Information processing apparatus, machine learning apparatus, inference apparatus, information processing method, machine learning method, and inference method

Abstract

The invention aims to provide a method and device. Provided are an information processing device, a machine learning device, an inference device, and an information processing method that are capable of verifying the amount of fluid passing through a valve mechanism, verifying the presence or absence of accidental leakage of fluid from between a valve body and a valve seat, verifying the amount of leakage thereof, and verifying the omen of accidental leakage of fluid from the valve mechanism. A machine learning method and a reasoning method. An information processing device (250) predicts the state of a valve mechanism, and obtains a valve mechanism state indicating the state of the valve mechanism on the basis of input information that is a predetermined variable between a valve body and a valve seat.

Description

Technical Field

[0001] This disclosure relates to information processing apparatus, machine learning apparatus, reasoning apparatus, information processing method, machine learning method, and reasoning method. Background Technology

[0002] Previously, a vent valve with a positive pressure side valve seat and a positive pressure side valve was known, wherein the positive pressure side valve seat had a discharge port for fluid inside the container; and the positive pressure side valve was capable of opening and closing the discharge port according to the pressure inside the container (see, for example, Patent Document 1).

[0003] Existing technical documents

[0004] Patent documents

[0005] Patent Document 1: Japanese Patent Application Publication No. 2018-21652 Summary of the Invention

[0006] The problem the invention aims to solve

[0007] In valve mechanisms consisting of a valve core and a valve seat positioned along a fluid flow path, the importance of managing the fluid flowing internally necessitates reliable closure of the flow path, verification of the fluid volume passing through the valve mechanism, verification of any accidental leakage between the valve core and valve seat, and verification of the leakage amount should fluid leak from the valve mechanism in the event of leakage. Furthermore, to prevent leakage from the valve mechanism due to unforeseen circumstances or to minimize leakage, verification of signs of accidental leakage from the valve mechanism is also required. This requirement is further strengthened in recent years due to increased concerns about flora, fauna, and the environment, especially when handling volatile, flammable, or harmful fluids. However, existing valve mechanisms only allow for the closure and opening of the flow path, and cannot verify the fluid volume passing through the valve mechanism, the presence or absence of accidental leakage between the valve core and valve seat, the leakage amount, or the signs of accidental leakage from the valve mechanism. This problem has not been solved in the valve mechanism used in the vent valve for handling volatile fluids, which is an example of an existing valve mechanism, and exists as a general problem of valve mechanisms.

[0008] This disclosure is made to solve the above-mentioned problems, and its purpose is to provide an information processing apparatus, a machine learning apparatus, an inference apparatus, an information processing method, a machine learning method, and an inference method, which are capable of verifying the amount of fluid passing through a valve mechanism, verifying whether there is any accidental leakage of fluid between the valve core and the valve seat, verifying the amount of such leakage, and verifying the signs of accidental leakage of fluid from the valve mechanism.

[0009] Solution for solving the problem

[0010] The information processing apparatus disclosed herein is an information processing apparatus for predicting the state of a valve mechanism. It obtains the valve mechanism state representing the state of the valve mechanism based on the specified variables between the valve core and the valve seat, i.e., the input information.

[0011] Invention Effects

[0012] By employing the information processing apparatus, machine learning apparatus, inference apparatus, information processing method, machine learning method, and inference method disclosed herein, it is possible to verify the amount of fluid passing through the valve mechanism, verify whether there is any accidental leakage of fluid between the valve core and the valve seat, verify the amount of leakage, and verify the signs of accidental leakage of fluid from the valve mechanism. Attached Figure Description

[0013] Figure 1 This is a schematic diagram showing the valve system of the first embodiment;

[0014] Figure 2 It is shown Figure 1 A schematic diagram of the vent valve;

[0015] Figure 3 It is shown Figure 2 Top view of the inlet section;

[0016] Figure 4 It is shown Figure 2 An enlarged view of the components included in the detection device;

[0017] Figure 5 It is shown Figure 2 A top view of the valve contact components;

[0018] Figure 6 It is shown Figure 1 A structural diagram of a machine learning device;

[0019] Figure 7 It is shown Figure 6 A conceptual diagram of the elements of machine learning implemented by a machine learning device;

[0020] Figure 8 It is shown Figure 1 A schematic diagram of an information processing device;

[0021] Figure 9 It is a hardware structure diagram of a computer;

[0022] Figure 10 It is shown Figure 1 A flowchart of the machine learning method of the machine learning device;

[0023] Figure 11 It is shown Figure 1 A flowchart of a valve mechanism state prediction method for a learning information processing device;

[0024] Figure 12 This is a top view showing the inlet section of the second embodiment. Detailed Implementation

[0025] Hereinafter, specific embodiments of the present disclosure will be described with reference to the accompanying drawings. It should be noted that the following description schematically illustrates the scope required to achieve the purpose of the present disclosure, mainly describing the scope required for the description of the corresponding parts of the present disclosure, while the parts omitted in the description employ known techniques.

[0026] First Implementation Method

[0027] As an illustration of the valve system of this disclosure, in the first embodiment, a valve system suitable for a vent valve is used. However, the valve system of this disclosure is not limited to vent valves and can be suitably used in general valve mechanisms. It should be noted that the same applies to the valve systems of the second and third embodiments described below.

[0028] Figure 1 This is a schematic diagram showing a valve system 200 according to a first embodiment. The valve system 200 includes a vent valve 5, a machine learning device 240, an information processing device 250, and a network 270 connecting them respectively.

[0029] Figure 2 It is shown Figure 1 A schematic diagram of the vent valve 5. The vent valve 5 is connected to the opening of the tank 4 via the upstream flange 4b. Figure 2 In the diagram, a portion of the vent valve 5 is represented by a cross-section cut along the axis containing the flow channel.

[0030] Tank 4 stores flammable gases or liquids. Furthermore, tank 4 also contains fluids that have evaporated from the stored fluids. Flammable gases or liquids include, for example, fossil fuels and volatile gases. Tank 4 can be spherical, cylindrical, cuboid, or cubic in shape.

[0031] In this first embodiment, the tank 4 is shaped such that its horizontal cross-sectional area decreases as it faces upwards. A mounting portion 4a is provided at the upper end of the tank 4. A through hole communicating with the interior of the tank 4 is formed in the mounting portion 4a. As a result, the fluid inside the tank 4 is efficiently discharged through the mounting portion 4a.

[0032] The base of the vent valve 5 is mounted on the mounting section 4a via the upstream flange 4b. In this first embodiment, the tank 4 is positioned on the upstream side, and the vent valve 5 is provided on the downstream side relative to the tank 4.

[0033] The vent valve 5 can discharge the fluid inside the tank 4 to the atmosphere according to the internal pressure of the tank 4, and allow the atmosphere, which is used as the suction fluid, to flow into the tank 4. That is, the vent valve 5 can regulate the pressure inside the tank 4.

[0034] In this first embodiment, the tank 4 and the vent valve 5 are located outdoors and exposed to the atmosphere.

[0035] The vent valve 5 includes a main section 10, an air inlet section 20, and an air outlet section 40. The main section 10 is a T-shaped piping component. The ends of the T-shaped piping component of the main section 10 are an inlet end 10a that opens downward in the vertical direction, an outlet end 10b that opens upward in the vertical direction, and an air inlet end 10c that opens horizontally.

[0036] The main section 10 is positioned such that it has an inlet end 10a on its lower side and an outlet end 10b on its upper side. A flow channel extending from the inlet end 10a to the outlet end 10b is designated as the main flow channel 110. The axis of the main section 10 is aligned vertically with the axis of the main flow channel 110. The main flow channel 110 extends toward the discharge section 40. The main flow channel 110 extending to the discharge section 40 will be described later.

[0037] A branch flow path 111 extends from the main flow path 110 to the intake side opening end 10c. The axis of the branch flow path 111 is horizontal. An intake section 20 is connected to the intake side opening end 10c of the main section 10.

[0038] The intake section 20 has an intake section body 21 and an intake valve mechanism 22 disposed on the intake section body 21. The intake section body 21 is a tubular component with two open ends.

[0039] One end of the intake body 21 is a connecting side opening end 21a that opens in the horizontal direction. An intake port 21b that opens downward in the vertical direction is formed at the other end of the intake body 21. The flow channel connecting the connecting side opening end 21a and the intake port 21b is designated as the intake flow channel 120. Here, the other end of the intake body 21 surrounding the intake port 21b is designated as the intake valve seat 21x.

[0040] The intake valve mechanism 22 includes an intake valve 23 serving as an on / off valve, an intake valve shaft 24 fixed to the intake valve 23, and an intake valve guide 25 disposed on the intake section body 21. The intake port 21b is closed by the intake valve 23 engaging with the intake valve seat 21x surrounding the intake port 21b. The intake port 21b is opened by creating a gap between the intake valve 23 and the intake valve seat 21x. In other words, the intake valve 23 can open / close the intake passage 120.

[0041] The intake valve guide 25 supports the intake valve shaft 24, enabling it to move vertically. Thus, the intake valve 23, moving vertically via the intake valve shaft 24, can move between an intake closed position and an intake open position. The intake valve 23 is an open valve. The intake valve 23, by virtue of its own weight and that of the intake valve shaft 24, falls along the intake valve guide 25 and can move towards the intake closed position.

