JP2002182736A - Facility diagnosis device and facility diagnosis program storage medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To precisely judge the presence or absence of an abnormality on a facility where the normal state is not uniquely decided and the operation state is temporally changed concerning a facility diagnosis device diagnosing the normality or abnormality of the facility in which one or more or a plurality of units are installed and they are operated while the operation mode is temporally switched. SOLUTION: The respective differences of the operation states of the respective units constituting the facility constituted of a plurality of units installed in one room are considered as the differences of the operation modes. Reference data of an inverse filter is obtained for the respective operation modes. For judging the presence or absence of abnormality, the presence or absence of abnormality is judged based on a signal obtained at that time and reference data obtained on the operation mode similar to the operation mode at that time.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a facility diagnostic apparatus for diagnosing a normal or abnormal condition of a facility in which one or more devices, typically a plurality of devices, are installed and operated while the operation mode is changed over time. The present invention relates to a facility diagnostic program storage medium storing a facility diagnostic program that causes a computer to operate as such a facility diagnostic apparatus.

[0002]

2. Description of the Related Art Conventionally, equipment diagnosis has been executed or proposed by various equipment diagnosis methods for judging the presence or absence of abnormality in equipment or equipment. In this equipment diagnosis, not only a serious failure that requires the equipment to be destroyed or needs to be stopped immediately is detected, but rather, for example, a bearing in a rotating machine is damaged before such a serious failure occurs. At present, it is still possible to continue operation sufficiently, such as entering or abrasion of a movable part, but if it is left as it is, an abnormality that may lead to a serious failure in the future is detected. There is a need.

As a typical example of such a facility diagnosis method,
For example, the acoustic vibration waveform when the device or equipment is in a normal state is obtained, the characteristic of the acoustic vibration waveform is analyzed by spectrum analysis, and the acoustic vibration waveform of the device or equipment is detected when an abnormality is detected. And perform a spectrum analysis to determine whether there is a peak of a specific frequency component that cannot be seen in the normal state in the spectrum or whether the combination of peaks is the same as that in the normal state. It is known to perform detection.

Japanese Patent Application Laid-Open No. 7-43259 discloses that
Obtain the acoustic vibration waveform when the equipment or equipment is in a normal state, create an inverse filter based on the acoustic vibration waveform, and obtain the acoustic vibration waveform of the equipment or equipment when detecting the presence or absence of an abnormality. It has been proposed that a residual signal is obtained by applying an inverse filter, which is obtained in advance, to the acoustic vibration waveform, and an abnormality of a device or equipment is detected by analyzing the residual signal.

Japanese Unexamined Patent Application Publication No. 8-304124 discloses a method of obtaining a plurality of acoustic vibration waveforms when the equipment or facility is in a normal state, and for example, converting one of the plurality of acoustic vibration waveforms into one. A plurality of residual signals are obtained by creating an inverse filter based on the plurality of residual signals, for example, by applying the inverse filter to the remaining plurality of acoustic vibration waveforms, and a plurality of statistical variables are obtained based on each of the plurality of residual signals. In addition, when detecting the presence or absence of an abnormality, a plurality of acoustic vibration waveforms of the device or equipment are obtained, and a plurality of residuals are obtained by applying the above-described inverse filter to the plurality of acoustic vibration waveforms in advance. Signals, obtain a plurality of statistical variables based on the plurality of residual signals, and obtain a plurality of statistical variables obtained in a normal state and a plurality of statistics obtained in detecting the presence or absence of an abnormality. In between the variables, for example the F-test and t
It has been proposed to detect the presence or absence of an abnormality in the device or equipment by performing verification or estimation by a method such as verification.

[0006]

The method of detecting an abnormality of a device or equipment by performing the above-described spectrum analysis is also quite effective depending on the properties of the device or equipment to be diagnosed. A method that is prepared or a method that performs a statistical test or the like is a more effective method.

However, all of the above methods detect the presence or absence of a difference as compared with a normal state, but there is a case where a state to be set to a normal state cannot be uniquely determined. .

[0008] For example, a compressor room in which a plurality of compressors such as a reciprocating compressor and a turbo compressor are installed will be described.

A large-sized reciprocating compressor has a plurality of pistons, and each of the plurality of pistons has a state in which air is actually compressed and a state in which the piston is simply running idle. Often, multiple compressors and other turbocompressors are installed in the compressor room of the above, and depending on the equipment, other devices such as motors and pumps may be installed in the same room. There is also. Since such equipment generates extremely loud noise, the noise is contained in the room by being divided into one room and installed in the room.

[0010] When such a large number of devices are operated or stopped in various states, it is extremely difficult to determine which state is to be a normal state. Alternatively, in a normal state, all of these states are all normal states.

The present invention provides a facility diagnostic apparatus capable of detecting, with high accuracy, the presence / absence of an abnormality in a facility, for example, in which the above-mentioned normal state changes sequentially, and a facility for operating a computer as such a facility diagnostic apparatus. It is an object of the present invention to provide an equipment diagnostic program storage medium in which a diagnostic program is stored.

