Notification device, notification system, learning device, inference device, and refrigeration cycle device

Abstract

This notification device (50) comprises: an electronic component (12) to which power is supplied; a power supply unit (13) that supplies power to the electronic component (12); a detector (11) that detects a physical quantity correlated with the voltage or current applied to the electronic component (12); a diagnosing unit (15) that diagnoses, on the basis of the physical quantity, whether the electronic component (12) or a movable component (26) driven by the electronic component (12) has failed or degraded; and a notification unit (16) that notifies the user or the repair person of the result of the diagnosis performed by the diagnosing unit (15).

Description

Notification device, notification system, learning device, inference device, and refrigeration cycle device

[0001] The present disclosure relates to a notification device, a notification system, a learning device, an inference device, and a refrigeration cycle device that notify of failure or deterioration of an electronic component to which power is supplied or a moving component driven by an electronic component.

[0002] There are a wide variety of electronic components to which power is supplied. Examples of electronic components include coils and capacitors. There are also a wide variety of moving parts driven by these electronic components. Examples of moving parts are four-way valves and electronic expansion valves used in refrigerant circuits. Refrigerant circuits are installed in air conditioners, freezers, and other refrigeration cycle devices. Other examples of moving parts are motors that drive actuators, relays, and the like. Because these moving parts physically move, they are more likely to fail or deteriorate than parts that do not physically move.

[0003] A four-way valve is a component that switches the direction of refrigerant flow in the refrigerant circuit when switching between cooling and heating in the refrigeration cycle. Because this four-way valve is a component that slides and moves, it is prone to failure due to initial defects or deterioration over time. Therefore, if the four-way valve fails, it will directly lead to problems such as being unable to switch between cooling and heating.

[0004] Furthermore, for example, if there is a problem with an electronic component that controls a moving part, it is merely a malfunction, but if, for example, a switching element disposed in a path through which a load current flows fails due to a short circuit, it may directly lead to a fatal malfunction such as a current short circuit. Therefore, when such a switching element fails, it is required to quickly detect the failure and stop the power supply.

[0005] Against the above technical background, Patent Document 1 below proposes a method of monitoring the current flowing through a solenoid coil for driving a four-way valve, and if an excessive current flows, stopping the application of voltage to the solenoid coil and storing information about the flow of overcurrent in a microcomputer.

[0006] Japanese Patent Application Laid-Open No. 2005-133997

[0007] However, the method proposed in Patent Document 1 could only detect a fault in the solenoid coil itself, such as a short circuit or an open circuit. Therefore, the method proposed in Patent Document 1 could not directly detect physical faults or deterioration in the four-way valve body, which is a moving part. Furthermore, as mentioned above, even if the electronic component is not a moving part, if it is an important electronic component such as a switching element located in a path through which a load current flows, the impact on the failure of the electronic component can be significant. Therefore, it is desirable to detect the fault or deterioration of these components and notify the user or repairer.

[0008] The present disclosure has been made in consideration of the above, and aims to provide a notification device that can detect failure or deterioration of electronic components or moving parts driven by electronic components and notify the user or repairer.

[0009] In order to solve the above-mentioned problems and achieve the objectives, the notification device of the present disclosure includes an electronic component to which power is supplied, a power supply unit that supplies power to the electronic component, a detector that detects a physical quantity that is correlated with the voltage or current applied to the electronic component, and a notification unit that notifies a user or repairer of a diagnosis result regarding failure or deterioration of the electronic component or a moving part driven by the electronic component.

[0010] The notification device according to the present disclosure has the advantage of being able to detect failure or deterioration of an electronic component or a moving part driven by an electronic component and notify the user or repair person.

[0011] 1 is a diagram showing an example of the configuration of an outdoor unit of an air conditioner including a notification device according to embodiment 1; sectional view showing the internal configuration of a typical four-way valve; diagram showing an example of the hardware configuration for realizing a control unit provided in a notification device according to embodiment 1; functional block diagram summarizing the functions of a notification device according to embodiment 1; diagram showing an example of the current waveform of a coil current flowing when driving the four-way valve shown in FIG. 1 during normal operation; diagram showing an example of the current waveform of a coil current flowing when driving the four-way valve shown in FIG. 1 during abnormal operation; diagram showing an example of the waveform of a coil current flowing when a step-like voltage is applied to the coil when driving the four-way valve shown in FIG. 1; diagram for explaining a first digitization process according to embodiment 1; FIG. 15 is a diagram illustrating an example of a neural network applicable to the learning device shown in FIG. 15; FIG. 16 is a diagram illustrating an example of a configuration of an inference device according to embodiment 4;

[0012] A notification device, a notification system, a learning device, an inference device, and a refrigeration cycle device according to embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. While this document will exemplify a case in which the notification device is installed in an air conditioner, which is an example of a refrigeration cycle device, application to other refrigeration cycle devices and devices other than refrigeration cycle devices is not intended to be excluded. In addition, the following description will simply refer to "connection" without distinguishing between electrical connection and physical connection. That is, the term "connection" includes both a direct connection between components and an indirect connection between components via other components. In addition, in the following description, when specific examples of electronic components to which power is supplied and moving parts driven by the electronic components are specifically mentioned, a coil will be used as an example of the electronic component, and a four-way valve will be used as an example of the moving part.

[0013] Embodiment 1. Fig. 1 is a diagram showing an example of the configuration of an outdoor unit in an air conditioner 100 including a notification device 50 according to embodiment 1. Fig. 1 shows an AC power supply 110 that supplies power to the outdoor unit, a compressor motor 120 and a four-way valve 2 that are actuators mounted on the outdoor unit, and a circuit board 10 on which electronic components are arranged to provide drive power to the compressor motor 120 and drive the four-way valve 2. The four-way valve 2, together with other components not shown, constitute a refrigerant circuit 5.

[0014] FIG. 2 is a cross-sectional view showing the internal structure of a typical four-way valve. The four-way valve 2 is a refrigerant circuit component designed to switch the flow direction of refrigerant. As shown in FIG. 2 , a typical four-way valve 2 includes a main valve 17 and a pilot valve 18. A coil 3 is disposed near the pilot valve 18. The pilot valve 18 is driven by the magnetic force of the coil 3. When the pilot valve 18 is driven, it moves left or right in the drawing. This movement causes refrigerant in the refrigerant circuit 5 to flow into the main valve 17. The inflowing refrigerant then moves the main valve 17, changing the flow direction of the refrigerant in the refrigerant circuit 5. In this manner, the main valve 17 and pilot valve 18 constituting the four-way valve 2 are driven directly or indirectly by the coil 3. In this document, the valve directly driven by the coil 3 is defined as the pilot valve 18. Note that this definition is provided for convenience and is not intended to limit the configuration of the controlled object. Furthermore, the configuration shown in FIG. 2 is an example, and the main valve 17 and the pilot valve 18 are not limited to the configuration shown in FIG.

