Air conditioners and air conditioning systems

Abstract

To provide an air conditioner which can achieve both the diversification of an AI function and a comfort of a user.SOLUTION: An air conditioner comprises an indoor machine and an outdoor machine. The air conditioner comprises a detection part for detecting an operation state amount of the air conditioner, an operation part for allowing a user to operate the air conditioner, and an inference part for inferring, on the basis of a detection result of the detection part, a succeeding operation method. The inference part has a first learning model for inferring the operation method related to a comfort of a user on the basis of a learning result of learning a tendency of an operation by the user. The operation part has an activation button for activating the inference part, and when the activation button is operated, the first learning model is activated.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to an air conditioner and an air conditioning system. [Background technology]

[0002] For example, some air conditioners are equipped with a learning model that learns people's operating tendencies for setting the temperature and predicts an appropriate set temperature. Furthermore, some air conditioners equipped with a prediction function using a learning model (hereinafter also referred to as an AI function) are known to have a button (hereinafter referred to as an AI start button) on the remote control that operates the air conditioner to start the learning model and activate the AI ​​function, thereby making it easier to operate the learning model and improving convenience. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2018-200127 [Patent Document 2] Japanese Patent Application Laid-Open No. 2015-117933 Summary of the Invention [Problem to be solved by the invention]

[0004] In recent years, learning models have emerged that not only ensure user comfort during air conditioning operation, but also provide user convenience in controlling air conditioning operation (e.g., controlling air conditioning operation on / off). Learning models related to user comfort include, for example, a perceived temperature prediction model that predicts the perceived temperature of a user in an air-conditioned space, reflecting the user's past operating tendencies, and automatically changes the set temperature. In contrast, learning models related to user convenience include, for example, a presence / absence prediction model that automatically stops air conditioning operation to save power if it predicts that the user will be absent from the air-conditioned space.

[0005] If an air conditioner has multiple learning models, for example, if a user operates the AI ​​start button to activate multiple AI functions at once, the learning model related to the user's comfort and the learning model related to the user's convenience in operating the air conditioning will be activated.

[0006] However, when a learning model related to user comfort and a learning model related to user convenience are activated, it is possible that user comfort will be affected. For example, if the presence / absence prediction model predicts a tendency for users to be absent from the air-conditioned space and stops the air conditioning, if the user returns to the indoor space in a short time, contrary to the trend, it may take time for the room temperature to return to a comfortable temperature for the user, which could result in a loss of user comfort. In this way, equipping air conditioners with a variety of AI functions and making them easy to execute can lead to situations where user comfort is reduced.

[0007] In view of these problems, the present invention aims to provide an air conditioner that can achieve both diversification of AI functions and user comfort. [Means for solving the problem]

[0008] One embodiment of the air conditioner is an air conditioner comprising an indoor unit and an outdoor unit. The air conditioner comprises a detection unit that detects operational state quantities of the air conditioner, an operation unit that allows a user to operate the air conditioner, and an inference unit that infers a subsequent operating method based on the detection results of the detection unit. The inference unit has a first learning model that infers an operating method related to the user's comfort based on learning results that have learned the user's operating tendencies. The operation unit has a start button that starts the inference unit, and when the start button is operated, the first learning model starts. [Effects of the Invention]

[0009] One aspect of the air conditioner of the present invention is that it can achieve both diversified use of AI functions and user comfort. [Brief explanation of the drawings]

[0010] [Figure 1] FIG. 1 is an explanatory diagram showing an example of an air conditioning system according to this embodiment. [Figure 2] FIG. 2 is a block diagram showing an example of the configuration of an air conditioner. [Figure 3] FIG. 3 is a block diagram showing an example of the configuration of a communications adapter. [Figure 4] FIG. 4 is a block diagram showing an example of the configuration of the server device. [Figure 5] FIG. 5 is a block diagram showing an example of the configuration of a remote controller. [Figure 6] FIG. 6 is an explanatory diagram showing an example of the display content on the touch panel of the remote controller. [Figure 7] FIG. 7 is an explanatory diagram showing an example of transition of display content on the touch panel. [Figure 8] FIG. 8 is a flowchart showing an example of the processing operation of the communication adapter related to the communication adapter side control processing. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, embodiments of the air conditioner and air conditioning system disclosed in the present application will be described in detail with reference to the drawings. Note that the disclosed technology is not limited to these embodiments. Furthermore, each embodiment described below may be modified as appropriate within the scope of not causing any contradiction. [Example]

[0012] <Air conditioning system configuration> Fig. 1 is an explanatory diagram showing an example of an air conditioning system 1 according to Example 1. The air conditioning system 1 shown in Fig. 1 includes an air conditioner 2, a communication adapter 3, a router 4, a server device 5, a relay device 6, a terminal device 7, and a communication network 8.

