JP2009180722A - Support method for optimal maintenance time determination of object facility, computer program, and support device for optimal maintenance time determination of object facility - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a support method for optimal maintenance time determination of an object facility, a computer program, and a support device for the optimal maintenance time determination of the object facility. <P>SOLUTION: This support method comprises an easy diagnostic step of diagnosing mechanical integrity at a certain time, a precision diagnostic step of estimating class and degradation grade of mechanical degradation, a tendency monitoring step of tracking evolution of the degradation state, a life prediction step of predicting the arrival time at a dangerous region of the failure from a degradation pattern in the tendency monitoring step, an energy loss evaluation step of estimating the amount of energy loss of the object facility from the evolution of the degradation grade in the tendency monitoring step, and an energy load evaluation step required for restoring the degradation component or degradation device, and optimizes the restoring time of the degradation state by adding the energy loss evaluation result to the energy load evaluation result. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a support method for determining the optimum maintenance time of a target facility, a computer program, and a support device for determining the optimal maintenance time of a target facility.

By acquiring vibration data of the target equipment, diagnosing the mechanical soundness of the target equipment, monitoring the tendency of the deterioration state, and predicting the lifetime, the maintenance time of the target equipment is determined.
For example, as the vibration data, data such as vibration displacement, vibration speed, vibration acceleration, vibration acceleration distortion, vibration acceleration kurtosis, vibration acceleration crest factor, and the like are acquired, and the acquired data is set as a predetermined threshold value. By comparing, it is determined whether or not there is an abnormality in the target equipment, and the state prediction is conventionally performed on the assumption that the data in a normal state follows a “normal distribution” (see, for example, Patent Document 1).

Alternatively, in order to solve the above-described problem of Patent Document 1, state determination and state prediction are performed using various information processing techniques by converting vibration waveform data and vibration feature values into a specified probability distribution, usually a normal distribution. (For example, refer to Patent Document 2).
Patent Publication 2000-171291 Patent WO2004-068078

  Furthermore, when deciding the maintenance time of the degraded part of the target equipment or the deteriorated equipment, in addition to the progress of the deteriorated state, it is necessary to input energy to restore the target facility's operating energy loss or the deteriorated state due to the deterioration. Taking into consideration the energy burden from the viewpoint of making, there is no consideration for determining the maintenance time of the deteriorated part or the deteriorated equipment.

  By the way, in the diagnosis and maintenance of the target equipment, the user predicts the life of the target component or device while performing trend monitoring as well as early detection of abnormality, but the accuracy of the trend monitoring and life prediction is insufficient, We do not have an objective criterion for determining when to actually stop the target equipment and recover the deteriorated part. The main reason is that in the conventional technology, the future is predicted from the past transition of the diagnostic index based on the vibration data in the normal state, but when the deterioration phenomenon occurs, the change in the index of the vibration data is past normal. This is because the state transition is completely different.

Furthermore, not only the mechanical maintenance time obtained from the trend monitoring and life prediction of the degradation event, but also the maintenance time comprehensively considering the energy loss due to the progress of degradation of the target equipment and the energy burden for restoration of the degraded part There was no knowledge of evaluation methods for decision optimization.

Originally, in order to determine the service life at the component level of the target equipment, the stress generated in the component or member is external force from three directions, frictional force, etc., and chemical damage (corrosion, etc.) of the component or member. Complex factors such as design / manufacturing specifications, product quality, career after operation, and changes in operating conditions in the future are intertwined, and it is necessary to examine them from various perspectives.
Here, taking up the factors from various aspects described above and accurately predicting the progress of the deterioration state and the life of the target part or member is the best measure, but it is not realistic. That being said, the transition of the degradation state with completely different progress mechanism is predicted by the progress, accumulation and analysis of vibration data or its feature value in the normal state where no degradation (abnormal) event has occurred as in the prior art. This method is not practical because it has little evidence.

The present invention has been made in view of such a problem, and its purpose is to provide a method for predicting a deterioration state that is simple, realistic, and can be put to practical use, although it is a suboptimal measure. is there. In other words, this is a technique characterized by the following.
・ Search for a more universal index corresponding to the type of abnormality deterioration ・ Search for an index monotonously corresponding to the degree of abnormality deterioration ・ Construction of the effect of vibration speed on the progress of abnormality deterioration from actual vibration data In addition to providing a diagnosis method for the target equipment, a method for tracking the progress of deterioration, and a method for predicting the service life, comprehensive evaluation of the amount of energy lost due to the mechanical deterioration and the amount of energy load required for restoration of the deteriorated part is performed. It is to support the determination of the optimal maintenance time.

