JPH07159231A - Deterioration diagnostic system for pump - Google Patents

Abstract

PURPOSE:To decide the extent of deterioration of a sliding oscillation source, e.g. the shaft sealing part or liner ring part, of a pump by comparing an observation spectral pattern with normal and abnormal spectral patterns. CONSTITUTION:Output from a vibration detector 6 comprising a vibration sensor 2 and a converter 4 is passed through filters 8, 10 where irrelevant high and low frequency components are cut off before being fed to an A/D converter 12. An analytic operating unit 14 samples the digital signals and calculates a vibration spectral pattern. A spectral feature pattern operated by a spectral feature pattern operating unit 18 is fed through a switch 20 to a normal or abnormal spectral feature pattern memory 22 or 24. At the time of diagnosis, the observed vibration spectral feature pattern of an object operated by the operating unit 18 is compared with normal and abnormal spectral feature patterns thus deciding the state of a pump.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention judges the progress of deterioration / damage of rotating equipment as an element machine constituting various production facilities, particularly sliding vibration parts such as a shaft sealing part and a liner ring part in a pump. Deterioration diagnosis system.

[0002]

2. Description of the Related Art A rotary machine is a one-element machine that constitutes various production facilities. Pumps, which make up the majority of rotating machines, are often used for water supply, water distribution, drainage, etc. related to water and sewerage. It is also used in the industrial world for delivery of main raw material liquids in chemical plants, circulation of cooling water to high-temperature equipment, and the like. There are three types of pumps: horizontal-axis spiral, vertical-axis spiral, and vertical-axis mixed flow. In rotating equipment such as pumps, it is important to quantitatively grasp the progress of deterioration and damage and determine the appropriate repair content and timing. "Periodic maintenance" for repair work of pumps (determine repair time, parts replacement, etc. based on years of operation, hours of operation, etc.)
The main focus was "urgent maintenance" when abnormal noise, abnormal odor, abnormal vibration, abnormal water leakage, etc. were found in the patrol inspection by the maintenance inspection personnel. In such maintenance, overhaul or replacement of parts may be performed due to the arrival of repair time, even though it is still in a usable state, resulting in excessive maintenance. On the contrary, the discovery may be delayed even though the abnormality progresses, and a sudden accident may occur.

As a deterioration diagnosis method corresponding to these problems, a method of measuring a vibration generated by a rotating device such as a pump and analyzing the vibration to set a reference value or a reference pattern for discriminating a normal state from an abnormal state is set. Alternatively, there has been a method of continuously monitoring the generated vibration of the rotating device to be diagnosed and diagnosing whether or not there is an abnormality depending on whether or not the judgment reference value (or reference pattern) is exceeded. Specifically, for example, JP-A-5
In the abnormality diagnosing device disclosed in Japanese Unexamined Patent Publication No. 6-70426 and Japanese Unexamined Patent Publication No. 62-93620, the frequency of vibration caused by various abnormal observation phenomena has been examined in advance. Frequency analysis is performed to store the normal spectrum components at a specific generation frequency and its harmonics, such as the shaft rotation speed, whirling vibration frequency (f ₀ ) or its higher order component (nf ₀ ). Then, during operation, the spectrum component at the specific frequency of the vibration signal obtained by the vibration detector in real time is detected, and compared with the above-mentioned spectrum component at the normal time by using an appropriate threshold value, and whether there is any abnormality in the rotating equipment. To diagnose. Further, in this diagnostic method or a diagnostic method similar to this, the analysis accuracy may decrease due to the relationship between the generated frequency and the resolution of the frequency spectrum analysis value. Therefore, there is a method for correcting the deterioration of accuracy and improving the diagnostic accuracy.
It is disclosed in the publication. In addition, JP-A-5-72026
According to the method disclosed in the publication, presence / absence of abnormality is diagnosed by causing a neural network model to learn normal state vibration information and abnormal state vibration information of equipment as learning data.

[0004]

DISCLOSURE OF THE INVENTION Problems to be Solved by the Invention
In the method disclosed in Japanese Patent Application Laid-Open No. 0426 and Japanese Patent Application Laid-Open No. 1-199127, or a conventional diagnostic technique similar thereto, the power levels of specific frequency spectrum components depending on the shaft speed obtained by vibration frequency analysis are compared. To do. Rolling bearing damage and gear damage to which these conventional abnormality diagnosis methods have been applied, and unbalance,
Due to the progress of deterioration due to shaft bending and the like, at the time of abnormality, a clear specific frequency component depending on the shaft rotation speed is generated, and the specific frequency does not change with time. The above-mentioned abnormality diagnosis method has a limitation that only a deterioration observation phenomenon in which such a specific frequency component is clearly calculated and found can be diagnosed. In addition, with respect to wear deterioration of sliding vibration sources of rotating equipment such as shaft seals (seals), linings, impeller rings, and intermediate bearings, which are pump components, the vibration level of The increase allows diagnosis of deterioration using conventional methods. However, heretofore, there has been no suitable method for diagnosing the initial abnormality of wear deterioration for the sliding vibration source. In the deterioration of the sliding vibration source, in contrast to the above-mentioned abnormalities of the mechanical part and rolling bearing damage, the vibration spectrum is not concentrated into specific frequency components and has a gentle mountain-shaped distribution. The distribution of components deteriorates and changes as the damage progresses. When the spectrum pattern changes as the deterioration progresses, it is impossible to diagnose the deterioration by the conventional method. This is for comparing and diagnosing the spectral components at specific frequencies.

