JPH0557528B2 - - Google Patents

Description

[Detailed Description of the Invention] [Objective of the Invention] (Industrial Application Field) The present invention is directed to the early detection and diagnosis of abnormal signs in the operating conditions of high-speed rotating machinery such as steam turbines and generator plants by detecting shaft vibration. This invention relates to a diagnostic device for rotating machines based on values.

(Prior Art) Generally, most abnormalities during operation of rotating machines appear as changes in shaft vibration. Particularly in steam turbines, power generation units, etc. for power generation plants, rotors weighing tens to hundreds of tons rotate at high speed, and even minute vibrations can lead to serious accidents. Therefore, in order to understand vibration phenomena during operation, vibrations are detected at each bearing part and continuously recorded by a recorder, and an alarm value is set to call for caution when driving. Based on the detected vibration value, operation restrictions are etc. is going on. Furthermore, recently, as shown in Japanese Patent Laid-Open No. 61-128128, a diagnostic device has been proposed that detects abnormality symptoms before the operating limit value is reached and further diagnoses the cause from characteristics such as vibration frequency.

FIG. 8 shows a conventional diagnostic device for the steam turbine/generator unit 1. Vibration detectors 6a, . . . 6i are provided in bearings 5a, . Additionally, there is a tachometer 7 that detects the rotational speed of the rotor, a phase reference pulse transmitter 8 that generates a reference pulse when detecting the phase of shaft vibration in each bearing section 5a,...5i, and a generator output ( A group of various sensors 9 are installed to detect the operating status of the plant, such as load), steam temperature, and bearing temperature. The outputs of the tachometer 7 and various sensor groups 9 are then sent to the operation monitoring device 11 provided in the central operation panel 10.
The outputs of the vibration detectors 6a, . . . 6i and the phase reference pulse transmitter 8 are sent to a vibration monitoring device 12 provided in the central operation panel 10, where they are constantly checked. If a predetermined limit value is exceeded, an alarm device 13 is activated. A command signal is issued, and a warning or automatic trip signal is output. The outputs of the vibration detectors 6a, ... 6i, the tachometer 7, the phase reference pulse transmitter 8, and the various sensor groups 9 are connected to an analog continuous recorder 14.
It will also be sent and recorded. In addition, vibration detectors 6a,...
6i and the output of the phase reference pulse transmitter 8 are sent to the A/D by the vibration monitoring device 12.
It is converted and sent to the shaft vibration diagnosis device 30 for diagnosis. This shaft vibration diagnosis device 30 is a system that uses a computer to perform abnormality diagnosis and support an operator. The A/D converted vibration data is compared with a predetermined threshold value, and an abnormality sign is detected by the abnormality sign detection device 16. In addition, the above A/
The frequency analyzer 17 analyzes the frequency component distribution of shaft vibration based on the D-converted vibration data. The diagnostic unit 19 then diagnoses the cause of the abnormality symptom based on the outputs of the various sensor groups 9, the abnormality symptom detection device 16, and the frequency analyzer 17, as well as past history data stored in the memory 21. The results are displayed on the display device 20, and various data used in the diagnosis and the diagnosis results are stored in the memory 21.

(Problem to be solved by the invention) Since the shaft vibration of rotating machinery can change rapidly depending on the degree of abnormality in the operating condition, early detection of abnormality symptoms and early diagnosis of the degree of abnormality are required. , real-time characteristics are required. However, since these diagnostic results are expected to directly contribute to plant operation, prevention of malfunctions in detecting abnormality symptoms is also an important issue. For example, Fig. 9 shows a conventional method for detecting abnormality symptoms. Based on the shaft vibration amplitude level and the value of the amplitude increase rate at that time, judgment areas are distinguished into normal range, alarm range, and trip range, and the degree of emergency is determined. Judging.
Furthermore, the amplitude level and amplitude increase rate used for the above judgment are determined by the measured value V _i at the detection time t _i as shown in FIG. The value obtained from the previous n pieces of data V _i-o+1 to _{V i} using the least squares method using the temporary approximation formula V=α・T+ρ...(1), V (vibration amplitude level), α (amplitude increase rate) ). Here, T is a sampling time determined by the sampling interval and the number of samples.

By performing such processing, it is possible to prevent malfunctions in abnormality symptom detection due to data variations. However, by averaging changes, abnormal signs that change minutely may not be detected.
In particular, among the unbalanced vibrations that account for the majority of vibration abnormalities in rotating machinery, those that can lead to serious accidents include those caused by "mechanical unbalance" changes in rotating parts such as flying blades, and "rubbing". Some of this is caused by bending of the rotor due to contact with a stationary part, etc., but since the vibration characteristics of the instantaneous value of the vibration are similar in both cases, the difference in the form of change before and after the vibration is the deciding factor.