[0015] Furthermore, the drive methods for the four-way valve 2 can be broadly classified into a non-latching method, in which current continues to flow through the coil 3 to maintain the state of the pilot valve 18 once it has been switched, and a latching method, as shown in Figure 2, in which a magnet 19 is attached to one side of the pilot valve 18 and the switched state is maintained by the magnetic force of the magnet 19, thereby eliminating the need for continuous current to the coil 3. With the latching method, when the polarity of the four-way valve 2 is switched again, the coil 3 must generate a magnetic force sufficient to break away from the holding force of the magnet 19.

[0016] Returning to the explanation of Fig. 1 , various electronic components and circuit units are mounted on the substrate 10. Specifically, the substrate 10 includes a noise filter 1, a reactor 4, a rectifier circuit 52, a smoothing capacitor 54, an inverter circuit 56, and a drive circuit 58. In Fig. 1 , the rectifier circuit 52, the inverter circuit 56, and the drive circuit 58 are examples of power supply units that supply power to the electronic components. For example, the drive circuit 58 is positioned as a power supply unit that supplies power to the coil 3, which is an electronic component.

[0017] The noise filter 1 removes noise from the AC voltage applied from the AC power supply 110. The rectifier circuit 52 rectifies the noise-removed AC voltage output from the noise filter 1 and outputs the rectified voltage to a smoothing capacitor 54 and an inverter circuit 56. The smoothing capacitor 54 smoothes the rectified voltage output from the rectifier circuit 52 and applies it to the inverter circuit 56. The inverter circuit 56 converts the DC voltage smoothed by the smoothing capacitor 54 into a drive voltage for the compressor motor 120 to drive the compressor motor 120.

[0018] The drive circuit 58 has a coil 3 and a polarity switching mechanism 7, and drives the four-way valve 2 by passing a drive current for driving the four-way valve 2 through the coil 3. The voltage to be applied to the coil 3 is applied via the polarity switching mechanism 7. Although FIG. 1 shows a configuration in which a DC voltage is applied to the polarity switching mechanism 7, the voltage applied to the polarity switching mechanism 7 may also be an AC voltage. However, in either case, the polarity of the voltage applied to the coil 3 is switched by the polarity switching mechanism 7. The voltage polarity is switched by a relay, a physical switch, a semiconductor switch, or the like.

[0019] Current detectors 8 are also disposed at multiple key locations on the circuit board 10. In the example of FIG. 1 , the current detectors 8 are disposed at locations that detect the amount of current flowing through the coil 3, the amount of current flowing through the compressor motor 120 (which is the load), and the total amount of current flowing through the circuit board 10. In the method described herein, the current detector 8 may be a mechanism that detects changes in a magnetic field caused by current, such as a DCCT (Direct Current Transformer), or a mechanism that detects current from voltage using a shunt resistor or the like. The current detector 8 may be any detector capable of detecting a physical quantity correlated with the amount of current flowing through key locations on the circuit board 10. In other words, the current detector 8 in this example may be a detector that detects a physical quantity correlated with the voltage or current applied to an electronic component.

[0020] Furthermore, the notification device 50 in this paper is assumed to acquire various information using various sensors, such as the voltage values ​​applied to the various actuators, and the temperature, humidity, and vibration of the substrate 10 and the actuators. These sensors are not shown in FIG. 1 . It is not essential to include all of these sensors. These sensors are appropriately selected depending on the conditions of the environment in which the device in which the notification device 50 is installed is installed, and are placed on the substrate 10 or in the vicinity of the substrate 10.

[0021] The substrate 10 also includes a display unit 9. The display unit 9 may be any means that can be visually confirmed by a user or repairer, such as a liquid crystal panel, an LED (Light Emitting Diode), a display, etc. In addition to the display unit 9, a detection means that can be audibly confirmed by a user or repairer may also be added.

[0022] The circuit board 10 also includes a control unit 6. The control unit 6 receives the detected value of the current detector 8 and various sensor values ​​detected by various sensors other than the current detector 8. The control unit 6 controls the operation of the inverter circuit 56 and the drive circuit 58 based on these detected values ​​and various sensor values. That is, the control unit 6 manages the operation of the compressor motor 120 driven by the inverter circuit 56 and the four-way valve 2 driven by the drive circuit 58. The control unit 6 also manages the operation of actuators other than the compressor motor 120 and the four-way valve 2, which are not shown in FIG. 1 . Furthermore, the control unit 6 manages the display processing on the display unit 9 within the circuit board 10 and the transmission and reception of information to the indoor unit 102 and an information processing terminal 80 on an external network. Examples of the information processing terminal 80 include a personal computer and a smartphone. The control unit 6 and the information processing terminal 80 are connected via a wired or wireless network. The network may include the Internet. Furthermore, the control unit 6 diagnoses failures or deterioration of electronic components and movable components based on the input detection values ​​of the current detector 8 and various sensor values, and displays the diagnosis results on the display unit 9. The control unit 6 may also display the diagnosis results on a display unit (not shown) of the indoor unit 102, or on a remote control (not shown) via the indoor unit 102, or may notify the outside via a network. In this way, it becomes possible to notify the user or repairer of the diagnosis results regarding failures or deterioration of electronic components and movable components in a timely manner. Note that although one control unit 6 is shown in FIG. 1, multiple control units 6 may be present.

[0023] 3 is a diagram showing an example of a hardware configuration that realizes the control unit 6 included in the notification device 50 according to embodiment 1. The control unit 6 is realized by a processor 91 and a memory 92.

[0024] The processor 91 is a CPU (Central Processing Unit, also called a central processing unit, processing unit, arithmetic unit, microprocessor, microcomputer, processor, or DSP (Digital Signal Processor)), or a system LSI (Large Scale Integration). Examples of memory 92 include non-volatile or volatile semiconductor memories such as RAM (Random Access Memory), ROM (Read Only Memory), flash memory, EPROM (Erasable Programmable Read Only Memory), and EEPROM (Electrically Erasable Programmable Read Only Memory). Memory 92 is not limited to these, and may also be a magnetic disk, optical disk, compact disk, minidisc, or DVD (Digital Versatile Disc).

[0025] Fig. 4 is a functional block diagram summarizing the functions of the notification device 50 according to the first embodiment. As shown in Fig. 4, the notification device 50 is configured inside the air conditioner 100. The air conditioner 100 contains a display unit 9, movable parts 26, and the notification device 50, and the notification device 50 contains the above-mentioned control unit 6 as well as a detector 11, electronic parts 12, and a power supply unit 13. Although the display unit 9 is shown in Fig. 1 as being inside the notification device 50, it may be outside the notification device 50 as shown in Fig. 4.

[0026] 4, the functions of the control unit 6 can be divided into a memory unit 14, a diagnosis unit 15, and a notification unit 16. Some of the electronic components 12 may be located outside the notification device 50. The detector 11 corresponds to the above-mentioned current detector 8 and various sensors other than the current detector 8. The power supply unit 13 corresponds to the above-mentioned rectifier circuit 52, inverter circuit 56, and drive circuit 58.