The main present invention diagnoses a plurality of deterioration states of a target facility using vibration data, predicts the progress and life of the deterioration state, and restores energy required for recovering a failure caused by the deterioration of the target facility. Determine the monthly recovery energy calculated from the amount of recovery and the recovery time of the failure, and at the same time, determine the optimal maintenance time for the target facility considering the increase in operating energy loss of the target facility due to the mechanical deterioration In the support system for determining the optimal maintenance time for equipment
Simple diagnosis step for diagnosing mechanical soundness at a certain point in time, precise diagnosis step for estimating the type and degree of deterioration of the mechanical deterioration, trend monitoring step for tracking the progress of the deterioration state, and trend monitoring step A life prediction step for predicting the arrival time of the failure from the deterioration pattern in the failure, an energy loss evaluation step for estimating the amount of energy loss of the target facility from the progress of the deterioration degree in the trend monitoring step, and the deteriorated component Alternatively, an energy load evaluation step necessary for restoring the deteriorated device, and an optimization of timing for restoring the deteriorated state by adding the energy loss evaluation result and the energy load evaluation result This is a maintenance time decision support system.

  Other features of the present invention will become apparent from the description of the present specification and the accompanying drawings.

  At least the following will be made clear by the description of the present specification and the accompanying drawings.

Diagnose multiple deterioration states of the target equipment using vibration data, predict the progress and life of the deterioration state, recover the amount of energy required to recover the failure caused by the deterioration of the target equipment, and the Optimizing the facility to determine the monthly maintenance energy calculated from the restoration time and making a decision on the optimum maintenance time of the target facility considering the increase in operating energy loss of the target facility due to the mechanical degradation In the maintenance time decision support system,
Simple diagnosis step for diagnosing mechanical soundness at a certain point in time, precise diagnosis step for estimating the type and degree of deterioration of the mechanical deterioration, trend monitoring step for tracking the progress of the deterioration state, and trend monitoring step A life prediction step for predicting the arrival time of the failure from the deterioration pattern in the failure, an energy loss evaluation step for estimating the amount of energy loss of the target facility from the progress of the deterioration degree in the trend monitoring step, and the deteriorated component Alternatively, an energy load evaluation step necessary for restoring the deteriorated device, and an optimization of timing for restoring the deteriorated state by adding the energy loss evaluation result and the energy load evaluation result Maintenance time decision support system.

  According to the diagnosis method, trend monitoring, life prediction, and optimum maintenance time determination method of the target equipment, it is practical to monitor the progress of the deterioration event of the target equipment and optimize the recovery time of the deteriorated portion.

In the simple diagnosis step of the target equipment, a plurality of waveform feature amounts based on vibration time waveform data include vibration displacement, vibration speed, vibration acceleration, vibration acceleration skewness, vibration acceleration kurtosis, and vibration acceleration. Taking into account at least one of the crest factors of the noise and a plurality of deterioration-specific dimensionless feature quantities that are specific to the deterioration obtained by dividing the spectrum of the vibration data and correlating the strength between the divided frequency bands on the time axis It is good.
In such a case, when performing a simple diagnosis, a reliable diagnosis result is obtained not only by the determination result by the vibration waveform but also by the combined determination with the determination result by the dimensionless feature amount in the spectrum of the vibration data. You can also.

Also, the ISO absolute determination result based on the current vibration speed average value and the vibration displacement, vibration speed, vibration acceleration, vibration acceleration skewness, and vibration acceleration kurtosis, which are a plurality of waveform feature values in the previous period based on the time waveform data of vibration And a plurality of deteriorations specific to the deterioration obtained from the determination result based on at least one of the crest factors of vibration acceleration and the correlation of the strength of the divided frequency bands on the time axis by dividing the spectrum of the vibration data It is good also as determining the management treatment in the next time combining the determination result based on a specific dimensionless feature-value, and three types of determination results.
In such a case, it is possible to determine the type of treatment in the subsequent trend monitoring according to the degree of progress of deterioration of the target facility.

In addition, a plurality of deterioration-specific non-dimensional feature quantities that are specific to the deterioration obtained from the increase / decrease tendency of the average value of the current vibration speed and the correlation of the vibration data spectrum on the time axis of the strength between the divided frequency bands It is good also as selecting the treatment classification of the monitoring level in the next time combining the increase / decrease tendency of the said degradation specific dimensionless feature-value corresponding to the said degradation event.
In such a case, it is possible to grasp the characteristics of the progress of the degradation state of the target equipment, the future prediction is simple, and the accuracy can be improved.

Alternatively, the deterioration state at the next time may be predicted based on the power law of the ratio between the average value of the vibration speed at the latest time and the average value of the vibration speed at the current time.
In such a case, since the influence of the vibration energy change at the deteriorated part due to the vibration speed is taken into account, the characteristics of the progress of fatigue, cracks, wear, etc. at the deteriorated part and the accuracy of prediction are improved. .

