Flow rate estimation system

Abstract

To improve the accuracy of estimating drain flow rate in a drain trap.SOLUTION: A flow rate estimation system 100 comprises: an accumulation section 21 that accumulates first site data obtained by associating vibration information on a steam trap 2 measured by a diagnosis personnel, drain discharging ability of the steam trap 2, and drain flow rate of the steam trap 2, which is determined by the diagnosis personnel from an operation state of the steam trap 2 when the vibration information is measured, in a site S of the steam trap 2; and an estimation device 10 having an estimation section 16 that estimates the drain flow rate of the steam trap 2 to be estimated on the basis of a probe 18 for detecting the vibration information on the steam trap 2 to be estimated, the first site data of the accumulation section 21, the vibration information detected by the probe 18, and the drain discharging ability of the steam trap 2 to be estimated.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The technology of the present disclosure relates to a flow rate estimation system.

Background Art

[0002] Conventionally, devices for estimating the flow rate of a fluid passing through a valve such as a drain trap have been known. For example, the estimation device disclosed in Patent Document 1 detects the vibration of a valve and estimates the actual flow rate from the correlation between the flow rate and the vibration stored in advance. Further, the measuring device disclosed in Patent Document 2 determines the operating condition of the valve from the relationship between the vibration of a valve such as a drain trap and the amount of steam leakage.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] By the way, in the estimation device such as Patent Document 1 described above, since the correlation between the flow rate and the vibration is mainly generated based on experimental data, the amount of data is small, and particularly, the data corresponding to various environmental conditions is small. Therefore, the estimation accuracy of the flow rate is not sufficient, and there is a desire to improve the estimation accuracy. Further, in the measuring device such as Patent Document 2, when detecting the vibration generated when the drain flow rate is large, there is a risk of misjudging that the state is a malfunction (that is, steam is leaking) even though the drain is being discharged normally. In such a case, if the flow rate of the drain can be grasped, misjudgment regarding the operating condition can be suppressed. From this as well, high estimation accuracy of the flow rate is desired.

[0005] The technology disclosed herein has been made in view of such circumstances, and its purpose is to provide a flow rate estimation system that can improve the accuracy of estimating the drain flow rate in a drain trap. [Means for solving the problem]

[0006] The technology disclosed herein is a flow rate estimation system for estimating the drain flow rate of a drain trap to be estimated. The flow rate estimation system comprises a storage unit and an estimation device. The storage unit associates vibration information of the drain trap measured by a diagnostician at the site of the drain trap, the drain discharge capacity of the drain trap, and the drain flow rate of the drain trap determined by the diagnostician from the operating state of the drain trap when the vibration information was measured. Field data is accumulated. The estimation device includes a detection unit that detects vibration information of the drain trap to be estimated, and an estimation unit that estimates the drain flow rate of the drain trap to be estimated based on the first field data from the accumulation unit, the vibration information detected by the detection unit, and the drain discharge capacity of the drain trap to be estimated. [Effects of the Invention]

[0007] According to the flow rate estimation system described above, the accuracy of estimating the drain flow rate in a drain trap can be improved. [Brief explanation of the drawing]

[0008] [Figure 1] Figure 1 is a block diagram showing the schematic configuration of the flow rate estimation system. [Figure 2] Figure 2 shows an example of the first and second field data in the storage unit. [Figure 3] Figure 3 is a conceptual diagram showing the input and output of the first estimation model. [Figure 4] Figure 4 is a conceptual diagram showing the input and output of the second estimation model. [Figure 5] Figure 5 is a flowchart showing the operation of the estimation device for drain flow rate. [Figure 6] Figure 6 shows an example of a display mode in the display unit. [Modes for carrying out the invention]

[0009] The following exemplary embodiments will be described in detail with reference to the drawings.

[0010] The flow rate estimation system 100 of this embodiment estimates the drain flow rate in a steam trap 2, for example, which is installed in a steam system. The steam trap 2 is an example of a drain trap, and is installed in a drain pipe 1, for example. The steam trap 2 is a so-called automatic valve that allows drain to flow downstream when drain flows in from the drain pipe 1, while preventing steam from flowing out when steam flows in from the drain pipe 1.

[0011] As shown in Figure 1, the flow rate estimation system 100 comprises an estimation device 10 and a server device 20. The estimation device 10 and the server device 20 can communicate with each other via a network N. Network N is a wide-area communication network such as the Internet.

[0012] The estimation device 10 is a portable device for estimating the drain flow rate in the steam trap 2, that is, the drain flow rate passing through the steam trap 2. Specifically, the estimation device 10 has a device body 11 and a probe 18.

[0013] The probe 18 is an example of a detection unit that detects vibration information (e.g., vibration level) of the steam trap 2 to be estimated. The probe 18 detects vibration information of the steam trap 2 by, for example, pressing it against the casing of the steam trap 2. The probe 18 is connected to the main body of the device 11 via cable 18a. The vibration information detected by the probe 18 is output to the main body of the device 11 via cable 18a. The steam trap 2 to be estimated is the steam trap 2 whose drain flow rate is estimated.

[0014] Note that the apparatus main body 11 and the probe 18 may be integrally formed. Also, the apparatus main body 11 and the probe 18 may be wirelessly connected by a wireless communication standard such as Bluetooth (registered trademark).

[0015] The apparatus main body 11 estimates the drain flow rate in the steam trap 2 based on the vibration information detected by the probe 18. The apparatus main body 11 can communicate with the server apparatus 20 via the network N. Specifically, the apparatus main body 11 includes an input unit 12, a storage unit 13, a display unit 14, a derivation unit 15, an estimation unit 16, and a correction unit 17.

[0016] The input unit 12 receives an input operation from a diagnostician who is a user. The input unit 12 outputs an input signal corresponding to the input operation. The input unit 12 is, for example, an input key or a touch panel superimposed on the display unit 14 described later.

[0017] The storage unit 13 is a computer-readable storage medium that stores various programs and various data. The storage unit 13 is formed of a magnetic disk such as a hard disk, an optical disk such as a CD-ROM and a DVD, or a semiconductor memory. The storage unit 13 stores a first estimation model M1 and a second estimation model M2 generated by the learning unit 22 of the server apparatus 20 described later. Also, the storage unit 13 stores the drain flow rate and the like output by the estimation unit 16 described later and the drain flow rate corrected by the correction unit 17.

