JP7153542B2 - Equipment maintenance support system - Google Patents

Description

The present invention relates to technology for equipment maintenance support systems.

Facilities and equipment in many infrastructure facilities such as bridges and roads, as well as water supply and sewage systems, are about to reach the end of their service lives. Many of these were developed and spread during the high economic growth period. Appropriate renewal is necessary to maintain the functions of these facilities and equipment. However, it is estimated that their maintenance will be costly. For this reason, the idea of adjusting and leveling the timing according to the situation, instead of concentrating the cost of updating in a specific period, is being considered in asset management. In asset management, it is necessary to diagnose the status of facilities and equipment, and determine whether or not the required performance can be maintained. One of the indicators is, for example, what is called soundness.

Water supply and sewage systems are a kind of plant, and have water collection facilities (intake of raw water and inflow of sewage through pipelines), facilities and equipment for water treatment and sludge treatment. A characteristic of this field is that there are a wide variety of facilities and equipment, and the environment and frequency of use vary from site to site. Therefore, even if the equipment is of the same type, the progression of deterioration and the degree of soundness differ depending on the site.

In addition, inspection items are not necessarily standardized except for inspections by laws and regulations. While some companies regularly perform minimum inspections and maintenance, others install additional sensors and use them for management. . In order to aim at the leveling of maintenance, maintenance, and renewal for such multiple sites, first, the soundness is calculated using the measurement items common to the equipment. A method of calculating the degree of health using common measurement items is disclosed, for example, in Patent Document 1 and the like.

In Patent Document 1, "The equipment maintenance and management work support system S is a monitoring cycle determined for each data from the water and sewage facility 200 or various equipment information, inspection / maintenance data, monitoring data, plant Data is taken into the computer system 100 and stored in the database 10. From the stored inspection/maintenance and monitoring data regarding the equipment, the quantitative soundness evaluation and the statistical service life by deterioration diagnosis prediction are calculated, and each equipment According to the risk level set in , leveling the equipment renewal demand cost that complies with budget constraints, updating it as appropriate and supporting demand planning. .

JP 2014-16691 A

In the technology described in Patent Document 1, soundness evaluation, deterioration diagnosis prediction, and risk assessment required for maintenance and management of a wide variety of equipment such as mechanical equipment, electrical equipment, various measuring instruments, and pipelines included in water and sewage facilities Evaluation becomes possible. As a result, in the technique described in Patent Document 1, cost optimization can be achieved by estimating the cost required for renewal based on the calculated renewal demand timing. However, further improvements are needed.

The present invention has been made in view of such a background, and an object of the present invention is to improve the accuracy of the soundness of equipment.

In order to solve the above-described problems, the present invention provides a health degree calculation unit that calculates a health degree, which is an index of the state of the device to be inspected, based on measurement information measured by the device to be inspected, a first time, And, a cluster analysis unit that cluster-analyzes the measurement information at a second time before the first time, the result of the cluster analysis at the first time, and the result of the cluster analysis at the second time according to the and a soundness correction unit that corrects the soundness.
Other solutions will be described as appropriate in the embodiments.

ADVANTAGE OF THE INVENTION According to this invention, the precision of the soundness in an apparatus can be improved.

It is a figure which shows the structural example of the water supply management system in 1st Embodiment. It is a figure which shows the structural example of the equipment maintenance assistance apparatus which concerns on 1st Embodiment. 5 is a flow chart showing the procedure of soundness correction processing performed in the present embodiment. It is a table showing soundness correction values and error values. 4 is a flowchart showing the procedure of soundness recovery processing; FIG. 11 is a diagram (part 1) showing the procedure of the soundness recovery process; FIG. 12 is a diagram (part 2) showing the procedure of the soundness recovery process; FIG. 13 is a diagram (part 3) showing the procedure of the soundness recovery process; FIG. 12 is a diagram (part 4) showing the procedure of the soundness recovery process; It is a figure which shows the example of the display screen displayed by this embodiment. It is a figure which shows the structural example of the water supply management system in 2nd Embodiment.

Next, modes for carrying out the present invention (referred to as "embodiments") will be described in detail with reference to the drawings as appropriate.