Also, the vibration data spectrum is divided, and the deterioration-specific non-dimensional features corresponding to the deterioration event among a plurality of deterioration-specific non-dimensional features that are specific to the deterioration obtained from the correlation in the time axis of the strength between the divided frequency bands. For a dimension feature, a secondary caution threshold is provided between a primary caution threshold corresponding to the start of occurrence of the degradation event and a danger threshold for shifting from the caution area to the danger area, and the primary feature threshold The deterioration state may be predicted between the attention threshold and the second attention threshold.
In such a case, it is possible to greatly improve the prediction accuracy of the deterioration state while simply defining the period for extracting the progress characteristics related to the deterioration event.

In addition, based on the degradation-specific dimensionless feature amount corresponding to the degradation event among a plurality of dimensionless feature amounts obtained by dividing the spectrum of vibration data and obtaining the correlation between the strengths of the divided frequency bands on the time axis. The amount of energy loss corresponding to the deterioration state may be estimated.
In such a case, the amount of energy loss corresponding to the deterioration event of the target facility can be estimated from the vibration data.

Further, in the energy loss evaluation step and the energy load evaluation step, the energy loss corresponding to the deterioration progress of the target facility is evaluated, and the recovery energy of the failure caused by the deterioration is evaluated by the environmental load due to the carbon dioxide emission to the environment. It is good as well.
In such a case, the energy loss amount and the recovery energy load amount can be expressed by a unified evaluation standard called environmental load.

In addition, a computer program for realizing an optimum maintenance time determination support system for the target equipment can be realized.
According to such a computer program, the progress of the deterioration event and the execution of the recovery time optimization of the deteriorated portion in the diagnosis, trend monitoring, life prediction, and optimum maintenance time determination of the target facility can be simplified.

In addition, it is possible to realize an optimum maintenance time determination support apparatus for a target facility characterized by having a computer including the computer program.
According to the apparatus for determining the optimum maintenance time of the target equipment, it is easy to perform the progress of the deterioration event and the optimization of the recovery time of the deteriorated part in the diagnosis, trend monitoring, life prediction, and optimum maintenance time determination of the target equipment. It will be a thing.

=== Overall Flow for Determining Optimal Maintenance Time for Target Equipment According to this Embodiment ===
FIG. 1 shows a flowchart of the steps of simple diagnosis, precise diagnosis, trend monitoring, life prediction, and optimum maintenance time determination, and FIG. 2 shows the procedure for comprehensive judgment based on the judgment results by the three judgment methods and the comprehensive judgment results. An example of classification of monitoring actions is shown.

The vibration data of the target equipment is collected, the waveform feature amount and the degradation-specific dimensionless feature amount are extracted from the vibration data, the determination result of the determination method 1 based on the ISO absolute determination criterion, and the high frequency (Hi) / medium frequency (Mid ) ・ Determination method 2 based on waveform feature quantity at low frequency (Lo), and judgment in high frequency (Hi), medium frequency (Mid), and low frequency (Lo) area based on degradation-specific dimensionless feature quantity Comprehensive determination with the determination result of method 3 is performed to determine the presence / absence of abnormality and a measure for future monitoring.
The determination result by the determination method 1 is a four-step determination of good / attention level 1 / attention level 2 / danger, and these are standardized by ISO 10816-1 according to the size of the drive motor and the basic state of the target equipment. .

Next, in the determination method 2, the peak value / average value / square average value of acceleration and the average value / square value of velocity at the high frequency (Hi) / medium frequency (Mid) / low frequency (Lo) in the determination method 1 are described. Dimensional feature quantities such as average values and peak values of displacement, and dimensionless feature quantities defined by the following [Equation 1] and [Equation 2], with each value at normal times as a reference, several times the reference value, Usually, 2 to 3 times is set as the caution threshold, and 2 to 3 times the caution threshold is set as the danger threshold.

  Next, in the determination method 3, the m-th principal component score is used as a determination index from the first principal component score obtained by Expression 4 disclosed in Japanese Patent No. 3382240. For the determination method using the principal component score as an index, the principal component score is obtained for normal vibration data for each frequency region of high frequency (Hi), medium frequency (Mid), and low frequency (Lo). This is specifically described in FIG. 6 and paragraphs [0037] to [0042] in Japanese Patent No. 3382240. In this determination method 3, the primary caution threshold, the secondary caution threshold, and the danger threshold are set, and the four stages of determination of good / caution level 1 / caution level 2 and dangerous area are performed.

