Learning device, sensor anomaly detection device, learning method, sensor anomaly detection method, and program

Abstract

Even when the air conditioner is only briefly shut down, the system detects any deviations in the detection of the sensors used in the air conditioner. [Solution] The learning device comprises an acquisition unit and a generation unit. The acquisition unit acquires operating data including a first detection value detected by a first sensor provided in the air conditioner for detecting ambient temperature, which is detected while the air conditioner is stopped, and a second detection value detected by a second sensor provided in the air conditioner, which is detected while the air conditioner is stopped. The generation unit uses the operating data to generate a trained model for calculating an estimated value of the second detection value from the first detection value to determine whether the second sensor is abnormal or not.

Description

Technical Field

[0001] The present invention relates to a learning device, a sensor abnormality determination device, a learning method, a sensor abnormality determination method, and a program.

Background Art

[0002] For example, in a refrigerant circuit of an air conditioner, temperature sensors for detecting the temperature of the refrigerant and the temperature of the air are provided, and the detection values of these temperature sensors are used for the operation control of the air conditioner. The detection values of various sensors such as temperature sensors may deviate from normal values (detection deviation) due to aging or the like. In such a case, the air conditioner may not be properly controlled and may cause problems in operation. Therefore, it is known to detect the detection deviation by comparing the outside air temperature with the detection value of the sensor by utilizing the fact that when a sufficient time (for example, 8 hours) has elapsed since the operation of the air conditioner stopped, the detection value of the temperature sensor approximates the outside air temperature (see Patent Document 1 below).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] By the way, in the detection method of the detection deviation described in Patent Document 1, a stop time of, for example, 8 hours or more is required. However, for example, in summer, the operation time of the air conditioner becomes long, and even if the air conditioner is stopped, the stop time may be short and a sufficient time may not be secured. In such a case, it is difficult to detect the detection deviation of the sensor by the detection method described in Patent Document 1.

[0005] The present invention has been made in view of the above-mentioned problems, and its objective is to provide a learning device, a sensor abnormality determination device, a learning method, a sensor abnormality determination method, and a program that can detect detection deviations of sensors used in an air conditioner, even when the downtime of the air conditioner is short. [Means for solving the problem]

[0006] One aspect of this disclosure is a learning device comprising an acquisition unit and a generation unit. The acquisition unit acquires operating data including a first detection value detected by a first sensor provided in the air conditioner for detecting ambient temperature, the first detection value detected while the air conditioner is stopped, and a second detection value detected by a second sensor provided in the air conditioner, the second detection value detected while the air conditioner is stopped. The generation unit uses the operating data to generate a trained model for calculating an estimated value of the second detection value from the first detection value to determine whether the second sensor is abnormal or not. [Effects of the Invention]

[0007] According to various aspects and embodiments of this disclosure, even when the air conditioner is shut down for a short period, it is possible to detect detection errors in the sensors used in the air conditioner. [Brief explanation of the drawing]

[0008] [Figure 1] Figure 1 shows an example of an air conditioner. [Figure 2] Figure 2 shows an example of an outdoor unit. [Figure 3] Figure 3 is a block diagram showing an example of the control unit for an outdoor unit. [Figure 4] Figure 4 shows an example of an indoor unit. [Figure 5] Figure 5 is a block diagram showing an example of the control unit for an indoor unit. [Figure 6] Figure 6 is a block diagram of an example server. [Figure 7]Figure 7 shows an example of the structure of the detection value table included in the driving data. [Figure 8] Figure 8 shows an example of the structure of the operating status table included in the operating data. [Figure 9] Figure 9 is a block diagram showing an example of the functional configuration of the server's control unit in the process of generating a trained model. [Figure 10] Figure 10 shows an example of the changes in the detected values ​​of each sensor after the operation of the first compressor has stopped. [Figure 11] Figure 11 shows an example of the relationship between sensor readings and ambient temperature during period T. [Figure 12] Figure 12 is a block diagram showing an example of the functional configuration of the server's control unit in the process of determining whether or not there is a sensor abnormality. [Figure 13] Figure 13 is a flowchart showing an example of the process for generating a trained model. [Figure 14] Figure 14 is a flowchart showing an example of a process for determining whether or not there is a sensor malfunction. [Modes for carrying out the invention]

[0009] The following describes in detail embodiments of the learning device and the like disclosed in this application, based on the drawings. However, these embodiments do not limit the disclosed technology. Furthermore, the embodiments described below may be modified as appropriate, provided that they do not contradict each other. [Examples]

[0010] <Configuration of air conditioner 10> Figure 1 shows an example of an air conditioner 10. The air conditioner 10 comprises a server 20, a plurality of outdoor units 30-1 to 30-m, and a plurality of indoor units 40-1 to 40-n (where m and n are natural numbers), as shown in Figure 1. Hereafter, when referring to the plurality of outdoor units 30-1 to 30-m without distinction, they will be referred to as outdoor unit 30, and when referring to the plurality of indoor units 40-1 to 40-n without distinction, they will be referred to as indoor unit 40.

[0011] Each outdoor unit 30 is connected to each indoor unit 40 via a liquid pipe 14, a diverter 12, and a liquid pipe 16. Also, each outdoor unit 30 is connected to each indoor unit 40 via a gas pipe 15, a diverter 13, and a gas pipe 17. In this way, the m outdoor units 30 and the n indoor units 40 are connected by the diverter 12, the diverter 13, the liquid pipe 14, the gas pipe 15, the liquid pipe 16, and the gas pipe 17, thereby forming the refrigerant circuit 19 of the air conditioner 10.

[0012] Also, each outdoor unit 30 is connected to the server 20 via a communication line 11. Each outdoor unit 30 communicates with the server 20 via the communication line 11 by means of wired communication or wireless communication.

[0013] The server 20 collects the detection values of the plurality of sensors each outdoor unit 30 has, and based on the collected detection values, generates, by machine learning, a learned model for determining the presence or absence of an abnormality of each sensor for each sensor. Then, the server 20 determines the presence or absence of an abnormality of each sensor using the learned model generated for each sensor and the detection value detected by each sensor. The server 20 is an example of a learning device and a sensor abnormality determination device.