[0013] <Air conditioner configuration> FIG. 2 is a block diagram illustrating an example of the configuration of an air conditioner 2. The air conditioner 2 illustrated in FIG. 2 includes an indoor unit 21, an outdoor unit 22, and a remote control 23. The indoor unit 21 is, for example, a part of the air conditioner 2 that is placed indoors and heats or cools the air in the air-conditioned space. An indoor unit 21 is provided for each air-conditioned space, such as a living room or a bedroom. The indoor unit 21 includes a main body 21A, a sensor 21B, a light-receiving unit 21C, a control unit 21D, and a memory 21E. The main body 21A includes an indoor fan, a wind direction panel, an indoor heat exchanger, and other components (not shown). The indoor air exchanges heat with a refrigerant supplied from the outdoor unit 22 in the indoor heat exchanger, and is then blown out by the indoor fan to heat, cool, dehumidify, or otherwise operate the room. The sensor 21B is a detection unit that detects operational state quantities, such as the indoor temperature, the outdoor temperature, the heat exchanger temperature, the compressor rotation speed, the user's operation of the air conditioner, and detection results indicating the presence or absence of a person in the air-conditioned space. The sensor 21B includes, for example, a human detection sensor that detects the presence or absence of a person in the air-conditioned space. The light receiving unit 21C receives a command signal from the remote control 23 and transmits the received command signal to the control unit 21D. The memory 21E is an area for storing various information. The control unit 21D controls the entire indoor unit 21. The control unit 21D executes various commands based on the command signal. The outdoor unit 22 is equipped with, for example, an outdoor fan, a compressor, etc. The remote control 23 is an operation unit that remotely controls the indoor unit 21 in response to operation by the user.

[0014] The communication adapter 3 of this embodiment has a communication function for wirelessly connecting the indoor units 21 of the air conditioner 2 to the router 4 and a control function for controlling the indoor units 21 using AI (Artificial Intelligence). A communication adapter 3 is provided for each indoor unit 21. The router 4 is a device (also called an access point) that wirelessly connects the communication adapter 3 to a communication network 8 using, for example, a wireless local area network (WLAN) and also wirelessly connects a terminal device 7 to the communication network 8. The terminal device 7 is a communication terminal such as a smartphone of a user who serves as an administrator among multiple users of the air conditioning system 1. The communication network 8 is, for example, a communication network such as the Internet. The server device 5 has a function for generating a learning model (described later) and a database for storing operation history data, etc. The server device 5 is located, for example, in a data center. The relay device 6 is connected to the communication network 8 and has a function for connecting to the server device 5 via communication. The relay device 6 transmits driving history data and the like used to generate or update a learning model from the communication adapter 3 to the server device 5 via the communication network 8. The relay device 6 also transmits the learning model generated or updated by the server device 5 to the communication adapter 3 via the communication network 8. The relay device 6 is located, for example, in a data center or the like.

[0015] The relay device 6 has a first relay unit 6A, a second relay unit 6B, and a third relay unit 6C. The first relay unit 6A transmits various data related to the learning model (hereinafter referred to as operation history data) from the communication adapter 3 to the server device 5 via the communication network 8, and transmits the learning model generated or updated by the server device 5 to the communication adapter 3 via the communication network 8. The second relay unit 6B acquires the operating conditions (such as the operating mode, e.g., cooling / heating, and the set temperature) of the indoor unit 21 set by the user using the terminal device 7 while away from home, and transmits these to the indoor unit 21. The third relay unit 6C acquires external data, such as weather forecasts and calendar information, from the communication network 8, e.g., the Internet, and transmits the acquired external data to the server device 5. The third relay unit 6C also transmits the external data to the communication adapter 3 via the communication network 8.

[0016] <Communication adapter configuration> FIG. 3 is a block diagram showing an example of the configuration of the communication adapter 3. The communication adapter 3 shown in FIG. 3 has a first communication unit 31, a second communication unit 32, a storage unit 33, and a CPU (Central Processing Unit) 34. The first communication unit 31 is a communication IF (Interface) such as a UART (Universal Asynchronous Receiver Transmitter) that communicatively connects the control unit 21D in the indoor unit 21 and the CPU 34. The second communication unit 32 is a communication unit such as a communication IF such as a WLAN that communicatively connects the router 4 and the CPU 34. The storage unit 33 has, for example, a ROM (Read Only Memory) or a RAM (Random Access Memory), and stores various information such as data and programs. The CPU 34 controls the entire communication adapter 3.