According to the result of method 1, that is, when the determination result of method 1 is good or attention level 1, Table 1 in FIG. 2 is applied, and when the determination result of method 1 is attention level 2 or a dangerous area. Comprehensive judgment as shown in Table 2 and subsequent monitoring treatment classification are performed.
That is, in Table 2, since the determination result of method 1 shows that there is a high probability that the average vibration speed value in the low frequency region is large and the deterioration tendency is serious, the overall determination of method 2 and method 3 and the subsequent monitoring procedure Compared to the case of Table 1, it is natural to ensure the safe operation by grasping the progress of deterioration of the target equipment by selecting strict monitoring measures, and the accuracy of life prediction by extracting the deterioration characteristics. It improves.
In the above-described determination method, the frequency is divided into three frequency bands of high frequency (Hi), medium frequency (Mid), and low frequency (Lo), and the determination is performed in each band. Consider the fact that many initial phenomena of wear and scratches occur, that self-excited vibration is dominant in the middle frequency range, and that structural abnormalities such as unbalance mainly occur in the low frequency range. This improves the detection performance of the abnormality and is effective in identifying the cause of the abnormality.
Incidentally, in this example, diagnosis was performed in a frequency band of 5 kHz to 30 kHz, a medium frequency region of 1 kHz to 10 kHz, and a low frequency region of 5 Hz to 1 kHz.

=== About Diagnosis Method, Trend Monitoring, and Life Prediction of Target Equipment According to this Embodiment ===
A diagnosis method, trend monitoring, and life prediction for the target facility according to the present embodiment will be described.
In this specification, the target equipment is described by taking a rotating machine, for example, a pump driven by a motor, as an example, but the target equipment is naturally not limited to this. Absent.
Here, an embodiment in which the present invention is applied to the detection of the occurrence of a bearing flaw, its progress, and life prediction in the diagnosis of a pump flaw, a misalignment (shaft misalignment), and a deterioration event of unbalanced rotation will be shown.

FIG. 3 shows the passage of time of deterioration-specific dimensionless feature quantities Z1, Z5, and Z8 since the start of equipment diagnosis based on vibration data.
A bearing flaw is a characteristic amount Z1, a misalignment is a characteristic amount Z5, and an unbalanced rotation is a characteristic amount Z8 corresponding to each deterioration event. In FIG. 3, when the feature amount Z1 has passed 22.0 months, it exceeds the primary attention threshold of 8.0, and then reaches the secondary attention threshold of 20 after 24.7 months. ing. On the other hand, as shown in FIG. 4, during this period, the feature quantity Z5 exceeds the primary attention threshold after 24.0 months, and the feature quantity Z8 does not reach the primary attention threshold. That is, when the deterioration state of the bearing flaw is in the late stage of the attention area, it can be understood that the deterioration state of the misalignment is also in the attention area.

Deterioration of bearing damage has progressed in advance, FIG. 4 shows the deterioration specific feature amount Z1 of the bearing damage, and FIG. 5 shows the transition of the vibration speed average value Vrms. FIG. 6 shows the future vibration monitoring class classification according to the combination of the increase / decrease sign of the feature quantity Z1 and the increase / decrease sign of the vibration speed average value Vrms. Here, the case of dividing into the following four classes is shown.
AA: Extending the diagnosis cycle interval at the most recent prediction time A: Follow-up as it is B: Follow-up by shortening the diagnosis cycle C: Examining the relationship between the feature amount Z1 and the vibration speed average value Vrms and reaching the risk threshold Predict arrival time

  FIG. 7 shows the classification of the monitoring class, the prediction of Z1 and the vibration according to the classification of FIG. 6 for each diagnostic period of the feature quantity Z1 and the vibration speed average value Vrms shown in FIGS. 4 and 5 after 22.0 months. The transition of the index n in the power law of the ratio of the speed average value is shown. The feature quantity Z1 is predicted only in the case of class C, and the exponent n of the power law is corrected by comparing with the measured value of Z1.

FIG. 8 shows how the exponent n of the power law is corrected.
That is, until the secondary caution threshold is reached after the deterioration specific dimensionless feature value Z1 of the bearing flaw exceeds the primary caution threshold (T1 = 22.0 months to T2 = 24.7 months in FIG. 3). During the period, the progress characteristic of the degradation event is extracted as an exponent n of the power law by predicting and measuring the feature quantity Z1 according to the class classification of FIG. 6 and the power law of FIG. 8 and actually measuring the vibration velocity average value Vrms. To do. FIG. 7 shows the change of the exponent n of the power law.