[0012]

According to the present invention, there is provided a facility diagnosis apparatus for diagnosing a normal or abnormal state of a facility in which at least one device is installed and which operates while its operation mode is temporally switched. In the diagnostic device, the signal obtaining unit obtains a time-series signal carrying a predetermined physical quantity obtained from the diagnostic target facility, the mode recognizing unit recognizing the current operation mode of the diagnostic target facility, and the signal obtaining unit. Based on the reference time-series signal of the time-series signal, a reference calculation unit for obtaining reference data corresponding to the operation mode when the reference time-series signal is obtained, and a reference data obtained by the reference calculation unit. A reference storage unit that stores the reference time-series signal based on which the reference data is obtained, in association with the operation mode obtained, and a plurality of operation modes stored in the reference storage unit. Based on the reference data corresponding to the current operation mode of the reference data, and the diagnostic time-series signal in the current operation mode obtained by the signal acquisition unit, the presence or absence of an abnormality in the diagnostic target facility is determined. A determination unit for determining whether there is an abnormality.

[0013] The equipment diagnosis apparatus of the present invention uses the operation mode for all of the differences in the operation state of any one of the equipment constituting the equipment comprising a plurality of equipments installed in close proximity, such as in one room. Considering the difference, for each operation mode, for example, the reference data such as the above-described spectrum analysis, inverse filter, and a plurality of statistical variables are obtained, and when determining whether there is an abnormality, the signal obtained at the time of the determination is used. And the reference data obtained for the same operation mode as the operation mode at that time is used to determine the presence or absence of an abnormality. Therefore, even if the operation states are sequentially switched, the presence or absence of an abnormality in each operation state is determined. Can be determined with high accuracy.

Here, the time-series signal carrying the predetermined physical quantity may be a signal obtained by recording sound in the vicinity of the facility, or a device constituting the facility or an installation location of the facility. The signal may be a signal obtained by picking up vibration or acceleration from a floor, a wall, or the like, and is not limited to a signal carrying a specific physical quantity.

In the equipment diagnosis apparatus of the present invention, it is determined whether or not reference data corresponding to the current operation mode has already been obtained. If the reference data corresponding to the current operation mode has not been obtained yet, And instruct the reference calculation unit to execute calculation for obtaining reference data when the time-series signal in the current operation mode acquired by the signal acquisition unit is used as the reference time-series signal, and correspond to the current operation mode. When the reference data has already been obtained, a signal instructing the abnormality presence / absence determination unit to determine the presence / absence of abnormality when the time series signal in the current operation mode acquired by the signal acquisition unit is used as the time series signal for diagnosis. It is preferable to provide a distribution unit.

Taking the above-described compressor chamber as an example, for example, in the case of a reciprocating compressor having six pistons whose operating state changes independently of actual operation and idle operation, theoretically, one Thus, there are 64 operation modes, and even if only two reciprocating compressors exist in the compressor chamber, 64 × 64 = 4096 operation modes exist.

According to the present invention, for example, even in a facility having such a large number of operation modes, each operation mode is one by one.
When it is possible to produce reference data corresponding to each operation mode, it is not denied that there is a preparatory stage for producing such reference data before performing equipment diagnosis. In practice, it is extremely difficult to create operating states one by one in each of a large number of operating modes. On the other hand, the above-mentioned equipment is in a normal state in most cases, and in most cases is trying to detect an abnormal state that may occur very rarely. Even if an abnormal condition occurs, it is almost always possible to find the abnormal condition by the attention of the operator. Under such a background, the provision of the above-described signal distribution unit obtains reference data in each operation mode based on a time-series signal obtained from equipment actually operating while the operation modes are sequentially switched. And equipment diagnosis can be performed without special preparation before actual operation.

Further, in the equipment diagnosis apparatus of the present invention,
The reference calculation unit obtains reference data by an operation including an operation for obtaining an inverse filter corresponding to an operation mode when the reference time-series signal is obtained, based on the reference time-series signal. The abnormality presence / absence determination unit performs an operation including an operation of obtaining a residual signal by applying an inverse filter corresponding to the same operation mode as the current operation mode to the diagnostic time-series signal in the current operation mode, and performs the operation. It is preferable to determine the presence or absence of an abnormality in the equipment to be diagnosed based on the result of the above.

Here, the above-mentioned "operation including an operation for obtaining an inverse filter" is a concept including a case where only an operation for obtaining an inverse filter is performed. can do. In addition, the “operation including an operation for obtaining an inverse filter” may include an operation for obtaining an inverse filter, and may be performed based on, for example, one time-series signal among a plurality of reference time-series signals as described above. An inverse filter may be created, and the inverse filter may be applied to a plurality of other time-series signals to calculate a plurality of statistical variables. In this case, the plurality of statistical variables thus determined may be used. Can be reference data.

