with vibration having a predetermined relationship with the original target. And

[0021]

[0022]

[0023]

A storage medium according to a twelfth aspect is a computer-readable storage medium storing a computer-executable program, and the program realizes the system according to any one of the first to eleventh aspects. There is a feature.

[0025]

According to a first aspect of the present invention, there is provided a signal control system comprising a control characteristic storing means for storing control characteristics of a controlled system, wherein a response of the controlled system driven by using a plurality of drive signals is controlled by a plurality of drive signals. And a plurality of main responses from the controlled system to the drive signal such that the response becomes substantially equal to the control target converted into a signal on the time axis. The control characteristic of the controlled system stored in the control characteristic storage means is updated based on the above, and the signal on the frequency axis obtained based on the control target converted into the signal on the time axis and the matrix in the control characteristic storage means Matrix of the transfer function matrix of the controlled system stored in the form
A drive signal on the time axis is generated based on the signal on the frequency axis obtained as a product of the control characteristic of the controlled system expressed by the following equation, and given to the controlled system. Therefore, even in a controlled system whose response is affected by a plurality of drive signals, a desired control target can be reproduced while adjusting the effect of each drive signal on the response.

The control target is a response and a power spectrum.
Generated based on the original goal given in density . Therefore, by controlling the control target to be adjusted according to the response, it is possible to control the response to have a desired relationship with the original target.

That is, by performing complementary control by combining two controls having different properties, the response of the controlled system driven using a plurality of drive signals has a desired relationship with the original target. Can be controlled precisely.

The signal control system according to claim 2 is characterized in that the control target is generated such that the amount obtained based on the response is substantially equal to the original target. Therefore,
In a controlled system driven using a plurality of drive signals, an original target can be accurately reproduced.

A signal control system according to a third aspect is characterized in that a control target is generated such that an amount obtained based on a response does not exceed a given limit value. Therefore, it is possible to accurately control the response of the controlled system driven using the plurality of drive signals so as to maintain the predetermined constraint.

In the signal control system according to the present invention, a temporary control target is generated based on the response and the original target so that the amount obtained based on the response does not exceed a given limit value. The control target is generated based on the response and the provisional control target so that the amount obtained in this manner is substantially equal to the provisional control target.

Therefore, it is possible to set a tentative control target so that the response of the controlled system maintains a predetermined constraint, and to set the control target so that the tentative control target can be accurately reproduced. Therefore, the control of the controlled system driven using the plurality of drive signals maintains the predetermined constraint condition, and the control is performed more accurately so that the response has a desired relationship with the original target. be able to.

A signal control system according to a fifth aspect of the present invention includes a control characteristic storing means for setting one or more control groups and storing control characteristics of a controlled system, wherein a main response belonging to each control group corresponds to a corresponding control group. In consideration of the possibility of being affected by a drive signal belonging to a control group other than the control group, the control signal is stored in the control characteristic storage means based on a plurality of drive signals and a plurality of main responses to the drive signal from the controlled system. The control characteristic of the controlled system is updated, the signal on the frequency axis obtained based on the control target converted into the signal on the time axis, and the transfer function of the controlled system stored in a matrix form in the control characteristic storage means. matrix
Calculates the drive signal on the time axis within the time required to guarantee real-time performance based on the signal on the frequency axis obtained as the product of the control characteristics of the controlled system expressed by the inverse matrix of By giving the control response to the controlled system, the main response belonging to each control group is controlled to be substantially equal to the control target converted into a signal on the time axis belonging to the corresponding control group. Therefore, by performing control so as to eliminate the influence of the control target belonging to another control group from the main response belonging to each control group, the control target belonging to each control group can be independently reproduced.

The control targets belonging to each control group are generated based on the responses belonging to the corresponding control group and the original targets. Therefore, the control targets belonging to each control group are generated only by the processing in the control group.

That is, a series of control processes are classified into processes that extend between control groups and processes that are completed within the control group. By performing these processes in combination, a controlled system that is driven using a plurality of drive signals is controlled. Signal control for the system can be performed more easily.

The signal control system according to claim 6, between the control group also see the response belonging to the control group other than one control group
By performing the process, a control target belonging to a corresponding control group is generated so as to be an appropriate control target . Therefore, when a response belonging to one control group is affected by a control target belonging to another control group, an appropriate control target can be generated by performing processing across control groups.

In the signal control system according to the present invention, at least some of the responses including the main response and the drive signal are substantially represented as signals on the time axis, and the original target and the control target are expressed on the frequency axis. As an amount corresponding to the amplitude spectrum of In addition, the control target is converted into a signal on the time axis, and the response expressed as the signal on the time axis is converted into an amount corresponding to the frequency characteristic.

Therefore, it is possible to control the response of the controlled system in consideration of the fact that the response of the controlled system can be affected by a plurality of drive signals by treating the quantity substantially expressed as a signal on the time axis. Become.

The signal control system of claim 8, the control system is stored as a control target that has been converted into a signal on the time axis
A drive signal is calculated within the time required for guaranteeing real-time performance based on the control characteristics of the control signal and given to the controlled system, and stored based on the drive signal and the main response from the controlled system to this drive signal. It is characterized in that the set control characteristics are updated.

Therefore, a drive signal can be given to the controlled system in real time. In addition, by updating the control characteristics as appropriate, the control characteristics can be reproduced more faithfully.

According to a ninth aspect of the present invention, in the signal control system, a predetermined inverse matrix of a transfer function matrix of the controlled system having a plurality of transfer functions of the controlled system as elements whose control characteristics indicate a relationship between the drive signal and the main response. It is characterized by being represented by Therefore, the control characteristics can be described in a relatively simple manner. Therefore, it is easy to grasp and analyze the control characteristics of the controlled system.

According to a tenth aspect of the present invention, a control target signal vector having a complex Fourier spectrum as an element is obtained by Fourier-transforming a control target signal vector having a control target converted into a signal on a time axis as an element. A drive spectrum vector which is a product of a predetermined inverse matrix of a transfer function matrix and a control target spectrum vector is obtained, and the drive spectrum vector is subjected to an inverse Fourier transform to output a drive signal vector having a drive signal as an element. .

That is, the control target represented by the signal on the time axis is converted into an amount on the frequency axis that is easy to calculate, and the product of the converted amount and the inverse matrix is obtained by calculation. Re-convert to the drive signal on the time axis. In this way, the calculation time can be reduced.

According to a eleventh aspect of the present invention, in the signal control system, the system to be controlled includes a specimen, a vibration generator that receives a drive signal and vibrates the specimen by a plurality of vibrators, and a vibration of the specimen. It is an object of the present invention to provide a test specimen with a vibration having a predetermined relationship with an original target when the apparatus has a vibration detector that detects the above response point.

Therefore, when a vibration test is performed using a plurality of exciters, it is possible to remove the distortion of the response due to the interference between the exciters. For this reason, the original target can be accurately reproduced, and vibration having a desired relationship with the original target can be accurately applied to the test sample.

A storage medium according to a twelfth aspect is characterized by storing a program for realizing the system according to any one of the first to eleventh aspects using a computer. Therefore, a signal control system or the like can be realized by causing a computer to read. That is, it is possible to provide a signal control system that can more easily and accurately control a response of a controlled system driven by using a plurality of drive signals so as to have a desired relationship with an original target. it can.