<First Embodiment>
FIG. 1 is a diagram showing a configuration example of a water supply management system Z in the first embodiment.
The target of management of the water supply management system Z is the water supply facility 3 . The water supply facility 3 has a water intake facility 310 , a water purification facility 320 , a water transmission and distribution facility 330 , and a water supply facility 340 . The water intake facility 310 is a facility that takes in water from a river or the like for drinking. The water purification facility 320 is a facility that purifies the water taken in so that it can be drunk. The water transmission and distribution facility 330 is a facility that distributes purified water to each district. The water supply facility 340 is a facility that supplies distributed water to the corresponding district.

The water treatment facility 320 has a receiving well 321 , a mixing basin 322 , a flocculation basin 323 , a sedimentation basin 324 , a filtration basin 325 and a clean water basin 326 . The receiving well 321 receives raw water taken from a river or the like. The mixing pool 322 is for adding a flocculant and rapidly stirring it. The floc forming pond 323 slowly agitates the water rapidly agitated in the mixing basin 322 to grow flocs. The sedimentation tank 324 sediments and separates the grown flocs. The filter basin 325 filters the water in which the flocs are sedimented and separated. The purified water reservoir 326 stores water that has been filtered and purified.

In this embodiment, the water transmission and distribution pump (not shown) included in the water transmission and distribution facility 330 is evaluated. The water transmission and distribution pump is one of the equipment to be evaluated in this embodiment. In the water transmission/distribution pump, measurements by a flow meter 331, a pressure gauge 332, a thermometer 333, a tachometer 334, and an ammeter 335 are performed at predetermined intervals. Data acquired by the measurement is stored in a storage unit (not shown) of the monitoring device 2 via a control LAN (Local Area Network) 4 . The data stored in the storage unit of the monitoring device 2 is used by the equipment maintenance support device (equipment maintenance support system) 1 .

<Equipment maintenance support device 1>
FIG. 2 is a diagram showing a configuration example of the equipment maintenance support device 1 according to the first embodiment. Please refer to FIG. 1 accordingly.
The device maintenance support apparatus 1 has a processing section (display processing section) 110 , a communication I/F (Interface) 121 , a display section 122 and an input section 123 . Furthermore, the equipment maintenance support apparatus 1 has an equipment information DB (Data Base) 131, a soundness DB 132, a cluster analysis result DB 133, a soundness correction DB 134, an inspection/maintenance information DB 135, a failure information DB 136, and a process DB 137.
The processing unit 110 has a soundness calculation unit 111 , a cluster analysis unit 112 and a soundness correction unit 113 .
The units 111 to 113 (110), 121 to 123 and the DBs 131 to 137 are connected to each other by a bus 141. FIG.

A program is stored in a storage unit (not shown) of the equipment maintenance support apparatus 1, and this program is loaded into a memory (not shown). When the loaded program is executed by a CPU (Central Processing Unit, not shown), the processing unit 110, and the health degree calculation unit 111, the cluster analysis unit 112, and the soundness correction unit 113 that constitute the processing unit 110 is embodied.

The communication I / F 121 is connected to the control LAN 4, and the operating conditions acquired by each equipment of the water supply facility 1 during inspection and maintenance work, and the measurement results (measurement data; measurement information) such as water quality are transmitted to the monitoring device 2. can be obtained via In this embodiment, the measurement data of the water transmission/distribution pump (not shown) installed in the water transmission/distribution facility 330 is acquired as described above.
These measurement data are stored in a storage unit (not shown) within the monitoring device 2 at a set frequency (predetermined cycle). The set frequency is the frequency of inspection/maintenance work.

The soundness calculation unit 111 calculates the soundness of the equipment based on the acquired measurement data, the data of the equipment information DB 131, the soundness DB 132, and the like. The method of calculating the degree of soundness is not particularly limited, but a method of ranking the values of each input item for category analysis of the predetermined equipment and determining the degree of soundness by combining them, or calculating from the Weibull hazard model. There are methods, etc.

The measurement data transmitted from the monitoring device 2 is stored in the equipment information DB 131 . The information stored in the equipment information DB 131 includes the “tag No.”, “input item name”, “measurement date and time”, “measurement result”, etc. of the measurement data. "Tag No." is an ID attached to the measurement data. For example, when temperature data and humidity data are obtained from the same sensor, different "tag numbers" are attached to the temperature data and humidity data, respectively. The input item is the name of the measurement data and the like.