As a conventional deterioration diagnosis method,
Only the method of detecting the increase or decrease of the sliding vibration level according to the progress of wear deterioration of seals, liner rings, etc.
However, with this method, since abnormalities in other mechanical components of the pump other than the sliding vibration source and vibration components generated by sources other than the sliding vibration source are also mixed and detected, only sliding parts such as seals and liner rings are detected. It is impossible to detect the vibration of. Therefore, even if the conventional method is used to make a diagnosis, the reliability of the diagnosis accuracy is low. In addition, the method of determining the increase of the specific vibration frequency component by the diagnostic frequency analysis, which is called the precise diagnosis method, can be used for the sliding vibration source such as the seal and the liner ring of the pump. This method could not be used because it was not aggregated into frequency components. In addition, the method disclosed in Japanese Patent Laid-Open No. 5-72026,
In a method similar to this, in order to diagnose the presence or absence of an abnormality by making the neural network model learn, it is necessary to ensure not only the normal state vibration information but also the abnormal state vibration information. A large number of effective learning patterns were necessary for the neural network model to output a sufficiently reliable decision.

It is an object of the present invention to provide a deterioration diagnosis system for a sliding vibration source such as a shaft seal portion, a liner ring portion, an impeller ring portion, an intermediate bearing portion of a pump.

[0007]

The object of diagnosis in the pump deterioration diagnosis system according to the present invention is a sliding vibration source. The vibration sensor that constitutes the vibration signal detection device is attached at a position where the vibration of the sliding signal source can be detected, and the vibration of the sliding vibration source is
It is detected as an electric signal and output as a digital signal. The frequency spectrum analysis calculator samples this digital signal to calculate a vibration spectrum pattern. The normal and abnormal spectrum patterns input from the frequency spectrum analysis calculator are stored in the first memory and the second memory, respectively. The pattern comparison device includes a normal-time spectrum pattern stored in the first memory and a second spectrum pattern.
The abnormal spectrum pattern stored in the memory is compared with the observed spectrum pattern input from the frequency spectrum analysis calculator, and the deterioration progress of the diagnostic target is judged from the similarity of the observed spectrum pattern to the normal spectrum pattern and the abnormal spectrum pattern. Diagnose. In particular,
This pattern comparison device calculates a power level, a center frequency mobility, and a sharpness from a spectrum pattern, and diagnoses deterioration progress using a discriminant function predetermined as a function of these three indexes. Preferably, the rotation speed detector detects the rotation speed of the pump. The spectrum feature pattern calculator inputs this rotation speed, inputs the vibration spectrum pattern calculated by the above frequency spectrum analysis calculator, and removes the rotation vibration spectrum generated depending on the shaft rotation speed of the diagnosis target. A pattern (hereinafter referred to as a spectrum characteristic pattern) is calculated and output to the above-mentioned first storage device, second storage device and pattern comparator. The first storage device and the second storage device store the spectral characteristic pattern, and the pattern comparison device calculates the similarity for the spectral characteristic pattern. Preferably, the comparison device is an observed spectrum feature pattern input from the spectrum feature pattern calculator, a normal-time spectrum feature pattern stored in the first storage unit, and an abnormal-time spectrum feature stored in the second storage unit. A similarity calculator for calculating the similarity with the pattern, a third storage for storing the criterion for determining the similarity with the normal spectrum characteristic pattern, and a criterion for determining the similarity with the abnormal spectrum characteristic pattern are stored. A fourth storage unit, and a determination unit that compares the similarity input from the similarity calculation unit with two determination criteria input from the third storage unit and the fourth storage unit to determine the deterioration progress state of the diagnosis target. Consists of. Preferably, the similarity calculator calculates the power level, the center frequency mobility, and the sharpness from the spectrum characteristic pattern, and the judger determines a discriminant function predetermined as a function of these three indices. It is used to diagnose deterioration progress. Preferably, the fourth memory stores a criterion of similarity lower than the criterion of similarity stored in the third memory,
The above-mentioned deciding device compares the degree of similarity inputted from the spectral characteristic pattern calculator with the decision criterion inputted from the third storage device and the fourth storage device, and determines the deterioration progress state of the diagnosis target. Preferably, the above pattern comparison apparatus has an observed spectrum input from the frequency spectrum analysis calculator for the normal spectrum pattern stored in the first storage and the abnormal spectrum pattern stored in the second storage. Pattern matching is performed on the patterns, and the deterioration progression of the diagnosis target is diagnosed based on the obtained similarity. Preferably, this pattern comparison device has an observed spectrum pattern input from a spectrum pattern calculator and a first spectrum
A pattern matching calculator for calculating the similarity between the normal spectrum pattern stored in the storage and the abnormal spectrum pattern stored in the second storage, and a criterion for determining the similarity with the normal spectrum pattern are stored. Third
A storage unit, a fourth storage unit that stores criteria for determining the degree of similarity with an abnormal-time spectrum pattern, a similarity level that is input from the pattern matching calculator, and two types of storage units that are input from the third storage unit and the fourth storage unit. And a judging device for judging the deterioration progress state of the diagnosis target by comparing with the judging standard.