[0027] The power supply unit 13 supplies power to the electronic component 12, and the detector 11 detects a physical quantity correlated with the voltage or current applied to the electronic component 12. The diagnosis unit 15 diagnoses whether the electronic component 12 or a moving component 26 driven by the electronic component 12 has failed or deteriorated based on the physical quantity detected by the detector 11. The notification unit 16 displays the diagnosis result regarding the failure or deterioration of the electronic component 12 or the moving component 26 on the display unit 9 and notifies the outside, if necessary. Note that the notification device 50 does not necessarily have the function of the diagnosis unit 15, and may be configured to utilize the function of an externally configured diagnosis unit, i.e., an externally configured diagnostic function. In this configuration, the notification device 50 may transmit the physical quantity detected by the detector 11 to the outside, receive the diagnosis result based on the physical quantity, and display it on the display unit 9.

[0028] Next, technical matters relating to a specific diagnostic method in embodiment 1 will be described. First, Fig. 5 is a diagram showing an example of a current waveform of the coil current that flows when the four-way valve 2 shown in Fig. 1 is driven during normal operation. Fig. 6 is a diagram showing an example of a current waveform of the coil current that flows when the four-way valve 2 shown in Fig. 1 is driven during abnormal operation. The coil current is a current that flows through the coil 3 when the four-way valve 2 is driven. In Figs. 5 and 6, the upper side shows the coil applied voltage, which is the voltage applied to the coil 3, and the lower side shows the coil current. Furthermore, the horizontal axis in Figs. 5 and 6 indicates time.

[0029] When a voltage is applied to the coil 3, a coil current flows as shown in the figure. The coil current generates a magnetic force in the coil 3, which drives the pilot valve 18 of the four-way valve 2. At this time, the pilot valve 18 moves inside the coil 3, changing the magnetic permeability within the coil 3, which in turn changes the inductance component of the coil 3. Depending on the change in the inductance component of the coil 3, the pilot valve 18 may not operate normally. There is a clear difference in the coil current waveform between the current waveform during normal operation shown in Figure 5 and the current waveform during abnormal operation shown in Figure 6, i.e., when the pilot valve 18 operates normally and when it does not operate normally.

[0030] 5 and 6 show examples of when the pilot valve 18 operates normally and when it does not operate normally. However, for example, wear on the moving parts of the pilot valve 18 can cause changes in the speed during operation and the position after operation. In this case, the amount of change over time in the magnetic permeability in the coil 3 differs from normal, resulting in a change in the current waveform. Therefore, it can be said that by making good use of information on the current waveform of the coil current, it is possible to diagnose deterioration or failure of the four-way valve 2.

[0031] However, in order to diagnose deterioration or failure of the four-way valve 2 based on information on changes in the current waveform, it is necessary to establish specific criteria for judgment. In the method described in the above-mentioned Patent Document 1, a typical technique is to set a threshold value for a current that does not flow under normal conditions, and to judge an abnormality to exist when the detected current exceeds the threshold value. However, this method has the disadvantages of only being able to detect short-circuit faults, and of not being able to detect deterioration before it leads to a fault, since a fault has already occurred when the fault is detected.

[0032] Furthermore, in the case of the current waveform during abnormal operation shown in Figure 6, the peak value of the current waveform during abnormal operation surrounded by the dashed frame and the peak value of the current waveform not surrounded by the dashed frame are approximately the same value, so it can be said that it is impossible to determine whether it is normal or abnormal by simple threshold judgment.

[0033] FIG. 7 is a diagram showing an example of the waveform of the coil current that flows when a stepped voltage is applied to the coil 3 to drive the four-way valve 2 shown in FIG. 1 . The horizontal axis of FIG. 7 represents time, with the solid line representing the current waveform when the pilot valve 18 is normal and the dashed line representing the current waveform when the pilot valve 18 is abnormal. The dashed-dotted line is an example of a threshold value set based on the solid-line current waveform that flows normally. This threshold value is set with consideration given to preventing the current that flows normally from being mistakenly determined to be abnormal, but it also prevents the dashed-line current that flows abnormally from being abnormal. In other words, with the current waveform shown in FIG. 7 , it is difficult to set a threshold value for determining whether the current is normal or abnormal.

[0034] The difficulty of setting the threshold value will be further explained. For example, if the moving part of the pilot valve 18 is worn, the pilot valve 18 moves faster than a normal pilot valve 18 with no worn moving part. In this case, the amount of change in the magnetic resistance inside the coil 3 increases, causing a change in the inductance component of the coil 3, resulting in a current waveform with a concave portion, as shown by the dashed line. Therefore, in the case of a current waveform such as that shown in Figure 7, it is difficult to determine the deterioration of the pilot valve 18, i.e., the four-way valve 2, using threshold determination based on the current waveform. Note that this is not limited to current waveforms, but also applies to all physical quantities, such as voltage, temperature, and pressure, that can be acquired by the air conditioner 100. Therefore, in the method of the first embodiment described below, the time waveform of the acquired physical quantity is digitized. This digitization process converts the acquired physical quantity into a value correlated with failure or deterioration. In this paper, the digitized physical quantity is appropriately referred to as a "feature quantity." Two specific examples of digitization are described below.

[0035] (First Quantification Process) FIG. 8 is a diagram illustrating the first quantification process in the first embodiment. FIG. 8 shows the results of frequency analysis of the time waveform of the coil current shown in FIG. 7. In FIG. 8, f2 to f6 represent the frequencies of the coil current that flows when a voltage of drive frequency f1 is applied to the coil 3. The larger the subscripts 2 to 6, the higher the frequency. That is, f2 < f3 < ... < f6. Although a current of frequency f1 also flows through the coil 3, this current is not shown. In FIG. 8, the unhatched vertical bars represent frequency components of the current waveform under normal conditions, and the hatched vertical bars represent frequency components of the current waveform under abnormal conditions. FIG. 8 also shows a threshold value for the frequency f2 component. The frequency f2 component is an example of a feature defined herein. If the feature value of frequency f2 component falls below the threshold value, the four-way valve 2 being diagnosed can be diagnosed as faulty or degraded.

[0036] Note that setting a threshold for the frequency f2 component is merely an example, and a threshold may be set for at least one of the frequency components f3 to f6 in addition to or instead of frequency f2. Furthermore, it is also possible to calculate the sum or average value of the frequency components f2 to f6, rather than just the frequency component f2 to f6 alone, and set a threshold for these calculated values. If these calculated values ​​are correlated with a tendency for failure or deterioration of the four-way valve 2, they are treated as feature quantities, and a threshold is set for the calculated values.