[0017] The storage unit 33 in the communication adapter 3 shown in FIG. 3 has a history memory 33A, an adapter-side learning storage unit 33B, a prediction result memory 33C, and an external memory 33D. The history memory 33A temporarily stores operation history data acquired from the indoor unit 21. The adapter-side learning storage unit 33B stores a learning model generated by the server device 5. The learning model is a learning model that infers a subsequent operation method based on the detection results of the sensor 21B. For example, the learning model stores a first learning model 33B1 that infers an operation method related to user comfort based on learning results of learning user operation tendencies, and a second learning model 33B2 that infers an operation method related to user convenience based on learning results of learning user presence / absence tendencies in the space where the air conditioner 2 is installed.

[0018] The first learning model 33B1 includes, for example, a sensible temperature prediction model 35 and a temperature unevenness prediction model 36.

[0019] The sensible temperature prediction model 35 is a learning model that predicts the sensible temperature that a user will feel comfortable at after, for example, five minutes in an air-conditioned space, using operational state quantities including operation history data obtained periodically (for example, every five minutes) from the indoor unit 21. The sensible temperature prediction model 35 is a program that is executed when adjusting the air conditioner 2 so that the user feels comfortable, for example, according to time-series indoor temperature, indoor humidity, outdoor temperature, etc. By controlling the air conditioner 2 based on the sensible temperature prediction model 35, the set temperature of the air conditioner 2 is changed so that the user feels comfortable. As a result, user comfort is maintained.

[0020] The temperature unevenness prediction model 36 is a learning model that predicts the location of temperature unevenness in the air-conditioned space, for example, 10 minutes from now, using operational state quantities including operation history data obtained periodically (for example, every 5 minutes) from the indoor unit 21. Temperature unevenness in the air-conditioned space is eliminated by controlling the air conditioner 2 based on the temperature unevenness prediction model 36. As a result, users will no longer feel uncomfortable (too hot or too cold) in areas where temperature unevenness occurs, and user comfort will be maintained.

[0021] The second learning model 33B2 includes, for example, a building load prediction model 37 and a presence / absence prediction model .

[0022] The building load prediction model 37 is a learning model that predicts the time it will take for the indoor temperature to reach the set temperature, taking into account the building load of the room in which the indoor unit 21 is installed. The building load refers to the energy required to maintain constant temperature and humidity in the room in which the indoor unit 21 is installed. In other words, the building load prediction model 37 is a learning model that predicts and determines the advance operation start time using operation state quantities including, for example, the current indoor temperature, the capacity at the time of operation start, and the extracted forecast outdoor temperature value, when, for example, a user makes a reservation for an on-timer that sets the operation start time for the indoor unit 21 to operate. By controlling the air conditioner 2 based on the building load prediction model 37, the user is no longer required to set the on-timer time taking into account the time it will take for the indoor temperature to reach the set temperature, thereby improving user convenience.

[0023] The presence / absence prediction model 38 predicts, for example, the presence or absence of a user in the air-conditioned space for 24 hours using operation state quantities including operation history data obtained periodically (for example, every five minutes) from the indoor unit 21 and detection results from the human detection sensor. The presence / absence prediction model 38 is a learning model that performs, for example, power-saving operation based on the prediction results of the user's presence or absence and the detection results from the human detection sensor. By controlling the air conditioner 2 based on the presence / absence prediction model 38, it is possible to perform power-saving operation, for example, during times when the user is predicted to be absent from the air-conditioned space. This allows the user to avoid unnecessary air-conditioning operation without having to frequently turn the air conditioner 2 on and off, thereby improving user convenience.

[0024] The presence / absence prediction result memory 33B stores the prediction results of the presence / absence of a user in the air-conditioned space for 24 hours, acquired from the presence / absence prediction model 38.

[0025] The CPU 34 can recognize the predicted results of the presence or absence of users in the air-conditioned space for 24 hours obtained by the presence / absence prediction model 38. The external memory 33D stores external data acquired from outside, such as the above-mentioned holiday information and weather forecast.

[0026] The CPU 34 includes a collection unit 34A, a transmission unit 34B, a reception unit 34C, and an inference unit 34D.