[0046]

BEST MODE FOR CARRYING OUT THE INVENTION

[System Configuration] FIG. 1 is an overall block diagram of a vibration control system which is a signal control system according to an embodiment of the present invention. The vibration control system includes Fourier transform means 16 and 26, which are second transform means, and PSD analysis means 1
8, 28, the PSD controllers 12, 2 which are the second control means
2. random wave generators 14, 2 as first conversion means
4. A multi-dimensional waveform controller 8 as a first control means is provided.

Fourier transform means 16, PSD analysis means 1
8, PSD controller 12 and random wave generator 14 belong to control group 10. Similarly, Fourier transform means 26, PSD
Analysis means 28, PSD controller 22, random wave generator 2
4 belongs to another control group 20.

An original target PSD which is an original target, a monitoring PSD which is a limit value, a control target PSD which is a control target, a drive signal for driving the controlled system 2, and a response which is a response of the controlled system 2 to the drive signal. It is defined to which control group each signal belongs. Although FIG. 1 illustrates the case where the number of control groups is two for convenience of explanation, the number of control groups may be one or three or more.

The response signal is roughly classified into a control response signal used as a control input of the vibration control system and a monitor response signal (not shown) for monitoring the system. Of the control response signal or the monitor response signal, a response signal used for limit control described later is particularly called a limit response signal. Therefore, the limit response signal and the control response signal are used as control inputs in a broad sense of the vibration control system shown in FIG. A response signal used as an input of the multidimensional waveform controller 8 among the control response signals is referred to as a main control response signal (corresponding to a main response). In addition,
It is sufficient that at least one control response signal exists in each control group, and the number of monitor response signals and limit response signals does not matter. Further, all of the control response signals may be the main control response signal, and some of the control response signals may be the main control response signal. The number of main control response signals in each control group may be one or more.

In this embodiment, each response signal, drive signal, and the like are substantially represented as signals on the time axis. In addition, the original target PSD, the monitoring PSD, the control target PS
D and the like are represented as power spectral densities.

The control group 10 will be described. The Fourier transform means 16 converts the control response signal and the limit response signal, both of which are signals on the time axis, into a complex Fourier spectrum, and based on each of the obtained complex Fourier spectra, calculates the power spectrum density. The control response PSD and the limit response PSD are calculated, and the limit response PSD is given to the PSD controller 12 among them.

The PSD analysis means 18 performs a predetermined averaging process on each control response PSD among the PSDs obtained by the Fourier transform means 16, thereby obtaining a representative control response PSD.
D is calculated and given to the PSD control means 12. In this embodiment, the thus obtained representative control response P
SD is called an averaging control response PSD.

The PSD analysis means 18 adds each control response PSD obtained by the Fourier transform means 16 to
By performing a predetermined maximum value extraction process, the representative control response P
It is also possible to configure so that SD is calculated and given to the PSD control means 12. The control response PSD obtained by the Fourier transform means 16 is subjected to a process combining the averaging process and the maximum value extraction process, thereby obtaining the representative control response PS.
D may be calculated and provided to the PSD control means 12.

The PSD controller 12 has an averaging control response PS
D, limit response PSD, original target PSD, monitoring PSD,
A control target PSD is generated based on a target level (described later), and a random wave generation PSD is generated.

The PSD controller 12 includes provisional control target generation means 30 and control target generation means 32. The provisional control target generation means 30 generates a provisional control target based on the limit response PSD, the averaging control response PSD, the supplied original target PSD, and the target level so that the limit response PSD does not exceed the supplied monitoring PSD. Generate. The control target generation means 32 generates the control target PSD based on the averaged control response PSD and the temporary control target so that the averaged control response PSD is substantially equal to the temporary control target.
Then, necessary correction is made to the control target PSD to generate a random wave generation PSD.

The random wave generator 14 converts the random wave generation PSD into a control target signal, which is a signal on the time axis, and supplies the control target signal to the multidimensional waveform controller 8. Although the control group 10 has been described above, the same applies to the control group 20.

The multidimensional waveform controller 8 generates each drive signal based on each control target signal given from the random wave generators 14 and 24, and supplies the drive signal to the controlled system 2. The multi-dimensional waveform controller 8 considers, for example, that the main control response signal belonging to the control group 10 can be affected by the drive signal belonging to the control group 20, and the main control response signal belonging to the control group 10 A drive signal is generated and supplied to the controlled system 2 while performing closed loop control so as to be substantially equal to the control target signal belonging to the group 10. Drive signals belonging to the control group 20 are similarly generated.

FIG. 2 is a detailed block diagram showing an example of the multidimensional waveform controller 8 shown in FIG. Multi-dimensional waveform controller 8
Is a transfer characteristic storage means 3 for storing transfer characteristics of the controlled system.
8, drive signal calculating means 40, transfer characteristic updating means 42 for updating the transfer characteristic of the controlled system, D / A converter 44, A
A / D converter 46 and a control target storage means 48 are provided.
A vector having each control target signal as an element is referred to as a control target signal vector, a vector having each drive signal as an element is referred to as a drive signal vector, and a vector having each main control response signal as an element is referred to as a main control response signal vector. Call.

The transfer characteristic storage means 38 stores the transfer characteristics of the controlled system. In this embodiment, the transfer characteristic of the controlled system is represented by an equalization matrix which is a predetermined inverse matrix of the transfer function matrix of the controlled system 2 representing the relationship between the drive signal vector and the main control response signal vector. You. The drive signal calculation means 40 performs drive based on the control target signal vector generated by the random wave generators 14 and 24 (see FIG. 1) and the transfer characteristic of the controlled system stored in the transfer characteristic storage means 38. The signal vector is calculated within the time required for guaranteeing the real-time property, converted into an analog signal by the D / A converter 44, and given to the controlled system 2.

The transfer characteristic updating means 42 is based on the drive signal vector output from the drive signal calculating means 40 and the main control response signal vector from the controlled system 2 for this drive signal vector.
The transfer characteristic of the controlled system stored in is updated. The main control response signal vector from the controlled system 2 is A /
The signal is converted into a digital signal by the D converter 46 and supplied to the transfer characteristic updating means 42.

The control target storage means 48 stores the control target signal vector and outputs it in correspondence with the response signal vector input from the A / D converter 46. These outputs are
The information is displayed on a CRT 88 of a personal computer 76 via a bus control board 72 described later (FIG. 5).
reference).

FIG. 3 shows the drive signal calculating means 40 of FIG.
FIG. 2 is a detailed block diagram showing an example of the first embodiment. The drive signal calculation means 40 includes a Fourier transform means 50, a multiplication means 5
2. Inverse Fourier transform means 54 is provided.

The Fourier transform means 50 performs a Fourier transform on a control target signal vector having a control target signal as a signal on the time axis as an element, and obtains a control target spectrum vector having a complex Fourier spectrum as an element. Multiplication means 5
2 obtains a drive spectrum vector, which is a product of the equalization matrix and the control target spectrum vector. The inverse Fourier transform unit 54 performs an inverse Fourier transform on the drive spectrum vector, and outputs a drive signal vector having a drive signal, which is a signal on the time axis, as an element.

FIG. 4 is a block diagram showing the drive signal calculating means 40 shown in FIG.
FIG. 10 is a detailed block diagram illustrating another example of the present invention. In the example of FIG. 4, a window operating means 56 is provided before the Fourier transform means 50 in the example of FIG.
Is provided with a superimposing means 58 after this.