[0037] In the above explanation, attention has been focused on the drive frequency f1 of the voltage applied to the coil, but this does not necessarily mean that the drive frequency f1 changes. The drive frequency f1 also varies depending on the operating environment conditions of the equipment in which the four-way valve 2 is installed, variations in the four-way valve 2 itself, the method of applying the voltage applied to the coil, and other factors. For this reason, in this method, there are no particular limitations on where in the frequency band the feature value occurs.

[0038] (Second quantification process) FIG. 9 is a diagram illustrating the second quantification process in the first embodiment. FIG. 9 illustrates a diagnosis result obtained when the time waveform of the coil current illustrated in FIG. 7 is converted into statistics and a diagnosis is performed. Typical statistics include the mean value, effective value, peak value, standard deviation, variance, kurtosis, and skewness. FIG. 9 illustrates the mean value, variance, and skewness. The column labeled "Normal" illustrates values ​​obtained by discretely sampling the current waveform in FIG. 7 during normal operation in the time direction and statistically processing the discrete values ​​within a specific period of sampling. Similarly, the column labeled "Abnormal" illustrates values ​​obtained by discretely sampling the current waveform in an abnormal operation in FIG. 7 and statistically processing the discrete values ​​within a specific period of sampling. Furthermore, the column labeled "Abnormal / Normal" illustrates the ratio of the statistical value during abnormal operation to the statistical value during normal operation, i.e., the value obtained by dividing the statistical value during abnormal operation by the statistical value during normal operation.

[0039] In Figure 9, the mean value and variance show no significant difference between normal and abnormal conditions. On the other hand, the skewness value during abnormal conditions is approximately half of the normal value, revealing a significant difference. Therefore, in the example of Figure 9, setting a threshold value for skewness is expected to enable accurate diagnosis of failure or deterioration of the four-way valve 2. Note that the results in Figure 9 are for a specific current waveform, and using skewness does not necessarily mean that all failures or deterioration can be diagnosed. Depending on the type or characteristics of the moving parts 26 or the environmental conditions under which the equipment is used, the mean value or variance may be highly sensitive. Therefore, by combining multiple highly sensitive statistics depending on the type or characteristics of the moving parts 26 or the environmental conditions under which the equipment is used, it becomes possible to accurately determine failure or deterioration.

[0040] The statistics described above are examples of representative values, and are not limited to these representative values. Any representative value may be used as long as one or more representative values ​​that represent the characteristics of a plurality of discrete values ​​obtained by sampling a time waveform within a specific period are obtained.

[0041] According to the first and second digitization processes described above, by converting the time waveform into a value or a statistical quantity in the frequency domain, it is possible to compress a large amount of data as a time waveform into a specific representative value. This makes it possible to reduce the storage capacity of the storage unit 14 in the control unit 6, and to reduce the communication time and communication costs for data communication with the outside.

[0042] Up to this point, the description has focused on the case where the moving part 26 is the four-way valve 2, but the diagnostic technique of the first embodiment is also useful for moving parts that move or vibrate when a current is passed through them, similar to the four-way valve 2. Examples of such moving parts include an electronic expansion valve, a relay, and a motor.

[0043] An electronic expansion valve is a moving part whose motor shaft is driven by a coil inside a stepping motor. Electronic expansion valves are also used in refrigeration cycle devices. The electronic expansion valve in a refrigeration cycle device adjusts the flow rate of refrigerant by changing the opening and closing state of a valve in the refrigerant circuit. For example, if the inside of a stepping motor becomes rusty and the rotating part becomes stuck, the current waveform flowing through the coil of the stepping motor will change as described above. Therefore, by applying the above-mentioned method to this change, it is possible to diagnose a malfunction or deterioration of the electronic expansion valve.

[0044] A relay is a moving part that switches between conductive and non-conductive states by passing a current through a coil and physically moving the contacts with the force of an electromagnet. As the relay deteriorates, for example, if the movement of the contacts changes, the waveform of the current flowing through the coil changes as described above. Therefore, by applying the above-described method to this change, it becomes possible to diagnose relay failure or deterioration.

[0045] A motor is a moving part that rotates a rotating part by passing a current through a coil to generate a rotating magnetic field. As the motor deteriorates, for example, if the rotating part becomes stuck, the waveform of the current flowing through the coil changes as described above. Therefore, by applying the above-described method to this change, it becomes possible to diagnose a motor failure or deterioration.

[0046] In addition to moving parts, similar changes in the current or voltage waveform occur when electronic components themselves fail or deteriorate. For example, when an electronic component experiences a short circuit or an open circuit, the shape of the current or voltage waveform inevitably changes. Examples of such electronic components include reactors and capacitors.

[0047] Reactors have a larger mass than other electronic components. Therefore, external vibrations can cause the reactor to come off its mounting, and erosion from wind and rain can cause the conductor to break. In other words, reactors are susceptible to failure or degradation due to external factors. If a reactor fails or deteriorates, the current waveform flowing through the reactor will also change as described above. Therefore, applying the above-described method to these changes makes it possible to diagnose reactor failure or deterioration.

[0048] Furthermore, capacitors are limited-life components whose capacitance value decreases over time. Electrolytic capacitors, which use electrolytes, are particularly susceptible to capacitance loss due to evaporation of the electrolyte. A decrease in capacitance can lead to unstable operation of the capacitor itself, and abnormal heat generation can cause the electrolyte to boil or overflow, potentially damaging or shorting the electrolytic capacitor and resulting in serious problems. When capacitance decreases, the current waveform of the current flowing through the electrolytic capacitor or the voltage waveform across the electrolytic capacitor also changes as described above. Therefore, applying the above-described techniques to these changes makes it possible to diagnose electrolytic capacitor failure or degradation.

[0049] Up to this point, only a brief explanation has been given of the usage environment conditions, such as the location, region, or season in which the air conditioner 100 is installed. Below, a diagnosis method that takes usage environment conditions into consideration will be described.

[0050] The air conditioner 100 is an electrical device used in various regions around the world and throughout the year. Therefore, the air conditioner 100's usage conditions, such as the operating conditions of its electronic components and the number of times that its moving parts, such as the four-way valve 2, operate, change depending on the installation location and environmental conditions, such as seasonal changes throughout the year. For this reason, in order to accurately diagnose a failure or deterioration of the four-way valve 2, it is desirable to anticipate all possible usage conditions before shipping from the factory and to understand in advance the relationship between the failure or deterioration of the four-way valve 2 and changes in the feature quantities. However, it is practically difficult to prepare a comprehensive list of all possible usage conditions.

[0051] Therefore, before shipping from the factory, data showing the relationship between the failure or deterioration of the four-way valve 2 and changes in the feature quantities under somewhat general use environmental conditions is stored in the control unit 6. FIG. 10 is a diagram illustrating a method for diagnosing failure or deterioration that takes into account use environmental conditions in the first embodiment. The horizontal axis of FIG. 10 is the time axis, and the vertical axis is the feature quantities defined above, i.e., the feature quantities calculated based on the acquired physical quantities. In FIG. 10, the feature quantities are divided into four seasons (spring, summer, autumn, and winter) so that seasonal changes in the feature quantities throughout the year can be easily understood.