[0008]

The function of the deterioration diagnosis system according to the present invention is as follows:
Among pumps, shaft seal (seal), liner ring,
It is a sliding vibration source such as the impeller ring section and intermediate bearing sleeve section.In the sliding vibration source, the initial wear deterioration and the bearing sleeve, bush, etc. do not raise the vibration level so much, and the characteristic of the vibration spectrum pattern Change will occur. In sliding vibration, not only the vibration spectrum power level increases, but the vibration spectrum is not concentrated into a specific component and has one or more wide mountain-shaped distributions. The frequency (calculated from the center of gravity and the second moment of inertia) and the sharpness of the spectrum change as the deterioration and damage progress. Especially when the center frequency moves, the conventional diagnostic method cannot be applied. Therefore, in the deterioration diagnosis system according to the present invention, these indexes are used as the diagnosis information, and the deterioration of the sliding vibration source is determined by the shift of the spectrum pattern from the normal time to the abnormal time. It is diagnosed whether the wear deterioration possibility is low or the wear deterioration possibility is high. Alternatively, since the spectrum pattern changes with the movement of the center frequency, the progress of deterioration can be diagnosed by performing matching with the spectrum pattern under normal conditions and abnormal conditions.

Specifically, before the diagnosis, the spectrum patterns at the normal time and at the abnormal time (or the spectrum characteristic patterns from which the component depending on the shaft rotation speed is removed) are observed, respectively, and the first storage device and the second storage device are observed. Store in a container. Then, when the rotating equipment is in operation, the observed spectrum pattern is compared with these normal and abnormal spectrum patterns. As a result, it is possible to quantitatively recognize the difference generated in the spectral pattern as the deterioration progresses. Here, when it is determined that the similarity of the observed spectrum pattern is closer to the similarity at the time of abnormality, it is determined that an abnormality has occurred. In this judgment, the similarity between the observed spectrum pattern and the normal and abnormal spectrum patterns is calculated and compared with the preset similarity judgment criteria for normal and abnormal, whichever is closer. It is determined whether the sliding vibration occurrence site has progressed due to the transition of the observed spectrum pattern from the normal state to the abnormal state by, for example, a tree method. The degree of similarity is calculated based on various indexes obtained for the observed spectrum pattern, or is obtained by pattern matching.

[0010]

EXAMPLE First, as an example of the vibration spectrum pattern of the sliding vibration part, an example of the vibration spectrum pattern of the shaft sealing part and the liner ring part will be described. Figure 1 shows the sliding vibration spectrum (acceleration G) of the shaft seal and liner ring in the normal state.
FIG. 2 shows an example of the shaft sealing portion in a state where deterioration has progressed,
An example of the sliding vibration spectrum of the liner ring portion is shown. In this example, there are peaks of a wide vibration spectrum having a vibration peak at f ₂ . In this vibration spectrum, as compared with the spectrum in the normal state shown in FIG. 1, the position of the frequency of the vibration peak shifts from f _{1 in} the normal state to the low frequency side, and the peak value also increases. The presence or absence of deterioration can be determined by comparing the spectrum pattern during operation with the spectrum patterns during normal operation and during abnormal operation.