指数ｎと予測係数Ｃとの関係を

It is an example that exceeded the second caution threshold of the bearing damage deterioration specific dimensionless feature quantity Z1 which is the deterioration event at 0.67, 0.17 and finally 0.34 starting from the initial value 0.8. .
The vibration speed average value at a certain time i is set to Vi, the deterioration-specific dimensionless feature quantity is set to Z1 (i), and prediction is performed from the predicted value and the actual measurement value of Z1 using the following [Formula 3] and [Formula 4]. The coefficient C is corrected.
As the prediction coefficient C of Z1

The relationship between the index n and the prediction coefficient C

Regarding the prediction of the deterioration-specific dimensionless feature value Z1 of the bearing flaw described above, the grounds for applying the power law of the vibration speed average value Vrms are that the fatigue phenomenon of the metal at the rotating part is proportional to Vrms, cracks and wear at the rotating part. There is a physical phenomenon in which Vrms squares, and the influence of the physical phenomenon on the progress of the degradation event does not reflect all vibration energy in the degradation progression. The problem is to solve the problem by correcting the index n.
Then, when the feature amount Z1 reaches the secondary attention threshold, the time when the feature amount Z1 reaches the danger threshold is predicted.

  By the way, the above-mentioned deterioration progression model of bearing flaws is basically applicable to deterioration events such as misalignment and unbalanced rotation. The reason is described below.

First, misalignment (axis misalignment) will be described. The factors behind the axis shift are:
(1) Rotation imbalance due to uneven wear on the driving side and driven side. (2) The external force and displacement applied from the connecting pipe and the basic part of the machine body.
Therefore, the increase in vibration energy indirectly causes the deterioration degree of the misalignment phenomenon through the increase in the degree of wear and unbalance, and thus has a deterioration prediction model and analogy in the damage progress of the bearing.

  Next, imbalance will be described. This deterioration factor is a mechanical / mechanical factor such as uneven wear or inclination of the rotating body or shaft, or a poor holding state at the bearing, so fatigue, defects, and wear progress due to increased vibration energy. It has a prediction model and analogy in the case of bearing damage development. In addition, when the direct cause of this imbalance is a chemical or electrochemical factor such as corrosion, the direct influence from the increase in vibration energy is small, but the increase in vibration energy, wear, eccentricity, etc. Since it enters a vicious circle through an increase in the degree of corrosion, the corrosion phenomenon also becomes an indirect factor, which is consistent with the life prediction model of the present invention in which vibration energy is considered.

Then, when the feature amount Z1 reaches the secondary attention threshold, the time when the feature amount Z1 reaches the danger threshold is predicted.

If such a trend monitoring step and a life prediction step are executed, it is possible to make a prediction in consideration of characteristics during the progress of bearing damage, and the accuracy of the life prediction becomes practical.
The relationship between Z1, Z5, and Z8, which are specific dimensionless characteristic quantities for bearing flaws, misalignment, and unbalanced rotation, and the degree of physical progress in the actual deterioration state is shown in FIGS. 11 shows. These are findings obtained from the results of vibration tests for each deterioration event in a test apparatus using a simulator for a rotating machine. In these figures, the primary attention threshold, the secondary attention threshold, and the danger threshold in each deterioration event are shown. Regarding these thresholds in actual machines, the threshold is applied because the fluid machine called a pump has generality in terms of the similarity in shape and the characteristics of the deterioration-specific dimensionless features. Is possible.

Now, the situation and results of applying the diagnosis method, trend monitoring, and life prediction method of the present invention to an actual pump with an on-site motor capacity of 800 kW will be described.
As shown in FIG. 3, the progress of each of the deterioration-specific dimensionless feature quantities Z1, Z5, and Z8 in the bearing damage, misalignment, unbalanced simple diagnosis, and trend monitoring, and the deterioration tendency monitoring result of the bearing damage, The result of life prediction that the risk threshold is reached after T3 = 31.8 months from the start was obtained. This is because when the feature value Z1 related to the bearing flaw reaches the first caution threshold value 8.0, the time T1 = 22.0 months after the feature value Z1 reaches the second caution threshold value T2 = 24.7. In the period up to a month, the actual value of the feature amount Z1 and the vibration velocity average value Vrms shown in FIG. 4 and FIG. 5 and the class classification of FIG. 6 and the prediction from the exponent change of the power law of FIG. 7 and FIG. Is performed, and when the second attention threshold value 20 of the feature amount Z1 corresponding to the bearing flaw is exceeded, the time when the risk threshold value 28 of the Z1 is reached is predicted.

  After executing the above life prediction, it was verified that the accuracy of the life prediction was sufficient by careful operation of the target pump. That is, when the target bearing portion was inspected after stopping and disassembling the pump after the predicted time T3 = 31.8 months, the depth of the outer ring flaw of the bearing was about 600 μ. It was revealed that the deterioration specific dimensionless feature amount Z1 shown was almost identical to the danger threshold 28, and the lifetime prediction method was effective.