[0014] <Configuration of the outdoor unit 30> FIG. 2 is a diagram showing an example of the outdoor unit 30. The outdoor unit 30 includes a first compressor 320-1, a second compressor 320-2, an oil separator 321, a four-way valve 322, an outdoor heat exchanger 323, an outdoor unit expansion valve 324, a shut-off valve 325, a shut-off valve 326, an accumulator 327, and an outdoor unit fan 328. These devices except the outdoor unit fan 328 are interconnected by the respective refrigerant pipes described in detail below, forming an outdoor unit refrigerant circuit 330 that is part of the refrigerant circuit 19.

[0015] The first compressor 320-1 is a variable-capacity compressor whose operating capacity can be varied by being driven by a motor (not shown) whose rotational speed is controlled by an inverter. The refrigerant discharge port of the first compressor 320-1 is connected to a discharge branch pipe 340-1 that branches off from one end of the discharge pipe 340. The other end of the discharge pipe 340 is connected to an oil separator 321, which will be described later. In other words, the refrigerant discharge port of the first compressor 320-1 is connected to the oil separator 321 via the discharge branch pipe 340-1 and the discharge pipe 340. The refrigerant inlet of the first compressor 320-1 is connected to a suction branch pipe 342-1 that branches off from one end of the suction pipe 342. The other end of the suction pipe 342 is connected to an accumulator 327, which will be described later. In other words, the refrigerant inlet of the first compressor 320-1 is connected to the accumulator 327 via the suction branch pipe 342-1 and the suction pipe 342.

[0016] The second compressor 320-2 is a variable-capacity compressor whose operating capacity can be varied by being driven by a motor (not shown) whose rotational speed is controlled by an inverter. The refrigerant discharge port of the second compressor 320-2 is connected to a discharge branch pipe 340-2 that branches off from one end of the discharge pipe 340. The other end of the discharge pipe 340 is connected to an oil separator 321, which will be described later. In other words, the refrigerant discharge port of the second compressor 320-2 is connected to the oil separator 321 via the discharge branch pipe 340-2 and the discharge pipe 340. The refrigerant inlet of the second compressor 320-2 is connected to a suction branch pipe 342-2 that branches off from one end of the suction pipe 342. The other end of the suction pipe 342 is connected to an accumulator 327, which will be described later. In other words, the refrigerant inlet of the second compressor 320-2 is connected to the accumulator 327 via the suction branch pipe 342-2 and the suction pipe 342.

[0017] In this embodiment, the first compressor 320-1 and the second compressor 320-2 are capable of performing the same function. Furthermore, during cooling operation, the rotational speeds of the first compressor 320-1 and the second compressor 320-2 are controlled so that the low-pressure saturation temperature, determined from the suction pressure detected by sensor 332, becomes a target low-pressure saturation temperature corresponding to the total cooling capacity required at various points in the indoor unit 40. Similarly, during heating operation, the rotational speeds of the first compressor 320-1 and the second compressor 320-2 are controlled so that the high-pressure saturation temperature, determined from the discharge pressure detected by sensor 331, becomes a target high-pressure saturation temperature corresponding to the total heating capacity required at various points in the indoor unit 40. The number of compressors is not limited to these; it can be one or three or more.

[0018] The oil separator 321 is a centrifugal separation type oil separator having a cylindrical sealed container. One end of the oil return pipe 347 is connected to the oil separator 321, and the other end of the oil return pipe 347 is connected to the suction pipe 342. A capillary tube 329 is provided in the oil return pipe 347. The oil separator 321 is also connected to port a of a four-way valve 322, which will be described later, via the outlet pipe 341. The oil separator 321 separates the refrigerant, including refrigerant oil, which is discharged from the first compressor 320-1 and the second compressor 320-2 respectively and flows in through the discharge branch pipe 340-1, the discharge branch pipe 340-2, and the discharge pipe 340, into refrigerant and refrigerant oil. The separated refrigerant oil is returned to the first compressor 320-1 and the second compressor 320-2, respectively, via the oil return pipe 347, the suction pipe 342, the suction branch pipe 342-1, and the suction branch pipe 342-2. The separated refrigerant flows out to the outlet pipe 341. Although refrigerant flows into the oil return pipe 347 along with the refrigerant oil, the amount of refrigerant returning to the first compressor 320-1 and the second compressor 320-2 is regulated by a capillary tube 329 provided in the oil return pipe 347.

[0019] The four-way valve 322 is a valve for switching the direction of refrigerant flow in the refrigerant circuit 19 and has four ports: port a, port b, port c, and port d. Port a is connected to the oil separator 321 via the outlet pipe 341 as described above. Port b is connected to one of the refrigerant inlets and outlets of the outdoor heat exchanger 323 via the refrigerant piping 343. Port c is connected to the refrigerant inlet of the accumulator 327 via the refrigerant piping 346. Port d is connected to the shut-off valve 326 via the outdoor unit gas pipe 345.

[0020] The outdoor heat exchanger 323 exchanges heat between the refrigerant and the outside air taken into the outdoor unit 30 by the rotation of the outdoor unit fan 328, which will be described later. As described above, one refrigerant inlet / outlet of the outdoor heat exchanger 323 is connected to port b of the four-way valve 322 via the refrigerant piping 343. The other refrigerant inlet / outlet of the outdoor heat exchanger 323 is connected to the shut-off valve 325 via the outdoor unit liquid pipe 344. The outdoor heat exchanger 323 functions as a condenser when the air conditioner 10 is in cooling operation and as an evaporator when the air conditioner 10 is in heating operation.

[0021] The outdoor unit expansion valve 324 is located on the outdoor unit liquid pipe 344. The outdoor unit expansion valve 324 is an electronic expansion valve driven by a pulse motor (not shown), and its opening degree is adjusted by the number of pulses supplied to the pulse motor, thereby adjusting the amount of refrigerant flowing into or out of the outdoor heat exchanger 323. The opening degree of the outdoor unit expansion valve 324 is adjusted so that when the air conditioner 10 is in heating operation, the superheating degree of the refrigerant flowing out of the outdoor heat exchanger 323 reaches a predetermined target refrigerant superheating degree (for example, 5°C). The opening degree of the outdoor unit expansion valve 324 is fully open when the air conditioner is in cooling operation.

[0022] As described above, the refrigerant inlet of the accumulator 327 is connected to port c of the four-way valve 322 via the refrigerant piping 346. The refrigerant outlet of the accumulator 327 is connected to the refrigerant inlet of the first compressor 320-1 via the suction pipe 342 and the suction branch pipe 342-1, and to the refrigerant inlet of the second compressor 320-2 via the suction pipe 342 and the suction branch pipe 342-2. The accumulator 327 separates the refrigerant that flows into the interior of the accumulator 327 from the refrigerant piping 346 into gaseous refrigerant and liquid refrigerant, and allows only the gaseous refrigerant to be drawn into the first compressor 320-1 and the second compressor 320-2.