[0018] In addition, the storage unit 13 stores the capacity characteristics D of the steam trap 2 for each type of the steam trap 2. The capacity characteristics D are those in which the drain discharge capacity of the steam trap 2 is determined according to the pressure difference between the inlet pressure and the outlet pressure of the steam trap 2. The drain discharge capacity is the maximum drain flow rate that can be discharged. The capacity characteristics D are, for example, a discharge capacity diagram in which the drain discharge capacity changes curvilinearly according to the pressure difference. Generally, the drain discharge capacity is larger as the pressure difference of the steam trap 2 is larger.

[0019] The display unit 14 displays the drain flow rate output by the estimation unit 16 and the confidence level of the drain flow rate. The confidence level of the drain flow rate is an example of information indicating the correct rates of the first estimation model M1 and the second estimation model M2 stored in the storage unit 13. The display unit 14 is, for example, a liquid crystal display or an organic EL display.

[0020] The derivation unit 15 derives the drain discharge capacity of the steam trap 2 to be estimated based on the performance characteristics D stored in the storage unit 13, the type of the steam trap 2 to be estimated, and the pressure difference. The pressure difference of the steam trap 2 to be estimated is, for example, as shown in FIG. 1, the inlet pressure measured by the first pressure gauge 3 provided in the drain pipe 1 on the inlet side of the steam trap 2 to be estimated, and the outlet pressure measured by the second pressure gauge 4 provided in the drain pipe 1 on the outlet side of the steam trap 2, and is the difference therebetween.

[0021] The estimation unit 16 estimates the drain flow rate from the vibration information detected by the probe 18 and the drain discharge capacity of the steam trap 2 to be estimated based on the first on-site data of the accumulation unit 21 described later. Specifically, the estimation unit 16 inputs the vibration information detected by the probe 18 and the drain discharge capacity of the steam trap 2 to be estimated into the first estimation model M1 generated by the learning unit 22, and outputs the estimated drain flow rate. More specifically, the estimation unit 16 inputs the vibration information detected by the probe 18, the drain discharge capacity of the steam trap 2 to be estimated, and the inlet pressure of the steam trap 2 to be estimated when the probe 18 detects, into the first estimation model M1, and outputs the estimated drain flow rate.

[0022] Furthermore, the estimation unit 16 estimates the drain flow rate from vibration information detected by the probe 18, based on second field data that does not associate the drain discharge capacity of the storage unit 21. Specifically, if the drain discharge capacity of the steam trap 2 to be estimated is unknown, the estimation unit 16 outputs the estimated drain flow rate by inputting the vibration information detected by the probe 18 into the second estimation model M2 generated by the learning unit 22. More specifically, if the drain discharge capacity of the steam trap 2 to be estimated is unknown, the estimation unit 16 outputs the estimated drain flow rate by inputting the vibration information detected by the probe 18 and the inlet pressure of the steam trap 2 to be estimated at the time detected by the probe 18 into the second estimation model M2.

[0023] The correction unit 17 corrects the drain flow rate output by the estimation unit 16 according to the diagnostician's input. More specifically, the correction unit 17 corrects the drain flow rate output by the estimation unit 16 to the drain flow rate input from the input unit 12, based on the drain flow rate determined by the diagnostician, which is input by the diagnostician from the input unit 12. The derivation unit 15, estimation unit 16, and correction unit 17 are implemented, for example, by a microcomputer or processor and various semiconductor memories.

[0024] The server device 20 stores first and second field data, including vibration information and drain flow rate of the steam trap 2 acquired at the site S. Furthermore, the server device 20 generates and updates the aforementioned first estimation model M1 using the stored first field data, and generates and updates the aforementioned second estimation model M2 using the stored second field data. The server device 20 is, for example, a cloud server. Specifically, the server device 20 has a storage unit 21 and a learning unit 22.

[0025] The storage unit 21 stores first field data, which is a correlation between vibration information of the steam trap 2 measured by a diagnostician at the steam trap 2 site S, the drain discharge capacity of the steam trap 2, and the drain flow rate of the steam trap 2 determined by the diagnostician from the operating state of the steam trap 2 when the vibration information was measured. The steam trap 2 site S is the location where the steam trap 2 is actually installed and operating.

[0026] Furthermore, the storage unit 21 stores second field data, which is a correlation between vibration information of the steam trap 2 measured by a diagnostician at the site S of the steam trap 2 for which the drain discharge capacity is unknown, and the drain flow rate of the steam trap 2 determined by the diagnostician from the operating state of the steam trap 2 when the vibration information was measured.

[0027] The diagnostician diagnoses the operating status of the steam trap 2 during inspections and maintenance at each site S and reports the diagnostic results to the customer. For example, the diagnostician measures the vibration information of the steam trap 2 using a vibration measuring instrument in order to diagnose the steam trap 2. The diagnostician also determines the drain flow rate from the operating status of the steam trap 2. The operating status of the steam trap 2 is determined by, for example, the operating sound of the steam trap 2 and the drain flow state. The operating sound of the steam trap 2 is measured by, for example, a stethoscope. The drain flow state is determined by, for example, visually inspecting the inside of the drain pipe 1 connected downstream of the steam trap 2. The diagnostician determines the drain flow rate from these operating statuses of the steam trap 2. In this way, various information about the steam trap 2 acquired at each site S is sequentially stored in the storage unit 21 as first site data and second site data.

[0028] While it is possible to measure the drain flow rate using a flow meter, these meters are expensive, and the installation of the flow meter on the drain piping is time-consuming. Therefore, it is more effective to have a diagnostician determine the drain flow rate.

[0029] Specifically, the first field data of the storage unit 21 is associated with vibration information, inlet pressure, drain discharge capacity, and drain flow rate, as shown in Figure 2. The second field data of the storage unit 21 is associated with vibration pressure, inlet pressure, and drain flow rate. In other words, the first field data is field data associated with drain discharge capacity, and the second field data is field data not associated with drain discharge capacity. The inlet pressure is the fluid pressure of the steam trap 2 when the diagnostician measures the vibration information, and is, for example, the drain pressure measured by the first pressure gauge 3 installed in the drain piping 1 on the inlet side.