= Support for determining the optimum maintenance time for the target equipment according to this embodiment =
First, a method for estimating the amount of energy loss caused by the progress of the deterioration event from vibration data will be described. FIG. 12 shows the misalignment amount and the loss ratio of the target pump motor power, and FIG. 13 shows the unbalance amount and the loss ratio of the motor power.

  When the deterioration event is misalignment, it is obtained from vibration data based on the relationship between the deterioration-specific dimensionless feature amount Z5 shown in FIG. 10 and the misalignment amount (axis shift amount) between the pump shaft and the motor shaft in the field. Further, the misalignment amount is estimated from the value of the feature amount Z5, and the energy loss amount due to the deterioration can be estimated from FIG.

Similarly, in the case of unbalance, the unbalance amount is estimated from the deterioration-specific dimensionless feature amount Z8 obtained from the vibration data using FIG. 11, and the energy loss amount due to the deterioration is estimated using FIG. It becomes possible.
The amount of energy loss in accordance with the progress of the bearing flaw is not considered because it is small compared to the deterioration events such as misalignment and imbalance described above, but the progress of the bearing flaw is less than that of other deterioration events. If it cannot be ignored, the energy loss ratio is obtained in advance by the same method as in the present invention.

  Next, the deterioration of the bearing damage progresses, and it becomes possible to predict the arrival time in the dangerous area by the trend monitoring step and the life prediction step. Regarding maintenance such as replacement and restoration of the bearing with a new part, An energy burden will occur. This energy burden can be quantitatively evaluated as the amount of carbon dioxide emissions required for manufacturing the bearing component.

Carbon dioxide emissions related to rolling bearing materials, production, and transportation are described in a report investigated and examined by the Japan Bearing Industry Association Global Environment Committee.
("Investigation and research on LCA (Life Cycle Assessment) of rolling bearings")

  In addition, regarding the carbon dioxide emissions related to the production of the target pumps, the annual CO2 emissions are developed / designed, machined and processed as the environmental impact of each manufacturing process at the production plant in the warming potential of large pumps. It is possible to refer to quantified values for assembly, testing, painting, logistics, etc. (For example, Ebara Manufacturing Co., Ltd., Kozuka, Eto “Case of LCA in Large Pump Production Activities” LCA Japan Forum News No. 43, pp. 2-5, March 31, 2007)

As a result of applying the diagnosis method, trend monitoring, and life prediction method of the present invention to the above-described actual pump with a motor capacity of 800 kW, as shown in FIG. It was 8 months.
Based on the results of the above deterioration monitoring and life prediction, recover misalignment, increase of power energy loss due to progress of deterioration of imbalance, energy burden due to replacement of bearing parts, and damage caused by misalignment to target pump internal parts. Depending on the energy burden, the increase in power energy loss due to the pump efficiency decreasing by 1.5% per year from the start of operation, and the number of days when the pump is stopped to restore the deteriorated part of the pump FIG. 14 shows the result of examining the elapsed months in order to determine the optimum maintenance time when considering the production loss.
From the results shown in FIG. 14, it can be seen that the total cost is minimized after 28 to 30 months from the start of diagnosis, and the replacement of the deteriorated bearing parts and the maintenance work for the alignment adjustment are executed during the period. It turned out to be optimal.

    In addition, the new replacement of the bearing parts described above is evaluated as the carbon dioxide emission amount related to the manufacture of the bearing, and the energy burden obtained by converting the emission amount into cost and the restoration of the internal parts of the pump corresponding to misalignment are supported. FIG. 15 shows an evaluation of the amount of carbon dioxide emitted and the energy burden obtained by converting the amount of emission into costs, and calculated as maintenance costs over time.

  Here, the deteriorated bearing part is model NU306, and the carbon dioxide emission amount for the production including the material of the part is 1, referring to the above-mentioned “Investigation / Research on Life Cycle Assessment (LCA) of Rolling Bearings”. 190kgCO2, carbon dioxide emissions related to the production of the target pump is 53,000kgCO2, referring to the above "Example LCA in large pump production activities", and the ratio of the misalignment to the internal damage of the pump is about 5.5% As (2,920 kgCO2), the total amount of discharge of the bearing parts is 4,100 kgCO2.

  The carbon dioxide emission was set at 170 yen / kg CO2 as a standard unit price for environmental load, and the energy burden of new bearing parts was estimated to be approximately 200,000 yen, and the energy burden related to misalignment adjustment was estimated to be approximately 500,000 yen.