[0023] The outdoor unit fan 328 is made of resin and is located near the outdoor heat exchanger 323. The outdoor unit fan 328 rotates with a fan motor (not shown) to draw outside air into the outdoor unit 30 through an intake port (not shown) provided in the housing of the outdoor unit 30. The outside air that has exchanged heat with the refrigerant in the outdoor heat exchanger 323 is discharged to the outside of the outdoor unit 30 through an outlet (not shown) provided in the housing of the outdoor unit 30.

[0024] In addition to the configuration described above, the outdoor unit 30 is equipped with various sensors. The first compressor 320-1 is equipped with a sensor 337-1 for detecting the temperature of the first compressor 320-1, and the second compressor 320-2 is equipped with a sensor 337-2 for detecting the temperature of the second compressor 320-2. The discharge pipe 340 is equipped with a sensor 331 for detecting the discharge pressure, which is the pressure of the refrigerant discharged from the first compressor 320-1 and the second compressor 320-2. Furthermore, the discharge branch pipe 340-1 is equipped with a sensor 333-1 for detecting the temperature of the refrigerant discharged from the first compressor 320-1. Furthermore, the discharge branch pipe 340-2 is equipped with a sensor 333-2 for detecting the temperature of the refrigerant discharged from the second compressor 320-2. Near the refrigerant inlet of the accumulator 327 in the refrigerant piping 346, there is a sensor 332 for detecting the suction pressure, which is the pressure of the refrigerant drawn into the first compressor 320-1 and the second compressor 320-2, and a sensor 334 for detecting the temperature of the refrigerant drawn into the first compressor 320-1 and the second compressor 320-2 (suction temperature).

[0025] Between the outdoor heat exchanger 323 and the outdoor unit expansion valve 324 in the outdoor unit liquid pipe 344, a sensor 335 is provided to detect the temperature of the refrigerant flowing into the outdoor heat exchanger 323, or the temperature of the refrigerant flowing out of the outdoor heat exchanger 323 (heat exchanger temperature). Furthermore, near the intake port of the housing (not shown) of the outdoor unit 30, a sensor 336 is provided to detect the temperature of the outside air flowing into the inside of the outdoor unit 30, i.e., the outside air temperature. The outside air temperature is an example of ambient temperature.

[0026] Furthermore, a sensor 338 for detecting the temperature of the refrigerant flowing through the outdoor unit liquid pipe 344 (first liquid pipe temperature) is provided between the outdoor unit expansion valve 324 and the shut-off valve 325 in the outdoor unit liquid pipe 344. Additionally, a sensor 339 for detecting the temperature of the refrigerant flowing through the outdoor unit liquid pipe 344 (second liquid pipe temperature) is provided between the outdoor heat exchanger 323 and the outdoor unit expansion valve 324 in the outdoor unit liquid pipe 344.

[0027] The outdoor unit 30 also includes a control device 300. The control device 300 is mounted on a control board housed in an electrical component box located inside the casing of the outdoor unit 30 (not shown). The control device 300 includes, for example, a communication unit 301, an acquisition unit 302, a control unit 303, and a storage unit 304, as shown in Figure 3. The communication unit 301 performs wired or wireless communication with the server 20 via the communication line 11. The communication unit 301 also performs wired or wireless communication with the control device 400 of the indoor unit 40.

[0028] The acquisition unit 302 acquires data such as detection values ​​detected by sensors placed in various parts of the outdoor unit 30 and outputs it to the control unit 303. The storage unit 304 has, for example, RAM (Random Access Memory), ROM (Read Only Memory), or SSD (Solid State Drive), and stores control programs 3040 and the like that are executed by the control unit 303.

[0029] The control unit 303, for example, has a CPU (Central Processing Unit) and controls each part of the outdoor unit 30 based on the detection values ​​of each sensor acquired by the acquisition unit 302 by executing the control program 3040 read from the storage unit 304. The control unit 303 also transmits operation data, including the detection values ​​of the sensors of each part of the outdoor unit 30 acquired by the acquisition unit 302, to the server 20 via the communication unit 301. The operation data includes data indicating the detection values ​​detected by each sensor, data indicating the operation status of operation data 230-1 and operation data 230-2, etc.

[0030] <Configuration of indoor unit 40> Figure 4 shows an example of an indoor unit 40. The indoor unit 40 includes an indoor heat exchanger 451, an indoor unit expansion valve 452, a liquid pipe connection 453, a gas pipe connection 454, and an indoor unit fan 455. Each of these components, excluding the indoor unit fan 455, is connected to each other by the refrigerant piping described in detail below, forming an indoor unit refrigerant circuit 450 which is part of the refrigerant circuit 19.

[0031] The indoor heat exchanger 451 exchanges heat between the refrigerant and indoor air drawn into the indoor unit 40 from an intake port (not shown) by the rotation of the indoor unit fan 455, which will be described later. One refrigerant inlet / outlet of the indoor heat exchanger 451 is connected to the liquid pipe connection part 453 via the indoor unit liquid pipe 471, and the other refrigerant inlet / outlet of the indoor heat exchanger 451 is connected to the gas pipe connection part 454 via the indoor unit gas pipe 472. The indoor heat exchanger 451 functions as an evaporator when the air conditioner 10 is in cooling operation, and functions as a condenser when the air conditioner 10 is in heating operation. The liquid pipe connection part 453 and the gas pipe connection part 454 are connected to each refrigerant pipe by welding or flare nuts, etc.

[0032] The indoor unit expansion valve 452 is located on the indoor unit liquid pipe 471. The indoor unit expansion valve 452 is an electronic expansion valve, and when the indoor heat exchanger 451 functions as an evaporator, i.e., when the indoor unit 40 is performing cooling operation, its opening is adjusted so that the degree of refrigerant superheat at the refrigerant outlet (gas pipe connection 454 side) of the indoor heat exchanger 451 becomes the target refrigerant superheat. Also, when the indoor heat exchanger 451 functions as a condenser, i.e., when the indoor unit 40 is performing heating operation, its opening is adjusted so that the degree of refrigerant subcooling at the refrigerant outlet (liquid pipe connection 453 side) of the indoor heat exchanger 451 becomes the target refrigerant subcooling. Here, the target refrigerant superheat and target refrigerant subcooling are the refrigerant superheat and refrigerant subcooling necessary for each of the indoor units 40 to exhibit sufficient cooling or heating capacity.