[0030] Drain flow rate is expressed as a degree rather than a numerical value. More specifically, the degree of drain flow rate is expressed as a percentage of the drain discharge capacity (i.e., the maximum drain flow rate that can be discharged). In this example, the degree of drain flow rate is expressed in three stages: "large," "medium," and "small." For example, "large" indicates a drain flow rate greater than 90% of the drain discharge capacity, "medium" indicates a drain flow rate between 10% and 90% of the drain discharge capacity, and "small" indicates a drain flow rate less than 10% of the drain discharge capacity. Note that the degree of drain flow rate may also be expressed in two or four or more stages.

[0031] Furthermore, the first and second field data in the storage unit 21 are associated with the diagnostician level, which indicates the skill level of the diagnostician who determined the drain flow rate of the steam trap 2 at site S. In this example, the diagnostician level is represented in three stages, "1," "2," and "3," in descending order of skill level. Naturally, the higher the diagnostician's skill level, the higher the accuracy of the drain flow rate determination. Note that the diagnostician level may also be represented in two or four or more stages.

[0032] In this way, the first and second field data, which associate vibration information, drain flow rate, diagnostician level, etc., are each categorized according to the type of steam trap 2. In this example, steam traps 2 of "Type A" and "Type B" are categorized as first field data, and steam trap 2 of "Type C" is categorized as second field data. In other words, in this example, the capacity characteristics of steam traps 2 of "Type A" and "Type B" are known, while the capacity characteristics of steam trap 2 of "Type C" are unknown.

[0033] Furthermore, the storage unit 21 stores the corrected drain flow rate by the correction unit 17 of the estimation device 10, and the vibration information input to the first estimation model M1 and the second estimation model M2 when the estimation unit 16 outputs the drain flow rate before correction by the correction unit 17, in association with each other. In other words, in this example, if the drain flow rate estimated by the estimation device 10 is corrected by the diagnostician's input operation, the corrected drain flow rate is stored in the storage unit 21 as the first and second field data. Therefore, if the diagnostician determines that the drain flow rate estimated by the estimation device 10 is inappropriate, the drain flow rate determined by the diagnostician is reflected as field data in the storage unit 21.

[0034] The learning unit 22 generates a first estimation model M1 by machine learning using training data that takes vibration information and drain discharge capacity from the first field data of the storage unit 21 as input and drain flow rate as output. The learning unit 22 also generates a second estimation model M2 by machine learning using training data that takes vibration information from the second field data of the storage unit 21 as input and drain flow rate as output. Specifically, the learning unit 22 includes a data selection unit 23 and a model generation unit 24.

[0035] The data selection unit 23 selects and creates the training data necessary for the model generation unit 24 to generate the first estimated model M1 and the second estimated model M2 from the first and second field data of the storage unit 21. More specifically, the data selection unit 23 creates training data for the first estimated model M1, which takes vibration information, drain discharge capacity, and inlet pressure from the first field data of the storage unit 21 as input and drain flow rate as output. The data selection unit 23 also creates training data for the second estimated model M2, which takes vibration information and inlet pressure from the second field data of the storage unit 21 as input and drain flow rate as output. In this embodiment, the data selection unit 23 also creates training data for each type of steam trap 2 so that the model generation unit 24 can generate the first estimated model M1 and the second estimated model M2 for each type of steam trap 2.

[0036] Furthermore, in this embodiment, the data selection unit 23 selects field data from the first field data etc. in the storage unit 21 that is associated with a diagnostician level of a predetermined skill level or higher as training data. For example, training data is created targeting field data from the first field data etc. in the storage unit 21 where the diagnostician level is "2" or higher. Therefore, compared to the case where field data of all diagnostician levels (in this example, diagnostician level "3" or higher) is targeted, training data with higher judgment accuracy regarding drain flow rate is created.

[0037] The model generation unit 24 generates a first estimation model M1 and a second estimation model M2, which have been trained using training data created by the data selection unit 23. As shown in Figure 3, the first estimation model M1 takes vibration information, drain discharge capacity, and inlet pressure as inputs and outputs an estimated drain flow rate. As shown in Figure 4, the second estimation model M2 takes vibration information and inlet pressure as inputs and outputs an estimated drain flow rate. The first estimation model M1 and the second estimation model M2 are trained models generated by supervised machine learning, which uses known machine learning algorithms such as neural networks to learn the relationship between vibration information, drain discharge capacity, etc., and drain flow rate.

[0038] In the learning unit 22, for example, the data selection unit 23 periodically updates each training data, and the model generation unit 24 updates the first estimated model M1 and the second estimated model M2.

[0039] In the flow rate estimation system 100 configured as described above, the diagnostician operates the estimation device 10, which then estimates the drain flow rate of the steam trap 2 to be estimated. The storage unit 13 of the estimation device 10 stores the capacity characteristics D for each type of steam trap 2. The storage unit 13 also stores the latest first estimation model M1 and second estimation model M2 generated by the model generation unit 24. For example, the estimation device 10 downloads the first estimation model M1, etc., from the server device 20 via the network N and stores it in the storage unit 13. More specifically, the storage unit 13 stores multiple types of first estimation models M1 and second estimation models M2 corresponding to each type of steam trap 2.

[0040] Specifically, the operation of the estimation device 10 for drain flow rate will be explained with reference to the flowchart in Figure 5.

[0041] In step S1, at site S, the diagnostician inputs the model, inlet pressure, and outlet pressure into the estimation device 10. Specifically, the diagnostician inputs the inlet and outlet pressures of the steam trap 2 to be estimated into the input unit 12, for example, based on the first pressure gauge 3 and second pressure gauge 4 installed in the drain piping 1. The diagnostician also inputs the model of the steam trap 2 to be estimated into the input unit 12.