The above-mentioned "operation including an operation for obtaining a residual signal by applying an inverse filter" is the same as described above, and may be constituted only by an operation for obtaining a residual signal. As described in JP-A-7-43259, data convenient for determining the presence or absence of abnormality by calculating the residual signal, such as calculating a moving average value of the power of the residual signal, is used. An arithmetic operation or a plurality of residual signals is obtained by applying an inverse filter to a plurality of diagnostic time-series signals as described in JP-A-8-304124 described above. May be used to calculate a plurality of statistical variables based on

When an inverse filter is used, stationary noise can be eliminated from the signal, and the presence or absence of an abnormality in the equipment to be diagnosed can be determined with higher accuracy.

According to the storage medium for diagnosing equipment of the present invention, which achieves the above object, a computer diagnoses the normality or abnormality of equipment in which one or more devices are installed and the operation modes of which are operated while being temporally switched. A facility diagnostic program storage medium storing a facility diagnostic program to be operated as a facility diagnostic apparatus, wherein the facility diagnostic program stored therein is obtained from a facility to be diagnosed, and is a time-series signal carrying a predetermined physical quantity. And a mode recognizing unit that recognizes the current operation mode of the equipment to be diagnosed, and a reference time series based on the reference time series signal among the time series signals obtained by the signal acquiring unit. A reference calculation unit for obtaining reference data corresponding to the operation mode when the signal is obtained; and a plurality of operation modes corresponding to the plurality of operation modes obtained by the reference calculation unit. Based on the reference data of the reference data corresponding to the current operation mode and the time series signal for diagnosis in the current operation mode obtained by the signal acquisition unit, it is determined whether or not there is an abnormality in the equipment to be diagnosed. And an abnormality determination unit.

Here, the facility diagnosis program further determines whether reference data corresponding to the current operation mode has already been obtained, and the reference data corresponding to the current operation mode has not yet been obtained. At the same time, the reference calculation unit is instructed to execute calculation for obtaining reference data when the time series signal in the current operation mode acquired by the signal acquisition unit is set as the reference time series signal, and the current operation mode is changed to the current operation mode. When the corresponding reference data has already been obtained, the abnormality determination unit instructs the abnormality determination unit to determine whether there is an abnormality when the time-series signal in the current operation mode acquired by the signal acquisition unit is used as a diagnostic time-series signal. It is preferable to have a signal distribution unit that performs the above.

Further, in the above-described facility diagnostic program storage medium of the present invention, the reference arithmetic unit constituting the facility diagnostic program stored therein obtains the reference time-series signal based on the reference time-series signal. The above-described reference data is obtained by an operation including an operation for obtaining an inverse filter corresponding to the operation mode at the time of the operation. It is preferable that an operation including an operation for obtaining a residual signal is performed by applying an inverse filter corresponding to the same operation mode, and based on a result of the operation, it is preferable to determine whether there is an abnormality in the equipment to be diagnosed. .

[0025]

Embodiments of the present invention will be described below.

FIG. 1 is a diagram showing an example of the equipment to be diagnosed in the present embodiment.

Here, five turbo compressors 12 to 15 and two reciprocating compressors 16 and 17 are installed in the compressor room 10. The operation mode of each of the turbo compressors 12 to 15 is switched between an actual operation for obtaining compressed air and an idle operation, and each of the reciprocating compressors 16 and 17 has its own piston 16a, 16b,.
The operation mode is switched between the actual operation and the idle operation for each of 16f; 17a, 17b,..., And 17f, and the entire facility of the compressor room 10 is operated in approximately 700 operation modes. Switching between these operation modes is performed by the control computer 20.

During operation, the equipment comprising a plurality of compressors 12 to 17 installed in the compressor room 10 is sequentially operated by the control computer 20 into any one of about 700 operation modes. Can be switched.

In this embodiment, two sound collectors 21 and 22 are provided in the compressor room 10.
Each of these two sound collectors 21 and 22 is disposed at a focal point of a parabolic sound collecting hood configured to detect only a sound wave from a specific direction and a parabolic inner surface of the sound collecting hood. And a sound sensor that picks up sound waves. Here, these two
Each of the sound collectors 21 and 21 is directed in a direction to capture a sound wave emitted from each of the reciprocating compressors 16 and 17. The sound signals obtained by these two sound collectors 21 and 22 are taken into the diagnostic computer 100, and the signal analysis in the diagnostic computer 100 causes the two reciprocating compressors 16 and 17 to detect the abnormality. The presence or absence is detected.

Here, two sound collectors 21 and 22 are shown, but one sound collector may be provided corresponding to each of the compressors 12 to 17, or some compressors may be provided. , Or, as the most extreme example, a single collection location and orientation for collecting sound generated by any compressor in the compressor room 10. A sound device may be arranged. this is,
It is determined from the viewpoint of how to detect an abnormality for each group into which the compressors installed in the compressor room 10 are divided.

FIG. 2 is an external perspective view of a diagnostic computer that operates as a facility diagnostic apparatus according to one embodiment of the present invention. The equipment diagnosis apparatus as one embodiment of the present invention is realized by a combination of hardware of the diagnostic computer 100 and software executed inside the computer.