Information stored in the equipment information DB 131 includes "equipment ID", "equipment name", "introduction date", "equipment specifications", and the like. "Equipment specifications" include the dimensions and power consumption of the water transmission and distribution pumps (not shown) installed in the water transmission and distribution facility 330, as well as the QH curve and the like for the water transmission and distribution pumps as information indicating performance. Each piece of information stored in the equipment information DB 131 is stored in advance by the user. The “equipment” here is equipment from which measurement data is acquired, and in this embodiment, it is a water transmission/distribution pump (not shown) installed in the water transmission/distribution facility 330 .

The information stored in the soundness DB 132 includes "equipment ID", "equipment name", "soundness calculation method", "input items" used for soundness calculation, and calculated and "soundness" and "error value". Error values will be described later.

The cluster analysis unit 112 performs cluster analysis on the measurement data. For cluster analysis, for example, Adaptive Resonance Theory (ART) is used. In ART, first, using learning data, a set of measurement data is classified for each region of a size set by the user. The learning data may be a set of only normal operating conditions, but it is desirable that the category indicated by the measurement data at the time of failure be associated with the type of failure and included as the learning data. In cluster analysis, input measurement data is classified into pre-learned categories. In the case of ART, since the size of the area of the category is determined in advance, a category newly found at the time of diagnosis in the category of the area that was not generated at the time of learning is generated as a new category. The cluster analysis unit 112 performs ART analysis on the acquired measurement data at a predetermined frequency, for example, the frequency at which a set of measurement data used for cluster analysis is newly stored in the monitoring device 2 . As a result of the ART analysis, the category to which the measurement data is assigned and the date and time of measurement are stored in the cluster analysis result DB 133 . In the ART example of this embodiment, "the state has changed" means moving from the previously obtained category to another category, or creating a new category and assigning it to the new category for which the measurement data was created. defined as being able to

The cluster analysis result DB 133 stores cluster analysis results of measurement data performed by the cluster analysis unit 112 . Here, cluster analysis is to determine to which category the acquired measurement data belongs.

As an evaluation method for cluster analysis, cluster analysis other than ART may be used. For example, a K-means method, a vector quantization method (VQC: Vector Quantization. Clustering), or the like may be used. In this case, since the category is determined during learning, the only state change is the transition between categories obtained during learning.

The soundness correction unit 113 uses the soundness correction DB 134, the result of the soundness calculation unit 111, and the result of the cluster analysis unit 112 to calculate a soundness correction value and an error value to correct the soundness.
A table described later with reference to FIG. 4 is stored in the soundness correction DB 134 .

The inspection/maintenance information DB 135 stores “inspection/maintenance details”, “target equipment”, “implementation date and time”, “implementer”, “inspection/maintenance results (measurement values, ○/× etc.)” and “maintenance work details” are stored. "Results of inspection/maintenance implementation" includes information such as measurement values measured when inspection/maintenance was performed, whether there were any problems (×), or no problems (○) as a result of inspection/maintenance. is stored. The "maintenance work content" stores greasing, replacement of consumables, overhaul, and the like. Moreover, the measured value is a value included in the measured data.
Information stored in the inspection/maintenance information DB 135 is information manually input by the user via the input unit 123 .

Further, the failure information DB 136 stores failure information of equipment. Specifically, "facility equipment ID", "problem category", "problem name", "problem occurrence date and time", "response details", "time required for response", "interim measures after response", " "severity of failure" and the like are stored in the failure information DB 136. FIG. Adjustment/replacement/switching of operation/communication with a trader, etc. are stored in the "response contents".

These pieces of information are metadata attached to data acquired from equipment, processing results by the cluster analysis unit 112, and information input manually by the user. The “failure category” and “failure name” stored in the failure information DB 136 are used during learning (during cluster analysis) by the cluster analysis unit 112 and during evaluation of state changes.

The process DB 137 stores the soundness level, cluster analysis results, input items used for soundness correction, and the like as a backup.
The display unit 122 presents the soundness level, the cluster analysis result, the soundness level correction result, and the like to the user, and is a display or the like. The input unit 123 is a keyboard, mouse, or the like. From the input unit 123, inspection results, information about failures, constants used in calculations such as soundness correction, and information used for editing input items used for soundness/cluster analysis results are input.

<Soundness Correction Processing>
FIG. 3 is a flow chart showing the procedure of soundness correction processing performed in this embodiment. Please refer to FIG. 2 accordingly.
First, the soundness correction unit 113 selects equipment to be managed (S101).
Then, the soundness correction unit 113 acquires the latest (current) soundness regarding the equipment from the soundness DB 132 (S102). It should be noted that the current health level has already been calculated by the health level calculation unit 111 . In addition, the soundness here is the soundness corrected by the method to be described later.