The diagnostic method of the present invention can also be used for determining damage to the impeller section. Here, the vibration sensor is attached to a bearing capable of detecting the vibration of the impeller portion that is the diagnosis target. FIG. 3 shows an example of spectra (velocity cm / s) of the impeller part, the impeller ring part, and the liner ring part in a normal state in which almost no deterioration occurs. This spectral pattern shows multiple vibration peaks, in contrast to the examples shown in FIGS. Relatively sharp vibration peaks are observed at f ₀ , f _{H, and the} like. Where f
₀ is the rotational frequency of the impeller, and f _H is the product of f ₀ and the number of blades (5 in this example). In addition, FIG.
4 shows a spectrum pattern of vibration of the impeller portion and the liner ring portion in a state where damage is advanced as compared with the case of FIG. 3. In this spectrum pattern, the levels of f ₀ and f _H can be considered to be almost the same as in the case of FIG. However, as the position of the peak shifts, the number of spectral components increases, and the peak of the vibration spectrum becomes wider. In such a case, the progress of deterioration cannot be diagnosed simply by comparing the peak levels. Therefore, the progress of deterioration can be determined by comparing the spectrum pattern during operation with the spectrum patterns in these states.

Embodiments according to the present invention will be described below with reference to the accompanying drawings. First, the first embodiment will be described. In the first embodiment, the index of the spectral characteristic pattern is obtained to obtain the degree of similarity, and the progress of deterioration is determined. Specific examples of the index of the spectrum characteristic pattern include spectrum power level, spectrum center frequency mobility, spectrum sharpness, and the like. Therefore, the progress of deterioration of the shaft seal portion, the liner ring portion, and the like from the normal state to the abnormal state is diagnosed by using the spectrum power level, the spectrum center frequency mobility, and the spectrum sharpness. A schematic description of these indicators is given in FIG.
Then, the spectrum power level in the normal state is S ₁ , but it increases to a large level S ₂ in the abnormal state. That is, the change in power level ΔS = S ₂ −S ₁ > 0. Also,
In FIG. 6, the spectrum center frequency changes from f ₁ at the normal time to f ₂ at the abnormal time. Further, in FIG. 7, the spectral cusp is larger in the abnormal state than in the normal state. The spectrum power level S, the spectrum center frequency F _c , and the spectrum sharpness (cootsis value) C are defined by the following equations.

[0013]

[Equation 1]

[Equation 2]

[Equation 3]

The discriminant function D is expressed by the following equation.

Equation 4] Here _{D = αS + βF c + γC} , S represents the spectral power level index (takes an integer of 0 to 5), F _c takes the integer spectral center frequency mobility index (0-5 ), C represents a spectral sharpness index (takes an integer of 0 to 5), and α, β, γ
Represents a weighting coefficient (takes a value of 0.1 to 1.0). Normally, the weighting factors are α = 1.0 and β =
It is set in the range of α>β> γ, such as 0.5 and γ = 0.2. Table 1 shows an example of the determination. Here, D <3.
It is determined to be normal in 4; it is determined that the deterioration possibility is small in 3.4 ≦ D <6.8; and the deterioration possibility is large in 6.8 ≦ D.

[Table 1]

FIG. 8 shows a block diagram of the pump deterioration diagnosis system of the first embodiment. The vibration sensor 2 is attached to a bearing portion of a pump to be diagnosed, and can detect signals (acceleration, AE (acoustic emission), etc.) from a shaft sealing portion (seal) and a liner ring portion. This detection signal is input to the vibration converter 4 and converted into an appropriate output signal by processing such as amplification, integration and level conversion. The vibration sensor 2 and the vibration converter 4 form a vibration detection device 6. The output signal of the vibration detecting device is then subjected to a high-frequency band and a low-frequency band unrelated to the sliding vibration frequency at the normal and abnormal times of the pump by the band filter including the high-frequency band cut-off filter 8 and the low frequency band cut-off filter 10. After the vibration component in the range is cut off, it is converted into a digital output signal in the A / D converter 12. This digital output signal is input to the frequency spectrum analysis calculator 14 of digital calculation formula. This frequency spectrum analysis calculator 14 is an A / D
The digital signal from the converter 12 is sampled and the vibration spectrum pattern is calculated by a fast Fourier transform (FFT) process.

By removing the vibration spectrum component depending on the shaft rotation speed of the pump from the signal in advance, it is possible to improve the S / N ratio of the sliding vibration spectrum (only the spectral characteristic pattern) independent of the shaft rotation speed. it can. The vibration component that depends on the shaft rotation speed includes the whirling vibration n of the mechanical part generated by unbalance, misalignment, or the like.
f ₀ , vibration Af ₀ generated by damage to the rolling bearing (A is a constant determined by the specifications of the rolling bearing), meshing vibration of the gear mechanism (frequency Zf ₀ ) (where Z is the number of teeth) There is) etc. Therefore, the rotation speed detector 16 is attached to the pump to detect the rotation speed of the pump. The spectral characteristic pattern calculator 18 receives the detected rotation speed and removes the vibration frequency spectrum of the above-described component depending on the shaft rotation speed from the entire spectrum. By this calculation, only the spectrum (that is, the spectrum characteristic pattern) due to the sliding vibration component remains.