The window operating means 56 fetches the control target signal vector for each frame, performs a window operation, and supplies it to the Fourier transform means 50. The superposition means 58 superimposes the output from the inverse Fourier transform means 54. As described above, by performing the window operation process and the superimposition process in combination, it is possible to prevent the drive signal vector from becoming discontinuous even if the transfer characteristic of the controlled system changes abruptly. It should be noted that Japanese Patent Application Laid-Open No. 2-213908 describes details of the window operation process and the overlay process.

[Hardware] FIG. 5 shows an example of a hardware configuration and a controlled system 2 when each function of the vibration control system shown in FIG. 1 is realized using a DSP (Digital Signal Processor). FIG. 5 shows an example in which the vibration control system shown in FIG. 1 has a total of three control groups. That is, the vibration control system has an X control group that controls vibration in the X direction in the figure, a Y control group that controls vibration in the Y direction, and a Z control group that controls vibration in the Z direction.

The controlled system 2 includes a specimen 90, a vibration generator 92 which receives the drive signal and vibrates the specimen 90, and an acceleration sensor 94 which is a vibration detector for detecting the vibration of the specimen 90. . The vibration generator 92 includes an X vibrator 92x that generates vibration in the X direction, and a Y that generates vibration in the Y direction.
Exciter 92y, Z exciter 92z that generates vibration in the Z direction
It has. The acceleration sensor 94 includes an X acceleration sensor 94x that detects vibration in the X direction, a Y acceleration sensor 94y that detects vibration in the Y direction, and a Z acceleration sensor 94z that detects vibration in the Z direction.

The X excitation group is formed by the X control group of the vibration control system, the X oscillator 92x and the X acceleration sensor 94x of the controlled system 2. Similarly, the Y control group, the Y vibrator 92y and the Y acceleration sensor 94y of the controlled system 2
Form a Y vibration group, and a Z control group and the controlled system 2
The Z vibrator 92z and the Z acceleration sensor 94z form a Z vibration group.

On the other hand, the vibration control system includes a DSP board 62, an output board 66, an amplifier 68, an input board 70,
A bus control board 72 and a personal computer 76 are provided. DSP board 62, output board 6
6, input board 70, bus control board 72,
They are interconnected via an internal bus 60.
A personal computer 76 is connected to the bus control board 72.

The personal computer 76 has a CPU 7
8, main memory 80, hard disk 82, FDD (flexible disk drive) 84, keyboard 86, C
RT88 is provided. The hard disk 82 stores a program for vibration control, data such as an original target PSD, a monitoring PSD, and the like.

The DSP board 62 includes a DSP 63, which is a microprocessor for performing high-speed operations, and a memory 64. The memory 64 has a bus control board 72
, A program or the like stored in the hard disk 82 of the personal computer 76 is loaded. DS
P63 executes a vibration control process according to the program loaded in the memory 64.

The drive signal output from the DSP 63 is converted into an analog signal by an output board 66 having a D / A converter 44, amplified by an amplifier 68, and supplied to a vibration generator 92. The amplifier 68
Are an X amplifier 68x for amplifying a drive signal given to the X shaker 92x, a Y amplifier 68y for amplifying a drive signal given to the Y shaker 92y, and a Z shaker 92z.
Amplifier 68z for amplifying the drive signal given to the
It has.

The response signal (analog signal) of the controlled system 2 detected by the acceleration sensor 94 is supplied to the A / D converter 46.
Is converted into a digital signal by the input board 70 having the

Note that a program start command, a program stop command, and the like are input from the keyboard 86 of the personal computer 76. These inputs are transmitted to the DSP 63 via the bus control board 72. Also, DSP
The processing result of the program by 63 is displayed on the CRT 88 of the personal computer 76 via the bus control board 72. The bus control board 72 includes a sampling clock generator 74 for giving a sampling timing to the A / D converter 46.
It has.

The exchange of programs and various data with an external device (not shown) is performed by a flexible disk (not shown) via an FDD (flexible disk drive) 84.

[System Operation] -Trial Excitation- FIG. 6 is a flowchart showing the operation of the vibration control system shown in FIG. The operation of the vibration control system will be described with reference to FIGS. First, trial excitation is performed to obtain the transfer function matrix [H] of the controlled system 2 (step S2). The instruction for trial excitation is given by the operator using the keyboard 86 of the personal computer 76.
Via the DSP 63.

FIG. 7 is a detailed flowchart of the trial excitation processing.
Shown in Upon receiving the instruction for the test excitation, the DSP 63 first applies simultaneously uncorrelated random waves to the X excitation group, the Y excitation group, and the Z excitation group, and simultaneously excites the specimen 90.
(Step S20).

The DSP 63 converts the drive signal vector {D (t)} given to the vibration generator 92 into the output board 6
6 and the main control response signal vector {Re} corresponding to the drive signal vector {D (t)}.
spP (t)} is fetched via the input board 70 (step S22).

DSP 63 takes in drive signal vector {D (t)} and main control response signal vector {Re
spP (t)} is Fourier-transformed to obtain a drive spectrum vector {D (f)} and a main control response spectrum vector {RespP (f)} (step S2).
4).

The DSP 63 calculates a transfer function matrix [H] of the controlled system 2 based on the obtained drive spectrum vector {D (f)} and the main control response spectrum vector {RespP (f)}. The vibration processing ends (step S26). The method of calculating the transfer function matrix [H] will be described later.

In this embodiment, the X excitation group, the Y excitation group, and the Z excitation group are provided with mutually uncorrelated random waves at the same time, so that the specimen 90 is simultaneously excited. (Step S20), X excitation group, Y excitation group, Z
It is also possible to provide a configuration in which random waves are sequentially applied to the excitation groups, and the specimen 90 is sequentially and individually excited for each excitation group. In this case, the random wave given to each excitation group is
It need not necessarily be uncorrelated with each other.

-Initial value setting and start-up When the trial excitation processing is completed, various initial values of the vibration control system are set, and then the system is started up (step S4 in Fig. 6). The operator can set the initial values and start the system by using the personal computer 76.
By giving an instruction to the DSP 63 through the keyboard 86 of FIG.

FIG. 8 shows a detailed flowchart 2 of the initial value setting and activation processing. The DSP 63, which has received the instruction of the initial value setting and the system startup processing, first calculates the system equalization matrix [G] based on the transfer function matrix [H] of the controlled system 2 calculated in the previous step (step S30). ). A method for calculating the equalization matrix [G] will be described later.

Next, the DSP 63 calculates the original target PSD vector {Ref ^PSD (f)} and the monitoring PSD vector {P
^PSD (f)｝ is set (step S32). Original goal P
FIG. 15 shows an example of the SD vector {Ref ^PSD (f)}. The original target PSD vector {Ref ^PSD (f)} is X
The original target PSD (Ref ^PSD (f) _X ) of the excitation group, the original target PSD (Ref ^PSD (f) _Y ) of the _Y excitation group, and the original target PSD (Ref ^PSD (f) _Z ) of the _Z excitation group It is a vector to be an element. FIG. 29 shows an example of the monitoring PSD vector {P ^PSD (f)}. In the example of FIG. 29, the monitoring P
Only SD (P ^PSD (f) _X ) is defined, and the monitoring PSD (P ^PSD (f) _Y ) of the _Y excitation group and the monitoring PSD of the Z excitation group
(P ^PSD (f) _Z ) is not defined.