[0042] In the following step S2, the derivation unit 15 determines whether or not there is a capability characteristic D corresponding to the model input in step S1. If the capability characteristic D corresponding to the model of the steam trap 2 to be estimated is stored in the memory unit 13, the derivation unit 15 determines that the capability characteristic D exists and proceeds to step S3.

[0043] In step S3, the derivation unit 15 derives the drain discharge capacity of the steam trap 2 to be estimated. Specifically, the derivation unit 15 reads the capacity characteristic D corresponding to the type of steam trap 2 to be estimated from the storage unit 12. The derivation unit 15 also calculates the pressure difference between the inlet pressure and outlet pressure input in step S1. Then, based on the capacity characteristic D read from the storage unit 13 and the calculated pressure difference, the derivation unit 15 derives the drain discharge capacity of the steam trap 2 to be estimated. This allows the maximum drain flow rate that the steam trap 2 can discharge when the inlet pressure and outlet pressure are measured to be determined.

[0044] In the following step S4, the diagnostician uses the estimation device 10 to detect vibration information of the steam trap 2 to be estimated. Specifically, the diagnostician presses the probe 18 of the estimation device 10 against the casing of the steam trap 2, causing the probe 18 to detect vibration information of the steam trap 2. The vibration information detected by the probe 18 is output to the main unit 11 of the device.

[0045] In the following step S5, the estimation unit 16 estimates the drain flow rate of the steam trap 2 to be estimated. Specifically, the estimation unit 16 reads a first estimation model M1 corresponding to the type of steam trap 2 input by the input unit 12 from the storage unit 13. Then, the estimation unit 16 inputs the vibration information output from the probe 18, the drain discharge capacity derived by the derivation unit 15, and the inlet pressure input in step S1 into the first estimation model M1 read from the storage unit 13, and outputs the estimated drain flow rate (more specifically, the degree of drain flow rate). Once the estimated drain flow rate is output, the process proceeds to step S8.

[0046] On the other hand, if the derivation unit 15 determines in step S2 that there is no capability characteristic D corresponding to the model input in step S1, the process proceeds to step S6. In other words, if the capability characteristic D corresponding to the model of the steam trap 2 to be estimated is not stored in the memory unit 13, it is determined that there is no such capability characteristic D.

[0047] In step S6, similar to step S4, the diagnostician uses the estimation device 10 to detect vibration information from the steam trap 2 to be estimated. In other words, the probe 18 detects vibration information from the steam trap 2. The detected vibration information is output to the main unit 11 of the device.

[0048] In the following step S7, the estimation unit 16 estimates the drain flow rate of the steam trap 2 to be estimated. Specifically, the estimation unit 16 reads a second estimation model M2 corresponding to the type of steam trap 2 input by the input unit 12 from the storage unit 13. Then, the estimation unit 16 inputs the vibration information output from the probe 18 and the inlet pressure input in step S1 into the second estimation model M2 read from the storage unit 13, and outputs the estimated drain flow rate (more specifically, the degree of drain flow rate). Once the estimated drain flow rate is output, the process proceeds to step S8.

[0049] In the subsequent step S8, the degree of drain flow rate output by the estimation unit 16 in step S5 or step S7 is displayed on the display unit 14 as shown in Figure 6. In this embodiment, the degree of drain flow rate is displayed as "high". Also in step S8, the confidence level of the drain flow rate (degree of drain flow rate) output by the estimation unit 16 is displayed on the display unit 14. In other words, the accuracy rates of the first estimation model M1 and the second estimation model M2 read by the estimation unit 16 from the storage unit 13 are displayed on the display unit 14. In this embodiment, the confidence level is displayed as "80%".

[0050] In the following step S9, the correction unit 17 determines whether or not the diagnostician has given a correction instruction. If the diagnostician determines that the drain flow rate displayed on the display unit 14 is appropriate, they do not give a correction instruction for the drain flow rate; if the diagnostician determines that the drain flow rate displayed on the display unit 14 is inappropriate, they give a correction instruction for the drain flow rate.

[0051] Specifically, in this example, the display unit 14 displays the drain flow rate and confidence level as described above, as well as an OK button 14a and a correct button 14b. The OK button 14a and correct button 14b are touch-panel buttons. If the diagnostician determines that the drain flow rate displayed on the display unit 14 is appropriate, they press the OK button 14a. In other words, if the drain flow rate displayed on the display unit 14 is about the same as the drain flow rate determined by the diagnostician themselves, they press the OK button 14a without issuing a corrective instruction. In this case, in step S9, the corrective unit 17 determines that there was no corrective instruction and proceeds to step S10.

[0052] In step S10, the memory unit 13 stores the drain flow rate and other information estimated by the estimation unit 16. Specifically, the degree of the drain flow rate output by the estimation unit 16 and the vibration information and other information input to the first estimation model M1 and the second estimation model M2 when the estimation unit 16 outputs the degree of the drain flow rate are associated and stored in the memory unit 13. More specifically, when moving from step S5 to step S8, the vibration information, drain discharge capacity, and inlet pressure input to the first estimation model M1, the degree of the drain flow rate output by the first estimation model M, and the diagnostician level input by the diagnostician are associated and stored in the memory unit 13. Also, when moving from step S7 to step S8, the vibration information and inlet pressure input to the second estimation model M2, the degree of the drain flow rate output by the second estimation model M, and the diagnostician level input by the diagnostician are associated and stored in the memory unit 13. Once the drain flow rate and other values ​​estimated by the estimation unit 16 are stored, the process proceeds to step S13.

[0053] Furthermore, if the diagnostician determines that the drain flow rate displayed on the display unit 14 is not appropriate, they press the correction button 14b. In other words, if the drain flow rate displayed on the display unit 14 differs from the drain flow rate determined by the diagnostician, they press the correction button 14b to issue a correction instruction. In step S9, the correction unit 17 determines that a correction instruction has been issued and proceeds to step S11.

[0054] In step S11, the correction unit 17 corrects the drain flow rate level displayed on the display unit 14 (i.e., the drain flow rate level output by the estimation unit 16). Specifically, the diagnostician inputs the drain flow rate level they have determined using the input unit 12. At the same time, the diagnostician also inputs their diagnostician level using the input unit 12. When the correction unit 17 receives this input from the diagnostician, the drain flow rate level displayed on the display unit 14 is corrected to the drain flow rate level input by the diagnostician using the input unit 12.