Next, from the cluster analysis result DB 133, the soundness correction unit 113 obtains the previous (time t−1; second time) cluster analysis result and the current (time t; first time) cluster analysis result for the equipment. Obtained as R_i(t-1) and R_i(t) (S103). That is, the soundness correction unit 113 acquires the category to which the previous (time t−1) and current (time t) measurement data belong.
In this embodiment, normal data sets and abnormal data sets are classified in the learning stage of cluster analysis performed in advance. These data sets are pre-divided into normal and abnormal categories, respectively.
It is also assumed that cluster analysis has already been performed using the latest measurement data. That is, it is known to which category the latest measurement data belongs.

Subsequently, the soundness correction unit 113 performs case classification based on the previous cluster analysis result (R_i(t-1)) (S111). That is, the soundness correction unit 113 determines to which category (for example, normal/abnormal/new) the target measurement data belongs to by the previous cluster analysis.

Next, the soundness correction unit 113 determines which category (normal/abnormal/new) the current cluster analysis result (R_i(t)) belongs to, and determines the transition from the previous category (S112 ). That is, the soundness correction unit 113 determines to which category (for example, normal/abnormal/new) the target measurement data belongs to by the previous cluster analysis.

Then, the soundness correction unit 113 acquires the soundness correction value (a_i(t)) and the error value (e_i(t)) from the soundness correction DB 134 according to the transition pattern (S113). Examples of soundness correction values and error values are shown in FIG. 4 and will be described later. The soundness correction value corrects the soundness according to the state change of the category analysis result. The error value also indicates the error range of the soundness (the range that the soundness can take) according to the state change of the category analysis result.
Then, the soundness correction unit 113 corrects the soundness using the soundness correction value (S114). Correction of soundness will be described later.

Next, the soundness level correction unit 113 acquires input items (soundness level input items) used for soundness level calculation and input items (cluster analysis input items) used for cluster analysis (S121). Then, the soundness correction unit 113 calculates the number of different items (Δitem_i) between the soundness input item and the cluster analysis input item (S122).

If the input items used in the soundness calculation and the category analysis are significantly different, it is conceivable that the targets evaluated by each (defective part, mechanical/electrical defect type, etc.) do not match. For this reason, when the input items used for the calculation of the soundness level and the category analysis are different, it is desirable to correct the error value e_i(t). Note that the soundness correction value a_i(t) does not have to be corrected.
Therefore, the soundness correction unit 113 corrects the error value (e_i(t)) using Δitem_i (S123). Specifically, for example, the soundness correction unit 113 calculates an error value (ec_i(t)) corrected by the following equation (1).

ec_i(t)=e_i(t)/(1+|Δitem_i|) (1)

Then, the soundness correction unit 113 outputs the corrected soundness to the soundness DB 132, and converts the error value (e_i(t)) or the corrected error value (ec_i(t)) into a new correction value ((e_i (t))) is stored in the soundness DB 132 (S124).
Then, the processing unit 110 displays the corrected health degree and the error value on the display unit 122 (S125).
Note that steps S121 to S123 can be omitted. Further, by performing the processing of steps S121 to S123, it is possible to improve the accuracy of the error value.

<Soundness correction value and error value>
FIG. 4 is a table showing soundness correction values (a_i(t)) and error values (e_i(t)).
As shown in FIG. 4, the soundness correction value and the error value are determined by comparing the previous (t-1) category and the current (t) category. Note that the table shown in FIG. 4 is set in advance by the user and stored in the soundness correction DB 134 .

(1) If the previous category is abnormal, the soundness correction value is set to take a value in the range of -1 to 0.

(1-1) If the current category is abnormal and is the same as the previous category, the soundness correction unit 113 determines the soundness correction value so that the soundness is −0.5. Further, since the abnormality continues, the soundness level may be even lower, so the soundness correction unit 113 sets the error value to -0.5.

(1-2) If the current category is abnormal and different from the previous category, the equipment is determined to be in an unstable state, and the soundness correction value is determined so that the soundness is -1. In addition, the error value is determined to be -1 in consideration of the possibility that the soundness level is even lower. This is because there is a possibility that a new anomaly has occurred since the anomaly continues and the category has changed since the previous time.