[0033] The indoor unit fan 455 is made of resin and is located near the indoor heat exchanger 451. The indoor unit fan 455 rotates with a fan motor (not shown) to draw indoor air into the indoor unit 40 through an intake port (not shown), and releases the indoor air, which has exchanged heat with the refrigerant in the indoor heat exchanger 451, into the room through an outlet (not shown).

[0034] In addition to the configuration described above, the indoor unit 40 is equipped with various sensors. Near the indoor heat exchanger 451 between the indoor heat exchanger 451 and the indoor unit expansion valve 452 in the indoor unit liquid pipe 471, there is a liquid-side temperature sensor 461 that detects the temperature of the refrigerant flowing into the indoor heat exchanger 451 during cooling operation. The liquid-side temperature sensor 461 also detects the temperature of the refrigerant flowing out of the indoor heat exchanger 451 during heating operation. Near the indoor heat exchanger 451 in the indoor unit gas pipe 472, there is a gas-side temperature sensor 462 that detects the temperature of the refrigerant flowing out of the indoor heat exchanger 451 during cooling operation, and the temperature of the refrigerant flowing into the indoor heat exchanger 451 during heating operation. Furthermore, near the intake port (not shown) of the indoor unit 40, there is an indoor temperature sensor 463 that detects the temperature of the indoor air flowing into the interior of the indoor unit 40.

[0035] The indoor unit 40 also includes a control device 400. The control device 400 is mounted on a control board housed in an electrical component box located inside the enclosure (not shown) of the indoor unit 40. The control device 400 includes, for example, a communication unit 401, an acquisition unit 402, a control unit 403, and a storage unit 404, as shown in Figure 5. The communication unit 401 communicates with the control device 300 of the outdoor unit 30 via wired or wireless communication. The communication unit 401 may also communicate with the server 20 via the communication line 11.

[0036] The acquisition unit 402 acquires detection values ​​from sensors located in various parts of the indoor unit 40 and outputs them to the control unit 403. The storage unit 404 has, for example, RAM, ROM, or SSD, and stores the control program 4040, etc., which is executed by the control unit 403. The control unit 403 has, for example, a CPU, and by executing the control program 4040 read from the storage unit 404, controls each part of the indoor unit 40 based on the detection values ​​of each sensor acquired by the acquisition unit 402.

[0037] <Server 20 Configuration> Figure 6 is a block diagram showing an example of a server 20. The server 20 includes a communication unit 21, a control unit 22, and a storage unit 23.

[0038] The communication unit 21 communicates with the control devices 300 of each outdoor unit 30 via the communication line 11. For example, the communication unit 21 receives operation data from the control devices 300 of each outdoor unit 30 via the communication line 11 and outputs the received operation data to the control unit 22. The communication unit 21 may also communicate with the control devices 400 of the indoor units 40 via the communication line 11.

[0039] The memory unit 23 is composed of, for example, RAM, ROM, or SSD, and stores operation data 230, a learning program 231, a learned model 232, and a judgment program 233. The operation data 230 stores a detection value table 2301 for each outdoor unit ID 2300 that identifies the outdoor unit 30, as shown in Figure 7, for example. The detection value table 2301 stores the detection values ​​of each sensor, associated with the detection time. The operation data 230 also stores an operation status table 2302 for each outdoor unit ID 2300 that identifies the outdoor unit 30, as shown in Figure 8, for example. The operation status table 2302 stores information indicating the start and stop times of operation of the first compressor 320-1 and the second compressor 320-2.

[0040] The control unit 22, for example, has a CPU and controls each part of the server 20 based on a program read from the storage unit 23. For example, the control unit 22 receives operation data from the outdoor unit 30 via the communication unit 21 and stores the received operation data as operation data 230 in the storage unit 23. The control unit 22 also generates a trained model using the operation data 230 by executing a learning program 231 read from the storage unit 23. The control unit 22 then stores the generated trained model as trained model 232 in the storage unit 23. The control unit 22 also determines whether there is an abnormality in each sensor by executing a determination program 233 read from the storage unit 23, using the operation data 230 and the trained model 232.

[0041] <Learning Process> Figure 9 is a block diagram showing an example of the functional configuration of the control unit 22 of the server 20 in the process of generating a trained model 232. The control unit 22 implements the operation data acquisition unit 220 and the learning unit 221 by executing the learning program 231 read from the storage unit 23. The operation data acquisition unit 220 is an example of an acquisition unit, and the learning unit 221 is an example of a generation unit.

[0042] Here, we will explain the changes in the detected values ​​of each sensor installed on the outdoor unit 30. Figure 10 is a diagram showing an example of the changes in the detected values ​​of each sensor after both the first compressor 320-1 and the second compressor 320-2 have stopped operating. When the operation of the first compressor 320-1 stops, the temperatures detected by each sensor gradually converge to a specific temperature, as shown in Figure 10, for example. Figure 10 illustrates the changes in the temperature of the first compressor 320-1, the temperature of the second compressor 320-2, the discharge temperature of the first compressor 320-1, the discharge temperature of the second compressor 320-2, the heat exchanger temperature, the first liquid pipe temperature, the second liquid pipe temperature, the outside air temperature, the suction temperature, and the discharge pipe pressure (saturation temperature). In the example in Figure 10, the second compressor 320-2 is stopped.

[0043] The temperature of the first compressor 320-1 is detected by sensor 337-1, the temperature of the second compressor 320-2 is detected by sensor 337-2, the discharge temperature of the first compressor 320-1 is detected by sensor 333-1, and the discharge temperature of the second compressor 320-2 is detected by sensor 333-2. In addition, the heat exchanger temperature is detected by sensor 335, the temperature of the first liquid pipe is detected by sensor 338, the temperature of the second liquid pipe is detected by sensor 339, the ambient temperature is detected by sensor 336, and the suction temperature is detected by sensor 334. Furthermore, the discharge pipe pressure (saturation temperature) is calculated using the detected value of the discharge pressure, for example, detected by sensor 331. Sensor 336 is an example of a first sensor, and the ambient temperature detected by sensor 336 is an example of a first detected value. Furthermore, sensors 333-1, 333-2, 334, 335, 337-1, and 337-2 are examples of a second or third sensor. Also, the detected values ​​detected by each of sensors 333-1, 333-2, 334, 335, 337-1, and 337-2 are examples of a second or third detected value.