Next, the DSP 63 sets a random number initial value for generating a random wave (control target signal), which will be described later (step S34), and sets the initial value {Rctrl ^PSD (f) ⁰ )} of the control target PSD and Random wave generation PS
The initial value of D {RDR ^PSD (f) ⁰ } is set (step S36).

As will be described later, the random wave generation PSD vector {RDR ^PSD (f)} is converted into the control target signal vector {Rctrl} on the time axis by the random wave generation processing.
(T) After being transformed into {(random wave vector)}, it is Fourier transformed into a control target spectrum vector {Rctrl (f)} on the frequency axis. That is, {RDR ^PSD (f)} = {| Rctrl (f) | ² / {f} (1) Further, the drive spectrum vector {D (f)} is given by the following equation: {D (f) )｝ = [G] ｛Rctrl (f)｝ (2) Therefore, from equations (1) and (2), {RDR ^PSD (f)} = ｛| [G] ⁻¹ {D (f)} | ² / {f} (3) In this embodiment, the initial value of the drive spectrum vector {D (f)} given to the controlled system 2 is set to white noise {D (f) ^w }. You have set. Therefore, the initial value of the random wave generation PSD vector ｛RDR
^From Equation (3), ^PSD (f) ⁰ is given by {RDR ^PSD (f) ⁰ } = {| [G] ^-1 {D (f) ^w } | ² / {f} (4) I can give it.

The initial value {Rctrl ^PSD (f) ⁰ } of the control target PSD vector is also set to the same value.

-Generation of Random Wave- When the initial value setting and the start-up process are completed, the DSP 63 performs a random wave (control target signal) generation process (FIG. 6, step S6). That is, based on the random wave generation PSD vector {RDR ^PSD (f)} having the given power spectrum density as the element, the DSP 63 provides the control target signal vector having the random wave (control target signal) on the time axis as the element. Generate {Rctrl (t)}. In addition, Japanese Patent Publication No. 6-7081 describes details of a random wave generation process.

In this embodiment, the initial value {Rctrl ^PSD (f) ^{0 of the} control target PSD vector is set such that the initial value of the drive spectrum vector {D (f)} becomes white noise {D (f) ^w }. ｝ And random wave generation P
The initial value of the SD vector {RDR ^PSD (f) ⁰ } is set, and the control target PSD vector {Rctrl} is set.
^PSD (f)} gradually becomes the original target PSD vector {Ref
^The control target PSD vector {Rctrl ^PSD (f)} and the random wave generation PSD vector {RDR ^PSD (f)} themselves are sequentially updated so as to approach the ^PSD (f)} (see step S16 described later).

FIG. 16 shows the control target PSD vector {Rctr} after such processing is repeated to some extent.
An example of l ^PSD (f)} is shown. FIG. 17 shows a random wave generation PSD vector {RDR} obtained from the control target PSD vector {Rctrl ^PSD (f)}.
^The control target signal vector {Rct} generated in the step (random wave generation processing) based on ^PSD (f)}
rl (t)}.

—Calculation and Output of Drive Signal— When the random wave generation processing ends, the DSP 63 next calculates a drive signal vector {D (t)} based on the control target signal vector {Rctrl (t)}. This is output (FIG. 6, step S8).

FIG. 9 shows a detailed flowchart of an example of the drive signal vector calculation and output processing. DSP63
Is a control target signal vector {Rct} having a signal on the time axis generated in the random wave generation processing as an element.
By subjecting rl (t)} to Fourier transform, a control target spectrum vector {Rctrl (f)} having a complex Fourier spectrum as an element is obtained (step S42).

FIG. 18 shows the control target spectrum vector {Rctrl (f)} obtained in this step. In FIG. 18, each element of the vector, that is, the control target spectrum of the X excitation group (Rctrl (f) _X ), the control target spectrum of the _Y excitation group (Rctrl (f) _Y ), Z
Control target spectrum of excitation group (Rctrl (f) _Z )
Are complex spectra each having an amplitude component and a phase component.

Next, the DSP 63 calculates the aforementioned equation (2) based on the system equalization matrix [G] obtained before and the control target spectrum vector {Rctrl (f)} obtained in the previous step. , A drive spectrum vector {D (f)} is calculated (step S44).

FIG. 19 shows the drive spectrum vector {D (f)} obtained in this step. In addition,
FIG. 25 shows the equalization matrix [G]. The equalization matrix [G] will be described later.

Next, the DSP 63 performs an inverse Fourier transform on the drive spectrum vector {D (f)} having the complex Fourier spectrum obtained in the previous step as an element, thereby obtaining a drive signal having a signal on the time axis as an element. A vector {D (t)} is obtained (step S46). FIG. 20 shows the drive signal vector {D (t)} obtained in this step.

The DSP 63 outputs the drive signal vector {D (t)} obtained in the previous step to the output board 6
6 and the output via the amplifier 68 (step S4
9). That is, each element D (t) _X , D of the drive signal vector {D (t)} given as a digital signal
(T) _Y and D (t) _Z are D / D of the output board 66, respectively.
The signal is converted into an analog signal by the A
After being amplified by 8, the X vibrator 92 x, the Y vibrator 92 y, and the Z vibrator 92 z constituting the vibration generator 92 are respectively driven. Thus, the drive signal vector calculation and output processing are performed.

FIG. 10 is a flowchart showing another example of the drive signal vector calculation and output processing. Steps S52 to S56 and step S59 in the processing of FIG.
This is the same as steps S42 to S46 and step S49 in the process of FIG. That is, in the process of FIG. 10, the control target signal vector {Rctrl (t)} is fetched for each frame, and the window operation is performed (step S
50), and perform Fourier transform and the like (steps S52 to S56). A drive signal vector {D (t)} is obtained by superimposing the resulting signals (step S).
58), and outputs it (step S59). As described above, even when the equalization matrix [G] changes abruptly, the drive signal vector {D (t)} becomes discontinuous as described above. Never be.

-Sampling of response, Fourier transform, PSD calculation, etc.- The DSP 63 drives the vibration generator 92 based on the drive signal vector {D (t)} obtained by drive signal vector calculation and output processing, and Shake 90. DSP6
3 receives, via the input board 70, a response signal vector {Resp (t)} having each response signal of the control system 2 corresponding to the drive signal vector {D (t)} as an element (FIG. 6, Step S10).

FIG. 21 shows the response signal vector {Resp (t)} obtained in this step. The response signal vector {Resp (t)} is an X acceleration sensor 94x, a Y acceleration sensor 94y,
The response signal (Resp (t) _X ) of the _X excitation group, the response signal (Resp (t) _Y ) of the _Y excitation group, and the response signal (Re of the Z excitation group) detected by the Z acceleration sensor 94z, respectively.
sp (t) _Z ) as an element.

In the vibration control system shown in FIG.
The response signal vector {Resp (t)} is a control response signal vector {RespC (t)} and also a main control response signal vector {RespP (t)}. Also,
Only the response signal (Resp (t) _X ) of the X excitation group is also the limit response signal (Lmt (t) _X ) of the _X excitation group.
Note that the monitor response signal is not set in the vibration control system shown in FIG.