The diagnostic computer 100 has a CP
U, RAM memory, magnetic disk, communication board, etc., main body 101, CRT display 10 for displaying a screen on its display screen 102a according to instructions from the main body
2. A keyboard 103 for inputting instructions and text information of a subject and other operators in the computer for diagnosis, an icon displayed at an arbitrary position on a display screen by designating the position, and the like. Is provided with a mouse 104 for inputting an instruction corresponding to the mouse.

A CD-ROM 105 (see FIG. 3) is removably loaded into the main body 101, and the loaded CD-ROM 105 is loaded.
A CD-ROM drive for driving the ROM 105 is also provided.

Here, the equipment diagnosis program is stored in the CD-ROM 105, and the CD-ROM 10
5 is loaded into the main body 101, and the equipment diagnostic program stored in the CD-ROM 105 is installed in the magnetic disk of the diagnostic computer 100 by the CD-ROM drive. When the equipment diagnostic program installed in the magnetic disk of the diagnostic computer 100 is started, the diagnostic computer 100
It operates as one embodiment of the equipment diagnosis device of the present invention.

That is, the CD-ROM in which the equipment diagnosis program is stored corresponds to an embodiment of the equipment diagnosis program storage medium of the present invention, and the equipment diagnosis program is installed in the magnetic disk of the diagnostic computer 100. Then, the magnetic disk in which the facility diagnostic program is installed also corresponds to an embodiment of the facility diagnostic program storage medium of the present invention.

FIG. 3 shows the diagnostic computer 1 shown in FIG.
FIG. 3 is a hardware configuration diagram of a hardware configuration 00.

The hardware configuration includes a central processing unit (CPU) 111, a RAM 112, a magnetic disk controller 113, a CD-ROM drive 115, a mouse controller 116, and a keyboard controller 11.
7, display controller 118, communication board 1
19 and two A / D conversion boards 120 and 121 are shown, which are interconnected by a bus 110.

As described with reference to FIG. 2, the CD-ROM drive 115 is loaded with the CD-ROM 105 and accesses the loaded CD-ROM 105.

The communication board 119 is connected to the control computer 20 shown in FIG.
From the compressor 1 in the compressor chamber 10 shown in FIG.
A code representing the current operation mode as a whole of 1 to 18 is input.

The two A / D conversion boards 120 and 12
1 is connected to the sound collectors 21 and 22 shown in FIG. These two A / D conversion boards 120 and 121 are:
Each of the sound collectors 21 and 22 has a role of inputting each sound signal, converting the sound signal into a digital signal, and taking in the digital signal.

FIG. 3 shows a magnetic disk 114 accessed by a magnetic disk controller 113 and a mouse 10 controlled by a mouse controller 116.
4. Keyboard 103 controlled by keyboard controller 117, and display controller 11
Also shown is a CRT display 102 controlled by 8.

FIG. 4 is a diagram showing a configuration of a facility diagnostic program stored in a facility diagnostic program storage medium according to an embodiment of the present invention.

The equipment diagnostic program storage medium 200 shown in FIG. 4 is representative of the CD-ROM 105, the magnetic disk 114, and the like shown in FIG. 2 in which the equipment diagnostic program 210 is stored.

The equipment diagnostic program 210 stored in the equipment diagnostic program storage medium 200 includes a signal acquisition unit 2
11, code acquisition unit 212, inverse filter presence / absence determination unit 21
3. Inverse filter creation unit 214 and abnormality presence / absence determination unit 2
15 is comprised.

The signal acquisition section 211 and the code acquisition section 2
12, an inverse filter presence / absence determination unit 213, an inverse filter creation unit 214, and an abnormality presence / absence determination unit 215 are respectively a signal acquisition unit and a mode recognition unit referred to in the equipment diagnosis program stored in the equipment diagnosis program storage medium of the present invention. , A signal distribution unit, a reference calculation unit, and an abnormality determination unit. The operation of each of the units 211 to 215 constituting the equipment diagnosis program 210 will be described later.

FIG. 5 shows functions of an embodiment of the equipment diagnosis apparatus of the present invention realized by installing and executing the equipment diagnosis program shown in FIG. 4 on the diagnostic computer shown in FIGS. It is a block diagram.

The equipment diagnostic apparatus 300 shown in FIG. 5 includes a signal acquisition section 311, a code acquisition section 312, an inverse filter presence / absence determination section 313, an inverse filter creation section 314, an abnormality presence / absence determination section 315, and an inverse filter storage section 316. It is configured.

Here, the equipment diagnosis device 30 shown in FIG.
0, the signal acquisition unit 311 and the code acquisition unit 31 except for the inverse filter storage unit 316 among the units 311 to 316.
2. Inverse filter existence determining unit 313, inverse filter creating unit 3
14 and the abnormality presence / absence determination unit 315 are elements corresponding to the respective elements constituting the equipment diagnosis program 210 to which the same names shown in FIG. 4 are assigned, and constitute the equipment diagnosis apparatus shown in FIG. The signal acquisition unit 311, the code acquisition unit 312, the inverse filter presence / absence determination unit 313, the inverse filter creation unit 314, and the abnormality presence / absence determination unit 315 refer to a combination of hardware and software, and perform the equipment diagnosis shown in FIG. A signal acquisition unit 211 constituting the program 210,
The code acquisition unit 212, the inverse filter presence / absence determination unit 213, the inverse filter creation unit 214, and the abnormality determination unit 215 indicate each program component of the equipment diagnosis program 210 as an application program.