[0042] When the intake valve 23 is in the intake closed position, the intake valve 23 closes the intake port 21b. That is, since the lower surface of the intake valve 23 is in contact with the intake valve seat 21x, the intake passage 120 is closed, and the connection between the intake passage 120 and the outside of the vent valve 5 is cut off.

[0043] When the intake valve 23 overcomes its own weight and moves upward from the intake closed position to the intake open position, a gap is created between the intake valve 23 and the intake valve seat 21xb. That is, the intake passage 120 opens, and the intake passage 120 becomes connected to the external atmosphere of the vent valve 5.

[0044] The intake-side opening end 10c is connected to the connecting-side opening end 21a, and the intake section 20 is connected to the intake-side opening end 10c of the main section 10. Thus, the intake air passage 120 inside the intake section body 21 communicates with the branch passage 111 inside the main section 10. That is, the intake air passage 120 extends horizontally from the branch passage 111.

[0045] A discharge section 40 is connected to the outlet opening end 10b of the main section 10. The discharge section 40 has a discharge section body 41, a discharge cover 47, a discharge valve mechanism 50 provided on the discharge cover 47, a detection device 60, and a valve contact member 70.

[0046] The discharge section body 41 consists of a tubular inlet section 42 and an atmospheric discharge section 43 that surrounds an open end of the inlet section 42. The discharge section body 41 is arranged in a vertical orientation with the inlet section 42 as a conduit.

[0047] Figure 3 It is shown Figure 2 Top view of the inlet section 42. Figure 4 It is shown Figure 2 An enlarged view of the components included in the detection device 60. Figure 3 Since the upper surface of the inlet portion 42 is shown, the valve contact component 70 is not shown. Figures 2 to 4 Let me continue explaining.

[0048] The downward-opening end of the inlet portion 42 is designated as the discharge portion inlet-side opening end 42a. The vertically upward-opening end of the inlet portion 42 is designated as the valve-side end 42b. The opening of the valve-side end 42b is designated as the valve-side opening 42c, and the end face of the valve-side end 42b is designated as the main body end face 42d. The discharge portion inlet-side opening end 42a is connected to the discharge portion outlet-side opening end 10b of the main portion 10.

[0049] A detector setting space 42e is formed in the inlet portion 42. The detector setting space 42e is a space that opens to the main body end face 42d and then opens to the outer periphery of the inlet portion 42 through the wall of the inlet portion 42. A portion of the detection device 60 is provided in the detector setting space 42e.

[0050] The detection device 60 includes a detector 61 and wiring 62 for transmitting signals from the detector 61 to an input / output device (not shown). The detector 61 is capable of detecting a specified variable of the object being detected.

[0051] At this time, detector 61 can detect the specified variable of the detection object in the direction in which detector 61 is facing, and this direction is taken as the detection direction of detector 61. That is, the so-called detection direction of detector 61 is the direction in which detector 61 can detect the specified variable of the detection object.

[0052] A detector 61 and a portion of a wiring 62 extending from the detector 61 are provided in the detector setting space 42e. The detector 61 is configured such that its detection direction extends from the detector setting space 42e toward the outer side of the main body end face 42d. That is, the detector 61 is capable of detecting a predetermined variable of the object to be detected, which is present on the end face of the valve side end 42b of the inlet portion 42.

[0053] Wiring 62 extends outward from the outer periphery of the inlet portion 42 through the detector setting space 42e. Wiring 62 is connected to an input / output device (not shown). The input / output device is capable of sending a specified variable detected by the detector 61 to the network 270.

[0054] Network 270 is an information transmission network constructed via wired or wireless means. Network 270 can connect to the Internet, thereby enabling information transmission. Devices connected to network 270 can acquire signals sent to detector 61 on network 270. Other devices connected to network 270 will be described later.

[0055] A valve contact component 70 is provided at the valve side end 42b. Figure 5 It is shown Figure 2 A top view of the valve contact component 70. According to... Figures 2 to 5 Let me continue explaining.

[0056] The valve contact member 70 is an annular component. The cross-section of the valve contact member 70, taken along a line in the radial direction of the annulus, is L-shaped. That is, the valve contact member 70 is a shape in which the entire circumference of the annular planar plate is bent perpendicular to the plane. The valve contact member 70 is formed to correspond to the shape of the body end face 42d.

[0057] The valve contact member 70 is provided in the inlet portion 42 such that it covers the main body end face 42d of the inlet portion 42 and a portion of the inner peripheral surface of the inlet portion 42 that is continuous from the main body end face 42d. Here, the surface of the valve contact member 70 facing the discharge valve 53 is referred to as the valve facing surface 70a. The discharge valve 53 will be described later.

[0058] The opening of the detector setting space 42e formed on the main body end face 42d is closed by providing a valve contact member 70 on the inlet portion 42. That is, the detector 61 is provided in the detector setting space 42e covered by the valve contact member 70. In addition, the detection direction of the detector 61 is toward the discharge valve 53 which is located across the valve facing surface 70a.

[0059] The valve contact component 70 is fixed to the inlet portion 42 by screws. A mounting portion for screw fixing is provided on the outer periphery of the valve contact component 70. However, besides screw fixing, known methods can also be used to fix the valve contact component 70 to the inlet portion 42. The valve contact component 70 is preferably made of a material that will not significantly reduce the detection capability of the detector 61. For example, the valve contact component 70 is made of stainless steel.

[0060] Preferably, efforts are made to optimize the shape of the valve contact component 70 to avoid a significant reduction in the detection capability of the detector 61. For example, by thinning the valve contact component 70, a significant reduction in the detection capability of the detector 61 can be prevented. It should be noted that, as... Figures 2 to 5 As shown, the inner peripheral surface of the portion of the inlet 42 where the detector mounting space 42e is formed protrudes further inward, and the wall thickness of the portion of the inlet 42 where the detector mounting space 42e is formed is thicker than the wall thickness of the other portions. Furthermore, the valve contact member 70 is formed in a shape corresponding to the shape of the main body end face 42d. However, this is not a limitation. For example, the wall thickness of the inlet 42 may be uniform, meaning that no portion protruding inward is formed on the inner peripheral surface of the inlet 42, and the inner peripheral surface of the inlet 42 viewed axially may also be circular. In this case, the valve contact member 70 is also formed in a shape corresponding to the shape of the main body end face 42d.

[0061] Here, the valve-side end 42b of the inlet portion 42 is used as the valve seat body 3. It should be noted that the valve seat body 3 can be cylindrical, as the valve-side end 42b of the inlet portion 42 forms part of the discharge flow channel 140, which is part of the flow channel closed by the valve core. Furthermore, the main body end face 42d of the inlet portion 42 is formed at the end of the flow channel of the valve seat body 3, and is the main body end face of the valve seat body 3 facing the valve core. A detector mounting space 42e is formed in the valve seat body 3.

[0062] Furthermore, the valve seat body 3, which serves as the valve-side end 42b of the inlet portion 42, and the valve contact member 70 are used as the valve seat 2. Therefore, the valve seat 2 constitutes part of the flow channel. Additionally, the valve seat 2 and the detection device 60 are used as the valve seat mechanism 1. Furthermore, in this case, the valve-facing surface 70a of the valve contact member 70 is used as the valve seat end face 2a.

[0063] That is, the valve seat end face 2a faces the lower surface of the discharge valve 53, which is the valve core, and is the part that the discharge valve 53 directly contacts when the discharge passage 140 is closed. Moreover, when a gap is formed between the discharge valve 53 and the valve seat end face 2a and the discharge passage 140 is open, the fluid in the tank 4 is discharged through the gap between the discharge valve 53 and the valve seat end face 2a.

[0064] Furthermore, with the detector 61 set in the detector setting space 42e, the detection direction is toward the discharge valve 53, which is a valve core located across the valve seat end face 2a.

[0065] return Figure 2 Continuing the explanation, the atmospheric discharge section 43 is formed as a valve-side end 42b surrounding the inlet section 42. An opening is formed on the upper end side of the atmospheric discharge section 43. The upper end side of the atmospheric discharge section 43 serves as the valve mechanism-side opening end 43b. On the lower side of the valve mechanism-side opening end 43b of the atmospheric discharge section 43, i.e., the body portion of the atmospheric discharge section 43, a plurality of discharge ports 43a opening in the horizontal direction are formed.

[0066] The vent cover 47 is connected to the valve mechanism side opening end 43b. The vent cover 47 closes the opening of the valve mechanism side opening end 43b. The connection between the vent cover 47 and the valve mechanism side opening end 43b can be achieved using a known structure. A discharge valve mechanism 50 is provided on the vent cover 47.

[0067] The discharge valve mechanism 50 has a valve core, namely a discharge valve 53, capable of closing the valve side opening 42c, a discharge valve shaft 54 ​​fixed to the discharge valve 53, and a discharge valve guide 55 provided on the vent cover 47.