Next, the DSP 63 obtains the response signal vector {R} with the signal on the time axis taken in the previous step as an element.
esp (t)} is Fourier-transformed and a response spectrum vector {Resp
(F)}, and a response PSD vector {Resp ^PSD } having the power spectral density of each response signal as an element.
(F)｝ is calculated. In this embodiment, the obtained response spectrum vector {Resp (f)} corresponds to the main control response spectrum vector {RespP (f)}, and the calculated response PSD vector {Resp (
^PSD (f) is the control response PSD vector {RespC ^PSD
(F) Corresponds to｝. Also, the control response PS of the X vibration axis group
D (RespC ^PSD (f) _X ) is the limit response PSD
(Lmt ^PSD (f) _X ). Further, the control response PS
An averaging control response PSD vector {RespM ^PSD (f)} is calculated based on the D vector {RespC ^PSD (f)}. In this embodiment, the control response PSD vector {RespC ^PSD (f)} is
It is also a D vector {RespM ^PSD (f)} (step S12).

FIG. 22 shows the main control response spectrum vector {RespP (f)} obtained in this step. FIG. 23 shows the averaging control response PSD vector {RespM} obtained in this step.
^PSD (f)｝.

-Equalization matrix [G]
Next, the DSP 63 calculates the main control response spectrum vector {RespP (f)} obtained in the previous step and the drive spectrum vector {D (f) obtained in step S44 of FIG. 9 (or step S54 of FIG. 10). ), An update process of the equalization matrix [G] is performed (step S14).

FIG. 11 shows a detailed flowchart of an example of the processing for updating the equalization matrix [G]. FIG. 13 is a block diagram showing the contents of the processing of each step shown in FIG. 11 and the like. First, the DSP 63 performs the above-described main control response spectrum vector {RespP
A temporary transfer function matrix [Htemp] of the controlled system 2 is calculated based on (f)} and the drive spectrum vector {D (f)} (step S60).

Temporary transfer function matrix [Htemp]
Is calculated by the following equation: [Htemp] = [[ ^{SD · D} ] ⁻¹ [ ^{SD · RespP} ]] ^T (5) where [ ^{SD · D} ] = ｛D (f) ｝ _BAR ｛D (f)｝ ^T (6) [ ^{SD · RespP} ] = ｛D (f)｝ _BAR ｝ RespP (f)｝ ^T (7) In this specification, [A] ^-1 is an inverse matrix of the matrix [A], [A] ^T ({A｝ ^T ) is a transposed matrix (vector) of the matrix [A] (vector {A}), [ A] _B A
_R ({A} _BAR ) is a complex conjugate matrix (vector) of the matrix [A] (vector {A}).

Next, the DSP 63 compares the temporary transfer function matrix [Htemp] calculated in the previous step with the current transfer function matrix [H], and converts the transfer function matrix [H] based on a predetermined criterion. Whether or not to update is determined for each frequency component (step S62).

If it is determined that the transfer function matrix [H] should be updated, the DSP 63 updates the transfer function matrix [H] using the temporary transfer function matrix [Htemp] (step S64).

FIG. 24 shows an updated new transfer function matrix [H]. Transfer function matrix [H]
Is a square matrix of 3 rows and 3 columns. Transfer functions H _XX , H _XY , H _XZ which are the elements of the first row of the transfer function matrix [H]
Are the respective components of the drive spectrum vector {D (f)} in FIG. 19, that is, the drive spectrum (D (f) _X ) of the _X excitation group, the drive spectrum (D (f) _Y ) of the _Y excitation group, Drive spectrum of Z excitation group (D
(F) _Z ) and the transfer function between the main control response spectrum (RespP (f) _X ) of the _X excitation group, which is one of the elements of the main control response spectrum vector {RespP (f)} in FIG. It is. Similarly, the transfer functions H _YX , H _YY , and the elements in the second row of the transfer function matrix [H] shown in FIG.
H _YZ is the drive spectrum vector ｛D
(F) A transfer function between each element of｝ and the main control response spectrum (RespP (f) _Y ) of the _Y excitation group, and the transfer functions H _ZX , H _ZY , and H _ZZ which are the elements in the third row Are the respective elements of the drive spectrum vector {D (f)} and the main control response spectrum (RespP (f)
_Z ).

Next, the DSP 63 updates the equalization matrix [G] based on the updated transfer function matrix [H] (step S66). The equalization matrix [G] is calculated by the following equation as the inverse matrix of the transfer function matrix [H]. [G] = [H] ⁻¹ (8) FIG. 25 shows an updated new equalization matrix. [G] is shown.

If it is determined in step S62 that the update of the transfer function matrix [H] is not appropriate, the DSP 63 performs control without updating the transfer function matrix [H] and the equalization matrix [G]. Move to step S16 (see FIG. 6).

-Control target PSD and random wave generation PS
Updating D- In this way, when the update processing of the equalization matrix [G] is completed, the DSP 63
The SD vector {Rctrl ^PSD (f)} and the random wave generation PSD {RDR ^PSD (f)} are updated (FIG. 6, step S16).

Control target PSD vector ｛Rctrl ^PSD
(F)｝ and random wave generation PSD ｛RDR
FIG. 12 shows a detailed flowchart of an example of the ^PSD (f) （update processing. FIG. 14 is a block diagram showing the contents of the processing of each step shown in FIG. 12 and the like. In this embodiment, the control target PSD vector {Rctrl}
^PSD (f)｝ and random wave generation PSD ｛RDR
^{When the PSD} (f)｝ is updated, the limit control is also performed in the X vibration group, and the limit control is not performed in the Y vibration group and the Z vibration group. Taking the X excitation group as an example, the control target PSD (Rctrl
^PSD (f) _x ) and random wave generation PSD (RDR)
The update process (including the limit control process) of the ^PSD (f) x) will be described. In the following, for convenience of explanation, a subscript <
m> and <m + 1> are used. The subscript <m> indicates the current control state, and the subscript <m + 1> indicates the next control state.

In the limit control process, the next limit response PSD (Lmt ^PSD (f)) among the responses of the controlled system 2
_{X <m + 1>} ) does not exceed the given monitoring PSD (P ^PSD (f) _X ) so that the next control target PSD (Rctrl)
^PSD (f) _{X <m + 1>} ).

FIGS. 26 and 27 are schematic diagrams for explaining the concept of the limit processing. The concept of the limit process will be described with reference to FIG. As shown in FIG. 26, the limit response PSD (Lmt ^PSD (f) _{X <m>} ) is currently the monitoring PS
D (P ^PSD (f) _X ) and the averaging control response P
SD (RespM ^PSD (f) _{X <m>} ) is also the original target PSD (R
ef ^PSD (f) _X ). In this case, the next predicted averaging control response PSD (RespM ^PSD (f)
_{X <m + 1>} ) approaches the original target PSD (Ref ^PSD (f) _X ) so that the next control target PSD (Rctrl ^PSD (f))
_{X <m + 1>} ) (not shown), the next predicted limit response PSD (Lmt ^PSD (f) _{X <m + 1>} ) will be monitored PSD (P ^PSD (f) _X ) at the protrusion. And exceed Therefore, the next control target PSD (Rctrl
Upon setting the _{^{PSD (f) X <m +}} 1>), that only a small set value of the portion corresponding to the projection of the next control target ^{PSD (Rctrl PSD (f) X} <m + 1>) Then, the next predicted limit response PSD (Lmt ^PSD (f) _{X <m + 1>} )
May not exceed the monitoring PSD (P ^PSD (f) _X ). This is the limit processing.