Hereinafter, each element of the equipment diagnosis apparatus 300 shown in FIG. 5 will be described, and each element of the equipment diagnosis program 210 shown in FIG. 4 will be described at the same time.

The signal acquiring section 311 has a function of acquiring sound signals picked up by the two sound collectors 21 and 22 provided in the compressor room shown in FIG. The A / D conversion boards 120 and 121 mainly correspond to this, and include program parts for taking in sound signals converted into digital signals by the A / D conversion boards 120 and 121 into a facility diagnosis program.

In this embodiment, as shown in FIG.
/ D conversion board, two sound collectors 2 shown in FIG.
The A / D conversion boards 120 and 121 convert the sound signals picked up by the channels 1 and 22 into digital signals by the A / D conversion boards 120 and 121. However, even if there are a plurality of sound collectors, the A / D conversion boards can be used. There may be only one, and a switching circuit for sequentially switching a plurality of sound signals obtained by a plurality of sound collectors and transmitting the sound signals to the A / D conversion board may be arranged in the middle.

The code acquiring unit 312 receives a code representing the current operation mode of the diagnostic target equipment including the plurality of compressors 12 to 17 installed in the compressor room 10 from the control computer 20 shown in FIG. It is.

The inverse filter presence / absence determination unit 313 determines whether an inverse filter corresponding to the current code obtained by the code acquisition unit 312 has already been created, and an inverse filter corresponding to the code is still created. If not,
The two-channel sound signal obtained by the signal acquisition unit 311 is sent to the inverse filter creation unit 314 to instruct creation of an inverse filter corresponding to the code. On the other hand, an inverse filter corresponding to the code has already been created. The signal acquisition unit 3
The two-channel sound signal obtained in step 11 is sent to the abnormality presence / absence determination unit, and an instruction is provided to determine whether there is an abnormality. The inverse filter presence / absence determination unit 313 mainly includes hardware
Inverse filter presence / absence determination unit 213 as a program component
The CPU 111 (see FIG. 3) that executes (see FIG. 4) corresponds thereto.

The inverse filter creating section 314 creates an inverse filter for each channel based on each of the sound signals of the two channels transmitted via the inverse filter presence / absence determining section 313. A detailed description of the inverse filter will be given later.
The inverse filter creation unit 314 is also mainly composed of the inverse filter creation unit 21 as a program component.
4 (see FIG. 4) corresponds to the CPU 111 (see FIG. 3). The inverse filter created by the inverse filter creation unit 314 is stored in the inverse filter storage unit 316. The magnetic disk 14 and the like correspond to the inverse filter storage unit 316 in hardware.

FIG. 6 is a diagram showing a memory structure of the inverse filter storage unit 316 constructed in the magnetic disk 114.

Here, a memory structure for one channel of two-channel sound signals (two sound collectors 21 and 22) is shown.

As shown in FIG. 6A, each code # 1,
A correspondence table is created between # 2, # 3,... And the storage start address of the inverse filter, and an inverse filter corresponding to each code is stored in a memory area after each start address (in the example shown here, address A and address B). Is stored. In the example shown here, an inverse filter has already been created for code # 1 and code # 3, and no inverse filter has been created for code # 2.

Therefore, in the inverse filter presence / absence determining section 313 shown in FIG. 5, when the current state of the inverse filter storage section 316 is the state shown in FIG. 6, the code received from the code acquisition section 312 is the code # 1 or the code # 1. In the case of code # 3, the sound signal received from the signal acquisition unit 311 at that time is sent to the abnormality presence / absence determination unit 315 to instruct the presence / absence determination of abnormality, while the code received from the code acquisition unit 312 is the code # 2. In this case, the sound signal received from the signal acquisition unit 311 at that time is sent to the inverse filter creation unit 314 to instruct creation of an inverse filter.

Upon receiving the code and the two-channel sound signal from the inverse filter presence / absence determination unit 313, the abnormality presence / absence determination unit 315 stores the inverse filter corresponding to the received code for each of the two-minute channels in the inverse filter storage unit. 316, and the inverse filter for each channel is applied to the sound signal of each channel to obtain each residual signal, and it is determined whether or not each residual signal is abnormal. If any one of them has an abnormality in the residual signal, it is notified that there is an abnormality. This notification may indicate that there is an abnormality on the display screen 102a (see FIG. 2) of the equipment diagnosis device 300, or may directly drive a buzzer or an alarm lamp with the equipment diagnosis device 300 to notify the operator. May be notified or the equipment diagnostic device 2
From 00, the control computer 20 (see FIG. 1) is notified that there is an abnormality, the operator is notified that the abnormality has occurred on the control computer 20 side, or the control mode is shifted to the control mode when the abnormality occurs. You may do so.