  In the above-described embodiment, the amount of energy burdened by the replacement of the bearing part with a new one and the alignment adjustment work for correcting misalignment (axial misalignment) is evaluated as the carbon dioxide emission generated by the maintenance, and the emission amount Was converted into the amount of maintenance cost because the production loss due to the suspension of the target pump for the maintenance work was evaluated as the cost (amount). For the purpose of optimization of the maintenance period, carbon dioxide emissions It is also possible to search for the time when the amount of emissions is minimized by using the amount as a unified index.

=== Device for determining the optimum maintenance time of the target equipment ===
Next, the optimum maintenance time determination support device for the target facility for realizing the above-described simple diagnosis, precise diagnosis, deterioration tendency monitoring, life prediction, and optimum maintenance time determination method described above (hereinafter also referred to as the maintenance support device) An example will be described with reference to FIG. FIG. 16 is a conceptual diagram illustrating an example of a maintenance support device.

  In the present embodiment, the maintenance support device includes the maintenance support device main body 2 and the vibration sensor 1. The maintenance support apparatus main body 2 includes an A / D converter 100 that converts an analog electrical signal from a vibration sensor into a digital signal, the simple diagnosis, the precise diagnosis, the trend monitoring, the life prediction, the energy loss described in FIG. Main computer 200 for executing evaluation, recovery energy burden evaluation, calculation and determination of determination of optimum maintenance time, and treatment, degradation energy loss database 300, recovery energy burden database 400, and results of displaying results in the main computer 200 Display 500.

The main computer 200 has a computer program 200a for realizing the above-described optimal maintenance time determination method, and the CPU provided in the main computer 200 processes the computer program 200a so that the above-described optimal maintenance time is described. A decision method is executed. The computer program 200a is composed of codes for executing the above-described optimal maintenance time determination method. The main computer 200 may be composed of a plurality of devices instead of a single device. At this time, the computer program 200a may be provided separately for each of a plurality of devices.
The result display 500 has a function of giving various types of information to the operator of the maintenance support apparatus 2 that executes quality determination, deterioration state monitoring, life prediction, and optimum maintenance time determination of the target equipment.

=== Other Embodiments ===
As described above, the simple diagnosis, precise diagnosis, deterioration tendency monitoring, life prediction, and optimum maintenance time determination method of the target facility according to the present invention have been described based on the above-described embodiments. The present invention is not intended to limit the present invention. The present invention can be changed and improved without departing from the gist thereof, and the present invention includes the equivalents thereof.

In the above-described embodiment of the present invention, among the values of (Vi / Vi-1) ≧ 1.0 (deterioration does not progress when <1.0) in the attention area in the deterioration tendency monitoring, 1) Maximum value ・ ・ ・ ・ When the vibration energy to the deteriorated part shows a steep progress ▲ 2 ▼ Average value ・ ・ ・ ・ When the fluctuation of the progress is moderate and can be represented by averaging ▲ 3 ▼ Minimum value ... One of the embodiments of the present invention is to adopt any one of cases where the influence on the deterioration progress of vibration energy is relatively small and to perform life prediction when the secondary attention threshold is exceeded. It is.

  That is, in the above-described embodiment, the prediction is made based on the most recent increase or decrease in Z1 and V when the secondary attention threshold is exceeded. However, when the progress of the deterioration state in the attention region varies, the prediction is made only in the latest situation. As the accuracy decreases, the deterioration characteristics in the entire attention area exceeding the primary attention threshold and up to the secondary attention threshold are averaged, or the maximum and minimum values are used to represent the deterioration progress. Is another embodiment of the present invention.

1 is an overall flowchart of the present invention. It is the classification of the comprehensive judgment procedure and the monitoring treatment of the three judgment methods. It is the figure which showed the transition of the deterioration specific dimensionless feature-value corresponding to a bearing crack, misalignment, and imbalance, and a lifetime prediction result. It is a transition of the deterioration specific feature amount corresponding to the bearing flaw. It is transition of the actual measurement value of the vibration speed average value corresponding to FIG. It is a class classification diagram of deterioration progress by a combination of a deterioration specific feature and a vibration speed average value. It is a change figure of the exponent of the power law in deterioration tendency monitoring. It is the transition between the predicted value and the actual measurement value according to the power law in the progress of deterioration of bearing damage. It is a figure which shows the relationship between the magnitude | size of a bearing flaw, and the said deterioration specific dimensionless feature-value. It is a figure which shows the relationship between the amount of misalignment and the said deterioration specific dimensionless feature-value. It is a figure which shows the relationship between an imbalance amount and the said degradation specific dimensionless feature-value. It is a figure which shows the relationship between the amount of misalignment, and the ratio of the said power loss. It is a figure which shows the relationship between the amount of imbalance and the ratio of the said electric power loss. It is a figure which shows the Example of the optimal maintenance time determination support of this invention. It is an evaluation figure of the total cost for the optimal maintenance time determination of this invention. It is a block diagram of the optimal maintenance time determination assistance apparatus of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Vibration sensor 2 Optimal maintenance time determination support apparatus 100 A / D converter 200 Main computer 200a Computer program 300 Degradation energy loss database 400 Restoration energy burden database 500 Display