[0068] With the vent cover 47 connected to the valve mechanism side opening end 43b, the discharge valve guide 55 extends vertically downward from the vent cover 47. The discharge valve guide 55 supports the discharge valve shaft 54 ​​so that it can move vertically.

[0069] By moving the discharge valve shaft 54 ​​vertically, the discharge valve 53 can move between a discharge closed position and a discharge open position. The discharge valve 53 is an open valve.

[0070] The discharge valve 53 falls along the discharge valve guide 55 by the weight of the discharge valve 53 and the discharge valve shaft 54, and can move toward the discharge closed position.

[0071] When the discharge valve 53 is in the discharge closed position, the discharge valve 53 closes the valve side opening 42c. Specifically, when the discharge valve 53 is in the discharge closed position, the surface of the discharge valve 53 is in contact with the valve facing surface 70a of the valve contact member 70. At this time, the surface of the discharge valve 53 is in continuous contact with the valve facing surface 70a around the valve side opening 42c.

[0072] In this first embodiment, the surface of the discharge valve 53 that contacts the valve-facing surface 70a is the lower surface of the discharge valve 53. That is, the detection device 60 is configured to face the lower surface of the discharge valve 53 and is capable of detecting the state up to the discharge valve 53.

[0073] When the discharge valve 53 overcomes its own weight and moves from the discharge closed position to the discharge open position, a gap is generated between the discharge valve 53 and the valve side opening 42c.

[0074] A discharge passage 140 is formed in the discharge section 40. The discharge passage 140 communicates with a plurality of discharge ports 43a from the discharge section inlet opening end 42a through the valve side opening 42c. When the discharge valve 53 is in the discharge closed position, the main channel 110 extending from the inside of the main section 10 is disconnected from the discharge passage 140.

[0075] When the discharge valve 53 moves from the discharge closed position to the discharge open position and a gap is generated between the discharge valve 53 and the valve side opening 42c, the main flow channel 110 extending from the interior of the main part 10 and the discharge flow channel 140 become connected.

[0076] Next, the operation of the vent valve 5 will be explained. The vent valve 5 operates based on the internal pressure of the tank 4.

[0077] First, with the internal pressure of tank 4 at normal pressure, inlet valve 23 and outlet valve 53 are both closed. In this state, air is not introduced through inlet valve 23 or discharged through outlet valve 53, thus maintaining the internal pressure inside tank 4.

[0078] Next, we will explain the case where the internal pressure of tank 4 is lower than the normal pressure. In the air intake section 20, because the internal pressure of the air intake section 20 is lower, the air intake valve 23 is pushed by the atmosphere. Therefore, the air intake valve 23 moves from the air intake closed position to the air intake open position.

[0079] Specifically, the intake valve 23 is pushed upwards by atmospheric pressure, overcoming its own weight. It should be noted that the combined weight of the intake valve 23 and the intake valve shaft 24 is set as the weight that is pushed upwards by atmospheric pressure when the internal pressure of the intake section 20 is below the normal pressure.

[0080] The air rises against its own weight by the intake valve 23, the intake passage 120 opens, and the atmosphere enters the interior of the intake section 20 from the intake port 21b.

[0081] On the other hand, due to the decrease in internal pressure of the discharge section 40, the discharge valve 53 moves downward due to its own weight, and the discharge valve 53 remains in the discharge closed position. That is, the valve side opening 42c remains closed by the discharge valve 53.

[0082] As a result, air flows in through the air inlet 21b, and soon the internal pressure of tank 4 increases. Due to its own weight, the air inlet valve 23 moves to the air inlet closed position, closing the air inlet 21b. In this way, with the help of the weight of the air inlet valve 23, the internal pressure of tank 4 becomes the set pressure.

[0083] Next, the situation where the internal pressure of tank 4 is higher than normal will be explained. In this case, due to the increased internal pressure in the air intake section 20, the air intake valve 23 remains in the closed position due to the internal pressure and its own weight. As a result, the air intake port 21b is closed.

[0084] In the discharge section 40, due to the increased internal pressure of the tank 4, main section 10, and air inlet section 20, the discharge valve 53 moves from the discharge closed position to the discharge open position. Specifically, the discharge valve 53 is pushed upwards by the fluid inside the tank 4, overcoming its own weight. It should be noted that the combined weight of the discharge valve 53 and the discharge valve shaft 54 ​​is set as the weight required to overcome its own weight when the internal pressure of the main section 10 reaches or exceeds the normal pressure.

[0085] Therefore, the fluid inside tank 4 is discharged from tank 4 through the main flow channel 110, the valve-side opening 42c, and the discharge flow channel 140. That is, when the internal pressure of the discharge section 40 is higher than the normal pressure, the discharge valve mechanism 50 connects the main flow channel 110 and the discharge flow channel 140. As a result, fluid can be discharged from the inside of tank 4 to the downstream side of the discharge valve mechanism 50 through the discharge flow channel 140.

[0086] When the fluid in tank 4 is discharged from the discharge section 40, soon the discharge valve 53 moves to the discharge closed position by its own weight based on the internal pressure of tank 4, closing the valve side opening 42c.

[0087] Therefore, the internal pressure of tank 4 becomes the pressure based on the weight of the discharge valve 53. In this way, the vent valve 5 can maintain the pressure inside tank 4 at a constant pressure.

[0088] It should be noted that, as mentioned above, the fluid inside tank 4 includes the fluid inside tank 4 and the fluid that evaporates and vaporizes. Therefore, the discharged fluid also includes the fluid inside tank 4 and the fluid that evaporates and vaporizes.

[0089] Next, the detection device 60 will be described. The detector 61 can continuously detect the state of the discharge valve 53 and the valve seat 2. The specified variable of the detection object detected by the detector 61 can be identified as the discharge valve 53, or the state between the discharge valve 53 and the valve seat 2, i.e., the state of the valve mechanism.

[0090] For example, as detector 61, a distance sensor capable of measuring the distance to the lower surface of discharge valve 53 can be used. In this case, the state of discharge valve 53 that can be detected based on the detection result of detector 61 is the current distance to discharge valve 53, that is, the current position of discharge valve 53.

[0091] It should be noted that detector 61 does not necessarily have to be a distance sensor. For example, a vibration sensor facing the lower surface of the discharge valve 53 can also be used as detector 61. This vibration sensor can also be used to measure the vibration of the valve-side end 42b facing the discharge valve 53. Even if a vibration sensor is used as detector 61, the state of the valve mechanism can be detected in the same way as a distance sensor.

[0092] Alternatively, an acoustic sensor facing the lower surface of the discharge valve 53 can be used as detector 61. This acoustic sensor can also measure the sound between the discharge valve 53 and the detector. Using the acoustic sensor, it is possible to detect the sound of the discharge valve 53 operating and the sound of the fluid between the valve seat and the discharge valve 53. Even when using an acoustic sensor as detector 61, the state of the valve mechanism can be detected in the same way as a distance sensor. In the above description, three types of sensors have been exemplified as detector 61, but the types of sensors used as detector 61 are not limited to these three; various sensors can be used as detector 61.

[0093] The machine learning device 240 is a device that operates as the main body of the learning phase of machine learning. Figure 6 It is shown Figure 1 A structural diagram of the machine learning device 240. Figure 7 It is shown Figure 6A conceptual diagram of the elements of machine learning implemented by the machine learning device 240. The machine learning device 240 includes a machine learning control unit 241, a machine learning communication unit 242, a learning data storage unit 243, and a machine learning model storage unit 244.

[0094] The machine learning control unit 241 generates a learning model 245, which obtains output information corresponding to the input information. The machine learning control unit 241 uses one or more learning data 246 to generate the learning model 245. The learning data 246 consists of learning input information 201a and learning valve mechanism state 201b corresponding to the learning input information 201a.

[0095] Here, the input information refers to the specified variables of the detected object detected by detector 61, showing the specified variables between the discharge valve 53 (which acts as the valve core) and valve seat 2 as detected by valve seat 2. The valve mechanism state representation includes the state of the valve mechanism, that is, the state of valve seat 2 and the discharge valve 53 (which acts as the valve core), as well as the state between the discharge valve 53 and valve seat 2. It should be noted that the valve mechanism consists of valve seat 2 or valve seat mechanism 1 and discharge valve 53 (which acts as the valve core).

[0096] The learning data 246 is used as teacher data in supervised learning, namely training data, validation data, and test data. Additionally, the learning valve mechanism state 201b is used as the positive solution label in supervised learning.

[0097] It should be noted that the learning input information 201a and the valve mechanism state corresponding to the learning input information 201a, i.e., the learning valve mechanism state 201b, are the states of the input information 200a and the valve mechanism state corresponding to the input information 200a, i.e., the valve mechanism state 200b, stored as learning data 246 in the learning data storage unit 243.

[0098] Input information 200a and valve mechanism status 200b are obtained through product testing such as various trials and tests on the vent valve 5. The learning data storage unit 243, which stores input information 200a, valve mechanism status 200b, learning input information 201a, learning valve mechanism status 201b, and learning data 246, will be described later.

[0099] The machine learning control unit 241 enables the learning model 245 to learn the correlation between the learning input information 201a and the learning valve mechanism state 201b in one or more learning data 246. Thus, the machine learning control unit 241 can generate the learned model 245.