[0055] In the following step S12, the memory unit 13 stores the corrected drain flow rate, etc. Specifically, the degree of the corrected drain flow rate and the vibration information etc. input to the first estimation model M1 and the second estimation model M2 when the estimation unit 16 outputs the degree of the drain flow rate before correction are associated and stored in the memory unit 13. More specifically, when moving from step S5 to step S8, the vibration information, drain discharge capacity and inlet pressure input to the first estimation model M1, the degree of the corrected drain flow rate and the diagnostician level input by the diagnostician are associated and stored in the memory unit 13. Also, when moving from step S7 to step S8, the vibration information and inlet pressure input to the second estimation model M2, the degree of the corrected drain flow rate and the diagnostician level input by the diagnostician are associated and stored in the memory unit 13.

[0056] In the subsequent step S13, the first and second field data in the storage unit 21 are updated. The drain flow rate and vibration information stored in the storage unit 13 in steps S10 and S12 are stored in the storage unit 21 of the server device 20 as new first field data, etc., via the network N, for example. As a result, the first and second field data in the storage unit 21 are updated with more accurate data. In particular, if the data is corrected by a diagnostician with a high diagnostic level, the first field data, etc. will be updated with even more accurate data. Once step S13 is completed, the drain flow rate estimation operation is finished.

[0057] As described above, in the flow rate estimation system 100 of the embodiment, the storage unit 21 stores first field data that associates vibration information of the steam trap 2 measured by a diagnostician at the site S of the steam trap 2, the drain discharge capacity of the steam trap 2, and the drain flow rate of the steam trap 2 determined by the diagnostician from the operating state of the steam trap 2 when the vibration information was measured. The estimation device 10 has a probe 18 (detection unit) that detects vibration information of the steam trap 2 to be estimated, and an estimation unit 16 that estimates the drain flow rate (degree of drain flow rate) of the steam trap 2 to be estimated based on the first field data from the storage unit 21, the vibration information detected by the probe 18, and the drain discharge capacity of the steam trap 2 to be estimated. In this configuration, since the drain flow rate is estimated based on first field data acquired at each site of the steam trap 2, a large amount of first field data that takes various environmental conditions into account is stored in the storage unit 21. By estimating the drain flow rate based on this initial field data, the accuracy of the drain flow rate (degree of drain flow rate) estimation can be improved. In particular, since the drain discharge capacity of steam trap 2 is included as part of the initial field data, the estimation accuracy can be further improved. That is, even if the drain flow rate itself is the same, the degree of drain flow rate changes depending on the drain discharge capacity of steam trap 2. Since vibration information changes depending on the drain flow rate, including the drain discharge capacity as part of the initial field data further improves the estimation accuracy. Therefore, for example, even an inexperienced diagnostician can easily obtain a drain flow rate with high estimation accuracy.

[0058] Furthermore, the flow rate estimation system 100 of the above embodiment further includes a learning unit 22 that generates a first estimation model M1 which is trained using training data that takes vibration information and drain discharge capacity from the first field data of the storage unit 21 as input and drain flow rate as output.The estimation unit 16 then inputs the vibration information detected by the probe 18 and the drain discharge capacity of the steam trap 2 into the first estimation model M1 and outputs the estimated drain flow rate.In this way, the drain flow rate is estimated using the first estimation model M1 which has been trained using the first field data of the storage unit 21 as training data, so the estimation accuracy can be further improved.

[0059] Furthermore, the storage unit 21 stores additional second field data, which associates vibration information of steam trap 2 measured by a diagnostician for steam trap 2 whose drain discharge capacity is unknown, with the drain flow rate of steam trap 2 determined by the diagnostician based on the operating state of steam trap 2 at the time the vibration information was measured. The learning unit 22 further generates a second estimation model M2, which has been trained using training data that takes vibration information from the second field data as input and drain flow rate as output. Then, if the drain discharge capacity of the steam trap 2 to be estimated is unknown, the estimation unit 16 inputs the vibration information detected by the probe 18 into the second estimation model M2 and outputs the estimated drain flow rate. In this configuration, the drain flow rate is estimated using the second estimation model M2 generated based on second field data that is not associated with drain discharge capacity. Therefore, a highly accurate estimate of the drain flow rate can be obtained even for steam trap 2 whose drain discharge capacity is unknown.

[0060] Furthermore, the flow rate estimation system 100 of the above embodiment has a capacity characteristic D for each type of steam trap 2. The estimation device 10 further has a derivation unit 15 that derives the drain discharge capacity of the steam trap 2 to be estimated based on the capacity characteristic D, the type of steam trap 2 to be estimated, and the pressure difference. The estimation unit 16 then outputs the estimated drain flow rate by inputting the vibration information detected by the probe 18 and the drain discharge capacity derived by the derivation unit 15 into the first estimation model M1. With this configuration, for example, the drain discharge capacity can be determined simply by inputting the type of steam trap 2 to be estimated and the pressure difference. Therefore, a highly accurate drain flow rate can be easily obtained at the site S.

[0061] Furthermore, in the flow rate estimation system 100 of the above embodiment, the first field data of the storage unit 21 is also associated with the inlet pressure of the steam trap 2 when the diagnostician measured the vibration information. The training data of the learning unit 22 takes the vibration information, drain discharge capacity, and inlet pressure from the first field data of the storage unit 21 as input and outputs the drain flow rate. The estimation unit 16 then inputs the vibration information detected by the probe 18, the drain discharge capacity of the steam trap 2 to be estimated, and the inlet pressure of the steam trap 2 to be estimated when detected by the probe 18 into the first estimation model M1 and outputs the estimated drain flow rate. In this configuration, since the inlet pressure of the steam trap 2 when the vibration information was measured is also added as first field data, it is possible to accumulate first field data that is more in line with the environmental conditions of the site S. Therefore, the estimation accuracy of the drain flow rate (degree of drain flow rate) can be further improved.