FIG. 7 is an operation flowchart of the equipment diagnosis apparatus 300 shown in FIG.

First, a sound signal is acquired by the signal acquiring section 311 (step a), a code representing the operation mode is acquired by the code acquiring section 312 (step b), and the code is obtained by the inverse filter presence / absence determining section 313. It is determined whether an inverse filter corresponding to the code has already been created (step c). If an inverse filter corresponding to the code has not been created yet, the inverse filter creation unit 314 performs an inverse filter based on the sound signal acquired this time. A filter is created and stored (step d). On the other hand, when an inverse filter corresponding to the current code has already been created, the abnormality presence / absence determination unit 315 determines whether or not the equipment is abnormal (step c). . When it is determined that an abnormality has occurred in the determination, an alarm is issued by any of the methods described above. On the other hand, if it is determined that it is normal, step a
And the above processing is repeatedly executed.

Next, an inverse filter and a method for detecting the presence or absence of an abnormality using the inverse filter will be described.

An arbitrary time-series signal can be regarded as an output when white noise is input to an appropriate linear system. Determining the corresponding linear system from a given time-series signal is called linear prediction analysis, and there is an established method.
An autoregressive model (AR model) is usually obtained as such. This means that when the sampled and discretized time-series signal is X (n), n = 1, 2,.
The signal X (n) at the time is determined as follows from the data at the preceding M times.

[0064]

(Equation 1)

Here, e (n) is a virtual input signal to the linear system and is white noise. Given a time series signal,
By obtaining a coefficient set {Ak} from the data, an autoregressive model for the time-series signal is determined.

When the set of coefficients {Ak} is determined, Y (n) is defined as follows using the time-series signal data {X (n)}. At this time, Y (n) is called a linear predicted value of X (n).

[0067]

(Equation 2)

Then, when the following quantities are calculated,
From equations (1) and (2), X (n) -Y (n) = e (n) (3), and the residual is white noise. That is, when the predicted value Y (n) obtained from the previous M data is subtracted from the time-series signal data X (n) at the n-th time point, the input white noise is obtained. Here, obtaining the residual e (n) by subtracting the predicted value Y (n) from X (n) is referred to as applying an inverse filter. If a certain time-series signal can be represented by an appropriate auto-regression model in this way, white noise is obtained by applying an inverse filter configured using the model to the original time-series signal. That is, the input signal is whitened by the inverse filter. In this case, the input time-series signal does not have to be the signal used when designing the inverse filter. If the autoregressive model has the same signal, that is, a signal having the same characteristics, a whitened signal is obtained as an output. Can be. However, if the characteristics of the time-series signal are different from those used in the design, even if an inverse filter is applied, whitening is not performed and white noise is not obtained.

Therefore, an inverse filter is pre-configured using the first time-series signal that carries the normal operation sound and vibration (operation sound and the like), and carries the operation sound and the like at any time. By obtaining a new second time-series signal and applying an inverse filter to the second time-series signal and monitoring the output, a time-series signal (residual signal) different from the normal time can be detected. I can do it.

In this embodiment, specifically, the following signal processing method can be adopted.

First, in the inverse filter creating unit 314,
FFT (Fast Fourier Transform) is performed using 1024 sound signal data obtained when the equipment is in a normal state, and then a power spectrum is obtained. Then IF it
FT (Inverse Fast Fourier Transform) to find the autocorrelation function,
Using it, Levinson's algorithm (for example, see Mikami, "Introduction to Digital Signal Processing", published by CQ Publishing)
To obtain the inverse filter coefficient {Ak}.

Thereafter, the abnormality presence / absence determination section 315 applies the inverse filter to obtain a residual signal. In the present embodiment, the calculation for obtaining the residual signal is a coefficient {a [k]}. Is performed by a moving average calculation using

Here, in order to obtain a moving average of the power of the residual signal, first, 128 data are extracted from the time series of the residual signal, and FFT, power spectrum calculation, IFFT
, An autocorrelation function is obtained, and power is obtained from the peak value at the origin. Thereafter, the power is sequentially obtained while shifting the starting point of the data by 50 points.

As an example of the inverse filter, the order M
= 27, and the coefficient {a [k]} is as shown in Table 1.

[0075]

Table 1 a [0] = 1.000000 a [1] = -2.887330 a [2] = 3.947344 a [3] = -3.5535249 a [4] = 2.447053 a [5] = -1.620133 a [6] = 1.135352 a [7] =-1.268161 a [8] = 0.937471 a [9] =-0.380573 a [10] =-0.040919a [ 11] = 0.284076 a [12] = -0.353665 a [13] = 0.397849 a [14] = -0.533185 a [15] = 0.501902 a [16] = -0.238178 a [17] = − 0.003048 a [18] = 0.92420 a [19] = − 0.1666854 a [20] = − 0.010498 a [21] = 0.061383 a [22] = 0.017332 a [23] =-0.014146 a [24] =-0.131247 a [25] = 0.239157 a [26] =-0.242444 a [27] = 0.115678

8 to 11 show examples of waveforms obtained from equipment in a normal state. FIG. 8 shows a signal waveform of a sound signal collected from equipment in a normal state, and FIG. 9 shows this sound signal. , The signal waveform of the residual signal after the inverse filter is applied to FIG. 10, FIG. 10 shows the power spectrum of this residual signal, and FIG. 11 shows the moving average of the residual signal power.