[0052] Using the example of FIG. 10 , the accumulation period for accumulating feature information is, for example, one year from the time the air conditioner 100 begins actual use by a user. Information not directly related to the feature information may also be accumulated. Also, in FIG. 10 , the feature information for the previous year, i.e., the feature information for the accumulation period, is shown shifted one year to the right by a dashed line, and the feature information for the current year is shown by a dashed-dotted line. The control unit 6 compares the feature information for the current year with the feature information for the same period in the accumulation period and diagnoses a malfunction or deterioration by determining the difference between the feature information and a threshold value.

[0053] The diagnostic method shown in FIG. 10 can be expressed in the form of a flowchart as shown in FIG. 11. FIG. 11 is a flowchart showing the flow of diagnostic processing by the control unit 6 of embodiment 1. In FIG. 11, the feature amount during the accumulation period is referred to as a "first feature amount" and is represented by "A1." Furthermore, the feature amount at the time of diagnosis is referred to as a "second feature amount" and is represented by "A2." Furthermore, in FIG. 11, the first and second thresholds in the threshold determination are represented by "B1" and "B2," respectively. The second threshold value B2 is a value greater than the first threshold value B1.

[0054] The control unit 6 calculates the absolute value |A1-A2| of the difference between the first feature amount A1 and the second feature amount A2 (step S11). If the absolute value |A1-A2| does not exceed the first threshold value B1 (step S12, No), the control unit 6 determines that the part to be diagnosed is normal (step S13). After this, the processing flow of FIG. 11 ends.

[0055] If the absolute value |A1-A2| exceeds the first threshold value B1 (step S12, Yes), the control unit 6 compares the absolute value |A1-A2| with the second threshold value B2 (step S14). If the absolute value |A1-A2| does not exceed the second threshold value B2 (step S14, No), the control unit 6 determines that the part being diagnosed has deteriorated (step S15). After this, the processing flow of FIG. 11 ends.

[0056] If the absolute value |A1-A2| exceeds the second threshold value B2 (step S14, Yes), the control unit 6 determines that the part to be diagnosed is faulty (step S16), and then ends the processing flow of FIG.

[0057] If the processing of step S15 determines that the diagnosed part is degraded, it is possible to encourage the replacement or repair of the part before the product in which the part is installed breaks down. If the user desires to replace the product in which the part is installed, it is possible to provide the user with sufficient time to prepare for the product replacement. Furthermore, if the processing of step S16 determines that the diagnosed part is broken down, it is possible to provide the user with appropriate and accurate information to encourage the replacement or repair of the part, or the replacement of the product.

[0058] The diagnostic method described above involves comparing the calculated feature quantity with a preset threshold value, but this threshold value may be updated as appropriate depending on the usage environment or frequency of use. Using an appropriately updated threshold value makes it possible to improve the accuracy of diagnosing component failures or deterioration.

[0059] As described above, the notification device according to the first embodiment includes an electronic component to which power is supplied, a power supply unit that supplies power to the electronic component, a detector that detects a physical quantity correlated with the voltage or current applied to the electronic component, and a notification unit that notifies a user or repairer of a diagnosis result regarding failure or deterioration of the electronic component or a moving component driven by the electronic component. Examples of the electronic component include a reactor or a capacitor, and examples of the moving component include a motor or a relay. In the case of a notification device mounted on a refrigeration cycle device, examples of the moving component include a four-way valve or an electronic expansion valve provided in the refrigerant circuit of the refrigeration cycle device. The notification device configured in this manner can directly detect failure or deterioration of the electronic component or a moving component driven by the electronic component and notify the user or repairer.

[0060] The notification device according to the first embodiment may include a diagnostic unit that diagnoses whether an electronic component or a moving component has failed or deteriorated based on a physical quantity correlated with the voltage or current applied to the electronic component. When the diagnostic unit is included, the notification device is capable of directly detecting failure or deterioration of an electronic component or a moving component driven by the electronic component in a self-contained manner. In addition, in the case of a notification device that does not include a diagnostic unit, the notification device can receive a diagnosis result regarding failure or deterioration of an electronic component or a moving component from an external device and notify a user or repairer. Therefore, even a notification device that does not include a diagnostic unit can indirectly detect failure or deterioration of an electronic component or a moving component driven by the electronic component and notify a user or repairer.

[0061] In the notification device according to the first embodiment, the diagnostic unit can determine whether an electronic component or a moving component has failed or deteriorated based on a threshold value. In this case, it is desirable that the threshold value used in the determination process be updated as appropriate depending on the environment in which the device is used or how often it is used. Using an appropriately updated threshold value can improve the accuracy of diagnosing component failure or deterioration.

[0062] Embodiment 2 In the first embodiment, the explanation has been centered on the diagnosis of failure or deterioration using feature quantities performed by the diagnosis unit 15 provided in the control unit 6. On the other hand, when the control unit 6 does not have sufficient or sufficient computing power, or when a large amount of data is collected to diagnose failure or deterioration, a system may be constructed in which the function of the diagnosis unit 15 in the control unit 6 is performed by an external processing device. This form will be explained in the second embodiment.

[0063] 12 and 13 are diagrams showing first and second configuration examples of a notification system 70 according to the second embodiment. The notification system 70 according to the second embodiment is a notification system that notifies a user or repairer of a diagnosis result regarding a failure or deterioration of an electronic component or a moving part driven by an electronic component. To realize this function, the notification system 70 is configured with the notification device 50 described in the first embodiment and a processing device 40 that cooperates with the notification device 50. However, only the control unit 6A of the notification device 50 is shown in FIGS. 12 and 13. Furthermore, FIG. 13 shows the information processing terminal 80 also shown in FIG. 1.

[0064] In the control unit 6A, the function of the diagnostic unit 15 is removed from the configuration of the control unit 6 shown in FIG. 4 , and the notification unit 16 is provided with a communication function. The function of the diagnostic unit 15 removed from the control unit 6 is provided by the processing device 40. Note that the control unit 6A may have the function of the diagnostic unit 15 or some of the functions, and in this case, the processing device 40 supports the processing of the control unit 6A by complementing some of the diagnostic function of the control unit 6A. The processing device 40 is a server device configured to be able to communicate with the control unit 6A. The processing device 40 may also be a cloud server configured on the Internet.

[0065] The control unit 6A exchanges data with the processing device 40 using the external communication means 42. The data transmitted by the control unit 6A are physical quantities related to the calculation of the above-mentioned feature quantities, or calculated values ​​calculated based on the physical quantities. The data received by the control unit 6A are values ​​resulting from calculations performed by the processing device 40, failure or deterioration diagnosis results performed by the diagnosis unit 15, etc.