This spectral feature pattern (eg 5
Through the switch 20, the spectrum values at 12 frequencies can be stored in the normal spectrum characteristic pattern storage 22 during normal operation and in the abnormal spectrum characteristic pattern storage 24 during abnormal operation. In this way, the spectrum characteristic patterns at the normal time and at the abnormal time are stored in advance in the two storage devices 22 and 24. The storage unit 24 stores the spectrum characteristic pattern measured for the diagnosis target in a deteriorated state effective for deterioration diagnosis.

Next, at the time of pump operation (at the time of diagnosis), regarding the observed vibration spectrum characteristic pattern of the observation target of the diagnosis object output from the spectrum characteristic pattern analysis calculator 22, the observed vibration spectrum characteristic pattern is set to normal and abnormal. To determine the state of the pump.

Specifically, the spectral characteristic pattern similarity calculator 28 inputs the observed spectral characteristic pattern of the diagnosis object from the spectral characteristic pattern calculator 18 through the switch 26, and the spectral power level S,
The spectrum center frequency F _c and the spectrum sharpness C are obtained, and the discriminant function D is calculated in real time. Similarly, the spectral feature pattern similarity calculator 28 uses the switch 2
6, the normal-time spectral characteristic pattern storage unit 28 or the abnormal-time spectral characteristic pattern storage unit 24.
The discriminant function D described above is also calculated for the normal or abnormal spectral feature pattern input from the. The determination section 30 stores these discriminant function values in the normal time spectral feature pattern similarity determination reference storage unit 32 stored in the normal time spectral characteristic pattern similarity determination reference storage unit 34. It is determined whether the observed spectral feature pattern is in a normal state or in an abnormal state, based on the similarity criterion for abnormalities.
Then, the diagnostic result is sent to the display unit (CRT, printer, etc.) 32 to display or print the diagnostic result. The user can judge the presence / absence of abnormality in the diagnosis target and the degree of progress of deterioration by looking at the display result. The control device 34 is
A / D converter 12, frequency spectrum calculation analyzer 14,
The control signal is sent to the spectrum characteristic pattern calculator 18, the switch 20, the normal spectrum characteristic pattern memory 22, the abnormal spectrum pattern memory 24, the switch 26, the spectrum characteristic pattern similarity calculator 28, and the determination unit 30, By controlling these units comprehensively,
The above operation is performed.

FIG. 9 shows a flow for determining the similarity of the spectral feature patterns in the spectral feature pattern similarity calculator 28 and the determining unit 30. In the similarity determination, the spectrum power level S, the spectrum center frequency F _c ,
For each of the spectral cusps C, the similarity between the normal and abnormal spectral feature patterns is determined by the tree method. When the similarity determination is started, the spectrum feature pattern similarity calculator 28 first calculates the similarity (discriminant function D) of the spectrum feature pattern for the observed spectrum feature pattern of the diagnosis target (step S2). Next, the determination unit 30 compares this similarity with the normal spectrum characteristic pattern similarity determination standard stored in the storage 32 (step S4). When it is determined that the degree of similarity is higher than the criterion for normality (YES in step S4), it is determined to be normal (step S6). Further, when it is judged that the similarity is smaller than the judgment criterion in the normal state as compared with the judgment criterion of the spectrum characteristic pattern similarity in the normal state (N in step S4).
O), the spectral feature pattern similarity calculator 28 calculates the similarity between the observed spectrum feature pattern and the spectrum pattern stored in the abnormal spectrum feature pattern storage 24 with the abnormal spectrum pattern (step S8). ). Next, the determination unit 30 compares this similarity with the spectrum pattern similarity determination criterion at the time of abnormality stored in the storage unit 30 (step S10). If it is determined that the degree of similarity is smaller than the criterion for abnormality (YES in step S10), it is determined that the possibility of deterioration is small (step S12). If it is determined in step S10 that the degree of similarity is higher than the criterion for abnormality (NO in step S14), it is determined that the possibility of deterioration is high (step S16), and this flow is ended. With the above flow, the progress of deterioration can be determined by the transition of the observed spectrum pattern from the normal state to the abnormal state.

Here, the weight coefficient of the discriminant function D of the spectral feature pattern similarity calculator 28 and the two storage devices 3 are used.
The judgment criteria of the spectrum characteristic pattern similarity in normal times and abnormal times stored in Nos. 2 and 34 can be changed by accumulating diagnostic cases. Thereby, the diagnostic accuracy can be improved. Further, by making the weighting factor and the judgment criterion variable, the judgment criterion can be adjusted in accordance with the characteristic peculiar to the diagnosis target (the shaft sealing portion, the liner ring portion), and the judgment can be made according to the actual condition.