[0044] Referring to Figure 10, from timing t1, which is 10 minutes after timing t0 when both the first compressor 320-1 and the second compressor 320-2 stopped operating, the temperature of the first compressor 320-1, the temperature of the second compressor 320-2, the discharge temperature of the first compressor 320-1, and the discharge temperature of the second compressor 320-2 gradually converge to their respective specific temperatures. Also referring to Figure 10, from timing t1 when both the first compressor 320-1 and the second compressor 320-2 stopped operating, the heat exchanger temperature, the first liquid pipe temperature, the second liquid pipe temperature, the suction temperature, and the discharge pipe pressure (saturation temperature) converge to the ambient temperature.

[0045] Furthermore, from timing t2 onward, 30 minutes after timing t0 when both the first compressor 320-1 and the second compressor 320-2 have stopped operating, the temperature of the first compressor 320-1, the temperature of the second compressor 320-2, the discharge temperature of the first compressor 320-1, the discharge temperature of the second compressor 320-2, the heat exchanger temperature, the temperature of the first liquid pipe, the temperature of the second liquid pipe, the suction temperature, and the discharge pipe pressure (saturation temperature) are even more converged to their respective specific temperatures or ambient temperatures compared to timing t1.

[0046] The temperature of the first compressor 320-1, the temperature of the second compressor 320-2, the discharge temperature of the first compressor 320-1, the discharge temperature of the second compressor 320-2, the heat exchanger temperature, the temperature of the first liquid pipe, the temperature of the second liquid pipe, the suction temperature, and the discharge pipe pressure (saturation temperature) are maintained at a converged state until operation of at least one of the first compressor 320-1 and the second compressor 320-2 is resumed. Hereinafter, period T will refer to the period from timing t1, when a predetermined time (e.g., 10 minutes) has elapsed since the operation of both the first compressor 320-1 and the second compressor 320-2 was stopped, until operation of at least one of the first compressor 320-1 and the second compressor 320-2 is resumed.

[0047] Figure 11 shows an example of the relationship between sensor readings and ambient temperature during a period T in which both the first compressor 320-1 and the second compressor 320-2 are stopped from timing t1 onwards. Note that the stopping time of the first compressor 320-1 and the second compressor 320-2 is not considered within period T. In the example in Figure 11, the intake temperature detected by sensor 334 is shown as an example of a sensor reading detected during period T. For example, as shown in Figure 11, there is a certain correlation between sensor readings and ambient temperature during period T. Therefore, by machine learning the relationship between sensor readings and ambient temperature during period T, it is possible to generate a trained model for estimating sensor readings during period T in relation to ambient temperature.

[0048] Returning to Figure 9, the explanation continues. The operation data acquisition unit 220 refers to the operation data 230 in the storage unit 23 and extracts the detected values ​​of each sensor during the period T after the operation of the first compressor 320-1 and the second compressor 320-2 has stopped. In this embodiment, the operation data acquisition unit 220 performs a filtering process to exclude detected values ​​from the extracted operation data that are unsuitable for machine learning. Detected values ​​unsuitable for machine learning include, for example, obviously invalid detected values ​​and detected values ​​that have been detected multiple times.

[0049] The operation data acquisition unit 220 then outputs operation data, including the filtered detection values ​​of each sensor, to the learning unit 221. The operation data 230 used in the learning process may be a collection of detection value data from sensors of multiple different outdoor units 30, or it may be a collection of detection value data from sensors of the outdoor unit 30 to which the system is applied.

[0050] The learning unit 221 generates a trained model for each sensor using operating data from the period T after the operation of the first compressor 320-1 and the second compressor 320-2 has stopped. Then, for each sensor, the learning unit 221 generates a trained model using machine learning with the operating data after filtering. For example, the learning unit 221 generates a trained model using regression analysis (e.g., simple linear regression) with the outside air temperature as the explanatory variable and the detected values ​​of sensors other than the outside air temperature as the dependent variable from the operating data for period T.

[0051] For example, the learning unit 221 generates a trained model for sensor 337-1 using the ambient temperature as the explanatory variable and the detected value of sensor 337-1, which detects the temperature of the first compressor 320-1, as the objective variable, based on the operating data during period T. The learning unit 221 also generates a trained model for sensor 337-2 using the ambient temperature as the explanatory variable and the detected value of sensor 337-2, which detects the temperature of the second compressor 320-2, as the objective variable, based on the operating data during period T. The trained model corresponding to sensor 337-1 is an example of the first trained model, and the trained model corresponding to sensor 337-2 is an example of the second trained model.

[0052] Furthermore, the learning unit 221 generates a threshold value for each sensor that will be judged as an abnormal value based on the sensor's detected value relative to the ambient temperature. The learning unit 221 then stores the threshold values ​​generated for each sensor in the storage unit 23. For example, the learning unit 221 calculates the difference between each detected value and the estimated value estimated by the trained model for each sensor. For example, if the estimated value is smaller than the detected value, the difference will be a positive value, and if the estimated value is larger than the detected value, the difference will be a negative value. Then, based on the calculated distribution of differences, a threshold value is calculated for a predetermined percentage of detected values ​​to be judged as normal. For example, a threshold value is calculated for 99.7% (μ ± 3σ) of the calculated distribution of differences to be judged as normal. In the calculated distribution of differences, μ is the mean value of the difference, and σ is the standard deviation of the difference. This allows for the calculation of threshold values ​​such as those shown in Figure 11.

[0053] <Decision Process> Figure 12 is a block diagram showing an example of the functional configuration of the control unit 22 of the server 20 in the process of determining whether or not there is a sensor abnormality. The control unit 22 implements the operation data acquisition unit 220, estimation unit 222, and determination unit 223 by executing the determination program 233 read from the storage unit 23. The determination process is performed at predetermined intervals (for example, every day) during actual operation after the outdoor unit 30 has been shipped.

[0054] The operation data acquisition unit 220 refers to the operation data 230 in the storage unit 23 and extracts the detected values ​​of each sensor during the period T after the operation of the first compressor 320-1 and the second compressor 320-2 has stopped, for a period from the present to a predetermined time (for example, 1 day) prior. The operation data acquisition unit 220 also reads the threshold values ​​for each sensor from the storage unit 23. Then, the operation data acquisition unit 220 outputs the detected value indicating the outside air temperature from the extracted detected values ​​of each sensor to the estimation unit 222, and outputs the detected values ​​of other sensors detected at the timing corresponding to the outside air temperature, along with their threshold values, to the determination unit 223.