[0062] Furthermore, the first field data in the storage unit 21 is associated with the diagnostician level, which indicates the skill level of the diagnostician who determined the drain flow rate of the steam trap 2 at the site S of the steam trap 2. The learning unit 22 selects the first field data from the storage unit 21 that is associated with a diagnostician level of a predetermined skill level or higher as training data. Therefore, it is possible to create training data that reflects the knowledge of more skilled diagnosticians. As a result, it is possible to construct highly accurate training data and, consequently, the first estimation model M1 regarding the drain flow rate. Consequently, the estimation accuracy of the drain flow rate (degree of drain flow rate) can be further improved.

[0063] Furthermore, the estimation device 10 includes a correction unit 17 that corrects the drain flow rate output by the estimation unit 16 upon receiving an input operation. The storage unit 21 stores the corrected drain flow rate by the correction unit 17 in association with the vibration information and drain discharge capacity input to the first estimation model M1 when the estimation unit 16 outputs the drain flow rate before correction by the correction unit 17. In this configuration, the knowledge of more skilled diagnosticians can be reflected in the first field data of the storage unit 21. As a result, the first field data of the storage unit 21 is updated as highly accurate data, so that a highly accurate first estimation model M1 can be generated and updated. Thus, the estimation accuracy of the drain flow rate (degree of drain flow rate) can be further improved.

[0064] Furthermore, the estimation device 10 also has a display unit 14 that displays the drain flow rate output by the estimation unit 16 and information indicating the accuracy rate of the first estimation model M1. Therefore, the estimated drain flow rate and its confidence level can be easily visually confirmed. Consequently, the reliability of the device is increased.

[0065] Furthermore, in the flow rate estimation system 100 of the above embodiment, the operating state of the steam trap 2 when vibration information is measured is the operating sound of the steam trap 2. While it is difficult to determine the drain flow rate based on vibration information alone, by considering the operating sound of the steam trap 2, the drain flow rate can be determined with high accuracy.

[0066] (Other embodiments) As described above, the embodiments described herein have been presented as examples of the technology disclosed in this application. However, the technology in this disclosure is not limited thereto and can be applied to embodiments that have been modified, replaced, added, or omitted as appropriate. Furthermore, it is possible to combine the components described in the embodiments above to create new embodiments. In addition, the components described in the attached drawings and detailed description may include not only components essential for solving the problem, but also components that are not essential for solving the problem, in order to illustrate the technology. Therefore, the mere presence of such non-essential components in the attached drawings and detailed description should not be immediately assumed to mean that those non-essential components are essential.

[0067] For example, the training data for the first estimation model M1 may omit the inlet pressure and only input vibration information and drain discharge capacity, with the drain flow rate (degree of drain flow rate) as the output. The training data for the second estimation model M2 may omit the inlet pressure and only input vibration information, with the drain flow rate (degree of drain flow rate) as the output.

[0068] Furthermore, the storage unit 21 may also store various types of information, such as that obtained in the laboratory, in addition to the first field data.

[0069] Furthermore, the number of operations may be added instead of the inlet pressure as the first and second field data. Specifically, the first field data of the storage unit 21 is associated with vibration information, number of operations, drain discharge capacity, and drain flow rate (degree of drain flow rate). The second field data of the storage unit 21 is associated with vibration information, number of operations, and drain flow rate (degree of drain flow rate). The number of operations is the number of times the steam trap 2 operates within a predetermined time when the diagnostician measures the vibration information. The number of operations is the number of times the steam trap 2 operates to discharge drain. The more operations there are within a predetermined time, the more the drain flow rate tends to increase.

[0070] In this case, the training data for the first estimation model M1 takes vibration information, drain discharge capacity, and number of operations from the first field data of the storage unit 21 as input and outputs the drain flow rate. The estimation unit 16 outputs the estimated drain flow rate by inputting the vibration information detected by the probe 18, the drain discharge capacity of the steam trap 2 to be estimated, and the number of operations of the steam trap 2 to be estimated within a predetermined time as detected by the probe 18 into the first estimation model M1 generated by the learning unit 22. The estimation unit 16 also outputs the estimated drain flow rate by inputting the vibration information detected by the probe 18 and the number of operations of the steam trap 2 to be estimated within a predetermined time as detected by the probe 18 into the second estimation model M2 generated by the learning unit 22. Even with this configuration, the accuracy of drain flow rate estimation can be improved.

[0071] Furthermore, both the inlet pressure and the number of operations of steam trap 2 may be used as first field data, etc.

[0072] Furthermore, the diagnostician level may be omitted when presenting the first field data, etc.

[0073] Alternatively, the temperature of steam trap 2 may be added as first-field data, and the aforementioned temperature may be added as input for training data.

[0074] Furthermore, the capability characteristic D may be stored in, for example, the server device 20, rather than in the memory unit 13.

[0075] The technology disclosed in this disclosure can be summarized as follows:

[0076] [1] The flow rate estimation system 100 estimates the drain flow rate of the steam trap 2 to be estimated. The flow rate estimation system 100 includes an accumulation unit 21 in which first field data is accumulated, relating vibration information of the steam trap 2 measured by a diagnostician at the site S of the steam trap 2, the drain discharge capacity of the steam trap 2, and the drain flow rate of the steam trap 2 determined by the diagnostician from the operating state of the steam trap 2 when the vibration information was measured; and an estimation device 10 having a probe 18 for detecting vibration information of the steam trap 2 to be estimated, and an estimation unit 16 for estimating the drain flow rate of the steam trap 2 to be estimated based on the first field data in the accumulation unit 21, the vibration information detected by the probe 18, and the drain discharge capacity of the steam trap 2 to be estimated.

[0077] In this configuration, the drain flow rate is estimated based on first-site data acquired at each site S of the steam trap 2. Therefore, a large amount of first-site data, taking into account various environmental conditions, is stored in the storage unit 21. Because the drain flow rate is estimated based on this first-site data, the estimation accuracy of the drain flow rate (the degree of drain flow rate) can be improved. In particular, since the drain discharge capacity of the steam trap 2 is included in the first-site data, the estimation accuracy can be further improved. That is, even if the drain flow rate itself is the same, the degree of drain flow rate changes depending on the drain discharge capacity of the steam trap 2. Since vibration information changes depending on the drain flow rate, including the drain discharge capacity as part of the first-site data further improves the estimation accuracy. Therefore, for example, even an inexperienced diagnostician can easily obtain a highly accurate estimate of the drain flow rate.