In the deterioration diagnosis system shown in FIG. 8, the abnormal spectrum characteristic pattern similarity determination criterion storage unit 34 is lower than the determination criterion stored in the normal spectrum characteristic pattern similarity determination criterion storage unit 32. The set similarity determination criterion may be stored. In this case, the abnormal spectrum characteristic pattern similarity determination criterion storage unit 34 functions as the second normal spectrum characteristic pattern determination criterion storage unit. For example, regarding the pump to be diagnosed, the vibration spectrum at the time of abnormality could not be obtained,
Therefore, this method can be adopted when the abnormal spectrum characteristic pattern is not calculated. At this time, the determination unit 30 sets the similarity of the observed spectrum feature pattern to 2
By comparing with the determination criteria stored in the individual similarity determination criteria storage devices 32 and 34, the diagnosis target is in a normal state, in a state in which wear degradation possibility is low, or in wear degradation possibility is high. It can be determined whether or not it is in a state.

Next, a second embodiment using pattern matching will be described. This embodiment is applied to the case where the center frequency moves as the deterioration progresses in the vibration spectrum pattern. FIG. 10 shows a block diagram of the pump deterioration diagnosis system of the second embodiment. This diagnostic system determines wear deterioration of the shaft seal portion, liner ring, etc. by shifting the observed spectrum pattern from a normal state to an abnormal state. The vibration sensor 2 is attached to a bearing portion of a pump to be diagnosed, and can detect signals (acceleration, AE, etc.) from the shaft sealing portion and the liner ring. This detection signal is
It is input to the vibration converter 4 and converted into an appropriate output signal by processing such as amplification, integration and level conversion. The vibration sensor 2 and the vibration converter 4 form a vibration detection device 6. The output signal of the vibration detecting device is then subjected to a high-frequency band and a low-frequency band unrelated to the sliding vibration frequency at the normal and abnormal times of the pump by the band filter including the high-frequency band cut-off filter 8 and the low frequency band cut-off filter 10. After the vibration component in the range is cut off, it is converted into a digital output signal in the A / D converter 12. This digital output signal is input to the frequency spectrum analysis calculator 14 of digital calculation formula. The frequency spectrum analysis calculator 14 is the A / D converter 1
The digital signal from 2 is sampled, and a vibration spectrum pattern is calculated by a fast Fourier transform (FFT) process. This vibration spectrum pattern (for example, 51
The spectrum values at the two frequencies are transmitted through the switch 56 in the normal state to the normal state spectrum pattern storage unit 5
8 and can be stored in the spectrum pattern memory 60 at the time of abnormality at the time of abnormality. In this way, two storage devices 5
The spectrum patterns of normal time and abnormal time are stored in advance in Nos. 8 and 60. The recorder 60 stores the vibration spectrum pattern measured for the diagnosis target in a deteriorated state effective for deterioration diagnosis.

Next, at the time of pump operation (at the time of diagnosis), regarding the observed vibration spectrum pattern of the observation object of the diagnosis object output from the frequency spectrum analysis calculator 14, the observed vibration spectrum pattern is the normal and abnormal spectrum patterns. To determine the state of the pump. Therefore, pattern matching is performed between the observed vibration spectrum pattern and the above-mentioned normal and abnormal spectrum patterns. There are various methods for pattern matching, but in the present embodiment, the pattern similarity is determined for each of a plurality of identification patterns (the above-mentioned normal and abnormal spectrum patterns) by calculating the pattern similarity.

Specifically, the spectrum pattern matching calculator 64 inputs the observed vibration spectrum pattern of the diagnosis object from the frequency spectrum analysis calculator 64 via the switch 66, and stores the normal time spectrum pattern selected by the switch 62. The similarity is compared with the normal or abnormal spectrum pattern from the device 58 or the abnormal spectrum pattern storage 60. The determination unit 66 determines whether the collation result is the normal-time similarity pattern determination criterion stored in the normal-time spectrum pattern similarity determination criterion storage unit 68 and the abnormal-time spectrum pattern similarity determination criterion storage unit 70 in the abnormal state. On the basis of the similarity criterion of (1), whether the observed vibration spectrum pattern is in the normal state or the abnormal state is determined in real time. The diagnostic result is displayed on the display unit (CRT, printer, etc.) 7
2 to display and print the diagnostic results. The user
The presence or absence of an abnormality in the diagnosis target and the degree of deterioration can be determined by looking at the display result. In addition, the control device 74 is an A / D
The converter 12, the frequency spectrum calculation analyzer 14, the switch 56, the normal spectrum pattern storage 58, the abnormal spectrum pattern storage 60, the switch 62, the spectrum pattern matching calculator 64, and the determination unit 6.
A control signal is sent to 6 to comprehensively control these units to perform a predetermined operation.