[0100] The learning model 245 employs a neural network structure. The learning model 245 includes an input layer 245a, an intermediate layer 245b, and an output layer 245c. The input layer 245a has a number of neurons corresponding to the number of parameters in the learning input information 201a. The output layer 245c has a number of neurons corresponding to the number of conditions in the learning valve mechanism state 201b.

[0101] Between each layer, synapses (not shown) are laid, each connecting to a neuron. Each synapse is weighted.

[0102] The machine learning control unit 241 adjusts the weight parameter set, which consists of the weights of each synapse, through machine learning. The weight parameter set is reflected in the learning model 245.

[0103] In the machine learning control unit 241, the parameters of the learning input information 201a are input to each neuron of the input layer 245a. Through the learning model 245, the learning result valve mechanism state 201c is output. This learning result valve mechanism state 201c represents the state of the valve mechanism corresponding to the learning input information 201a. The machine learning control unit 241 studies the learning result valve mechanism state 201c and adjusts the weights of each synapse based on the study results.

[0104] When the learning model 245 is a regression model, the learning valve mechanism state 201b is output as a standardized value within a specified range (e.g., 0 to 1). Alternatively, when the learning model 245 is a classification model, the learning valve mechanism state 201b is output as a score (accuracy) for each level, also as a standardized value within a specified range (e.g., 0 to 1).

[0105] The learning data storage unit 243 can store multiple learning data 246 as a database. It should be noted that the specific structure of the database constituting the learning data storage unit 243 can be appropriately designed.

[0106] Input information 200a, obtained in advance through product testing, and the corresponding valve mechanism state 200b are input into the learning data storage unit 243. The input information 200a and valve mechanism state 200b input to the learning data storage unit 243 are stored as learning input information 201a and learning valve mechanism state 201b, respectively. Therefore, the learning valve mechanism state 201b corresponds to the learning input information 201a. Thus, one set of learning data 246 is stored.

[0107] The learning input information 201a and the learning valve mechanism state 201b are based on the input information 200a obtained by the detector 61 in the product test of the vent valve 5 and the valve mechanism state 200b actually observed and measured in the product test.

[0108] The aforementioned product tests are conducted during the design, trial production, and pre-shipment inspection of the vent valve 5. The specified variables of the test object, as detected by the detector 61, which have passed various product tests, are obtained as input information 200a.

[0109] Additionally, the valve mechanism state 200b is obtained by observing or measuring the amount of fluid passing between the discharge valve 53 and the valve seat 2, the presence or absence of accidental fluid leakage between the valve core and the valve seat, the amount of leakage, and any signs of accidental fluid leakage during product testing. During product testing, the vent valve 5 and the machine learning device 240 are connected in a manner capable of information exchange. Variables detected by the detector 61 during product testing are sequentially input as input information 200a to the learning data storage unit 243. Then, the valve mechanism state 200b corresponding to the input information 200a determined based on operator judgment is input to the learning data storage unit 243. It should be noted that the variables detected during product testing can also be pre-stored in a storage device (not shown). After determining the corresponding valve state, the operator inputs this information as input information 200a and the valve mechanism state 200b corresponding to the input information 200a to the learning data storage unit 243. In this case, the vent valve 5 and the machine learning device 240 may not necessarily be connected in a way that allows for information exchange.

[0110] During product testing, even when the discharge valve 53 is open to zero, meaning the flow path is closed by the discharge valve 53, the detector 61 still detects the specified variables of the test object. In other words, because the detector 61 continuously detects the variables of the test object during product testing, the timing data of these variables can be obtained.

[0111] In product testing, the opening degree of the discharge valve 53 and the flow rate of the fluid discharged from the discharge valve 53 are measured using measuring instruments. The amount of fluid passing between the discharge valve 53 and the valve seat 2 can be obtained based on the measurement results, which is used as valve mechanism state 200b.

[0112] In addition, the observation results of the operator who can determine the state of valve seat 2 and discharge valve 53 by observing the discharge valve 53 and valve seat 2 can be obtained as valve mechanism state 200b, for example, whether the state of valve seat 2 and discharge valve 53 is normal or abnormal, and what kind of abnormal state it is in the case of abnormal state.

[0113] Furthermore, when the state of the discharge valve 53 and the valve seat 2 is abnormal, the presence or absence of fluid passing between the discharge valve 53 and the valve seat 2 can be used as the presence or absence of accidental fluid leakage, and this presence or absence of accidental fluid leakage can be used as the valve mechanism state 200b.

[0114] Furthermore, when the state of the discharge valve 53 and the valve seat 2 is abnormal, the amount of fluid passing through the discharge valve 53 and the valve seat 2 can be taken as the amount of unexpected fluid leakage, and this amount of unexpected fluid leakage can be obtained as the valve mechanism state 200b.

[0115] Furthermore, when the states of the discharge valve 53 and valve seat 2 are abnormal and the amount of fluid passing through the discharge valve 53 and valve seat 2 is not zero, an unexpected fluid leakage warning can be obtained as valve mechanism state 200b. That is, when the states of the discharge valve 53 and valve seat 2 are abnormal and the amount of fluid passing through the discharge valve 53 and valve seat 2 is not zero, valve mechanism state 200b is obtained as an unexpected fluid leakage warning based on the input information of a time series over a certain period of time, relative to when the amount of fluid passing through the discharge valve 53 and valve seat 2 was zero before this.

[0116] Product testing will also include tests that reproduce the occurrence of failures. For example, there are product tests that assume a foreign object is trapped between valve seat 2 and discharge valve 53, or product tests that cause at least one of valve seat 2 and discharge valve 53 to corrode. Furthermore, product tests that simulate actual operating conditions and allow the valve mechanism to operate for extended periods can also be considered to confirm the condition of aging degradation and durability.

[0117] In such product testing, if the state of the discharge valve 53 and valve seat 2 is judged to be abnormal due to malfunctions or deterioration over time, the presence or absence of accidental fluid leakage, the amount of accidental fluid leakage, and signs of accidental fluid leakage can be obtained as the valve mechanism state 200b.

[0118] Furthermore, in addition to product testing, the valve mechanism status 200b can be obtained from the input information 200a acquired from the currently operating vent valve 5 and information that can be obtained through maintenance checks, etc. Alternatively, data generated through simulation can also be used.

[0119] In this way, the correlation between the input information 200a obtained in product testing and the valve mechanism state 200b is derived from the results of actual testing, which is appropriate. Therefore, it is preferable to use such data as learning data 246 for machine learning.

[0120] The variables used as input information 200a are specifically determined based on the type of sensor used as detector 61. Since detector 61 can employ distance sensors, vibration sensors, or acoustic sensors, a specified variable corresponding to the aforementioned sensor type can be obtained as input information 200a.

[0121] Valve mechanism status 200b can be represented using numerical values ​​and error codes. For example, if valve mechanism status 200b can be quantitatively obtained, such as flow rate or leakage, it can also be represented based on its specific quantity. For instance, under normal valve mechanism conditions, valve mechanism status 200b can be represented as 1, and under abnormal conditions, it can be represented as 0.

[0122] For example, when representing signs of an unexpected fluid leak, the case of having no signs of the leak at all can be represented as 0, and the case of being able to detect the signs of the leak can be represented as 1. Alternatively, the case of having no signs of an unexpected fluid leak can be set to 0, and the case of frequently detecting the signs of the leak and being able to detect the signs of the unexpected fluid leak with a high probability can be set to 100, using values ​​in between.

[0123] The machine learning control unit 241 can extract one or more learning data 246 from multiple learning data 246 stored in the learning data storage unit 243 for machine learning.

[0124] The machine learning model storage unit 244 is a database that stores the learned model 245 generated by the machine learning control unit 241 after learning, i.e., the adjusted weight parameter group.

[0125] The machine learning communication unit 242 is a communication interface unit. The machine learning communication unit 242 connects to external devices via network 270 and can send and receive various types of data. The learned model 245, stored in the machine learning model storage unit 244, is provided to the information processing device 250 via network 270, storage medium, etc.

[0126] It should be noted that, in Figure 6 In this context, the learning data storage unit 243 and the machine learning model storage unit 244 are represented as independent storage units, but they can also be composed of a single storage unit.

[0127] Figure 8 It is shown Figure 1 A schematic diagram of the information processing device 250. The information processing device 250 is a device that operates as the main body of the inference stage of machine learning. The information processing device 250 uses the learning model 245 generated by the machine learning device 240 to predict the new valve mechanism state 202b corresponding to the prediction input information 202a of the new input. The prediction input information 202a is a variable obtained by the detector 61 in the currently operating vent valve 5. That is, the information processing device 250 determines the new valve mechanism state 202b based on the prediction input information 202a in the currently operating vent valve 5.

[0128] The information processing device 250 includes an information processing control unit 251, an information processing storage unit 255, and an information processing communication unit 256. The information processing control unit 251 includes an information acquisition unit 252, an information prediction unit 253, and an output processing unit 254.