As shown in FIGS. 12 and 14, the DSP 63
Calculates the transmissibility (Trans _<< f >> X _<m> ) (step S70). Transmission rate (Trans _<< f >> _{X <m>} )
Is based on the averaging control response PSD (RespM ^PSD (f) _{X <m>} ) and the limit response PSD (Lmt ^PSD (f) _{X <m>} ) obtained in step S12 (see FIG. 6). Trans _< f _{> X <m>} = Lmt ^PSD (f) _{X <m>} / RespM ^PSD (f) _{X <m>} (9)

The transmissibility (Trans _<< f >> _{X <m>} ) is
It is a dimensionless quantity expressed as a real spectrum on the frequency axis. Therefore, the next averaging control response PSD (Re
Given _{^{spM PS D (f) X <}} m + 1>) is, (using a _{Trans "f" X <m>} ), next limit response PSD (Lmt ^PSD (f) the current transmission rate _{X <m +1>} ) can be estimated. In this specification, "B << f >>" indicates that B is a dimensionless quantity represented as a real spectrum on the frequency axis.

In the system shown in FIG. 5, as described above, the limit response PSD (Lmt ^PSD (f) _X )
Is the same as the averaging control response PSD (RespM ^PSD (f) _X ). Therefore, in this case, from equation (9), Tr
ans _<< f >> _{X <m>} = 1.

Next, the DSP 63 sets the next suppression coefficient (L
im _<< f >> _{X <m + 1>} ) is calculated (step S72). A procedure for calculating the suppression coefficient (Lim _<< f >> _{X <m + 1>} ) will be described.

First, the next predicted limit response PSD (L
mt ^PSD (f) _{X <m + 1>} ). The next predicted limit response PSD (Lmt ^PSD (f) _{X <m + 1>} ) is the transmission rate (Trans _<< f >> _{X <m>} ) obtained in the previous step and the original target P
SD (Ref ^PSD (f) _X ), next target level (Lvl
_{X <m + 1>} ), and is calculated according to the following equation: Lmt ^PSD (f) _{X <m + 1>} = Ref ^PSD (f) _X × Lvl _{X <m + 1>} × Trans _<< f >> _{X <m>} (10) In the above expression, the next target level (Lvl _{X <m + 1>} ) is
Next time averaging control response PSD (RespM ^PSD (f)
_{X <m + 1>} ) is a scalar quantity (0 <Lvl) for specifying how close the original target PSD (Ref ^PSD (f) _X ) is to be.
_{X <m + 1>} ≦ 1), which is determined automatically by the program or manually entered via the keyboard 86.

Next, the next suppression coefficient (Lim << f >>
_{X <m + 1>} ). Next suppression coefficient (Lim << f >>
_{X <m + 1>} ) is the next predicted limit response PSD (Lmt
Lim _<< f >> _{X <m + 1>} = 1 (if P ^PSD (f) calculated based on the ^PSD (f) _X < _{m + 1>} ) and the monitoring PSD (P ^PSD (f) _X ) according to the following equation. ) _X / Lmt ^PSD (f) _{X <m + 1>} ≧ 1) Lim _< f _{> X <m + 1>} = P ^PSD (f) _X / Lmt ^PSD (f) _{X <m + 1>} (if ^PSD (F) _X / Lmt ^PSD (f) _{X <m + 1>} <1) (11) FIG. 27A shows the suppression coefficient (Lim
(F) _{X <m + 1>} ).

In this way, the suppression coefficient (Lim << f >>
After calculating _{X <m + 1>} , the DSP 63 calculates the next temporary control target PSD (Rtemp ^PSD (f) _{X <m + 1>} ) (step S74). Next temporary control target PSD (Rt
emp ^PSD (f) _{X <m + 1>} ) is the suppression coefficient (Lim _<< f >> _{X <m + 1>} ) calculated in the previous step, and the original target PSD described above.
(Ref ^PSD (f) _X ), next target level (Lvl
_{X <m + 1>} ), and is calculated according to the following equation: Rtemp ^PSD (f) _{X <m + 1>} = Ref ^PSD (f) _X × Lvl _{X <m + 1>} × Lim _<< f >> _{X < m + 1>} (12) FIG. 27B shows a provisional control target PSD in the case of FIG.
(Rtemp ^PSD (f) _X < _{m + 1>} ). Step S7
Steps S0 to S74 are limit control processing.

Next, the DSP 63 calculates the next correction coefficient k.
_<< f >> _{X <m + 1>} is calculated (step S76). The correction coefficient k _<< f >> _{X <m + 1>} is the current control target PSD (Rctrl
^PSD (f) _{X <m>} ) and the corresponding averaging control response P
K (f) _{X <m + 1>} = (Rctrl ^PSD (f) _{X <m>} / RespM ^PSD (f) Calculated based on SD (RespM ^PSD (f) _{X <m} >) according to the following equation: _{X <m>} ) ^{1 / Q} (13) In equation (13), Q is a coefficient for adjusting the convergence of the control of the system, and for example, Q = 4.

Next, the DSP 63 sets the next control target PS.
D (Rctrl ^PSD (f) _{X <m + 1>} ) is calculated (step S78). Control target PSD (Rctrl ^PSD (f)
_{X <m + 1>} ) is the correction coefficient k _<< f >> calculated in the previous step.
_{X <m + 1>} and the tentative control target PS calculated in step S74
D (Rtemp ^PSD (f) _{X <m + 1>} ), and is calculated according to the following equation: Rctrl ^PSD (f) _{X <m + 1>} = Rtemp ^PSD (f) _{X <m + 1>} × k _<< f >> _{X <m + 1>} (14).

Further, the current averaging control response PSD (Re
spM ^PSD (f) _{X <m>} ) and the corresponding random wave generation PSD (RDR
^PSD (f) _X <_m> ) and the control response transmission rate (r
_<< f >> _{X <m + 1>} ) is calculated by the following equation. R _<< f >> _{X <m + 1>} = RespM ^PSD (f) _{X <m>} / RDR ^PSD
(F) _{X <m>} .

Then, the next random wave generation PSD (R
DR ^PSD (f) _{X <m + 1>} ) is calculated.

Random wave generation PSD (RDR ^PSD (f)
_{X <m + 1>} ) is the control target PSD (Rctrl) calculated earlier.
^PSD (f) _{X <m + 1>} ) and the control response transmission rate (r
_<< f >> _{X <m + 1>} ) according to the following equation: RDR ^PSD (f) _{X <m + 1>} = Rctrl ^PSD (f) _{X <m + 1>} /
r _<< f >> _{X <m + 1>} .

Thus, the control target PSD (Rctrl ^PSD (f) _X ) and the random wave generation P
Update processing of SD (RDR ^PSD (f) x) is performed. Similar processing is performed for the Y excitation group and the Z excitation group. However, as described above, in the system illustrated in FIG. 5, the limit control is not performed for the Y excitation group and the Z excitation group. That is, for the Y excitation group and the Z excitation group, the following equation is used instead of equation (11) in step S72: Lim _<< f >> _{Y <m + 1>} = 1 (15) Lim _<< f >> _{Z <m + 1>} = 1 (16) FIG. 28 shows the suppression coefficient vector {Lim _<< f >>_{<m + 1>} } in the system shown in FIG.

The DSP 63 sets the control target P
After updating the SD vector {Rctrl ^PSD (f)} and the random wave generation PSD {RDR ^PSD (f)}, as shown in FIG. 6, the control returns to step S6, and the same operation (steps S6 to S16) repeat.