Here, the criteria for determining the spectral pattern similarity in the normal state and the abnormal state, which are stored in the two storage devices 68 and 70, can be changed by accumulating diagnostic cases. This allows
The diagnostic accuracy can be improved. Further, by making the judgment criterion variable, the judgment criterion can be adjusted according to the unique characteristics of the diagnosis target (shaft seal portion, liner ring), and the judgment can be made according to the actual condition.

FIG. 10 shows a flow of judgment in the judgment unit 66. When the determination is started, first, the degree of similarity between the observed vibration spectrum pattern of the diagnostic object and the spectrum pattern stored in the normal-time spectrum pattern storage unit 58 is calculated (step S). S32). Next, this similarity is stored in the storage unit 5.
It is compared with the normal time spectrum pattern similarity determination criterion stored in step 8 (step S34). When it is determined that the similarity is higher than the criterion for normality (Y in step S34)
ES), it is determined to be normal (step S36), and this flow ends. Further, when it is judged that the similarity is smaller than the judgment criterion in the normal state as compared with the judgment pattern in the normal time spectrum pattern similarity (NO in step S34), the observed vibration spectrum pattern and the abnormal time spectrum pattern storage unit 60 are stored. The similarity between the stored spectrum pattern and the spectrum pattern at the time of abnormality is calculated (step S
38). Next, this similarity is compared with the spectrum pattern similarity determination criterion at the time of abnormality stored in the storage device 60 (step S40). If it is determined that the similarity is smaller than the criterion for abnormality (YES in step S40), it is determined that neither normality nor abnormality can be determined (step S4).
2) Then, this flow ends. Also, step S40
Then, if it is determined that the degree of similarity is larger than the criterion for abnormality (NO in step S44), it is determined to be abnormal (step S46), and this flow ends. With the above flow, the progress of deterioration can be determined by the transition of the observed spectrum pattern from the normal state to the abnormal state.

[0028]

According to the pump deterioration diagnosis system of the present invention,
It is possible to judge the progress of deterioration of sliding vibration sources such as a shaft seal portion, a liner ring portion, an impeller ring portion, and an intermediate bearing portion of a pump, which has been difficult to diagnose in the past. In particular, the progress of deterioration can be determined even when the center frequency moves in the spectrum pattern. As a result, it becomes possible to more appropriately determine the repair time of the pump and to examine the repair content in advance, and it is possible to reduce the maintenance cost. The pump deterioration diagnosis system of the present invention narrows the index of the characteristic parameter to a specific characteristic parameter as compared with the diagnosis method using the neural network model, and facilitates the practical operation of the system.

[Brief description of drawings]

FIG. 1 is a diagram schematically showing a vibration spectrum pattern of a normal seal and a liner ring.

FIG. 2 is a diagram schematically showing a vibration spectrum pattern of a seal and a liner ring whose deterioration is progressing.

FIG. 3 is a diagram schematically showing vibration spectrum patterns of an impeller and a liner ring under normal conditions.

FIG. 4 is a diagram schematically showing vibration spectrum patterns of an impeller and a liner ring at the time of damage.

FIG. 5 is a diagram showing a change in spectrum power level from a normal state to an abnormal state.

FIG. 6 is a diagram showing a shift of a spectrum center frequency from a normal state to an abnormal state.

FIG. 7 is a diagram showing a change in spectral sharpness from a normal state to an abnormal state.

FIG. 8 is a block diagram of a system of a first embodiment of a deterioration diagnosis system.

9 is a flowchart of the determination of the determination unit of the deterioration diagnosis system shown in FIG.

FIG. 10 is a block diagram of a second embodiment of the deterioration diagnosis system.

11 is a flowchart of the determination of the determination unit of the deterioration diagnosis system shown in FIG.

[Explanation of symbols]

2 ... Vibration sensor, 4 ... Vibration converter, 12 ... A /
D converter, 14 ... Frequency spectrum analysis calculator, 22 ...
Normal-time spectrum feature pattern storage, 24 ... Abnormal-time spectrum feature pattern storage, 28 ... Spectral feature pattern similarity calculator, 30 ... Judgment unit, 32 ... Normal-time similarity determination reference storage, 34 ... Abnormal-time similarity Judgment standard storage device, 58 ... Spectral pattern storage device under normal conditions, 60 ...
Abnormal spectrum pattern storage, 64 ... Spectral pattern matching calculator, 66 ... Judgment unit, 68 ... Normal time similarity determination reference storage, 70 ... Abnormality similarity determination reference storage.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Kazumi Fukuda 2-3-2 Uchisaiwaicho, Chiyoda-ku, Tokyo Kawasaki Steel Corporation Tokyo headquarters (72) Inventor Isao Nagai 14th Nihonbashi Kodenmacho, Chuo-ku, Tokyo No. 4 Kawatetsu Advantech Co., Ltd. Tokyo Branch (72) Inventor Yoshio Akiba 14-4 Nihonbashi Kodenmacho, Chuo-ku, Tokyo Kawatetsu Advantech Co., Ltd. Tokyo Branch