FIGS. 12 to 15 show examples of waveforms obtained when the equipment is in an abnormal state, and each figure shows a signal waveform of the same format as FIGS. 8 to 11, respectively.
8 and 12 showing the waveforms of the sound signals obtained from the equipments in the normal state and the abnormal state, respectively,
It is difficult to judge normality / abnormality immediately from these. However, by analyzing the residual signals obtained by applying an inverse filter to these signals in FIG. 9 and FIG. 13, normal / abnormal judgment can be made.

As can be easily understood by comparing FIGS. 9 and 13, when the equipment is in a normal state, the amplitude of the residual signal is extremely small. Becomes extremely large. Therefore, it is possible to determine normal / abnormal based on the maximum value of the power in the residual signal. For example, a case where the amplitude of the residual signal is 10 dB or more larger than the maximum power of the signal obtained when the facility is in a normal state is abnormal, and a case where the amplitude of the residual signal is less than this is normal. Normal / abnormal judgment can be made.

As shown in FIGS. 10 and 14, in the spectrum obtained by performing Fourier transform on the residual signal, when an abnormality occurs, the power spectrum increases. For example, while the peak of the power in FIG. 10 is 100 dB or less, the peak of the power is almost 120 dB in FIG.
dB has been reached.

Further, comparing FIGS. 11 and 15 showing the moving average of the power of the residual signal, it can be seen that the moving average increases due to the abnormality. For example, when the moving average data is larger than the maximum value of the moving average by 20 dB or more when the equipment is in a normal state, it can be determined that the moving average is abnormal, and when the data indicates less than this, it can be determined to be normal. When this method is employed, it is particularly easy to determine the presence or absence of an abnormality, and it is possible to perform abnormality detection in a short time. Depending on the type of abnormality, the presence of a defect can be detected more accurately by the power spectrum analysis than by the power moving average analysis.

In the embodiment described above, an inverse filter is prepared as reference data in the present invention, a residual signal is obtained, and the presence or absence of an abnormality is detected based on the residual signal. The present invention does not necessarily need to employ this detection method, and may employ, for example, the above-described spectrum analysis method, or a method of performing statistical estimation or test.

In the above-described embodiment, the reference data (inverse filter) is created based on the sound signal obtained for the first time for each operation mode (each code), and the second and subsequent times are set as objects for determining the presence or absence of abnormality. However, when each operation mode (each code) can be created in advance, a preparation stage for creating reference data (inverse filter) for all the operation modes (codes) in advance may be provided.

Further, in the above-described embodiment, the presence or absence of abnormality is detected based on the sound signal obtained by picking up the noise in the compressor room. The presence or absence of an abnormality may be detected based on a signal obtained by picking up vibration or the like.

Further, the present invention can be applied not only to the determination of the presence / absence of the abnormality of the compressor, but also to general diagnosis of equipment having a plurality of operation modes and sequentially switching the operation modes.

[0085]

As described above, according to the present invention,
It is possible to accurately determine the presence or absence of an abnormality in equipment whose normal state is not uniquely determined and whose operating state changes one after another.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating an example of a diagnosis target facility.

FIG. 2 is an external perspective view of a diagnostic computer that operates as a facility diagnostic device according to an embodiment of the present invention.

FIG. 3 is a hardware configuration diagram of the diagnostic computer shown in FIG. 2;

FIG. 4 is a diagram showing a configuration of a facility diagnostic program stored in a facility diagnostic program storage medium as one embodiment of the present invention.

FIG. 5 is a functional block diagram of an embodiment of the equipment diagnosis device.

FIG. 6 is a diagram showing a memory structure of an inverse filter storage unit built in a magnetic disk.

FIG. 7 is an operation flowchart of the equipment diagnostic device shown in FIG. 5;

FIG. 8 is a waveform diagram of a sound signal obtained from equipment in a normal state.

9 is a signal waveform diagram obtained by applying an inverse filter to the signal of FIG. 8;

FIG. 10 is a power spectrum diagram obtained from the signal of FIG. 9;

FIG. 11 is a diagram illustrating a moving average of power obtained from the signal of FIG. 10;

FIG. 12 is a waveform diagram of a sound signal obtained from equipment in an abnormal state.

13 is a signal waveform diagram obtained by applying an inverse filter to the signal of FIG.

FIG. 14 is a power spectrum diagram obtained from the signal of FIG.