(1-3) If the current category is normal, it changes from abnormal to normal, so there is no need to correct the soundness level. It is determined. In addition, the error value is determined to be +0.5, taking into account the possibility that the soundness level is actually further improved due to the change from abnormal to normal.

(1-4) If the current category is new, a soundness correction value (“0”) is determined so that the soundness does not change. Also, the error value is determined to be ±0.25. This is because a new category cannot be judged as normal or abnormal.

(2) If the previous category was normal, the soundness correction value is set to take a value in the range of -1 to 0.

(2-1) If the current category is abnormal, it is determined that this is a sign that the soundness level will decrease, and the soundness correction value is determined so that the soundness level becomes -1. In addition, since there is a change from normal to abnormal, the error value is determined to be -1 in consideration of the possibility that the soundness level is even lower.

(2-2) If the current category is normal and the same as the previous category, the soundness correction value (“0”) is determined so that the soundness does not change. In addition, since the category does not change in addition to the change from normal to normal, the need for soundness correction is low, so the error value is determined to be 0.

(2-3) If the current category is normal and different from the previous category, the change from normal to normal means that the soundness level correction value (“0”) is determined so that the soundness level does not change. However, even if it is normal, the category has changed, so the soundness level may have improved or decreased, so the error value is determined to be ±0.25. Note that both the previous and this time are in the normal category, but they are in different categories. This is the case where, as a result of the analysis, the above-mentioned abnormality is found.

(2-4) If the current category is new, as described above, the new category cannot be judged normal or abnormal, so the soundness correction value ("0") is determined so that the soundness does not change. be done. Also, since the category has changed between the previous time and this time, there is a possibility that the soundness level has decreased, so the error value is determined to be -0.25.

(3) If the previous category was a new category, the new category cannot be determined to be normal or abnormal.

(3-1) If the current category is Abnormal, the transition to Abnormal is a clear drop in soundness. However, since it is not clear whether the previous category is normal or abnormal, the soundness correction value is determined so that the soundness is −0.5. Also, since the normality/abnormality of the previous category is not clear, the error value is determined to be -0.5 considering the possibility that the soundness level is actually low.

(3-2) If the current category is normal, the soundness level correction value (“0”) is determined so that the soundness level does not change. In addition, the error value is determined to be +0.25, taking into account the possibility that the degree of soundness is actually improving.

(3-3) If the current category is new, it is unclear whether it has deteriorated or improved, so the soundness correction value (“0”) is determined so that the soundness does not change. Also, since the soundness may have improved or decreased, the error value is determined to be ±0.25.

To sum up the error value, the soundness correction unit 113 sets only a negative error value when the previous category analysis result changes to abnormal. In addition, the soundness correction unit 113 basically sets a positive error value when the transition is normal from the previous category analysis result, but depending on the transition state, the error is plus or minus. Both positive and negative errors are considered for new categories. By calculating such an error value, it is possible to improve the evaluation accuracy of the degree of health.

Further, in step S114 of FIG. 3, the soundness correction unit 113 calculates the soundness corrected by the following equation (2). In equation (2), S_i(t) indicates the soundness level of equipment i, and Sc_i(t) indicates the corrected soundness level of equipment i.

Sc_i(t) = S_i(t) + a_i(t) (2)

Also, when plotting the degree of soundness on a trend graph or the like, the processing unit 110 displays the range of the error value (e_i(t)) together with this value. This display will be described later.

3 and 4, the calculation of the soundness level and the method of correcting the soundness level each time cluster analysis is performed have been described. However, in the actual operation of the plant and the maintenance work associated with it, not only does the soundness drop monotonously, but also measures are taken to restore the soundness through maintenance activities and the like.
Here, processing for restoring the soundness level by maintenance, overhaul, etc. will be described with reference to FIGS. 5 and 6A to 6D. Let T be the time when maintenance or overhaul is performed.

Incidentally, the table shown in FIG. 4 is an example, and each value does not have to be the value shown in FIG.