Claims (8)

[Claims] 1. A vibration signal detecting device for detecting vibration of a sliding vibration source as a diagnostic object as an electric signal and outputting it as a digital signal, and a vibration spectrum pattern by sampling a digital signal inputted from the vibration signal detecting device. A frequency spectrum analysis calculator for calculating, a first memory for storing a normal spectrum pattern input from the frequency spectrum analysis calculator, and an abnormal spectrum pattern input from the frequency spectrum analysis calculator The normal spectrum pattern stored in the second storage unit, the first storage unit and the abnormal spectrum pattern stored in the second storage unit are compared with the observed spectrum pattern input from the frequency spectrum analysis calculator, and the observed spectrum pattern is observed. Normal spectrum pattern and abnormal spectrum pattern A pump deterioration diagnosis system including a pattern comparison device for diagnosing deterioration progress of a diagnosis target based on similarity to the above. 2. The pump deterioration diagnosing system according to claim 1, wherein the pattern comparison device receives the observed spectrum pattern and the spectrum patterns stored in the first storage device and the second storage device. A pump deterioration diagnosis system characterized by calculating a power level, a center frequency mobility, and a sharpness, and using a predetermined discriminant function as a function of these three indexes to diagnose deterioration progress. 3. The pump deterioration diagnosis system according to claim 1, further comprising: a rotation speed detector for detecting the rotation speed of the pump, and the rotation speed of the pump inputted from the rotation speed detector, and the frequency spectrum described above. The vibration spectrum pattern is input from the analysis calculator, and a spectrum pattern (hereinafter, referred to as a spectrum characteristic pattern) in which the rotating vibration spectrum generated depending on the shaft rotation speed of the diagnosis target is removed is calculated, and the first pattern described above is calculated.
A storage unit, a second storage unit, and a spectrum feature pattern calculator for outputting to the pattern comparator, wherein the first storage unit and the second storage unit store the spectrum feature pattern, and the pattern comparison apparatus described above , A pump deterioration diagnosis system characterized by calculating a degree of similarity for a spectral characteristic pattern. 4. The pump deterioration diagnosis system according to claim 3, wherein the comparison device includes the observed spectrum feature pattern input from the spectrum feature pattern calculator and the normal spectrum stored in the first storage device. Feature pattern and second
A similarity calculator for calculating the similarity with the abnormal spectrum feature pattern stored in the memory, a third memory for storing the criterion for determining the similarity with the normal spectrum feature pattern, and an abnormal spectrum feature pattern And a fourth storage unit that stores a determination criterion of the similarity degree between the first and second storage units, and a diagnosis target by comparing the similarity degree input from the similarity degree computing unit and two determination criteria input from the third storage unit and the fourth storage unit. And a judging device for judging the deterioration progressing state of the pump. 5. The pump deterioration diagnosis system according to claim 4, wherein the similarity calculator calculates the power level, the center frequency mobility and the sharpness from the spectral feature pattern,
The pump deterioration diagnosis system, wherein the above-mentioned judging device diagnoses deterioration progress using a predetermined judging function as a function of these three indexes. 6. The pump deterioration diagnosis system according to claim 4, wherein the fourth memory stores a similarity criterion lower than a similarity criterion stored in the third memory. The above judging device compares the discriminant function value input from the spectral characteristic pattern calculator with the judgment criterion inputted from the third memory device and the fourth memory device, and judges the deterioration progress state of the diagnosis target. A pump deterioration diagnosis system. 7. The pump deterioration diagnosis system according to claim 1, wherein the pattern comparison device has a normal spectrum pattern stored in the first storage device and an abnormal spectrum pattern stored in the second storage device. On the other hand, a pump deterioration diagnosis system characterized by performing pattern matching of an observed spectrum pattern input from a frequency spectrum analysis calculator and diagnosing deterioration progress of a diagnosis target based on the obtained similarity. 8. The pump deterioration diagnosis system according to claim 7, wherein the pattern comparison device includes an observed spectrum pattern input from a spectrum pattern calculator and a normal-time spectrum pattern stored in the first storage unit. And a pattern matching calculator for calculating the similarity with the abnormal spectrum pattern stored in the second memory, a third memory for storing the criterion for determining the similarity with the normal spectrum pattern, and an abnormal spectrum pattern And a fourth storage unit that stores the determination criterion of the similarity degree with the similarity between the similarity input from the pattern matching calculator and the two determination criteria input from the third storage unit and the fourth storage unit. And a judging device for judging the deterioration progressing state of the pump.