[0129] The information acquisition unit 252 acquires the data obtained by the detector 61 and output from the detection device 60 as prediction input information 202a. The prediction input information 202a is input information used to determine the new valve mechanism state. Here, the valve mechanism state corresponding to the prediction input information 202a is taken as the new valve mechanism state 202b.

[0130] The information prediction unit 253 inputs the prediction input information 202a obtained by the information acquisition unit 252 into the learning model 245, thereby predicting the new valve mechanism state 202b. The output processing unit 254 outputs the new valve mechanism state 202b predicted by the information prediction unit 253 to the information processing and communication unit 256.

[0131] The information prediction unit 253 can select one learning model 245 from the multiple learning models 245 stored in the information processing storage unit 255 for use.

[0132] The information processing and storage unit 255 is a database that stores the learned learning models 245 used by the information prediction unit 253. The information processing and storage unit 255 is capable of storing multiple learned learning models 245 that are input from the machine learning device 240.

[0133] The multiple learning models 245 are, for example, machine learning methods, the types of data contained in the input information 201a, the types of data contained in the valve mechanism state 201b, and other different learned models.

[0134] The information processing storage unit 255 can also be replaced by the storage unit of an external computer, such as a server computer or a cloud computer. In this case, the information prediction unit 253 can access the storage unit of the external computer to obtain the learning model 245.

[0135] The information processing and communication unit 256 can be communicatively connected to devices outside the valve system 200 via the network 270. The information processing and communication unit 256 is a communication interface unit for transmitting and receiving various types of data. The information processing and communication unit 256 can output the new valve mechanism status 202b output by the output processing unit 254 to itself.

[0136] Figure 9 This is a hardware structure diagram of computer 900. The machine learning device 240 and information processing device 250 of valve system 200 are composed of general-purpose or special-purpose computer 900.

[0137] The computer 900 includes a bus 910, a processor 912, a memory 914, an input device 916, an output device 917, a display device 918, a storage device 920, a communication interface unit 922, an external device interface unit 924, an input / output device interface unit 926, and a media input / output unit 928. It should be noted that the above-mentioned components may be appropriately omitted depending on the intended use of the computer 900.

[0138] The processor 912 consists of one or more arithmetic processing units (CPU (Central Processing Unit), MPU (Micro-processing unit), DSP (digital signal processor), GPU (Graphics Processing Unit), etc.) and operates as the control unit that oversees the entire computer 900.

[0139] The memory 914 stores various data and programs 930, and may be composed of volatile memory (DRAM, SRAM, etc.) that functions as main memory, non-volatile memory (ROM), flash memory, etc.

[0140] Input device 916, for example, consists of a keyboard, mouse, numeric keypad, electronic pen, etc., and functions as an input unit. Output device 917, for example, consists of a sound output device including voice output, a vibration device, etc., and functions as an output unit. Display device 918, for example, consists of a liquid crystal display, an organic EL display, electronic paper, a projector, etc., and functions as an output unit.

[0141] Input device 916 and display device 918 can be integrated as a single unit, similar to a touch panel display. Storage device 920, such as HDD (Hard Disk Drive) or SSD (Solid State Drive), functions as a storage unit. Storage device 920 stores various data required for the operating system and program 930 to execute.

[0142] The communication interface unit 922 connects to a network 940, such as the Internet or an intranet, via wired or wireless means, and functions as a communication unit for sending and receiving data with other computers according to prescribed communication standards. Network 940 may be the same as network 270.

[0143] The external device interface unit 924 connects to external devices 950 such as cameras, printers, scanners, and readers via wired or wireless means, and functions as a communication unit that transmits and receives data between the external devices 950 and the external devices 950 in accordance with the prescribed communication standards.

[0144] The input / output device interface section 926 is connected to various input / output devices 960 such as sensors and actuators, and functions as a communication section for transmitting and receiving various signals and data, such as sensor detection signals and actuator control signals, between itself and the input / output devices 960.

[0145] The media input / output unit 928 is composed of a drive device such as a DVD drive or a CD drive, and reads and writes data to the media 970, which is a storage medium such as a DVD or CD.

[0146] In the computer 900 with the above structure, the processor 912 loads the program 930 stored in the storage device 920 into the memory 914 for execution, and controls various parts of the computer 900 via the bus 910. It should be noted that the program 930 can also be stored in the memory 914 instead of the storage device 920.

[0147] Program 930 can be recorded in media 970 in an installable or executable file format and provided to computer 900 via media input / output unit 928. Program 930 can also be provided to computer 900 by downloading via network 940 through communication interface unit 922.

[0148] In addition, the computer 900 can also implement various functions that are achieved by the processor 912 executing the program 930 through hardware such as FPGA (Field Programmable Gate Array) and ASIC (Application-Specific Integrated Circuit).

[0149] Computer 900 may be a fixed-type computer or a portable computer, and is any form of electronic device. Computer 900 may be a client computer, a server computer, or a cloud computer. Computer 900 may also be used in devices other than the machine learning device 240 and the information processing device 250 of valve system 200.

[0150] Next, we will explain the machine learning methods. Figure 10 It is shown Figure 1 A flowchart of the machine learning method of the machine learning device 240.

[0151] First, the operator inputs one or more sets of learning data 246 in advance and stores them in the learning data storage unit 243. The number of sets of learning data 246 stored is determined by considering the inference accuracy required by the final learning model 245. It should be noted that it is preferable to store multiple sets of learning data 246 in advance.

[0152] In the machine learning method, a learning model preparation process is performed as step S100. The machine learning control unit 241 prepares a learning model 245 before learning. In the prepared learning model 245, the weights of each synapse are set to initial values.

[0153] Next, a machine learning process is performed as step S110. In the machine learning process, first, a learning data acquisition process is performed as step S111. The machine learning control unit 241 randomly acquires one piece of learning data 246 from a plurality of learning data 246 stored in the learning data storage unit 243.

[0154] Next, the inference result output process, which is step S112, is implemented. The machine learning control unit 241 inputs the learning input information 201a contained in the acquired learning data 246 into the input layer 245a of the prepared learning model 245. As a result, the learning result valve mechanism state 201c is output from the output layer 245c of the learning model 245 as the inference result.

[0155] The learned valve mechanism state 201c, which is the output of the inference result, is generated by the learning model 245 before or during learning. Therefore, the learned valve mechanism state 201c is different from the learned valve mechanism state 201b, which is the positive solution label contained in the learning data 246.

[0156] Next, a weight adjustment process is performed as step S113. The machine learning control unit 241 compares the learning valve mechanism state 201b in the learning data 246 obtained in step S111, which serves as the positive solution label, with the learning result valve mechanism state 201c, which is output as the inference result in step S112. Based on this comparison, the machine learning control unit 241 performs a process of adjusting the weights of each synapse, i.e., backpropagation, and performs machine learning.

[0157] Therefore, the machine learning control unit 241 enables the learning model 245 to learn the correlation between the learning input information 201a and the learning valve mechanism state 201b.

[0158] Next, a machine learning termination determination process is performed as step S114. The machine learning control unit 241 determines whether the prescribed learning termination conditions are met. This determination is performed, for example, based on the evaluation value of the error function of the learning valve mechanism state 201b (which serves as the positive solution label) and the learning result valve mechanism state 201c, and the remaining number of unlearned learning data 246 stored in the learning data storage unit 243.

[0159] In step S114, when the machine learning control unit 241 determines that the learning termination condition is not met and continues machine learning, i.e., when the result in step S114 is "No", the process returns to step S111. In this way, for the learning model 245 that is learning, the processes from step S111 to step S114 are performed multiple times on the unlearned learning data 246.

[0160] On the other hand, in step S114, when the machine learning control unit 241 determines that the learning termination condition is met and ends the machine learning, that is, when it is "yes" in step S114, the process proceeds to step S120.

[0161] Then, the learning model storage process is performed as step S120. The machine learning control unit 241 stores the learned model 245, which has adjusted the weights of each synapse, i.e., the learned model 245 reflecting the adjusted weight parameter group, in the machine learning model storage unit 244. Thus, the machine learning method ends.

[0162] Next, the prediction method for the new valve mechanism state 202b of the prediction valve system 200 of the information processing device 250 will be explained. Figure 11 It is shown Figure 1 The flowchart of the learning valve mechanism state prediction method of the information processing device 250.

[0163] The valve mechanism state prediction method, which is used as an information processing method, is a method for determining the valve mechanism state that represents the state of the valve mechanism. It can obtain the valve mechanism state based on the specified variables, i.e., the input information, between the discharge valve 53, which is the valve core, and the valve seat 2.

[0164] First, the process of acquiring prediction input information is performed as step S200. By inputting the data of the variables acquired by the detection device 60 as prediction input information 202a into the information processing device 250, the information acquisition unit 252 acquires the prediction input information 202a.

[0165] Next, a prediction process is performed as step S210. The information prediction unit 253 inputs the prediction input information 202a obtained in step S200 into the learning model 245. As a result, the information prediction unit 253 predicts the new valve mechanism state 202b corresponding to the prediction input information 202a.

[0166] Next, the output processing step is performed as step S220. The output processing unit 254 displays the new valve mechanism state 202b generated in step S210 on a display screen (not shown) or the like, as output processing. Thus, the prediction and output of the new valve mechanism state 202b are completed. It should be noted that the new valve mechanism state 202b can also be sent to an email address specified in an email or similar document.