[0066] The external communication means 42 may be any means as long as it allows the control unit 6A to send and receive data to and from the processing device 40. For example, it may be a wired LAN cable or wireless LAN device that can connect to the Internet, a communication station for a portable device, or a device that can communicate with satellite communication.

[0067] The processing device 40 receives data transmitted from the control unit 6A via the external communication means 42. The processing device 40 performs the aforementioned digitization process using the received data to calculate feature quantities. The diagnosis unit 15 of the processing device 40 diagnoses failure or degradation of the diagnostic component using the calculated feature quantities. In the configuration shown in FIG. 13 , the notification system 70 can display the failure or degradation diagnosis results, etc., via the information processing terminal 80. This makes it possible to notify a repair person, etc., in a remote location of the failure or degradation diagnosis results, etc.

[0068] Furthermore, the diagnosis results from the diagnosis unit 15 of the processing device 40 can be transmitted to the control unit 6 via the external communication means 42. In this case, the control unit 6A can display the received diagnosis results on the display unit 9. This makes it possible to notify the user and on-site repair personnel of the failure or deterioration diagnosis results, etc.

[0069] 14 is a diagram showing a third example configuration of notification system 70 according to embodiment 2. In the third example configuration, a control unit 6B is shown that is installed in an air conditioner different from the air conditioner having control unit 6A.

[0070] By collecting data from multiple air conditioners, the processing device 40 can improve the accuracy of diagnosing failures or deterioration. As will be described in the embodiments below, if the notification device 50 or the processing device 40 has a learning function, it can learn about feature quantities based on data collected from multiple air conditioners and use the learning results, thereby further improving the accuracy of diagnosing failures or deterioration.

[0071] 14, if the control unit 6A is installed in an air conditioner that has already been used for more than a year and the control unit 6B is installed in a brand new air conditioner, only information about the air conditioner in which the control unit 6B is installed exists before it leaves the factory. For example, if the methods shown in FIGS. 10 and 11 are used, learning will be performed over the course of a new year, while updating the trends in the feature quantities. In contrast, with the configuration of FIG. 14, the learning results of other air conditioners can be used, making it possible to diagnose malfunctions or deterioration with high accuracy without waiting a year.

[0072] As described above, the notification system according to the second embodiment includes a processing device that acquires, from the notification device, a physical quantity detected by the detector provided in the notification device according to the above-described first embodiment or a calculated value calculated based on the physical quantity. In the notification system according to the second embodiment, the notification device is configured to be able to upload the physical quantity or the value calculated based on the physical quantity to the processing device, and is configured to be able to acquire data from the processing device.

[0073] In the notification system according to the second embodiment, the processing device includes a diagnostic unit that diagnoses, based on physical quantities, whether an electronic component provided in the notification device or a moving component driven by the electronic component has failed or deteriorated. According to the notification system according to the second embodiment, even if the control unit of the notification device does not have sufficient or sufficient computing power, the processing device can make up for the shortfall, thereby enabling more accurate diagnosis of failure or deterioration than in the first embodiment.

[0074] Furthermore, the notification system according to embodiment 2 can diagnose failures or degradation using information from devices other than the device being diagnosed, thereby enabling more accurate failure or degradation diagnosis than embodiment 1. In particular, the notification system according to embodiment 2 is configured to be suitable for learning feature quantities used in diagnosing failures or degradation. Therefore, if the notification system learns feature quantities based on data collected from multiple air conditioners and uses the learning results, it is possible to further improve the accuracy of failure or degradation diagnosis.

[0075] Embodiment 3 In embodiment 3, a learning device that cooperates with the notification device 50 according to embodiment 1 described above will be described. FIG. 15 is a diagram showing an example of the configuration of a learning device 20 according to embodiment 3. The learning device 20 includes a data acquisition unit 21 and a model generation unit 22. The storage unit 23 shown in FIG. 15 may be included in the learning device 20, or may be the storage unit 14 included in the control unit 6 of the notification device 50. Furthermore, the learning device 20 shown in FIG. 15 may be configured within the notification device 50, or may be configured within the processing device 40 described in embodiment 2. Furthermore, the learning device 20 shown in FIG. 15 may be configured to cooperate with an inference device described in embodiment 4.

[0076] The operation of the learning device 20 will be further described with reference to Fig. 16. Fig. 16 is a flowchart illustrating the operation of the learning device 20 according to the third embodiment.

[0077] The data acquisition unit 21 acquires learning data including the usage environment conditions of the equipment in which the notification device 50 is installed and a failure or abnormality determination result corresponding to the result of the quantification process of the physical quantities detected by the detector of the notification device 50 (step S21). The usage environment conditions of the equipment include information about the outside air temperature of the usage environment, humidity changes, and the location of use. Information such as the movement frequency of moving parts can also be included in the usage environment conditions. The learning data acquired by the data acquisition unit 21 is passed to the model generation unit 22. Note that the data acquisition unit 21 may acquire the information about the usage environment conditions of the equipment and the data regarding the failure or abnormality determination result corresponding to the result of the quantification process of the physical quantities at different times.

[0078] The model generation unit 22 uses the learning data to generate a trained model for inferring a threshold value for failure or deterioration determination corresponding to the result of the quantification process of physical quantities that change depending on the use environment conditions (step S22). Specifically, the model generation unit 22 uses at least one of information on use environment conditions, such as the outside air temperature of the use environment, humidity changes, the use location, and the movement frequency of moving parts, combines this information with the failure or abnormality determination result corresponding to the result of the quantification process of physical quantities, and generates a trained model by learning a threshold value for failure or deterioration determination that is correlated with the use environment conditions based on data created by associating them with each other.

[0079] The learning algorithm used by the model generation unit 22 can be any known algorithm such as supervised learning, unsupervised learning, semi-supervised learning, reinforcement learning, etc. Here, supervised learning using a neural network will be described as an example.

[0080] The model generation unit 22 learns a threshold value for determining failure or deterioration that correlates with information on the usage environment conditions by so-called supervised learning in accordance with a neural network. Here, supervised learning is a method of providing a set of input data and result (label) data to the learning device 20, learning the features of the data, and inferring the result from the input.

[0081] Fig. 17 is a diagram showing an example of a neural network applicable to the learning device 20 shown in Fig. 15. The neural network is composed of an input layer consisting of a plurality of neurons, an intermediate layer (hidden layer) consisting of a plurality of neurons, and an output layer consisting of a plurality of neurons. The intermediate layer may be one layer, or two or more layers.

[0082] For example, in a three-layer neural network as shown in Figure 17, when multiple inputs are input to the input layer (X1-X3), the values ​​are multiplied by weight W1 (w11-w16) and input to the intermediate layer (Y1-Y2). The result is then further multiplied by weight W2 (w21-w26) and output from the output layer (Z1-Z3). This output result varies depending on the values ​​of weights W1 and W2.