[0078] [2] The flow rate estimation system 100 described in [1] further includes a learning unit 22 that generates a first estimation model M1 which is machine-trained using training data that takes vibration information and drain discharge capacity from first field data of the storage unit 21 as inputs and drain flow rate as output. The estimation unit 16 outputs an estimated drain flow rate by inputting the vibration information detected by the probe 18 and the drain discharge capacity of the steam trap 2 to be estimated into the first estimation model M1 generated by the learning unit 22.

[0079] With this configuration, the drain flow rate is estimated using a first estimation model M1 that has been trained using first field data from the storage unit 21 as training data, thereby improving the estimation accuracy. Furthermore, because the estimation accuracy of the drain flow rate can be improved, measuring devices such as the one described in Patent Document 2 can accurately grasp the drain flow rate. Therefore, it is possible to suppress the misjudgment that a malfunction (i.e., steam leakage) is occurring even when the drain is being discharged normally. Thus, the accuracy of determining whether the device is functioning correctly can be improved.

[0080] [3] In the flow rate estimation system 100 described in [1] or [2], the storage unit 21 further stores second field data, which associates vibration information of the steam trap 2 measured by a diagnostician at the site of the steam trap 2 for a steam trap 2 whose drain discharge capacity is unknown, with the drain flow rate of the steam trap 2 determined by the diagnostician from the operating state of the steam trap 2 when the vibration information was measured. The learning unit 22 further generates a second estimation model M2 that has been trained using training data that takes vibration information in the second field data as input and drain flow rate as output. If the drain discharge capacity of the steam trap 2 to be estimated is unknown, the estimation unit 16 outputs an estimated drain flow rate by inputting the vibration information detected by the probe 18 into the second estimation model M2 generated by the learning unit 22.

[0081] With this configuration, the drain flow rate can be estimated using a second estimation model M2 generated based on second field data that does not associate drain discharge capacity. Therefore, even for steam trap 2 whose drain discharge capacity is unknown, a highly accurate estimate of the drain flow rate can be obtained.

[0082] [4] In the flow rate estimation system 100 described in any one of [1] to [3], each steam trap 2 has a capacity characteristic D which determines the drain discharge capacity of the steam trap 2 according to the pressure difference between the inlet pressure and outlet pressure of the steam trap 2, and the estimation device 10 further has a derivation unit 15 which derives the drain discharge capacity of the steam trap 2 to be estimated based on the capacity characteristic D, the type of steam trap 2 to be estimated and the pressure difference, and the estimation unit 16 outputs the estimated drain flow rate by inputting the vibration information detected by the probe 18 and the drain discharge capacity derived by the derivation unit 15 into a first estimation model M1 generated by the learning unit 22.

[0083] With this configuration, for example, the drain discharge capacity can be determined simply by inputting the type and pressure difference of the steam trap 2 to be estimated. Therefore, a highly accurate estimate of the drain flow rate can be easily obtained at the site S.

[0084] [5] In the flow rate estimation system 100 described in any one of [1] to [4], the estimation device 10 further includes a correction unit 17 that corrects the drain flow rate output by the estimation unit 16 upon receiving an input operation, and the storage unit 21 stores the corrected drain flow rate by the correction unit 17 in association with the vibration information and drain discharge capacity input to the first estimation model M1 when the estimation unit 16 outputs the drain flow rate before correction by the correction unit 17.

[0085] This configuration allows the expertise of more skilled diagnosticians to be reflected in the first field data stored in the storage unit 21. As a result, the first field data in the storage unit 21 is updated with highly accurate data, enabling the generation and updating of a highly accurate first estimation model M1. Therefore, the estimation accuracy of drain flow rate (degree of drain flow rate) can be further improved.

[0086] In the flow rate estimation system 100 described in any one of [6] [1] to [5], the first field data of the storage unit 21 is associated with the number of times the steam trap 2 operates within a predetermined time when the diagnostician measures vibration information, the training data of the learning unit 22 takes vibration information, drain discharge capacity and number of operations in the first field data of the storage unit 21 as input and outputs drain flow rate, and the estimation unit 16 outputs the estimated drain flow rate by inputting the vibration information detected by the probe 18, the drain discharge capacity of the steam trap 2 to be estimated and the number of times the steam trap 2 to be estimated operates within a predetermined time when detected by the probe 18 into the first estimation model M1 generated by the learning unit 22.

[0087] With this configuration, the inlet pressure of steam trap 2 when vibration information is measured is also added as first-site data, allowing for the accumulation of first-site data that more closely reflects the environmental conditions of site S. Therefore, the accuracy of estimating the drain flow rate (degree of drain flow rate) can be further improved.

[0088] In the flow rate estimation system 100 described in any one of [7] [1] to [6], the first field data of the storage unit 21 is associated with the inlet pressure of the steam trap 2 when the diagnostician measures the vibration information, the training data of the learning unit 22 takes the vibration information, drain discharge capacity and inlet pressure of the first field data of the storage unit 21 as input and outputs the drain flow rate, and the estimation unit 16 outputs the estimated drain flow rate by inputting the vibration information detected by the probe 18, the drain discharge capacity of the steam trap 2 to be estimated and the inlet pressure of the steam trap 2 to be estimated when detected by the probe 18 into the first estimation model M1 generated by the learning unit 22.

[0089] With this configuration, the inlet pressure of steam trap 2 when vibration information is measured is also added as first-site data, allowing for the accumulation of first-site data that more closely reflects the environmental conditions of site S. Therefore, the accuracy of estimating the drain flow rate (degree of drain flow rate) can be further improved.

[0090] In the flow rate estimation system 100 described in any one of [8] [1] to [7], the first field data of the storage unit 21 is associated with a diagnostician level indicating the skill level of the diagnostician who determined the drain flow rate of the steam trap 2 at the site S of the steam trap 2, and the learning unit 22 selects from the first field data of the storage unit 21 that is associated with a diagnostician level of a predetermined skill level or higher as training data.