Claims (10)

Diagnose multiple deterioration states of the target equipment using vibration data, predict the progress and life of the deterioration state, recover the amount of energy required to recover the failure caused by the deterioration of the target equipment, and the Optimizing the facility to determine the monthly maintenance energy calculated from the restoration time and making a decision on the optimum maintenance time of the target facility considering the increase in operating energy loss of the target facility due to the mechanical degradation In the maintenance time decision support system,
Simple diagnosis step for diagnosing mechanical soundness at a certain point in time, precise diagnosis step for estimating the type and degree of deterioration of the mechanical deterioration, trend monitoring step for tracking the progress of the deterioration state, and trend monitoring step A life prediction step for predicting the arrival time of the failure from the deterioration pattern in the failure, an energy loss evaluation step for estimating the amount of energy loss of the target facility from the progress of the deterioration degree in the trend monitoring step, and the deteriorated component Alternatively, an energy load evaluation step necessary for restoring the deteriorated device, and an optimization of timing for restoring the deteriorated state by adding the energy loss evaluation result and the energy load evaluation result Maintenance time decision support system. In the simple diagnosis step of the target equipment according to claim 1,
A plurality of waveform feature values based on vibration time waveform data include at least one of vibration displacement, vibration speed, vibration acceleration, vibration acceleration distortion, vibration acceleration kurtosis, and vibration acceleration crest factor, and the vibration A diagnostic method characterized by dividing a spectrum of data and considering a plurality of deterioration-specific dimensionless feature quantities specific to a deterioration event obtained from a correlation in the time axis of strength between the divided frequency bands. In the simple diagnosis step according to claim 1,
ISO absolute determination result based on current vibration speed average value, and vibration displacement, vibration speed, vibration acceleration, vibration acceleration skewness, kurtosis of vibration acceleration, which are a plurality of waveform feature values in the previous period based on time waveform data of vibration, and Multiple degradations specific to degradation events, which are obtained by dividing the spectrum of the vibration data by the determination result based on at least one of the crest factors of vibration acceleration and the correlation on the time axis of the strength between the divided frequency bands A diagnostic method characterized by performing comprehensive determination in consideration of three types of determination results, i.e., determination results based on unique dimensionless feature values, and further determining a monitoring treatment type at the next time. In the trend monitoring step and the life prediction step according to claim 1,
The deterioration event among a plurality of dimensionless feature quantities obtained by dividing the vibration data spectrum by dividing the vibration data spectrum and calculating the correlation between the divided frequency bands on the time axis. A prediction method characterized by selecting a treatment type of a monitoring level at the next time in combination with an increase / decrease tendency of the deterioration-specific dimensionless feature amount corresponding to the above. In the trend monitoring step and the life prediction step according to claim 1,
A prediction method characterized by predicting the deterioration state at the next time based on the power law of the ratio between the average value of the vibration speed at the latest time and the average value of the vibration speed at the current time. In the trend monitoring step according to claim 4 and 5,
The degradation-specific dimensionless feature corresponding to the degradation event among a plurality of degradation-specific dimensionless feature quantities obtained by dividing the spectrum of vibration data and obtaining from the correlation in the time axis of the strength between the divided frequency bands For the feature amount, a secondary caution threshold is provided between the primary caution threshold corresponding to the start of occurrence of the deterioration event and the danger threshold for shifting from the caution area to the danger area, and the primary caution threshold is set. A prediction method, wherein the deterioration state is predicted between a threshold value and the secondary attention threshold value. In the energy loss evaluation step according to claim 1,
Based on the degradation-specific dimensionless feature quantity corresponding to the degradation event among a plurality of dimensionless feature quantities obtained by dividing the spectrum of vibration data and obtaining the correlation between the strengths of the divided frequency bands on the time axis. An energy loss evaluation method characterized by estimating an energy loss amount corresponding to a deteriorated state. In the energy loss evaluation step and the energy load evaluation step according to claim 1,
An optimum maintenance time determination support system characterized by evaluating an energy loss corresponding to the progress of degradation of the target equipment and a recovery energy of a failure caused by the degradation based on an environmental load due to carbon dioxide emission to the environment.   The computer program for implement | achieving the optimal maintenance time determination support system of the object installation of Claim 1.   An optimal maintenance time determination support apparatus for a target facility, comprising: a computer comprising the computer program according to claim 9.

Images