FIG. 29 shows the averaging control response PSD vector {Re} when limit control is performed for the X excitation group.
Show spM ^PSD (f)｝. As shown in FIG. 29, of the averaging control response PSD vector {RespM ^PSD (f)}, the averaging control response PSD (RespM ^PSD (f) _Y ) of the _Y excitation group not performing the limit control is shown in FIG. 15 original target PSDs (Ref ^PSD (f) _Y ) multiplied by the current target level (Lvl _Y ) (“Original target PS of target level”
D ". ) And almost the same. The same applies to the averaging control response PSD (RespM ^PSD (f) _Z ) of the _Z excitation group. On the other hand, it can be seen that the averaging control response PSD (RespM ^PSD (f) X) of the X excitation group that has performed the limit control does not always match the original target PSD of the target level in the X excitation group. Note that the averaging control response PSD vector {RespM ^PSD (f)} in FIG. 29 is the target level (Lvl _X ), (Lvl _Y ), (Lv
l _Z ) is the data when 0 dB (= 1).

The DSP 63 performs a vibration test on the specimen 90 by repeating the above-described operation (steps S6 to S16) for a predetermined time. However,
Among the operations described above, an operation related to the update processing of the equalization matrix [G] or the update processing of the control target PSD and the random wave generation PSD (steps S10 to S10).
Step S16) may be performed for each loop, but may be configured to be performed only once in several loops.

[Other Embodiments] In the embodiment shown in FIG. 5, a monitor response is not set, but a monitor response can be set. In this case, a monitor response can be specified as the limit response.

Although the limit response is set only for the X excitation group, the limit response can be set for the Y excitation group and the Z excitation group.

Further, two or more limit responses can be set for one excitation group. In this case, the control target of the excitation group can be adjusted so that any limit response of the excitation group does not exceed the corresponding monitoring PSD.

As the limit response, the acceleration at the response point was used as in the case of the control response.
The amount different from the control response, such as the speed, displacement,
Stress or the like can also be used.

The monitoring PS is used as the limit value of the limit control.
Although D is used, a spectrum value other than PSD (a value given as a function of frequency) can be used as a limit value of the limit control. As the limit value of the limit control, for example, an rms value (root mean square value) or the like can be used in addition to the spectrum value. Further, these can be used in combination.

It is also possible to configure so that the limit control is not performed.

Further, in the embodiment shown in FIG. 5, the control response and the main control response in each excitation group are configured to be the same, but a plurality of control responses are provided in one excitation group, and An appropriate one can be used as the main control response.

When a plurality of control responses are provided in one excitation group, a maximum value, an average value, or a combination of these control responses can be used as an input to the second control means. When calculating the average of multiple control responses,
Each control response can also be weighted.

Although one main control response is set for each excitation group, two or more main control responses can be set for one excitation group. In this case, each main control response can be weighted.

Further, in the embodiment shown in FIG. 5, one drive signal (vibrator) is set in each excitation group, but two or more drive signals may be set in one excitation group. it can.

In the embodiment shown in FIG. 5, the number of drive signals and the number of main control responses are set to be the same, but the number of drive signals and the number of main control responses may be different. . In this case, to be applicable even when the number of drive signals and the number of main control responses are different,
The following equation may be used instead of the above equation (8) for finding the equalization matrix [G]. [G] = [H] ⁺ ... (17) In equation (17), [H] ⁺ This is a "Moore-Penrose" general inverse matrix of the function matrix [H]. For details of the "Moore-Penrose" general inverse matrix,
It is described in "Projection Matrix General Inverse Matrix Singular Value Decomposition" (Haruo Yanai / Hiroshi Takeuchi: University of Tokyo Press).

In the above embodiment, the control target PSD belonging to each excitation group is generated based on the response (control response, limit response) belonging to the corresponding excitation group. The control target PSD belonging to the vibration group may be generated with reference to limit responses belonging to vibration groups other than the corresponding vibration group.

Further, in the above-described embodiment, the drive signal calculating means temporarily converts the control target signal vector, which is a signal on the time axis, into a signal on the frequency axis, performs a predetermined matrix product operation, and then performs Although the drive signal vector is obtained by converting the calculation result into a signal on the time axis again, the convolution calculation is performed as it is without converting the control target signal vector into a signal on the frequency axis. It can be configured to obtain a drive signal vector. In addition, Japanese Patent Application Laid-Open No. 2-213908
The publication discloses details of the convolution operation.

In the above embodiment, the DSP
63 controls each unit according to a program stored in the hard disk 82 and downloaded to the memory 64. This program is read out from the flexible disk storing the program via the FDD 84 and installed on the hard disk 82.
In addition to the flexible disk, a CD-ROM,
The program may be installed on the hard disk 82 from a computer-readable storage medium storing a program such as a C card. Furthermore, you may make it download using a communication line.

In the above-described embodiment, the program stored in the flexible disk is indirectly executed by the DSP 63 by installing the program from the flexible disk to the hard disk 82 and downloading the program to the memory 64.
However, without being limited to this, the program stored in the flexible disk may be directly executed from the FDD 84. Note that, as a program executable by a computer, not only a program that can be directly executed by simply installing it as it is, but also a program that needs to be temporarily converted into another form (for example, decompresses a data-compressed program) Etc.), and also includes those that can be executed in combination with other module parts.

In the above embodiment, the DSP
Although the description has been given of an example in which each function of the vibration control system shown in FIG. 1 is realized using the above, each function of the vibration control system shown in FIG. 1 may be realized using a computer system other than the DSP. Further, a part or all of the functions may be configured to be realized by hardware logic.

Also, in the above-described embodiment, the case where the present invention is applied to a vibration control system has been described as an example. However, the present invention is not limited to such a system. Is composed of a positioning actuator and a position detection sensor (or speed sensor), the controlled system is composed of a speaker, a space, and a microphone, or the controlled system is a container or a part of a container. The present invention can also be applied to a case where an actuator having a piston that moves in and out, a fluid (a gas, a liquid, or the like) filling a container, and a pressure sensor are used. That is,
The present invention can be applied to a signal control system in general.

[Brief description of the drawings]

FIG. 1 is an overall block diagram of a vibration control system which is a signal control system according to an embodiment of the present invention.

FIG. 2 is a detailed block diagram illustrating an example of a multidimensional waveform controller 8 of FIG.

FIG. 3 is a detailed block diagram showing an example of a drive signal calculating means 40 of FIG.

FIG. 4 is a detailed block diagram showing another example of the drive signal calculating means 40 of FIG.

FIG. 5 is a diagram illustrating the functions of the vibration control system shown in FIG.
3 is a diagram illustrating an example of a hardware configuration and an example of a controlled system 2 realized by using a (Digital Signal Processor).

6 is a diagram showing a flowchart of an operation in the vibration control system shown in FIG.

FIG. 7 is a drawing showing a detailed flowchart of a trial excitation process.

FIG. 8 is a drawing showing a detailed flowchart of initial value setting and activation processing.

FIG. 9 is a drawing showing a detailed flowchart of an example of a drive signal calculation and output process.

FIG. 10 is a drawing showing a detailed flowchart of another example of drive signal calculation and output processing.

FIG. 11 is a drawing illustrating a detailed flowchart of an example of an update process of an equalization matrix [G].

FIG. 12 is a drawing illustrating a detailed flowchart of an example of a process of updating a control target PSD and a random wave generation PSD.