[0091] This configuration allows for the creation of training data that reflects the knowledge of more skilled diagnosticians. Therefore, it is possible to construct highly accurate training data and, consequently, the first estimation model M1 regarding drain flow rate. Thus, the estimation accuracy of drain flow rate (the degree of drain flow rate) can be further improved.

[0092] In the flow rate estimation system 100 described in any one of [9] [1] to [8], the estimation device 10 further includes a display unit 14 that displays the drain flow rate output by the estimation unit 16 and information indicating the accuracy of the first estimation model M1.

[0093] This configuration allows for easy visualization of the estimated drain flow rate and its confidence level. Therefore, the reliability of the device is enhanced.

[0094] In the flow rate estimation system 100 described in any one of [1] to [9], the operating state of the steam trap 2 when vibration information is measured is the operating sound of the steam trap 2.

[0095] With this configuration, while it is difficult to determine the drain flow rate based solely on vibration information, the operating sound of steam trap 2 can be taken into consideration, allowing for highly accurate determination of the drain flow rate. [Industrial applicability]

[0096] As described above, the technology of this disclosure is useful for flow rate estimation systems. [Explanation of Symbols]

[0097] 100 Flow rate estimation system 2. Steam trap (drain trap) 10 Estimation device 14 Display section 15 Derivation part 16 Estimation part 17 Correction section 18. Probe (detection unit) 21 Storage section 22 Learning Department S-site D ability characteristics M1 First Estimated Model M2 Second Estimated Model

Claims

1. A flow rate estimation system for estimating the drain flow rate of a drain trap to be estimated, A storage unit accumulates first field data, which is a correlation between vibration information of the drain trap measured by a diagnostician at the site of the drain trap, the drain discharge capacity of the drain trap, and the drain flow rate of the drain trap determined by the diagnostician from the operating state of the drain trap when the vibration information was measured. The estimation device comprises a detection unit for detecting vibration information of the drain trap to be estimated, and an estimation unit for estimating the drain flow rate of the drain trap to be estimated based on the first field data from the storage unit, the vibration information detected by the detection unit, and the drain discharge capacity of the drain trap to be estimated. The system further comprises a learning unit that generates a first estimation model trained using training data that takes vibration information and drain discharge capacity from the first field data of the storage unit as input and drain flow rate as output. The estimation unit inputs the vibration information detected by the detection unit and the drain discharge capacity of the drain trap to be estimated into the first estimation model generated by the learning unit, thereby outputting the estimated drain flow rate. The storage unit further stores second field data, which associates vibration information of a drain trap measured by a diagnostician at the site of a drain trap whose drain discharge capacity is unknown, with the drain flow rate of the drain trap determined by the diagnostician from the operating state of the drain trap when the vibration information was measured. The learning unit further generates a second estimation model that has been trained using training data that takes vibration information from the second field data as input and drain flow rate as output. If the drain discharge capacity of the drain trap to be estimated is unknown, the estimation unit inputs the vibration information detected by the detection unit into the second estimation model generated by the learning unit, thereby outputting an estimated drain flow rate. A flow rate estimation system characterized by the following features.

2. In the flow rate estimation system according to claim 1, The drain trap has a defined drain discharge capacity characteristic for each type of drain trap, which is determined according to the pressure difference between the inlet pressure and outlet pressure of the drain trap. The estimation device further includes a derivation unit that derives the drain discharge capacity of the drain trap to be estimated based on the capacity characteristics, the type of drain trap to be estimated, and the pressure difference. The estimation unit outputs an estimated drain flow rate by inputting the vibration information detected by the detection unit and the drain discharge capacity derived by the derivation unit into the first estimation model generated by the learning unit. A flow rate estimation system characterized by the following features.

3. In the flow rate estimation system according to claim 1, The estimation device further includes a correction unit that, upon receiving an input operation, corrects the drain flow rate output by the estimation unit. The storage unit stores the vibration information and drain discharge capacity input to the first estimation model when the estimation unit outputs the drain flow rate after correction by the correction unit and the drain flow rate before correction by the correction unit, in association with each other. A flow rate estimation system characterized by the following features.

4. In the flow rate estimation system according to claim 1, The first field data of the storage unit is also associated with the number of times the drain trap operated within a predetermined time when the diagnostician measured the vibration information. The training data for the learning unit takes vibration information, drain discharge capacity, and number of operations from the first field data of the storage unit as input and outputs drain flow rate. The estimation unit outputs an estimated drain flow rate by inputting the vibration information detected by the detection unit, the drain discharge capacity of the drain trap to be estimated, and the number of times the drain trap to be estimated operates within a predetermined time as detected by the detection unit, into the first estimation model generated by the learning unit. A flow rate estimation system characterized by the following features.

5. In the flow rate estimation system according to claim 1, The first field data from the storage unit is also associated with the inlet pressure of the drain trap when the diagnostician measured the vibration information. The training data for the learning unit takes vibration information, drain discharge capacity, and inlet pressure from the first field data of the storage unit as input and outputs drain flow rate. The estimation unit outputs an estimated drain flow rate by inputting the vibration information detected by the detection unit, the drain discharge capacity of the drain trap to be estimated, and the inlet pressure of the drain trap to be estimated at the time detected by the detection unit into the first estimation model generated by the learning unit. A flow rate estimation system characterized by the following features.

6. In the flow rate estimation system according to claim 1, The first field data from the storage unit is also associated with the diagnostician level, which indicates the skill level of the diagnostician who determined the drain flow rate of the drain trap at the drain trap site. The learning unit selects from the first field data in the storage unit the first field data associated with a diagnostician level of a predetermined skill level or higher as the training data. A flow rate estimation system characterized by the following features.

7. In the flow rate estimation system according to claim 1, The estimation device further includes a display unit that displays the drain flow rate output by the estimation unit and information indicating the accuracy of the first estimation model. A flow rate estimation system characterized by the following features.

8. In the flow rate estimation system according to claim 1, The operating state of the drain trap when the aforementioned vibration information is measured is the operating sound of the drain trap. A flow rate estimation system characterized by the following features.

Images