[0083] In this paper, the neural network learns by adjusting the weights W1 and W2 so that the threshold for determining failure or deterioration approaches the optimal condition. More specifically, the neural network inputs at least one of information on usage environment conditions, such as the outside air temperature of the usage environment, humidity changes, the location of use, and the frequency of movement of moving parts, into the input layer, and adjusts the weights W1 and W2 so that the result output from the output layer approaches the threshold for determining failure or deterioration corresponding to the result of digitizing the time waveform of the physical quantity, thereby learning the threshold for determining failure or deterioration that is correlated with the usage environment conditions.

[0084] The model generation unit 22 generates a trained model by executing the above-described learning process, and stores the generated trained model in the storage unit 23 (step S23).

[0085] In this paper, the learning algorithm used by the model generation unit 22 has been described as supervised learning using a neural network, but the present invention is not limited to this. The learning algorithm may also be deep learning, which learns to extract features themselves. Furthermore, the learning process may be performed according to other well-known methods, such as genetic programming, functional logic programming, or support vector machines.

[0086] The model generation unit 22 may also use learning data input from multiple notification devices to learn thresholds for determining failure or degradation that are correlated with usage environment conditions. The model generation unit 22 may acquire learning data collected from multiple devices used in the same area, or may acquire learning data collected from multiple devices operating independently in different areas to learn thresholds for determining failure or degradation that are correlated with usage environment conditions. Devices that are the targets of thresholds for determining failure or degradation that are correlated with usage environment conditions may be added later or deleted midway. When multiple learning devices are used, the multiple notification devices and the multiple learning devices may be freely combined and used according to various conditions.

[0087] As a specific example of learning in this case, the case where the first quantification process shown in FIG. 8 is applied will be described below. For example, from multiple air conditioners 100 in which the notification device 50 is previously installed, information on the air conditioner 100's usage environment, such as outdoor temperature, humidity changes, usage location, and frequency of movement of operating components, is collected in advance, and the learning device 20 learns a combination of the results of frequency analysis of the time waveform of the coil current that operates the internal operating components of the air conditioner 100, as learning data. In this example, the learning device 20 learns that the values ​​of the unhatched vertical bars shown in FIG. 8 are normal values, and that the values ​​of the hatched vertical bars are abnormal values. The learning device 20 can then learn a predetermined threshold for determining deterioration by repeatedly learning from multiple sets of data from normal and abnormal times.

[0088] Alternatively, an autoencoder model that reproduces input data with output data is used to train the learning device 20 using only normal data. When data other than normal (abnormal data) is input to an autoencoder, the reproduction accuracy of the output data deteriorates, so if the difference between the input data and the output data is large, it can be determined to be abnormal. In this way, the autoencoder allows the learning device 20 to train using only normal data. Note that the learning method of the learning device 20 is not limited to these descriptions, and any means can be used as long as it can determine deterioration based on multiple pieces of data.

[0089] As described above, the learning device according to embodiment 3 is a learning device that cooperates with or is configured within the notification device according to the above-described embodiment 1, or a learning device that cooperates with or is configured within the processing device according to the above-described embodiment 2. When the learning device and the processing device cooperate with the notification device according to embodiment 1, the notification device is configured to be able to upload physical quantities or values ​​calculated based on physical quantities to the learning device and the processing device, and is configured to be able to acquire data from the learning device and the processing device.

[0090] The learning device according to the third embodiment includes a data acquisition unit and a model generation unit. The data acquisition unit acquires learning data including the usage environment conditions of the equipment in which the notification device is installed and a failure or abnormality determination result corresponding to the result of quantification processing of physical quantities detected by the detector of the notification device. The model generation unit uses the learning data to generate a trained model for inferring a failure or deterioration determination threshold corresponding to the result of quantification processing of physical quantities that change depending on the usage environment conditions. By using the trained model possessed by the learning device configured in this way, it is possible to configure a notification device and a processing device with high accuracy in diagnosing failures or deterioration.

[0091] Fourth Embodiment In the fourth embodiment, an inference device that cooperates with the notification device 50 according to the first embodiment will be described. FIG. 18 is a diagram illustrating an example of the configuration of an inference device 30 according to the fourth embodiment. The inference device 30 includes a data acquisition unit 31 and an inference unit 32. The storage unit 33 illustrated in FIG. 18 may be included in the inference device 30, or may be the storage unit 14 included in the control unit 6 of the notification device 50. The inference device 30 illustrated in FIG. 18 may be configured within the notification device 50, or may be configured within the processing device 40 described in the second embodiment. The inference device 30 illustrated in FIG. 18 may be configured to cooperate with the learning device 20 according to the third embodiment, or may be configured within the learning device 20. In this configuration, the storage unit 33 illustrated in FIG. 18 may be the storage unit 23 included in the learning device 20.

[0092] The data acquisition unit 31 acquires data related to the usage environment conditions of the device in which the notification device 50 is installed. As described above, the usage environment conditions of the device include information such as the outside air temperature of the usage environment, humidity changes, the location of use, and the frequency of movement of moving parts.

[0093] The inference unit 32 receives the trained model stored in the storage unit 33. The inference unit 32 infers and outputs a threshold value for determining failure or deterioration according to the use environment conditions, using the trained model and the use environment conditions input from the data acquisition unit 31. Through this process, the notification device 50 can obtain a threshold value for determining failure or deterioration that is correlated with the use environment conditions.

[0094] The processing flow relating to the threshold value for determining a failure or deterioration using the inference device 30 will be described with reference to Fig. 19. Fig. 19 is a flowchart showing the processing flow relating to the output of the threshold value for determining a failure or deterioration using the inference device 30 according to the fourth embodiment.

[0095] The data acquisition unit 31 acquires data related to the usage environment conditions of the equipment in which the notification device 50 is installed (step S31). The inference unit 32 performs an inference calculation using the trained model and the usage environment conditions input from the data acquisition unit 31 (step S32). This process infers a threshold value for determining failure or degradation according to the usage environment conditions. The inference device 30 outputs the inference result calculated in step S32, i.e., the inferred threshold value for determining failure or degradation, to the notification device 50 (step S33). The control unit 6 of the notification device 50 determines failure or degradation of the equipment using the inference result output from the inference device 30, i.e., the threshold value for determining failure or degradation inferred by the inference unit 32 (step S34).

[0096] In addition, if there are one or more other learned models other than the learned model being used, the inference device 30 may use the other learned models or combine the learned model being used with the other acquired learned models to infer and output a threshold value for determining failure or deterioration.

[0097] As described above, the inference device according to embodiment 4 is an inference device that cooperates with or is configured within the notification device according to the above-mentioned embodiment 1, and an inference device that cooperates with or is configured within the processing device according to the above-mentioned embodiment 2. When the inference device and the processing device cooperate with the notification device according to embodiment 1, the notification device is configured to be able to upload physical quantities or values ​​calculated based on physical quantities to the inference device and the processing device, and is configured to be able to acquire data from the inference device and the processing device.