FIG. 13 is a block diagram showing the contents of processing in each step shown in FIG. 11 and the like.

FIG. 14 is a block diagram showing the contents of processing in each step shown in FIG. 12 and the like.

FIG. 15: Original target PSD vector {Ref ^PSD (f)}
It is a drawing which shows an example.

FIG. 16 is a diagram illustrating an example of a control target PSD vector {Rctrl ^PSD (f)} in a state where repetition processing has progressed to some extent.

FIG. 17 is a diagram showing a control target signal vector {Rctrl (t)} generated in the random wave generation processing.

FIG. 18 shows a control target spectrum vector ｛Rctrl.
(F) Drawing showing 図 面.

FIG. 19 is a diagram showing a drive spectrum vector {D (f)}.

FIG. 20 is a diagram illustrating a drive signal vector {D (t)}.

FIG. 21 is a diagram illustrating a response signal vector {Resp (t)}.

FIG. 22 shows a main control response spectrum vector ΔRespP.
(F) Drawing showing 図 面.

FIG. 23 shows an averaging control response PSD vector ｛RespM.
⁶ is a drawing showing ^PSD (f)｝.

FIG. 24 is a diagram showing a transfer function matrix [H] of the controlled system 2;

FIG. 25 is a diagram showing an equalization matrix [G] of the system.

FIG. 26 is a schematic diagram for explaining the concept of limit processing.

FIG. 27 is a schematic diagram for explaining the concept of limit processing.

FIG. 28 is a drawing showing a suppression coefficient vector {Lim << f >>}.

FIG. 29 illustrates an averaging control response PSD vector ｛RespM when limit control is performed on the X excitation group.
^PSD (f)｝, and the monitoring PSD (P
It is a drawing showing ^PSD (f) _X ).

FIG. 30 is a view for explaining a conventional multidimensional vibration test.

FIG. 31 is a drawing showing a conventional so-called multi-point excitation system.

[Explanation of symbols]

 2 ... controlled system 8 ... multidimensional waveform controller 10 ... control group 12 ... PSD controller 14 ... random wave generator 20 Control group 30 Temporary control target generation means 32 Control target generation means

──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-5-34245 (JP, A) JP-A-5-79950 (JP, A) JP-A-59-208615 (JP, A) JP-A-52-1982 79374 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) G01M ⁷ /00-7/06 G01M 17/00-17/06 G05D 19/02 B06B 1/00 G05B 11/36 G05B 13/02

Claims (12)

(57) [Claims] 1. A control characteristic storage means for storing control characteristics of a controlled system, wherein a response of the controlled system driven by using a plurality of drive signals may be influenced by a plurality of drive signals. The control characteristic storage means stores the plurality of drive signals and the plurality of main responses to the drive signal from the controlled system so that the response becomes substantially equal to the control target converted into the signal on the time axis. The control characteristic of the controlled system is updated, and the signal on the frequency axis obtained based on the control target converted into the signal on the time axis and the control characteristic of the controlled system stored in matrix form in the control characteristic storage means are updated . Transfer function mat
First control means for generating a drive signal on a time axis based on a signal on a frequency axis obtained as a product of a control characteristic of the controlled system expressed by an inverse matrix of Rix and giving the drive signal on the time axis to the controlled system; Second control means for generating a control target based on the response and the original target given by the power spectral density. 2. The signal control system according to claim 1, wherein said second control means generates a control target such that an amount obtained based on the response becomes substantially equal to the original target. What to do. 3. The signal control system according to claim 1, wherein said second control means generates a control target such that an amount obtained based on a response does not exceed a given limit value. thing. 4. The signal control system according to claim 1, wherein said second control means controls said signal based on said response and an original target so that an amount obtained based on said response does not exceed a given limit value. Temporary control target generation means for generating a temporary control target, generating a control target based on the response and the temporary control target such that an amount obtained based on the response is substantially equal to the temporary control target. Control target generating means, 5. The signal control system according to claim 1, wherein at least one control group is set, and an original target belonging to each control group is set.
A second control unit, a control target, a drive signal, and a response are defined, and one or more predetermined responses among the responses belonging to each control group are defined as a main response of the control group; Comprises control characteristic storage means for storing the control characteristics of the controlled system, and the main response belonging to each control group is
In consideration of being able to be affected by a drive signal belonging to a control group other than the corresponding control group, the control characteristic storage means is based on a plurality of drive signals and a plurality of main responses from the controlled system to the drive signal. The control characteristic stored in the control system is updated in the form of a matrix on the frequency characteristic signal obtained based on the control target converted into the signal on the time axis and the control characteristic stored in the control system. Transfer function matrix
Drive signal on the time axis based on the signal on the frequency axis obtained as the product of the control characteristics of the controlled system expressed by the inverse matrix of the By performing the calculation and giving it to the controlled system, control is performed such that the main response belonging to each control group becomes substantially equal to the control target converted into a signal on the time axis belonging to the corresponding control group. Wherein the second control means generates a control target belonging to the corresponding control group based on the response belonging to the corresponding control group and the original target. 6. A signal control system according to claim 5, the second control means belonging to the respective control group, between the control group also see the response belonging to a control group of non-corresponding control group
By performing a process, a control target belonging to a corresponding control group is generated so as to be an appropriate control target . 7. The signal control system according to claim 5, wherein at least some of the responses including the main response and the drive signal are represented as signals on a time axis. Wherein the original target and the control target are represented as an amount corresponding to an amplitude spectrum on a frequency axis, and the first conversion means for converting the control target into a signal on the time axis and providing the signal to the first control means; Second conversion means for converting a response expressed as a signal on a time axis into an amount corresponding to a frequency characteristic and providing the converted signal to a second control means. 8. The system of claim 7 signal control system, the first control means, the control characteristic memory means for storing the control characteristics of the control system, is converted into a signal on the time axis by the first conversion means Based on the control target and the control characteristics stored in the control characteristic storage means, calculates a drive signal within a time required for guaranteeing real-time performance, and provides drive signal calculation means to the controlled system; A drive signal output from the drive signal calculation means,
Control characteristic updating means for updating the control characteristic stored in the control characteristic storage means based on the main response from the controlled system to the drive signal. 9. The signal control system according to claim 8, wherein the control characteristic is a predetermined value of a transfer function matrix of the controlled system having a plurality of transfer functions of the controlled system representing a relationship between a drive signal and a main response. Characterized by the inverse matrix of. 10. The signal control system according to claim 9, wherein the drive signal calculation means performs a Fourier transform on a control target signal vector having a control target converted into a signal on a time axis as an element, and converts a complex Fourier spectrum into an element. Fourier transform means for obtaining a control target spectrum vector to be used, multiplication means for obtaining a drive spectrum vector which is a product of a predetermined inverse matrix of the transfer function matrix and the control target spectrum vector, and inverse Fourier transform of the drive spectrum vector, Inverse Fourier transform means for outputting a drive signal vector having a drive signal as an element. 11. The signal control system according to any one of claims 1 to 10, wherein the controlled system comprises: a specimen; a vibration generator which receives the drive signal and applies vibration to the specimen by a plurality of vibrators; A vibration detector that detects the vibration of the specimen at one or more response points, wherein the object is to apply vibration having a predetermined relationship to the original target to the specimen. 12. A computer-readable storage medium storing a computer-executable program, wherein the program realizes the system according to any one of claims 1 to 11. Storage medium.