[0055] The estimation unit 222 reads a learned model 232 from the storage unit 23 for each sensor, and uses the read learned model 232 to calculate an estimated value corresponding to the outside air temperature output from the operation data acquisition unit 220. Then, the estimation unit 222 outputs the calculated estimated value for each sensor to the determination unit 223.

[0056] For example, the estimation unit 222 inputs the ambient temperature output from the operation data acquisition unit 220 into a trained model corresponding to the sensor 337-1 that detects the temperature of the first compressor 320-1, thereby calculating an estimated value of the sensor 337-1's detection value. The estimation unit 222 also inputs the ambient temperature output from the operation data acquisition unit 220 into a trained model corresponding to the sensor 337-2 that detects the temperature of the second compressor 320-2, thereby calculating an estimated value of the sensor 337-2's detection value. The estimated value of the sensor 337-1's detection value is an example of an estimated value of the second sensor's detection value, and the estimated value of the sensor 337-2's detection value is an example of an estimated value of the third sensor's detection value.

[0057] The determination unit 223 determines whether there is an abnormality in each sensor based on the detected value and threshold output from the operation data acquisition unit 220 and the estimated value calculated by the estimation unit 222. For example, the determination unit 223 calculates the difference between the detected value output from the operation data acquisition unit 220 and the estimated value calculated by the estimation unit 222 for each sensor. Then, the determination unit 223 determines whether the calculated difference exceeds the threshold for each sensor.

[0058] For example, if the calculated difference for all sensors is below the threshold, the determination unit 223 determines that all sensors are normal. If the calculated difference for some sensors exceeds the threshold, the determination unit 223 determines that some of those sensors are abnormal. On the other hand, if the calculated difference for all sensors exceeds the threshold, the determination unit 223 determines that the sensor 336 that detects the outside air temperature is abnormal. The determination unit 223 then outputs the determination result to equipment at a maintenance site for the air conditioner 10.

[0059] <Steps for the learning process> Figure 13 is a flowchart illustrating an example of the process for generating a trained model 232. Each process illustrated in Figure 13 is implemented by the control unit 22 of the server 20 executing the training program 231 read from the storage unit 23. The training process illustrated in Figure 13 is performed using data collected in advance before the outdoor unit 30 is shipped. Furthermore, the training process illustrated in Figure 13 is performed for each outdoor unit 30. Note that the training process illustrated in Figure 13 may also be performed as additional training using data collected sequentially during actual operation after the outdoor unit 30 has been shipped. The training process illustrated in Figure 13 is an example of a training method.

[0060] First, the operation data acquisition unit 220 acquires operation data 230 from the storage unit 23 (step S100). Then, the operation data acquisition unit 220 extracts operation data from the acquired operation data 230 for the period T after the operation of the first compressor 320-1 and the second compressor 320-2 has stopped (step S101).

[0061] Next, the operation data acquisition unit 220 filters the operation data (step S102). In step S102, the operation data acquisition unit 220 excludes detection values ​​from the extracted operation data that are unsuitable for machine learning for each sensor. Then, the operation data acquisition unit 220 outputs the filtered operation data to the learning unit 221.

[0062] Next, the learning unit 221 sequentially selects the sensors of the outdoor unit 30 one by one and performs the process of generating a trained model and generating thresholds. First, the learning unit 221 selects one unselected sensor from among the sensors of the outdoor unit 30 (step S103). Then, the learning unit 221 generates a trained model 232 using machine learning with the filtered operating data for the sensor selected in step S103 (step S104). For example, the learning unit 221 generates a trained model 232 using regression analysis (e.g., simple linear regression) with the outside air temperature as the explanatory variable and the detected value of the sensor selected in step S103 as the dependent variable from the filtered operating data. Then, the learning unit 221 stores the generated trained model 232 in the storage unit 23.

[0063] Next, the learning unit 221 generates a threshold value based on the difference between the detected value included in the filtered driving data and the estimated value calculated using the trained model 232 generated in step S104 (step S105). Then, the learning unit 221 stores the generated threshold value in the storage unit 23.

[0064] Next, the learning unit 221 determines whether or not all sensors on the outdoor unit 30 have been selected (step S106). If there are unselected sensors (step S106: No), the process shown in step S103 is executed again. On the other hand, if all sensors have been selected (step S106: Yes), the learning process shown in this flowchart is terminated.

[0065] <Procedure for the judgment process> Figure 14 is a flowchart illustrating an example of a process for determining whether or not a sensor is abnormal. Each process illustrated in Figure 14 is implemented by the control unit 22 of the server 20 executing a determination program 233 read from the storage unit 23. The determination processes illustrated in Figure 14 are performed at predetermined intervals (for example, every day). The determination processes illustrated in Figure 14 are an example of a sensor abnormality detection method.

[0066] First, the operation data acquisition unit 220 acquires operation data 230 from the storage unit 23 along with threshold values ​​for a period from the present to a predetermined time (for example, 1 day) prior (S200). The operation data 230 includes multiple detection values ​​from each sensor. Then, the operation data acquisition unit 220 extracts operation data from the acquired operation data 230 for the period T after the operation of the first compressor 320-1 and the second compressor 320-2 has stopped (step S201).

[0067] Next, the operation data acquisition unit 220 filters the operation data (step S202). In step S202, the operation data acquisition unit 220 excludes detection values ​​from the extracted operation data that are unsuitable for estimating the detection value for each sensor. Then, the operation data acquisition unit 220 outputs the operation data containing the detection value indicating the outside air temperature from the filtered operation data to the estimation unit 222, and outputs the operation data containing the detection values ​​of other sensors detected at the timing corresponding to the outside air temperature, along with the threshold value, to the determination unit 223.

[0068] Next, the estimation unit 222 selects one unselected sensor from among the sensors (excluding the sensor that detects outside air temperature) of the outdoor unit 30 (step S203). Then, the estimation unit 222 calculates an estimated value of the detected value for the sensor selected in step S203 (step S204). In step S204, the estimation unit 222 reads the learned model 232 corresponding to the sensor selected in step S203 from the storage unit 23. Then, the estimation unit 222 uses the read learned model 232 to calculate an estimated value corresponding to each detected outside air temperature included in the operating data output from the operating data acquisition unit 220. Finally, the estimation unit 222 outputs the calculated estimated value to the determination unit 223.