FIG. 5 is a flow chart showing the procedure of the soundness recovery process. 6A to 6D are diagrams showing the procedure of the soundness recovery process. Please refer to FIG. 2 accordingly.
First, the soundness calculation unit 111 selects equipment to be restored for soundness (S201). Here, the equipment to be selected is the equipment to be maintained or overhauled.
Next, the soundness calculator 111 calculates the past soundness S0(t) using parameters other than the operating time (S202). In other words, the soundness is calculated using a plurality of parameters including the operating time, but in step S202, the past soundness S0(t) is calculated using a formula that does not include the operating time. In addition, the operating time is a factor that contributes to the deterioration of the soundness, using the service life as one criterion. In FIG. 6A, what is indicated by symbol H1 is the past soundness level S0(t) calculated in step S202. Note that time T in FIGS. 6A to 6D indicates the time when the maintenance work was performed. Here, the calculated soundness level may or may not be corrected by the methods of FIGS. 3 and 4 .

Next, after the maintenance work, the soundness calculator 111 calculates the soundness S2(t0) using parameters other than the operating time after the maintenance work (S211). Here, t0 is the time when the measurement data was acquired after the maintenance work. What is indicated by symbol H2 in FIG. 6B is the health degree S2(t0) calculated in step S211. Incidentally, maintenance work includes maintenance, maintenance, and overhaul.
Then, the health degree calculation unit 111 plots the past health degree S0(t) on a graph in which the horizontal axis is time and the vertical axis is the health degree. Then, the health degree calculation unit 111 determines whether or not the plotted past health degree S0(t) can be fitted with a function f1 that monotonically decreases with time (S212). That is, the health degree calculation unit 111 determines whether the fitting by the monotonically decreasing function f1 reproduces past data of the soundness degree S0(t) reasonably and whether the fitting error is sufficiently small. In the diagram shown in FIG. 6B, since the past soundness level S0(t) (symbol H1) is fitted to the monotonically decreasing function f1, "Yes" is determined in step S212.

As a result of step S212, if the past soundness level S0(t) can be fitted with a function f1 that monotonously decreases with time (S212→Yes), the soundness level calculation unit 111 advances the process to step S221.
In step S221, the soundness calculator 111 calculates the operating time t1 corresponding to the state of soundness after maintenance or overhaul from the values of the function f1 and S2(t0). Specifically, as shown in FIG. 6C, in the function f1, the soundness calculator 111 calculates the movable time t1 (third time) having the same value as S2(t0) (symbol H2). Here, the operating time t1 indicates that the time t1 has elapsed since the previous maintenance work was performed.
Subsequently, the health level calculation unit 111 recalculates the post-maintenance health level S1(t1) using the operating time t1 calculated in step S221 and the parameters acquired after the maintenance work (S222). . The soundness S1(t1) calculated in step S222 is indicated by symbol H3 in FIG. 6D.
The soundness calculation unit 111 uses the calculated soundness S1(t1) as a new soundness after maintenance and overhaul. That is, the health level calculation unit 111 sets the calculated health level S1(t1) as the health level after the maintenance work (at time t0) (S223).

On the other hand, as a result of step S212, if the past soundness level S0(t) cannot be fitted with the function f1 that monotonically decreases with time (S212→No), the soundness level calculation unit 111 determines the preset value The soundness level is increased by only (S231). Here, the soundness calculation unit 111 sets the soundness to 20% (1 in 5 stages) when maintenance work is performed, and 40% (5 stages in 5 stages) when overhaul is performed. In the case of 2) Raise. However, the maximum value of soundness shall be the upper limit.
After steps S223 and S231, the soundness calculator 111 stores the calculated soundness in the soundness DB 132 (S241).
Then, the processing unit 110 displays the recalculated soundness level on the display unit 122 (S242).

The post-maintenance state of health calculated by the process of FIG. 5 is basically not corrected based on FIGS. 3 and 4, but may be corrected.
By performing the processing in FIG. 5, the soundness after the maintenance work is completed can be calculated.

FIG. 7 is a diagram showing an example of a display screen 500 displayed in this embodiment.
The display screen 500 has an equipment setting area 510 , a display period setting area 520 , a measurement data display area 530 , a cluster analysis result display area 540 , a soundness level display area 550 and a maintenance history display area 560 .
In the equipment setting area 510, the equipment displayed on the display screen 500 is selected. As shown in FIG. 7, equipment is specified in the order of "major classification", "minor classification", and "equipment name".
In the display period setting area 520, the period displayed on the display screen 500 is set.

In the measurement data display area 530, changes over time of the measurement data stored in the equipment information DB 131 are displayed. In the measured data display area 530, the cluster analysis result display area 540, the soundness display area 550, and the maintenance history display area 560, the horizontal axis indicates time as shown in FIG. When the user selects and inputs the data update button 531 via the input unit 123, the latest measurement data is acquired and displayed. In the measurement data display area 530, different types of measurement data are displayed. In the example of FIG. 7, the types of measurement data are indicated by white squares, filled diamonds, and filled squares.