Conventionally, in addition to alarms using digital values, this information was recorded using an analog continuous recorder 14.
Judgments have been made based on the operator's experience and expertise. However, in computer-based driving support diagnostic systems, intervening human judgment at the time of initial diagnosis impairs online performance and real-time performance, resulting in the problem that the system does not meet its intended purpose.

The present invention has been made in consideration of the above-mentioned problems, and an object of the present invention is to provide a rotating machine diagnostic device that is excellent in online performance and real-time performance.

[Structure of the invention]

(Means for Solving the Problems) A diagnostic device for a rotating machine according to the present invention includes a detector for detecting shaft vibration of the rotating machine, an A/D converter for A/D converting a detection signal of the detector, This A/D
An abnormality indication means for detecting whether or not there is an abnormality sign in the operating state of the rotating machine based on the vibration data converted from A/D by the converter; and a frequency analysis means based on the vibration data converted from A/D. a frequency analysis means for detecting an abnormality, a sudden event detection means for detecting whether a sudden event occurs based on vibration data before and after the time when an abnormality symptom is detected, a memory for storing abnormal operation data of the rotating machine, and a memory for storing abnormal operation data of the rotating machine; When an abnormality symptom is detected by the symptom detection means, based on the output of the frequency analysis means, the output of the sudden event detection means, the current operating data of the rotating machine, and the past abnormal operation data stored in the memory. The present invention is characterized by comprising a diagnostic unit for diagnosing abnormal signs and storing data used in the diagnosis and diagnostic results in a memory, and display means for displaying the diagnostic results diagnosed by the diagnostic unit. do.

(Function) In the rotating machine diagnostic device according to the present invention configured as described above, the abnormality sign detection means determines whether or not there is an abnormality sign in the operating state of the rotary machine based on the A/D converted vibration data. detected.

Furthermore, frequency analysis is performed by the frequency analysis means based on the A/D converted vibration data.
When an abnormality sign is detected by the abnormality sign detection means, the trend of changes in shaft vibration before and after the time when the abnormality sign is detected is detected based on the vibration data before and after the time when the abnormality sign is detected. Whether or not there is an unusual event is detected by the unusual event detection means,
Based on this detection result, the output of the frequency analysis means, the current operating data of the rotating machine, and the past abnormal operating data stored in the memory, the diagnostic section diagnoses the abnormality symptoms, and also By displaying the information on the display means, the rotating machine diagnostic apparatus according to the present invention does not involve human judgment, and is excellent in online performance and real-time performance.

(Embodiment) An embodiment of a rotating machine diagnostic apparatus according to the present invention will be described with reference to FIGS. 1 and 2. The diagnostic device of this embodiment is similar to the conventional diagnostic device shown in FIG.
3b and a generator 4, and vibration detectors 6a,...6.
i, tachometer 7, phase reference pulse transmitter 8, various sensor groups 9, operation monitoring device 11, vibration monitoring device 1
2, an alarm device 13, a recorder 14, and a shaft vibration diagnosis device 15. Since the components other than the shaft vibration diagnosis device 15 have already been explained in the section of the prior art, their explanation will be omitted.

The shaft vibration diagnosis device 15 is an abnormality symptom detection device 16
, a frequency analyzer 17 , and a sudden event detection device 18
, a diagnostic section 19 , a display device 20 , and a memory 21
It has In the vibration monitoring device 12, A/
The D-converted vibration data is sent to the shaft vibration diagnosis device 15.
The frequency components are transmitted to the frequency analyzer 17 in the device 1, and the distribution of frequency components is determined.
6 is also transmitted. Based on the A/D converted vibration data and the output of the frequency analyzer 17, the values of the following items (1) Vibration amplitude level V (2) Amplitude increase rate α (3) Amount of amplitude change in instantaneous value (each (difference between sampling data and previous data) △V _p = V _i −V _i-1 (4) Change in frequency components for each frequency band (5) Deviation from standard data (for example, past performance data) ( 6) The abnormality sign detection device 16 determines whether the phase change (vector change amount), etc. exceeds each limit value, and the presence or absence of an abnormality sign is detected based on the result of this determination. For example, if at least one of the values in items (1) to (6) above exceeds the limit value, the abnormality sign detection device 16 determines that there is an abnormality sign as shown in FIG. Ru. Note that items (4) to (6) can be added or deleted depending on the configuration method of the diagnostic system and knowledge base, and are combined into one item in FIG. 2.