[0069] Furthermore, if the operating data filtered in step S202 includes detected values ​​for different ambient temperatures, an estimated value of the detected value from the selected sensor is calculated in step S203 for each detected ambient temperature.

[0070] Next, the determination unit 223 calculates the difference between the sensor's detected value and the estimated value output from the estimation unit 222 for each ambient temperature for the sensor selected in step S203 (step S205). If the operating data filtered in step S202 contains multiple different detected values ​​for the same ambient temperature, the difference between the detected value and the estimated value is calculated for each detected value.

[0071] Next, the determination unit 223 averages the differences calculated in step S205 (step S206). Then, the determination unit 223 determines whether or not all sensors on the outdoor unit 30 have been selected (step S207). If there are unselected sensors (step S207: No), the process shown in step S203 is executed again.

[0072] On the other hand, if all sensors are selected (step S207: Yes), the determination unit 223 determines whether or not there are any sensors whose average difference exceeds a threshold (step S208). If there are no sensors whose average difference exceeds a threshold (step S208: No), the determination unit 223 determines that all sensors are normal (step S210). The determination unit 223 then outputs the determination result to the equipment at the maintenance site for the air conditioner 10 (step S213), and the determination process shown in this flowchart ends.

[0073] On the other hand, if there is a sensor whose average difference exceeds the threshold (step S208: Yes), the determination unit 223 determines whether the average difference exceeds the threshold for all sensors on the outdoor unit 30 (step S209). If the average difference exceeds the threshold for all sensors on the outdoor unit 30 (step S209: Yes), the determination unit 223 determines that the sensor 336 that detects the outside air temperature is abnormal (step S211) and executes the process shown in step S213.

[0074] On the other hand, if the average difference value of some of the sensors on the outdoor unit 30 exceeds a threshold value (step S209: No), the determination unit 223 determines that the sensor whose average difference value exceeds the threshold value is abnormal (step S212) and executes the process shown in step S213.

[0075] <Effects of the Example> In this embodiment, by using a trained model, the presence or absence of a sensor abnormality is determined from the detected value during the period T of operation stoppage, starting from timing t1, which is 10 minutes after both the first compressor 320-1 and the second compressor 320-2 have stopped. Therefore, the presence or absence of a sensor abnormality can be determined with sufficient accuracy in a very short time of 10 minutes after both the first compressor 320-1 and the second compressor 320-2 have stopped. As a result, even when the operating time of the air conditioner 10 is long and the stoppage time of the air conditioner 10 is short, such as in the summer, the presence or absence of a sensor abnormality can be determined. Furthermore, since the presence or absence of a sensor abnormality can be determined even when the stoppage time of the air conditioner 10 is short, the presence or absence of a sensor abnormality can be determined during operational checks after the completion of installation work.

[0076] [others] Furthermore, the technology disclosed in this application is not limited to the embodiments described above, and numerous modifications are possible within the scope of its essence.

[0077] For example, in the embodiment described above, a trained model is generated for each sensor by simple linear regression analysis, with the ambient temperature as the explanatory variable and the sensor's detected value as the dependent variable. However, the disclosed technology is not limited to this. As another example, for each sensor, a trained model may be generated by multiple linear regression analysis, with the ambient temperature and elapsed time (the time elapsed since both compressors stopped) as explanatory variables and the sensor's detected value as the dependent variable, for calculating the sensor's detected value for each elapsed time from the ambient temperature. The estimation unit 222 may then use the trained model generated by multiple linear regression analysis to calculate an estimated value of the sensor's detected value from the ambient temperature and elapsed time. This allows for a determination of whether or not there is a sensor abnormality that more closely reflects the actual trend of the detected value, thereby improving the accuracy of the determination.

[0078] Furthermore, in the above-described embodiment, a trained model is generated for each sensor using the sensor detection values ​​collected in advance before the outdoor unit 30 is shipped, but the disclosed technology is not limited to this. As another example, the trained model may be further trained using the detection values ​​of sensors that are determined to be normal among the detection values ​​of each sensor detected during actual operation after the outdoor unit 30 has been shipped. This makes it possible to determine whether or not there is a sensor abnormality in a way that is more in line with the actual sensor detection values.

[0079] Furthermore, in the above-described embodiment, abnormality detection was explained using sensors that detect the temperature of the first compressor 320-1, the temperature of the second compressor 320-2, the discharge temperature of the first compressor 320-1, the discharge temperature of the second compressor 320-2, the heat exchanger temperature, the temperature of the first liquid pipe, the temperature of the second liquid pipe, the ambient temperature, the suction temperature, and the discharge pipe pressure (saturation temperature). However, sensors other than the sensor that detects the ambient temperature are not necessarily required, and the type of sensor is not limited to these; any sensor that detects temperature or pressure can be used to perform abnormality detection.

[0080] Furthermore, in the embodiments described above, the server 20 and the outdoor unit 30 are implemented as separate devices, but the disclosed technology is not limited to this. As another example, some or all of the functions of the server 20 may be incorporated into the outdoor unit 30. For example, the function of estimating the detected value of a sensor and the function of determining whether or not there is a sensor abnormality may be incorporated into each outdoor unit 30. In this case, the server 20 provides the corresponding outdoor unit 30 with a trained model generated for each sensor included in each outdoor unit 30. Each outdoor unit 30 may use the trained model for each sensor provided by the server 20 to determine whether or not there is a sensor abnormality.

[0081] Furthermore, in the above-described embodiment, the presence or absence of abnormalities in the sensors inside the outdoor unit 30 is determined, but the disclosed technology is not limited to this. As another example, the presence or absence of abnormalities in the sensors inside the indoor unit 40 may also be determined in a similar manner by using the indoor temperature sensor 463 as the first sensor.