FIG. 15 is a diagram showing a moving average of power obtained from the signal of FIG. 13;

[Explanation of symbols]

 Reference Signs List 10 compressor room 11, 12, 13, 14, 15 turbo compressor 16, 17, 18 reciprocating compressor 20 control computer 21, 22 sound collector 100 diagnostic computer 200 equipment diagnostic program storage medium 210 equipment diagnostic program 211 signal acquisition unit 212 Code acquisition unit 213 Inverse filter presence / absence determination unit 214 Inverse filter creation unit 215 Abnormality presence / absence determination unit 300 Equipment diagnostic device 311 Signal acquisition unit 312 Code acquisition unit 313 Inverse filter storage unit 314 Inverse filter creation unit 315 Abnormal existence determination unit 316 Inverse filter storage Department

Claims (6)

[Claims] 1. A facility diagnostic apparatus for diagnosing the normal or abnormal state of a facility in which one or more devices are installed and the operation mode of which is operated while being temporally switched, carries a predetermined physical quantity obtained from the facility to be diagnosed. A signal acquisition unit that acquires a time-series signal, a mode recognition unit that recognizes a current operation mode of the diagnostic target facility, and a time-series reference signal among the time-series signals obtained by the signal acquisition unit, A reference calculation unit for obtaining reference data corresponding to an operation mode when the reference time series signal is obtained; and a reference time series based on which the reference data obtained by the reference calculation unit is obtained. A reference storage unit that stores the signal in association with the operation mode at the time of obtaining the signal, and the current operation mode among the reference data corresponding to the plurality of operation modes stored in the reference storage unit. A corresponding reference data, and an abnormality presence / absence determination unit that determines presence / absence of abnormality in the diagnostic target facility based on the diagnostic time-series signal in the current operation mode obtained by the signal acquisition unit. A facility diagnostic device characterized by the following. 2. It is determined whether reference data corresponding to the current operation mode has already been obtained. If the reference data corresponding to the current operation mode has not yet been obtained, the reference operation unit includes: While instructing the execution of the calculation for obtaining the reference data when the time series signal in the current operation mode acquired by the signal acquisition unit is used as the reference time series signal, the reference data corresponding to the current operation mode has already been obtained. If the time series signal in the current operation mode acquired by the signal acquisition unit is used as a diagnostic time series signal, the signal distribution unit instructs the abnormality presence / absence determination unit to determine whether there is an abnormality. The equipment diagnosis apparatus according to claim 1, further comprising: 3. The reference calculation unit calculates the reference data based on the reference time series signal, the calculation including obtaining an inverse filter corresponding to an operation mode when the reference time series signal is obtained. The abnormality presence / absence determining unit performs an operation to obtain a residual signal by applying an inverse filter corresponding to the same operation mode as the current operation mode to the diagnostic time-series signal in the current operation mode. The equipment diagnosis apparatus according to claim 1 or 2, wherein the equipment diagnosis apparatus performs an operation including the calculation, and determines whether there is an abnormality in the equipment to be diagnosed based on a result of the operation. 4. A facility storing a facility diagnostic program for operating a computer as a facility diagnostic apparatus for diagnosing the normal or abnormal state of a facility in which one or more devices are installed and the operation modes are operated while being temporally switched. A diagnostic program storage medium, wherein the facility diagnostic program recognizes a current operation mode of the facility to be diagnosed, the signal acquiring unit acquiring a time-series signal carrying a predetermined physical quantity obtained from the facility to be diagnosed. A mode recognition unit, and a reference calculation unit that obtains reference data corresponding to an operation mode when the reference time-series signal is obtained based on the reference time-series signal among the time-series signals obtained by the signal acquisition unit. And, among the reference data corresponding to the plurality of operation modes, obtained by the reference calculation unit, the reference data corresponding to the current operation mode, An equipment for determining whether or not there is an abnormality in the equipment to be diagnosed, based on the diagnostic time-series signal in the current operation mode obtained by the signal acquisition unit. Diagnostic program storage medium. 5. The facility diagnosis program further determines whether reference data corresponding to the current operation mode has already been obtained, and when the reference data corresponding to the current operation mode has not yet been obtained. Instructs the reference calculation unit to execute calculation for obtaining reference data when the time-series signal in the current operation mode acquired by the signal acquisition unit is used as the reference time-series signal. When the reference data corresponding to has already been obtained, the abnormality presence / absence determination unit determines whether there is an abnormality when the time series signal in the current operation mode acquired by the signal acquisition unit is used as a diagnostic time series signal. 5. The storage medium according to claim 4, further comprising a signal distribution unit for instructing the determination. 6. The reference calculation unit calculates the reference data based on the reference time series signal, the calculation including an operation of obtaining an inverse filter corresponding to an operation mode when the reference time series signal is obtained. The abnormality presence / absence determining unit performs an operation to obtain a residual signal by applying an inverse filter corresponding to the same operation mode as the current operation mode to the diagnostic time-series signal in the current operation mode. The equipment diagnosis program storage medium according to claim 4, wherein an operation including the operation is performed, and based on a result of the operation, the presence or absence of an abnormality in the equipment to be diagnosed is determined.