[0167] In addition, since the specified variable is always output from the detection device 60, a process for obtaining the input information for prediction is then performed.

[0168] This disclosure can also be provided as a machine learning program, which is a program for enabling the computer 900 to function as the various parts of the machine learning device 240, and a program for enabling the computer 900 to perform the various steps of the machine learning method.

[0169] In addition, this disclosure can also be provided as a valve mechanism status output program, which is a program that enables the computer 900 to function as the various parts of the information processing device 250, and a program that enables the computer 900 to execute the various steps of the information processing method of the first embodiment.

[0170] Furthermore, this disclosure can be provided not only as the information processing apparatus 250, information processing method, or information processing program of the first embodiment, but also as a reasoning apparatus, reasoning method, or reasoning program for reasoning valve mechanism states. In this case, the reasoning apparatus, reasoning method, or reasoning program includes a memory 914 and a processor 912, wherein the processor 912 is capable of performing a series of processes.

[0171] This series of processes includes: information acquisition processing (information acquisition step), in which input information is acquired; reasoning processing (reasoning step), in which the learning model 245 stored in memory 914 is used to reason about the valve mechanism state; and output processing (output step), in which the reasoned valve mechanism state is output to external device 950. It should be noted that in the output processing, the valve mechanism state can also be output to network 940 via communication interface unit 922.

[0172] The information processing device 250 of the first embodiment predicts the state of the valve mechanism and can determine the valve mechanism state based on the input information, which is a predetermined variable between the discharge valve 53 (which is the valve core) and the valve seat 2. This allows for verification of the amount of fluid passing through the valve mechanism, verification of whether there is any accidental leakage of fluid between the valve core and the valve seat, verification of the amount of leakage, and verification of any signs of accidental leakage of fluid from the valve mechanism.

[0173] The information processing apparatus 250 of the first embodiment includes one or more detectors 61 capable of detecting a predetermined variable of the object to be detected in the detection direction. Furthermore, the valve seat 2 forms part of a flow channel, and a valve seat end face 2a is formed on the valve seat 2 to close the flow channel by connecting with a discharge valve 53, which serves as a valve core. Additionally, one or more detector placement spaces 42e are formed on the valve seat 2, and a detector 61 is installed in each detector placement space 42e. With the detector 61 installed in the detector placement space 42e, the detection direction is towards the discharge valve 53, which serves as a valve core, located across the valve seat end face 2a. Therefore, a predetermined variable between the valve seat 2 and the discharge valve 53, which serves as a valve core, can be detected by the detector 61. Thus, it is possible to more accurately verify the amount of fluid passing through the valve mechanism, verify whether there is any accidental leakage of fluid between the valve core and the valve seat, verify the amount of leakage, and verify the signs of accidental leakage of fluid from the valve mechanism.

[0174] The information processing apparatus 250 of the first embodiment uses at least one of the more than one detectors 61, which employs a distance sensor, a vibration sensor, and an acoustic sensor. While various types of sensors can be used, the detector 61 can utilize a commonly available and readily commercially available sensor by employing any of the distance sensor, vibration sensor, and acoustic sensor. Therefore, the design, manufacture, and maintenance of the valve seat mechanism 1 can be easily addressed.

[0175] Using the information processing device 250 of the first embodiment, the valve mechanism status includes at least one of the following: the flow rate of fluid flowing between the discharge valve 53 (which serves as the valve core) and the valve seat 2; whether there is any accidental leakage of fluid between the discharge valve 53 (which serves as the valve core) and the valve seat 2; the amount of leakage of fluid from the discharge valve 53 (which serves as the valve core) and the valve seat 2; and signs of accidental leakage of fluid from the discharge valve 53 (which serves as the valve core) and the valve seat 2. Therefore, based on the output of the information processing device 250, the valve mechanism status can be easily and quickly verified.

[0176] The information processing apparatus 250 of the first embodiment, when input information is received, uses a learning model that has learned the correlation between the input information and the valve mechanism state through machine learning to determine the valve mechanism state and outputs the determined valve mechanism state. Therefore, machine learning methods based on accumulated information can be used to verify the valve mechanism state. Consequently, it is possible to more accurately verify the amount of fluid passing through the valve mechanism, verify the presence or absence of accidental fluid leakage between the valve core and valve seat, verify the amount of leakage, and verify any signs of accidental fluid leakage from the valve mechanism.

[0177] The machine learning apparatus 240 of the first embodiment generates a learning model 245 for reasoning about the state of the valve mechanism. Additionally, a learning data storage unit 243 is provided to store multiple sets of learning data 246, which consists of input information (a defined variable between the discharge valve 53, which is the valve core, and the valve seat 2) and the valve mechanism state corresponding to the input information. A machine learning model storage unit is also provided, which, by being fed multiple sets of learning data 246, enables the learning model to learn the correlation between the input information and the valve mechanism state, and stores the learned learning model 245. Thus, a learning model 245 that has learned the correlation between the input information and the valve mechanism information can be generated and stored in advance. Therefore, the learned learning model 245 can be used to more accurately verify the amount of fluid passing through the valve mechanism, verify whether there is any accidental leakage of fluid between the valve core and the valve seat, verify the amount of leakage, and verify the signs of accidental leakage of fluid from the valve mechanism.

[0178] The inference apparatus employing the first embodiment includes a memory 914 and at least one processor 912, which infers the valve mechanism state representing the state of the valve mechanism. Furthermore, the at least one processor 912 executes: an information acquisition process, in which predetermined variables, i.e., input information, between the discharge valve 53 (which serves as the valve core) and the valve seat 2; an inference process, in which, upon acquiring the input information, the valve mechanism state is inferred using a machine learning-based learning model 245 stored in the memory 914; and an output process, in which the inferred valve mechanism state is output. Thus, it is possible to verify the amount of fluid passing through the valve mechanism, verify whether there is any accidental leakage of fluid between the valve core and the valve seat, verify the amount of leakage, and verify any signs of accidental leakage of fluid from the valve mechanism.

[0179] The reasoning method of the first embodiment is executed by a reasoning device equipped with a memory 914 and at least one processor 912 to reason about the state of the valve mechanism. In addition, at least one processor 912 executes an information acquisition step, in which it acquires input information, a predetermined variable between the discharge valve 53 (which is the valve core) and the valve seat 2. Furthermore, a reasoning step is executed, in which, upon acquiring the input information, a machine learning-based learning model 245 stored in the memory 914 is used to reason about the valve mechanism state. Finally, an output step is executed, in which the reasoned valve mechanism state is output. Thus, it is possible to verify the amount of fluid passing through the valve mechanism, verify whether there is any accidental leakage of fluid between the valve core and the valve seat, verify the amount of leakage, and verify any signs of accidental leakage of fluid from the valve mechanism.

[0180] Using the information processing method of the first embodiment, the valve mechanism state, representing the state of the valve mechanism, is determined. Furthermore, the valve mechanism state is obtained based on the input information, which is a predetermined variable between the discharge valve 53 (which is the valve core) and the valve seat 2. This allows for verification of the amount of fluid passing through the valve mechanism, verification of whether there is any accidental fluid leakage between the valve core and the valve seat, verification of the amount of leakage, and verification of any signs of accidental fluid leakage from the valve mechanism.

[0181] Using the machine learning method of the first embodiment, a learning model 245 is generated for reasoning about the valve mechanism state, which represents the state of the valve mechanism. Furthermore, a machine learning process is executed, in which, with one or more sets of learning data 246 consisting of input information (defined variables between the discharge valve 53 (valve core) and valve seat 2) and the valve mechanism state corresponding to the input information stored, the learning model learns the correlation between the input information and the valve mechanism state by inputting multiple sets of learning data 246. A learning-completed model storage process is then executed, in which the learning model 245 learned in the machine learning process is stored in the machine learning model storage unit 244. This enables verification of the amount of fluid passing through the valve mechanism, verification of whether there is accidental fluid leakage between the valve core and valve seat, verification of the amount of leakage, and verification of signs of accidental fluid leakage from the valve mechanism.

[0182] It should be noted that in the first embodiment, a neural network-based model is used as the learning model 245. However, it is not limited to this. Other machine learning models can also be used as the learning model 245. Examples of other machine learning models include tree-based models such as decision trees and regression trees, ensemble learning methods such as Bagging and Boosting, neural networks such as recurrent neural networks, convolutional neural networks, and neural networks such as LSTM (Long Short Term Memory) (including deep learning), hierarchical clustering, non-hierarchical clustering, clustering methods such as k-nearest neighbors and k-means, multivariate analysis such as principal component analysis, factor analysis, and logistic regression, and support vector machines.

[0183] Second Implementation Method

[0184] As an illustration of the valve system 200 of this disclosure, the valve system 200 used in the vent valve 5 is also used in the second embodiment. However, the valve system 200 of this disclosure is not limited to the vent valve 5 and can be suitably used in general valve mechanisms.

[0185] The information processing device 250 of the second embodiment differs from the valve system 200 of the first embodiment in that it does not use a learning model obtained through machine learning to output a new valve mechanism state 202b.