A cluster analysis result display area 540 displays the results of the cluster analysis. The example of FIG. 7 shows the result of cluster analysis of the measurement data displayed in the measurement data display area 530 . In addition, in the example of FIG. 7, the upper side of the threshold value L1 indicates the abnormal state, and the lower side indicates the normal state. The value of the threshold L1 is a value set as a result of learning. Abnormalities are indicated by solid triangles, and normality is indicated by solid circles. Furthermore, when the user selects and inputs a recalculation button 541 via the input unit 123 , cluster analysis is re-executed based on the latest measurement data, and the results are displayed in the cluster analysis result display area 540 .

Note that the example of FIG. 7 shows an example in which the cluster analysis result is divided into two, abnormal/normal. However, as described above, a plurality of clusters are generated even in abnormal and normal cases, and new clusters are generated. Therefore, the actual example is more complicated than the example shown in FIG.

The soundness display area 550 displays the soundness corrected by the process of FIG. 3 and the soundness recalculated by the process of FIG. Also, what is indicated by reference numeral 532 is indicated by an error value. When the user selects and inputs a recalculation button 551 via the input unit 123, the health level is calculated based on the latest cluster analysis, and the health level after maintenance work is recalculated and displayed.

The maintenance history display area 560 shows the timing at which maintenance work was performed. As described above, maintenance work here includes maintenance, maintenance, and overhaul. In the example of FIG. 7, line L2 indicates the time when the maintenance work was performed. In addition, in the example of FIG. 7, the type of maintenance work is indicated by white asterisks and solid asterisks. Filled stars indicate maintenance and maintenance, and white stars indicate overhaul. Further, when the user selects and inputs a maintenance content detail button 561 via the input unit 123, details of each maintenance work are displayed. For example, the detailed content of the corresponding maintenance work is displayed on a separate screen for the white asterisks and solid asterisks shown in the example of FIG. 7 .

By displaying the display screen 500 as shown in the example of FIG. 7, it becomes easier for the user to understand the interrelationship between the pieces of information.

According to this embodiment, the soundness level is corrected by reflecting the evaluation results obtained by the past cluster analysis method. By doing so, the accuracy of the soundness can be improved. Accordingly, the user can accurately grasp the service life of facilities and equipment. In other words, according to the present embodiment, asset management, which plans maintenance work and update timing of equipment, can be performed appropriately.
In addition, this makes it possible to improve the accuracy of priority setting in formulating an update plan for facilities and equipment.

<Second embodiment>
A second embodiment of the present invention will now be described with reference to FIG.
FIG. 8 is a diagram showing a configuration example of the water supply management system Za in the second embodiment.
In FIG. 8, the same components as in FIG. 1 are denoted by the same reference numerals, and descriptions thereof are omitted.
The water supply management system Za shown in FIG. 8 differs from the water supply management system Z shown in FIG. A vibrometer 336 is provided to determine the condition of the bearings of a rotating machine such as a pump.

In the second embodiment, the measurement data of the vibrometer 336 is subjected to ART analysis (cluster analysis) after preprocessing.
In application to ART analysis, first, as preprocessing, FFT (Fast Fourier Transform) is performed on waveform data obtained from the vibrometer 336, and frequency analysis is performed. The selection of the frequency range to be input in ART depends on the type of pump and bearing, and the installation state of the main body itself, but at least two wavelengths should be selected: the wavelength that increases in intensity when an abnormality occurs, and the wavelength that is less affected by it. be done. That is, the range of measurement data is divided. The ratio or difference of these intensities is taken and used as input data for ART analysis. By adding local vibration data as an input in this manner, the accuracy of diagnosis is improved. As a result, the accuracy of the soundness correction is improved.

In addition, the method of dividing the range of measurement data and performing the soundness level and the ART analysis respectively is applied to a pump or the like controlled at a variable speed. In a pump whose rotation speed is controlled, current, rotation speed, discharge pressure, etc. change with time. In other words, the vibration meter 336 can detect defects specific to equipment (here, pumps, etc.).