[0082] It should be noted that the embodiments disclosed herein are illustrative and not restrictive in all respects. Indeed, the embodiments described above can be embodied in a variety of forms. Furthermore, the embodiments described above may be omitted, replaced, or modified in various ways without departing from the scope and spirit of the appended claims. [Explanation of Symbols]

[0083] 10. Air conditioner 20 servers 21 Communications Department 22 Control Unit 220 Operation data acquisition unit 221 Learning Department 222 Estimation Department 223 Judgment section 23 Memory section 230 Driving Data 231 Learning Programs 232 Pre-trained Models 233 Judgment Program 30 Outdoor unit 300 Control device 320-1 First Compressor 320-2 Second Compressor 331 Sensor 332 sensors 333-1 Sensor 333-2 Sensor 334 Sensors 335 Sensor 336 Sensors 337-1 Sensor 337-2 Sensor 338 sensors 339 Sensors 40 Indoor unit

Claims

1. An acquisition unit that acquires operating data including a first detection value detected by a first sensor provided in the air conditioner for detecting ambient temperature, the first detection value detected while the air conditioner is stopped, and a second detection value detected by a second sensor provided in the air conditioner, the second detection value detected while the air conditioner is stopped. A generation unit generates a trained model for calculating an estimated value of the second detected value from the first detected value to determine whether the second sensor is abnormal or not, using the aforementioned operating data. A learning device characterized by being equipped with the following features.

2. The learning device according to claim 1, characterized in that the generation unit generates the trained model by regression analysis.

3. The acquisition unit acquires operating data including the second detected value for each elapsed time, which is the time elapsed since the air conditioner stopped operating. The learning device according to claim 1, characterized in that the generation unit generates a trained model for calculating the second detected value for each elapsed time from the first detected value using the operating data by multiple regression analysis.

4. The acquisition unit acquires operating data which further includes a third detection value detected by a third sensor provided in the air conditioner, and which includes a third detection value detected while the air conditioner is stopped. The generating unit is Using the aforementioned operating data, a first trained model is generated for calculating the second detected value based on the first detected value. The learning device according to claim 1, characterized in that it generates a second trained model for calculating the third detected value based on the first detected value using the aforementioned operating data.

5. An acquisition unit that acquires from the air conditioner a first detection value detected by a first sensor provided in the air conditioner for detecting ambient temperature, the first detection value detected while the air conditioner is stopped, and a second detection value detected by a second sensor provided in the air conditioner, the second detection value detected while the air conditioner is stopped. The acquisition unit inputs the first detected value acquired by the acquisition unit into a trained model, and the estimation unit calculates an estimated value of the second detected value corresponding to the first detected value. A determination unit determines whether the second sensor is abnormal or not based on the second detection value acquired by the acquisition unit and the estimated value of the second detection value calculated by the estimation unit. A sensor abnormality detection device characterized by comprising the following features.

6. The acquisition unit further acquires a third detection value detected by a third sensor provided in the air conditioner, which is a third detection value detected while the air conditioner is stopped from operation. The aforementioned trained model, The first trained model used to calculate the estimated value of the second detected value, The second trained model used to calculate the estimated value of the third detected value and Includes, The estimation unit, The acquisition unit inputs the first detected value acquired by the acquisition unit into the first trained model to calculate an estimated value of the second detected value. The acquisition unit inputs the first detected value acquired by the acquisition unit into the second trained model to calculate an estimated value of the third detected value. The determination unit, The sensor abnormality determination device according to claim 5, further characterized in that it determines whether the third sensor is abnormal or not based on the third detection value acquired by the acquisition unit and the estimated value of the third detection value calculated by the estimation unit.

7. The acquisition unit acquires the second detection value for each elapsed time, which is the time elapsed since the air conditioner stopped operating. The estimation unit inputs the first detected value acquired by the acquisition unit and the elapsed time into the trained model to calculate an estimated value of the second detected value at the elapsed time. The determination unit, The sensor abnormality determination device according to claim 5, characterized in that it determines whether the second sensor is abnormal or not based on the second detection value for each elapsed time obtained by the acquisition unit and the estimated value of the second detection value for each elapsed time calculated by the estimation unit.

8. Computers A step of acquiring operating data including a first detection value detected by a first sensor provided in the air conditioner for detecting ambient temperature, the first detection value detected while the air conditioner is stopped, and a second detection value detected by a second sensor provided in the air conditioner, the second detection value detected while the air conditioner is stopped. A step of generating a trained model for calculating an estimated value of the second detected value for determining whether the second sensor is abnormal or not, using the aforementioned operating data, from the first detected value. A learning method characterized by performing the following.

9. Computers A step of obtaining from the air conditioner a first detection value detected by a first sensor provided in the air conditioner for detecting ambient temperature, the first detection value detected while the air conditioner is stopped, and a second detection value detected by a second sensor provided in the air conditioner, the second detection value detected while the air conditioner is stopped. The process involves inputting the acquired first detected value into a trained model to calculate an estimated second detected value corresponding to the first detected value, A step of determining whether the second sensor is abnormal or not based on the acquired second detection value and the calculated estimated value of the second detection value. A sensor anomaly detection method characterized by performing the following.

10. The aforementioned acquisition step is to acquire a third detection value detected by a third sensor provided in the air conditioner, and further acquire the third detection value detected while the air conditioner is stopped from operation. The aforementioned trained model, The first trained model used to calculate the estimated value of the second detected value, The second trained model used to calculate the estimated value of the third detected value and Includes, The calculation process described above is: The process involves inputting the acquired first detected value into the first trained model to calculate an estimated value of the second detected value, The process involves inputting the acquired first detected value into the second trained model to calculate an estimated value of the third detected value. Includes, The aforementioned determination process is: A step of determining whether the third sensor is abnormal or not based on the acquired third detection value and the calculated estimated value of the third detection value, If both the second sensor and the third sensor are determined to be abnormal, the first sensor is determined to be abnormal. A sensor abnormality determination method according to claim 9, characterized by including the following:

11. On the computer, A step of acquiring operating data including a first detection value detected by a first sensor provided in the air conditioner for detecting ambient temperature, the first detection value detected while the air conditioner is stopped, and a second detection value detected by a second sensor provided in the air conditioner, the second detection value detected while the air conditioner is stopped. A step of generating a trained model for calculating an estimated value of the second detected value for determining whether the second sensor is abnormal or not, using the aforementioned operating data, from the first detected value. A program characterized by causing the execution of a specific action.

12. On the computer, A step of obtaining from the air conditioner a first detection value detected by a first sensor provided in the air conditioner for detecting ambient temperature, the first detection value detected while the air conditioner is stopped, and a second detection value detected by a second sensor provided in the air conditioner, the second detection value detected while the air conditioner is stopped. The process involves inputting the acquired first detected value into a trained model to calculate an estimated second detected value corresponding to the first detected value, A step of determining whether the second sensor is abnormal or not based on the acquired second detection value and the calculated estimated value of the second detection value. A program characterized by causing the execution of a specific action.

Images