[0186] The information processing device 250 is able to output a new valve mechanism state 202b corresponding to the obtained predictive input information 202a.

[0187] The information prediction unit 253 receives prediction input information 202a obtained by the information acquisition unit 252 and determines a new valve mechanism state 202b using a prescribed decision method. Since machine learning methods are not used in this determination method, no learning model is employed. The output processing unit 254 outputs the determined new valve mechanism state 202b to the information processing and communication unit 256.

[0188] In the information processing apparatus 250, the prediction input information 202a acquired by the information acquisition unit 252 can also be sequentially stored in the information processing storage unit 255. Therefore, the stored prediction input information 202a can be processed as time-series data. In the first embodiment, the information processing storage unit 255 functions as a database for the learning model 245, but in the second embodiment, the information processing storage unit 255 functions as a storage device capable of storing information.

[0189] Among the decision-making methods implemented in the Information Prediction Department 253, there are methods based on comparison with stored waveform data, methods based on thresholds to determine whether something is normal or abnormal, methods that calculate physical quantities by using coefficients, and other known statistical analysis calculations.

[0190] The information processing and storage unit 255 pre-stores waveform data, thresholds, or coefficients, etc., required for the decision-making method implemented by the information prediction unit 253. The stored waveform data may include, for example, time-series data of a specified variable detected by the detector 61 that could be a precursor to an unexpected fluid leak.

[0191] The information prediction unit 253 can use the waveform data to implement a prescribed decision method. For example, it can determine the corresponding valve mechanism state based on whether the waveform data or the characteristics of the waveform data stored as described above are consistent with the waveform data or the characteristics of the waveform data based on the prediction input information 202a. If they are consistent, it can determine a new valve mechanism state 202b, such as a sign of unexpected fluid leakage.

[0192] In addition, the stored thresholds include setting the state where the discharge valve 53 opens the flow channel for a certain period of time as an abnormal state and storing that time as a threshold, or setting the state where the micro-action of the discharge valve 53 continues to occur for a certain period of time as an abnormal state and storing that time as a threshold.

[0193] The information prediction unit 253 can use this threshold to implement a prescribed decision method. For example, based on the prediction input information 202a, it can determine whether the opening of the flow channel of the discharge valve 53 or the micro-movement of the discharge valve 53 has lasted for a time exceeding the threshold. If the time exceeds the threshold, a new valve mechanism state 202b, such as a valve mechanism malfunction, can be determined.

[0194] In addition, the stored coefficients include those required for calculating physical quantities. For example, these are coefficients needed to calculate the flow rate of the fluid flowing out from between the discharge valve 53 and the valve seat 2.

[0195] The information prediction unit 253 can use this coefficient to implement a prescribed decision method. For example, it can use the stored coefficient to calculate and determine the flow rate. The calculated flow rate can be used as the flow rate of the fluid flowing out of the valve mechanism to determine the new valve mechanism state 202b.

[0196] The information processing control unit 251 can use pre-stored waveform data, thresholds, and coefficients to determine the new valve mechanism state 202b based on the predicted input information 202a. Furthermore, in this second embodiment, the pre-stored waveform data, thresholds, and coefficients are not limited to those described above and can be appropriately selected. Additionally, to determine the new valve mechanism state 202b, the determination method is not limited to those described above; known arithmetic and statistical methods can be used.

[0197] In the second embodiment, since machine learning is not used, the structures associated with the learning model 245, machine learning device 240, and inference device in the first embodiment are not used. The other structures in the second embodiment are the same as those in the first embodiment, and therefore descriptions are omitted.

[0198] The information processing apparatus 250 of the second embodiment, when input information is input, determines the valve mechanism state based on a predetermined decision method that does not use machine learning methods, and outputs the determined valve mechanism state. Therefore, the state of the currently configured valve mechanism can be verified based on past insights into the valve mechanism. Consequently, it is possible to more accurately verify the amount of fluid passing through the valve mechanism, verify the presence or absence of accidental fluid leakage between the valve core and valve seat, verify the amount of leakage, and verify any signs of accidental fluid leakage from the valve mechanism.

[0199] Third Implementation Method

[0200] As an illustration of the valve system 200 of this disclosure, the valve system 200 used in the vent valve 5 is also used in the third embodiment. However, the valve system 200 of this disclosure is not limited to the vent valve 5 and can be suitably used in general valve mechanisms.

[0201] In the third embodiment, the difference from the first or second embodiment is that three detectors 61 are provided on the valve seat 2. Figure 12 This is a top view showing the inlet section 42 of the third embodiment.

[0202] The detection device 60 has three detectors 61, an input / output device (not shown), and wiring 62 for transmitting signals to each detector 61 to the input / output device. Three detector mounting spaces 42e are formed in the inlet section 42. One detector 61 and wiring 62 connected to the detector 61 are provided in one detector mounting space 42e.

[0203] The valve contact member 70 is formed in a shape corresponding to the shape of the main body end face 42d. By providing the valve contact member 70 to the inlet portion 42, the openings of the plurality of detector placement spaces 42e formed on the main body end face 42d are closed. That is, a plurality of detectors 61 are disposed in the detector placement spaces 42e covered by the valve contact member 70. The plurality of detectors 61 are respectively disposed along the valve facing surface 70a. It should be noted that, as... Figure 12 As shown, the inner peripheral surface of the portion of the inlet 42 where each detector mounting space 42e is formed protrudes further inward, and the wall thickness of the portion of the inlet 42 where the detector mounting space 42e is formed is thicker than the wall thickness of the other portions. Furthermore, the valve contact member 70 is formed in a shape corresponding to the shape of the main body end face 42d. However, this is not a limitation. For example, the wall thickness of the inlet 42 may be uniform, that is, no inwardly protruding portion may be formed on the inner peripheral surface of the inlet 42, and the inner peripheral surface of the inlet 42 viewed axially may also be circular. Even in this case, the valve contact member 70 is formed in a shape corresponding to the shape of the main body end face 42d.

[0204] The multiple detectors 61 are all sensors of the same type; for example, all three detectors 61 can be set as distance sensors. Other structures in this third embodiment are the same as those disclosed in the first embodiment, and therefore descriptions are omitted.

[0205] It should be noted that in this third embodiment, three detectors 61 are provided in the inlet section 42. However, this is not a limitation. The number of detectors 61 provided in the inlet section 42 can also be two or more, i.e., multiple. Alternatively, one or more detectors 61 can be provided in a single detector placement space 42e. In this third embodiment, multiple detectors 61 of the valve seat mechanism 1 are provided along the valve seat end face 2a. As a result, minute tilts, vibrations, etc., of the discharge valve 53 can be studied in more detail. Therefore, the movement of the valve core can be observed in more detail.

[0206] In the third embodiment, the multiple detectors 61 of the valve seat mechanism 1 are of a single type of sensor. Therefore, it is easy to compare and study the signals obtained from each detector 61. Consequently, minute tilts, vibrations, etc., of the discharge valve 53 can be detected easily and in more detail.

[0207] It should be noted that the three detectors 61 in the third embodiment are all sensors of the same type. However, this is not a limitation. For example, the three detectors 61 may use more than one distance sensor and more than one vibration sensor. That is, multiple sensors can be used as multiple detectors 61.

[0208] In the valve seat mechanism 1 of the third embodiment, at least two detectors 61 are sensors of different types. As a result, the operation of the discharge valve 53 can be studied from multiple angles.

[0209] Furthermore, the valve system 200 and valve seat mechanism 1 of the first to third embodiments are provided in the vent valve 5. However, it is not limited to this. Any valve system 200 and valve seat mechanism 1 of the first to third embodiments can be provided as long as it closes the flow passage by closing the valve seat end face 2a of the valve seat 2 with the valve component and opens the flow passage by the gap between the vent valve component and the valve seat 2. Examples of such valve mechanisms include rotary valves such as ball valves and diaphragm valves.

[0210] Furthermore, in the valve system 200 of the first to third embodiments, a predetermined variable detected by the detector 61, which is installed in the detector installation space 42e formed in the valve seat 2, is used as input information. However, it is not limited to this. The detector 61 can be installed anywhere in the valve seat 2 or the valve mechanism.

[0211] (Other implementation methods)

[0212] This disclosure is not limited to the above-described embodiments, and various modifications can be made to implement it without departing from the spirit of this disclosure. Furthermore, all of these modifications are encompassed within the technical concept of this disclosure.

[0213] The following is a summary of various aspects of this disclosure as an appendix.

[0214] (Note 1)

[0215] An information processing device that predicts the state of a valve mechanism, wherein,

[0216] Based on the specified variables between the valve core and the valve seat, i.e., the input information, the valve mechanism state representing the state of the valve mechanism is obtained.

[0217] (Note 2)

[0218] According to the information processing apparatus described in Appendix 1, wherein,

[0219] A detector having one or more detectors that detect the specified variables of the object to be detected, oriented in the detection direction.

[0220] The valve seat forms part of the flow channel.

[0221] The valve seat has a valve seat end face that closes the flow passage by connecting with the valve core.