When an abnormality symptom is detected by the abnormality symptom detection device 16, the sudden event detection device 18 is activated, and the value of each of the following items (1) Amount of amplitude change in instantaneous value △V _p =V _i −V _{i -1} (2) Difference in average value △V _n (3) Difference in predicted value △V _c is judged whether it exceeds each limit value, and the values of the above three items all exceed their respective limit values. In this case, as shown in FIG. 2, the sudden event detection device 18 determines that there is a sudden event. Here, the difference in average value △V _n is the average value of n pieces of amplitude data V _io ,...V _i-1 before abnormality symptom detection (t _i ) and n pieces of amplitude data V _i ,... after detection. It shows the difference from the average value of V _{i + o-1} , and the predicted value difference △V _c is the vibration at the detection time t _i calculated by the least squares method from the vibration data before the abnormality symptom detection. amplitude value
It is the difference between V′ and the vibration amplitude value V″ at time t _i , which is obtained from the data after detection using the least squares method using a first-order approximation formula. Here, V′ and V″ are calculated using equation (1). is expressed as follows.

V'=α'・T+β'; (V _io ,...V _i-1 ) V''=β''; (V _i ,...V _i+o-1 )...(2) In addition, in the calculation of the above average value, A simple average or a weighted average is used.

Next, it is determined whether or not the vibration is synchronous based on the distribution of frequency components determined by fast Fourier transform (FFT) by the frequency analyzer 17 and the output of the tachometer 7 transmitted via the operation monitoring device 11. The determination is made by the diagnostic unit 19. Based on this determination result, the output of the abnormality sign detection device 16, and the output of the sudden event detection device 18, it is determined whether the abnormality sign detected by the abnormality sign detection device 16 is due to mechanical imbalance or rubbing. The diagnostic unit 19 diagnoses whether the rotor is bent due to contact with a stationary part or the like. If a mechanical imbalance is diagnosed, it is determined based on the operating data of the steam turbine/power generation unit 1 transmitted via the operation monitoring device 11 and past history data stored in the memory 21 whether it is due to coupling misalignment. The diagnosis section 19 further diagnoses whether the problem is due to , scattering of the blade part, or the scattering of the blade part. The results diagnosed by the diagnostic section 19 are displayed on the display device 20 and are stored in the memory 21 together with the data used for diagnosis.

Next, a method for diagnosing whether an abnormality symptom is caused by mechanical imbalance or rubbing when synchronous vibration is diagnosed will be explained using FIGS. 3 and 4. First, as shown in FIG. 3, a case where there is a so-called sudden event in which the vibration amplitude changes greatly at time t _i will be described. time t _i-1
Up to, the amplitude value V, the amplitude increase rate α, and the amount of change in instantaneous value △V _p are all set to their respective thresholds.
V _A and α _A and A ₁ are not exceeded. time t _i
Then, the amount of change ΔV ₆ in the amplitude value V and the instantaneous value exceeds the respective threshold values V _A and A ₁ , and the abnormality sign detection device 16 detects an abnormality sign.
Then, at time t _i+o-1 , the average value of n pieces of vibration amplitude data V _io , ...V _i-1 before detection time t _i and n pieces of vibration amplitude data V _i , ...V after detection time The absolute value of the difference in mean value from _i+o-1 △V _n exceeds the corresponding threshold A ₂ , the difference in predicted values △V _c also exceeds the corresponding threshold A ₃ , and △V _n When ΔV _c is reached, the sudden event detection device 18 determines that the event is a sudden event. Then, the diagnostic section 19 diagnoses that the abnormal symptom is due to mechanical imbalance. Next, FIG. 4 shows how the vibration amplitude rapidly increases when rubbing occurs. Before time t _i , the amplitude value V and the amount of change ΔV _p in the instantaneous value do not exceed the respective thresholds V _A and A ₁ , but the amplitude increase rate α exceeds the threshold α _A. After time t _i , the amplitude value V, the amplitude increase rate α, and the amount of change in instantaneous value ΔV _p exceed their respective thresholds.

On the other hand, the difference △V _n between the average values before and after time t _i exceeds the threshold A ₂ , but the difference △V _c between the predicted values is small and does not exceed the threshold A ₃ , and △V _n ≫△ It becomes V _c .
Therefore, the sudden event detection device 18 determines that there is no sudden event. Based on this determination result, the diagnosis unit 19 diagnoses that the abnormality sign detected by the abnormality sign detection device 16 is caused by rubbing or the like.

From the above, the diagnostic device of this embodiment can detect the presence or absence of an unexpected event without human intervention.
It becomes possible to judge in real time.

In the embodiment illustrated in FIG. 1, the frequency analyzer 17 is included in the shaft vibration diagnosis device 15, but an independent configuration is also possible as shown in FIG.

That is, the shaft vibration diagnosis device 15a includes the abnormality symptom detection device 16, the sudden event detection device 18, and the diagnosis section 1.
9, a display device 20, and a memory 21.