[0098] The inference device according to the fourth embodiment includes a data acquisition unit and an inference unit. The data acquisition unit acquires data related to the usage environment conditions of the equipment in which the notification device is installed. The inference unit infers and outputs a threshold value for determining failure or degradation according to the usage environment conditions, using a trained model for inferring a threshold value for determining failure or degradation corresponding to the results of a process for quantifying physical quantities that change according to the usage environment conditions, and the usage environment conditions input from the data acquisition unit. Using an inference device configured in this manner, it is possible to configure a notification device and a processing device with high accuracy in diagnosing failure or degradation.

[0099] The configurations shown in the above embodiments are merely examples, and may be combined with other known technologies, or different embodiments may be combined with each other. It is also possible to omit or modify parts of the configurations as long as they do not deviate from the gist of the invention.

[0100] REFRIGERATION METHOD, 1 Noise filter, 2 Four-way valve, 3 Coil, 4 Reactor, 5 Refrigerant circuit, 6, 6A, 6B Control unit, 7 Polarity switching mechanism, 8 Current detector, 9 Display unit, 10 Board, 11 Detector, 12 Electronic components, 13 Power supply unit, 14, 23, 33 Memory unit, 15 Diagnosis unit, 16 Notification unit, 17 Main valve, 18 Pilot valve, 19 Magnet, 20 Learning device, 21, 31 Data acquisition unit, 22 Model generation unit, 26 Moving parts, 30 Inference device, 32 Inference unit, 40 Processing device, 42 External communication means, 50 Notification device, 52 Rectifier circuit, 54 Smoothing capacitor, 56 Inverter circuit, 58 Drive circuit, 70 Notification system, 80 Information processing terminal, 91 Processor, 92 Memory, 100 Air conditioner, 102 Indoor unit, 110 AC power supply, 120 compressor motor.

Claims

1. A notification device comprising: an electronic component to which power is supplied; a power supply unit that supplies power to said electronic component; a detector that detects a physical quantity that is correlated with the voltage or current applied to said electronic component; and a notification unit that notifies a user or repairer of a diagnosis result regarding a failure or deterioration of said electronic component or a moving part driven by said electronic component.

2. The notification device according to claim 1, further comprising a diagnostic unit that diagnoses whether the electronic component or the movable component has failed or deteriorated based on the physical quantity.

3. The notification device according to claim 2, wherein the diagnostic unit performs a process of determining whether the electronic component or the movable component has failed or deteriorated based on a threshold value, and the threshold value is updated as appropriate.

4. The notification device according to any one of claims 1 to 3, wherein the electronic component is a reactor or a capacitor.

5. The notification device according to any one of claims 1 to 4, wherein the moving part is a motor or a relay.

6. The notification device according to any one of claims 1 to 5, wherein the notification device is mounted on a refrigeration cycle device, and the movable part is a four-way valve or an electronic expansion valve provided in a refrigerant circuit of the refrigeration cycle device.

7. A notification system that notifies a user or repairer of a diagnosis result regarding failure or deterioration of an electronic component or a moving component driven by said electronic component, comprising: the notification device according to claim 1; and a processing device that acquires from said notification device the physical quantity detected by said detector or a calculated value calculated based on said physical quantity, said processing device comprising a diagnostic unit that diagnoses whether said electronic component provided in said notification device or said moving component driven by said electronic component has failed or deteriorated based on said physical quantity.

8. A learning device that cooperates with or is configured within the notification device according to claim 1, comprising: a data acquisition unit that acquires learning data including the usage environment conditions of a device in which the notification device is installed and failure or abnormality determination results corresponding to the results of quantification processing of the physical quantities detected by the detector of the notification device; and a model generation unit that uses the learning data to generate a trained model for inferring a failure or deterioration determination threshold value corresponding to the results of quantification processing of the physical quantities that change depending on the usage environment conditions.

9. The learning device according to claim 8, wherein at least one of the outside air temperature, humidity changes, and location of use in the environment in which the device is used is taken into consideration as the usage environmental conditions.

10. The learning device according to claim 9, wherein the notification device is mounted on a refrigeration cycle device, the movable part is a four-way valve or an electronic expansion valve provided in the refrigeration cycle device, and the operating environment conditions further take into account the frequency of movement of the four-way valve or the electronic expansion valve.

11. The learning device according to any one of claims 8 to 10, wherein the physical quantity is voltage, temperature, or pressure.

12. A learning device according to any one of claims 8 to 11, wherein the result of the digitization process is a result of frequency analysis of the time waveform of the physical quantity.

13. A learning device according to any one of claims 8 to 11, wherein the quantification process is a conversion process into a statistical quantity including at least one of the mean value, effective value, peak value, standard deviation, variance, kurtosis, and skewness.

14. An inference device that cooperates with the notification device described in claim 1 or is configured within the notification device, comprising: a data acquisition unit that acquires data related to the usage environment conditions of a device in which the notification device is installed; a trained model for inferring a threshold value for determining failure or deterioration corresponding to the result of quantification processing of the physical quantity that changes depending on the usage environment conditions; and an inference unit that infers and outputs a threshold value for determining failure or deterioration depending on the usage environment conditions using the usage environment conditions input from the data acquisition unit.

15. The inference device according to claim 14, wherein at least one of the outside air temperature, humidity changes, and location of use in the environment in which the device is used is taken into consideration as the usage environmental conditions.

16. The inference device according to claim 15, wherein the notification device is mounted on a refrigeration cycle device, the movable part is a four-way valve or an electronic expansion valve provided in the refrigeration cycle device, and the operating environment conditions further take into account the frequency of movement of the four-way valve or the electronic expansion valve.

17. An inference device according to any one of claims 14 to 16, wherein the physical quantity is voltage, temperature, or pressure.

18. An inference device according to any one of claims 14 to 17, wherein the result of the digitization process is a frequency analysis result of the time waveform of the physical quantity.

19. An inference device described in any one of claims 14 to 17, wherein the quantification process is a conversion process into a statistical quantity including at least one of the mean value, effective value, peak value, standard deviation, kurtosis, and skewness.

20. A notification device as set forth in any one of claims 1 to 6, configured to be capable of uploading the physical quantity or a value calculated based on the physical quantity to an external processing device having at least one of the learning device as set forth in any one of claims 8 to 13 or the inference device as set forth in any one of claims 14 to 19.

21. A notification device according to any one of claims 1 to 6, configured to be capable of acquiring data from an external processing device having at least one of a learning device according to any one of claims 8 to 13 or an inference device according to any one of claims 14 to 19.

22. A notification device according to any one of claims 1 to 6, wherein the notification device is configured to be able to notify an external information processing terminal of a diagnosis result relating to a failure or deterioration of the electronic component or the movable component.

23. A refrigeration cycle device comprising a notification device according to any one of claims 1 to 6 or a notification device according to any one of claims 20 to 22.

Images