[0222] The valve seat has one or more space for installing detectors.

[0223] The detectors are provided in each of the detector placement spaces.

[0224] With the detector positioned in the detector placement space, the detection direction is toward the valve core, which is located across the valve seat end face.

[0225] (Note 3)

[0226] According to the information processing apparatus described in Appendix 2, wherein,

[0227] At least one of the more than one detectors uses any one of a distance sensor, a vibration sensor, and an acoustic sensor.

[0228] (Note 4)

[0229] The information processing apparatus according to any one of Annexes 1 to 3, wherein,

[0230] The valve mechanism status includes at least one of the following: the flow rate of fluid flowing between the valve core and the valve seat; whether there is any accidental leakage of fluid between the valve core and the valve seat; the amount of accidental leakage of fluid between the valve core and the valve seat; and any signs of accidental leakage of fluid between the valve core and the valve seat.

[0231] (Note 5)

[0232] The information processing apparatus according to any one of Annexes 1 to 4, wherein,

[0233] When the input information is input, a learning model that has learned the correlation between the input information and the valve mechanism state through machine learning is used to determine the valve mechanism state and output the determined valve mechanism state.

[0234] (Note 6)

[0235] A machine learning apparatus generates a learning model for reasoning about the states of a valve mechanism, wherein it comprises:

[0236] The learning data storage unit stores multiple sets of learning data consisting of predetermined variables between the valve core and the valve seat, i.e., input information, and the valve mechanism state corresponding to the input information; and

[0237] The machine learning model storage unit is equipped with a learning model that learns the correlation between the input information and the state of the valve mechanism by being input with multiple sets of the learning data, and stores the learned learning model.

[0238] (Note 7)

[0239] A reasoning device comprising a memory and at least one processor, for reasoning about the state of a valve mechanism, wherein,

[0240] The at least one processor performs:

[0241] Information acquisition and processing: In this process, the specified variables between the valve core and the valve seat, i.e., the input information, are obtained.

[0242] The reasoning process involves, upon receiving the input information, using a machine learning-based learning model stored in the memory to infer the state of the valve mechanism.

[0243] Output processing, in which the inferred valve mechanism state is output.

[0244] (Postscript 8)

[0245] An information processing method is used to determine a valve mechanism state representing the state of the valve mechanism, wherein,

[0246] Determine the valve mechanism state corresponding to the specified variable between the valve core and the valve seat, i.e., the input information.

[0247] (Note 9)

[0248] A machine learning method generates a learning model for reasoning about the states of a valve mechanism, wherein...

[0249] In a state where there is one or more learning data consisting of input information, defined variables between the valve core and valve seat, and the valve mechanism state corresponding to the input information, execute:

[0250] A machine learning process, in which a learning model learns the correlation between the input information and the state of the valve mechanism by being fed the learning data; and

[0251] After the model storage process is completed, the learned model learned in the machine learning process is stored in the machine learning model storage unit.

[0252] (Postscript 10)

[0253] A reasoning method, executed by a reasoning device having a memory and at least one processor, infers the state of a valve mechanism representing the state of the valve mechanism, wherein...

[0254] The at least one processor performs:

[0255] The information acquisition process involves acquiring the specified variables, i.e., the input information, between the valve core and the valve seat.

[0256] The reasoning process involves, upon receiving the input information, using a machine learning-based learning model stored in the memory to infer the state of the valve mechanism.

[0257] The output process involves outputting the deduced valve mechanism state.

[0258] Explanation of reference numerals in the attached figures

[0259] 1: Valve seat mechanism; 2: Valve seat; 2a: Valve seat end face; 3: Valve seat body; 4: Tank; 4a: Mounting part; 4b: Upstream flange; 5: Vent valve; 10: Main part; 10a: Inlet opening end; 10b: Outlet opening end; 10c: Inlet opening end; 20: Inlet part; 21: Inlet part body; 21a: Connecting side opening end; 21b: Inlet port; 21x: Inlet valve seat; 22: Inlet valve mechanism; 23: Inlet valve (valve core); 24: Inlet valve shaft; 25: Inlet valve guide; 40: Discharge part; 41: Discharge part body; 42: Inlet part; 42a: Discharge part inlet opening end ; 42b: Valve side end; 42c: Valve side opening; 42d: Main body end face; 42e: Detector mounting space; 43: Atmospheric exhaust section; 43a: Exhaust port; 43b: Valve mechanism side opening end; 47: Exhaust cover; 50: Exhaust valve mechanism; 53: Exhaust valve (valve core); 54: Exhaust valve shaft; 55: Exhaust valve guide; 60: Detection device; 61: Detector; 62: Wiring; 70: Valve contact component; 110: Main flow channel; 111: Branch flow channel; 120: Inlet flow channel; 140: Exhaust flow channel; 200: Valve system; 200a: Input information; 200b: Valve mechanism status; 201 201a: Input information for learning; 201b: Valve mechanism state for learning; 201c: Valve mechanism state as a learning result; 202a: Input information for prediction; 202b: New valve mechanism state; 240: Machine learning device; 241: Machine learning control unit; 242: Machine learning communication unit; 243: Data storage unit for learning; 244: Machine learning model storage unit; 245: Learning model; 245a: Input layer; 245b: Intermediate layer; 245c: Output layer; 246: Data for learning; 250: Information processing device; 251: Information processing control unit; 252: Information acquisition unit; 253: Information prediction unit. 254: Output processing unit; 255: Information processing and storage unit; 256: Information processing and communication unit; 270: Network; 900: Computer; 910: Bus; 912: Processor; 914: Memory; 916: Input device; 917: Output device; 918: Display device; 920: Storage device; 922: Communication interface unit; 924: External device interface unit; 926: Input / output device interface unit; 928: Media input / output unit; 930: Program; 940: Network; 950: External device; 960: Input / output device; 970: Media (non-temporary storage medium).

Claims

1. An information processing device for predicting the state of a valve mechanism, characterized in that, Based on the specified variables between the valve core and the valve seat, i.e., the input information, the valve mechanism state representing the state of the valve mechanism is obtained.

2. The information processing device according to claim 1, characterized in that, It has one or more detectors capable of detecting the specified variables of the object to be detected in the direction of detection. The valve seat forms part of the flow channel. The valve seat has a valve seat end face that closes the flow passage by connecting with the valve core. The valve seat has one or more space for installing detectors. The detectors are provided in each of the detector placement spaces. With the detector set in the detector setting space, the detection direction is toward the valve core that exists across the valve seat end face.

3. The information processing device according to claim 2, characterized in that, At least one of the more than one detectors uses any one of a distance sensor, a vibration sensor, and an acoustic sensor.

4. The information processing apparatus according to claim 1, characterized in that, The valve mechanism status includes at least one of the following: the flow rate of fluid flowing between the valve core and the valve seat; whether there is any accidental leakage of fluid between the valve core and the valve seat; the amount of accidental leakage of fluid between the valve core and the valve seat; and any signs of accidental leakage of fluid between the valve core and the valve seat.

5. The information processing apparatus according to claim 1, characterized in that, When the input information is input, a learning model that has learned the correlation between the input information and the valve mechanism state through machine learning is used to determine the valve mechanism state and output the determined valve mechanism state.

6. A machine learning device that generates a learning model for inferring a valve mechanism state representing a state of a valve mechanism, characterized by, have: The learning data storage unit stores multiple sets of learning data consisting of predetermined variables between the valve core and the valve seat, i.e., input information, and the valve mechanism state corresponding to the input information; and The machine learning model storage unit is equipped with a learning model that learns the correlation between the input information and the state of the valve mechanism by being input with multiple sets of the learning data, and stores the learned learning model.

7. A reasoning device comprising a memory and at least one processor, for reasoning about a valve mechanism state representing the state of a valve mechanism, characterized in that, The at least one processor performs: Information acquisition and processing: In this process, the specified variables between the valve core and the valve seat, i.e., the input information, are obtained. Inference processing, in which, upon receiving the input information, a machine learning-based learning model stored in the memory is used to infer the state of the valve mechanism; and Output processing, in which the deduced valve mechanism state is output.

8. An information processing method for determining a valve mechanism state representing the state of a valve mechanism, characterized in that, Determine the valve mechanism state corresponding to the specified variable between the valve core and the valve seat, i.e., the input information.

9. A machine learning method for generating a learning model for reasoning about the states of a valve mechanism, characterized in that, In a state where there is one or more learning data consisting of input information, defined variables between the valve core and valve seat, and the valve mechanism state corresponding to the input information, execute: A machine learning process in which a learning model learns the correlation between the input information and the state of the valve mechanism by being fed the learning data. and After the model storage process is completed, the learned model learned in the machine learning process is stored in the machine learning model storage unit.

10. A reasoning method, executed by a reasoning device having a memory and at least one processor, for reasoning about a valve mechanism state representing the state of a valve mechanism, characterized in that, The at least one processor performs: The information acquisition process involves acquiring the specified variables, i.e., the input information, between the valve core and the valve seat. The reasoning process involves, upon receiving the input information, using a machine learning-based learning model stored in the memory to infer the state of the valve mechanism. The output process involves outputting the deduced valve mechanism state.

Images