In addition, in order to prevent malfunctions due to variations in data after A/D conversion, the amount of change in instantaneous value △V _p is changed as shown in Figure 6.
After V _p exceeds the threshold and a sudden event is confirmed by the sudden event detection device 18, that is, when the difference in average values and the difference in predicted values exceed their respective thresholds. It is also possible to have the abnormality sign detection device 16 determine that there is an abnormality sign.

In addition, sudden events can be detected without using the amount of change in instantaneous values, the difference in average values, and the difference in predicted values.
It is also possible to detect it based on the characteristics of the change in the amplitude increase rate α, that is, the pattern of “sudden increase followed by sudden decrease” as shown in FIG.

FIG. 7a is a graph of amplitude values when there is a sudden event, and FIG. 7b is a graph showing changes in the amplitude increase rate α corresponding to FIG. 7a. At this time, the change in the amplitude increase rate α shows a pattern in which it increases rapidly until a certain time, but then decreases rapidly.

FIG. 7c is a graph of amplitude values when rubbing etc. occur, and FIG. 7d is a graph of the corresponding amplitude increase rate α. At this time, the amplitude increase rate α is gradually increasing. FIG. 7e is a graph of the amplitude value V when no sudden event or rubbing occurs, and FIG. 7f is a graph of the corresponding amplitude increase rate α. At this time, the amplitude increase rate α is gradually decreasing. When detecting a sudden event from such a pattern of changes in the amplitude increase rate α, it is necessary to install a prevention circuit that prevents changes from repeating to prevent malfunctions.

〔Effect of the invention〕

According to the present invention, it is possible to provide a diagnostic device for a rotating machine that can determine the presence or absence of a sudden event in real time without human intervention.

[Brief explanation of drawings]

Fig. 1 is a block diagram showing an embodiment of the rotating machine diagnostic device according to the present invention, Fig. 2 is a block diagram illustrating the operation of the embodiment shown in Fig. 1, and Fig. 3 is a diagram illustrating the operation when a sudden event occurs. FIG. 4 is a graph showing changes in amplitude when rubbing etc. occur; FIG. 5 is a block diagram showing a first modification of the embodiment shown in FIG. 1; FIG. FIG. 7 is a block diagram showing a second modification of the embodiment shown in FIG. 1, and FIG. 7 is an amplitude value V and its corresponding amplitude, illustrating another detection method of the sudden event detection device according to the present invention. FIG. 8 is a block diagram showing a conventional diagnostic device, and FIG. 9 and FIG. 10 are graphs illustrating a shaft vibration monitoring method of a conventional diagnostic device. 1...Steam turbine/generator unit, 2...
High and intermediate pressure turbine, 3a, 3b...low pressure turbine,
4... Generator, 5a-5i... Bearing, 6a-6i
... Vibration detector, 7 ... Tachometer, 8 ... Phase reference pulse transmitter, 9 ... Various sensor groups, 10 ... Central operation panel, 11 ... Operation monitoring device, 12 ... Vibration monitoring device, 13 ...Alarm device, 14...Recorder,
15... Shaft vibration diagnosis device, 16... Abnormal symptom detection device, 17... Frequency analyzer, 18... Sudden event detection device, 19... Diagnosis section, 20... Display device,
21...Memory.

Claims (1)

[Claims] 1. A detector that detects shaft vibration of a rotating machine, an A/D converter that A/D converts the detection signal of this detector, and a frequency analysis means that performs frequency analysis based on the A/D converted vibration data. and an abnormality sign detection means for detecting whether or not there is an abnormality sign in the operating state of the rotating machine based on the vibration data A/D converted by the A/D converter and the output of the frequency analysis means. When abnormal symptoms are detected, A/
Based on the D-converted vibration data, the amount of change in the amplitude of the instantaneous value of shaft vibration before and after the abnormal symptom detection time, the difference between the average values of the shaft vibration amplitude data before and after the abnormal symptom detection, and each prediction. A sudden event detection means that calculates the difference between the values and determines that there is a sudden event only when the amplitude change amount of the instantaneous value, the difference between the average value, and the difference between the predicted values each exceed a limit value. , a memory for storing abnormal operation data of the rotary machine, and a determination as to whether or not the vibration is synchronous based on the output of the frequency analysis means and the detection output of the rotational speed of the rotary machine, and the determination result and the abnormality symptom detection. diagnosing the abnormality symptom based on the output of the means, the output of the sudden event detection means, current operation data of the rotating machine, and past abnormal operation data stored in the memory; A diagnosing device for a rotating machine, comprising: a diagnosing section that stores used data and diagnostic results in the memory; and a display means that displays the diagnostic results diagnosed by the diagnosing section.