Predetermined measurement data needs to be acquired as data used for the degree of soundness and ART analysis, as described above. However, if there is some relationship between the measurement data and the occurrence of a problem, for example, if the QH curve deviates from the optimum operating conditions, the load applied to the machine or electronic equipment will change. In such a case, variations in soundness and ART analysis results increase. Therefore, the measurement data is divided into two or more regions, and the median value (the most frequent measurement data) of the data set included in each region is used to calculate ART analysis (category analysis) and health degree. is done. The two or more regions are, as described above, a wavelength at which the intensity increases when an abnormality occurs and a wavelength at which the intensity is less susceptible to it. By doing so, it is possible to calculate the optimum soundness level and perform ART analysis.

According to the second embodiment, in addition to the effects of the first embodiment, for example, ART analysis and calculation of the degree of soundness can be performed for a wavelength at which the intensity increases when an abnormality occurs and a wavelength that is less affected by it. By doing so, it becomes easier to grasp the progress of deterioration. As a result, it is possible to construct an operation method that frequently uses operating conditions that can suppress deterioration.

Note that in the second embodiment, instead of or in addition to the vibrometer 336, an acoustic measuring device, a camera, or the like may be provided.

When acquiring new measurement data, the device maintenance support apparatus 1 may perform fitting using the past soundness level each time the soundness level is calculated.

The equipment maintenance support device 1 in this embodiment can also be applied to other plants such as chemical plants.

The present invention is not limited to the above-described embodiments, and includes various modifications. For example, the above-described embodiments have been described in detail in order to explain the present invention in an easy-to-understand manner, and are not necessarily limited to those having all the configurations described. Also, part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment. Moreover, it is possible to add, delete, or replace part of the configuration of each embodiment with another configuration.

Also, the above configurations, functions, units 110 to 113, DBs 131 to 137, etc., may be realized by hardware, for example, by designing a part or all of them using an integrated circuit. Further, each configuration, function, etc. described above may be realized by software by a processor such as a CPU interpreting and executing a program for realizing each function. Information such as programs, tables, files, etc. that realize each function is stored in a memory, a recording device such as an SSD (Solid State Drive), or an IC (Integrated Circuit) card, in addition to being stored in an HD (Hard Disk) , an SD (Secure Digital) card, a DVD (Digital Versatile Disc), or other recording media.
Further, in each embodiment, control lines and information lines are those considered necessary for explanation, and not all control lines and information lines are necessarily shown on the product. In fact, it can be considered that almost all configurations are interconnected.

1 Equipment maintenance support device (equipment maintenance support system)
110 processing unit (display processing unit)
111 soundness calculation unit 112 cluster analysis unit 113 soundness correction unit 132 soundness DB
133 Cluster analysis result DB
134 Soundness Correction DB
Z, Za Water supply management system

Claims (5)

a health degree calculation unit that calculates a health degree, which is an index of the state of the device to be inspected, based on measurement information measured by the device to be inspected;
a cluster analysis unit that cluster-analyzes the measurement information at a first time and at a second time prior to the first time;
a soundness correction unit that corrects the soundness according to the result of the cluster analysis at the first time and the result of the cluster analysis at the second time;
A device maintenance support system characterized by having: The health degree calculation unit
calculating a plurality of first soundness degrees, which are soundness degrees using the measurement information acquired before the maintenance work of the device to be inspected but not using the operating time parameter;
Plot the first degree of soundness at each time on coordinates with time as the horizontal axis and the soundness degree as the vertical axis, and set a curve that fits the plotted first degree of soundness,
calculating a second health degree that is a health degree using the measurement information acquired after the maintenance work but not using the operating time parameter;
In the coordinates, plotting the second degree of soundness at the time when the second degree of soundness was calculated,
In the curve, calculate a third time that is the time when the soundness is the same as the second soundness,
Using the third time as a parameter of the operating time, and using the parameter used in calculating the second soundness and the third time, calculating a third soundness,
The equipment maintenance support system according to claim 1, wherein the third soundness level is set as the soundness level after the maintenance work. 2. The method according to claim 1, further comprising: an error calculation unit that calculates an error, which is a possible value of the soundness, according to the results of the cluster analysis at the first time and the cluster analysis at the second time. equipment maintenance support system. 4. The equipment maintenance according to claim 3, further comprising an error correction unit that corrects the error based on a difference between the number of parameters used for the cluster analysis and the number of parameters used for calculating the degree of soundness. support system. 2. The equipment maintenance support system according to claim 1, further comprising a display processing unit that displays the corrected soundness level on a display unit.

Images