[0027] The collection unit 34A acquires operation state quantities, such as the indoor temperature, outdoor temperature, heat exchanger temperature, compressor rotation speed, user operation of the air conditioner, and detection results of the human detection sensor, as operation history data from the indoor unit 21 at a predetermined interval, for example, every 10 minutes. The indoor temperature is the temperature inside the air-conditioned space of the indoor unit 21, i.e., the indoor temperature, detected by an indoor temperature sensor. The outdoor air temperature is the temperature of the outdoor air flowing into the outdoor unit 22, i.e., the outdoor air temperature, detected by an outdoor air temperature sensor located near an air intake (not shown) of the outdoor unit 22. The heat exchanger temperature is detected, for example, by the temperature of the refrigerant flowing into (or out of) a heat exchanger (not shown). For example, when the heat exchanger (hereinafter also referred to as the outdoor heat exchanger) of the outdoor unit 22 functions as a condenser, the temperature of the refrigerant flowing out of the outdoor heat exchanger is detected as the heat exchanger temperature. Specifically, a temperature sensor (hereinafter also referred to as the outdoor heat exchanger outlet sensor) is placed at the outlet of the outdoor heat exchanger (not shown), and the outdoor heat exchanger outlet sensor detects the heat exchanger temperature. The compressor rotation speed is detected, for example, by a compressor rotation speed sensor (not shown). The collection unit 34A stores the acquired operation history data, such as the operating state quantities, in the history memory 33A.

[0028] The transmitting unit 34B transmits the driving history data stored in the history memory 33A to the server device 5 via the communication network 8. The server device 5 generates a learning model using the driving state quantities in the driving history data sequentially received from the communication adapter 3. The receiving unit 34C receives the learning model from the server device 5 via the communication network 8, and stores the received learning model in the adapter-side learning memory unit 33B.

[0029] The inference unit 34D uses, for example, the sensible temperature prediction model 35 to predict the sensible temperature that a user in the air-conditioned space will feel comfortable at, for example, five minutes from now, and transmits the predicted change in sensible temperature to the control unit 21D in the indoor unit 21 via the first communication unit 31. Specifically, if the inference unit 34D predicts that the sensible temperature will increase, it transmits a "sensible temperature increase" notification to the control unit 21D in the indoor unit 21 via the first communication unit 31. If the inference unit 34D predicts that the sensible temperature will decrease, it transmits a "sensible temperature decrease" notification to the control unit 21D in the indoor unit 21 via the first communication unit 31. If the inference unit 34D predicts that the sensible temperature will not change, it does not transmit a notification to the control unit 21D in the indoor unit 21.

[0030] In addition, the inference unit 34D predicts the location of temperature unevenness in the air-conditioned space 10 minutes from now, for example using a temperature unevenness prediction model 36, and transmits control content to eliminate the temperature unevenness to the control unit 21D in the indoor unit 21 via the first communication unit 31.

[0031] In addition, the inference unit 34D predicts the pre-operation start time at the operation start time of the on timer reservation, for example, using the building load prediction model 37, and transmits the pre-operation start time to the control unit 21D in the indoor unit 21 via the first communication unit 31.

[0032] The inference unit 34D also predicts the presence or absence of a user in the air-conditioned space for 24 hours using, for example, a presence / absence prediction model 38, and generates a presence / absence trend of a user in the air-conditioned space from the prediction result. Furthermore, the inference unit 34D uses the presence / absence trend of a user and the current detection result of the human detection sensor to predict the presence or absence of a user in the air-conditioned space for, for example, 60 minutes from the time when the absence of a person in the air-conditioned space is detected. The inference unit 34D transmits the prediction result of the presence or absence of a user to the control unit 21D in the indoor unit 21 via the first communication unit 31.

[0033] <Operation of the indoor unit control unit using learning model> Control unit 21D changes the current indoor temperature to the set temperature based on the sensible temperature increase notification or sensible temperature decrease notification received from sensible temperature prediction model 35. When control unit 21D receives a sensible temperature increase notification, it changes the set temperature, for example, to increase the set temperature by 0.5 degrees. Furthermore, when control unit 21D receives a sensible temperature decrease notification, it changes the set temperature, for example, to decrease the set temperature by 0.5 degrees. In other words, control unit 21D controls each device constituting main body 21A in indoor unit 21 so that the air-conditioned space reaches the set temperature.

[0034] Based on the control content of the received temperature unevenness prediction model 36, the control unit 21D controls each device that constitutes the main body 21A of the indoor unit 21, such as the indoor unit fan and air deflector, so as to eliminate temperature unevenness in the air-conditioned space.

[0035] The control unit 21D starts the air conditioning operation of the indoor unit 21 at the advance operation start time of the received building load prediction model 37. As a result, the temperature of the room in which the indoor unit 21 is installed will reach the set temperature through advance air conditioning operation at the operation start time set by the user in the on-timer reservation. In other words, the building load prediction model 37 controls the advance operation of the on-timer based on the air conditioning load.

[0036] The control unit 21D refers to the prediction result of the user presence / absence from the presence / absence prediction model 38 for, for example, 60 minutes from the point in time when the human detection sensor detects the absence of a person during air conditioning operation. For example, if the prediction result that the control unit 21D refers to also shows that a user is absent, the control unit 21D determines that the absence of a user will continue for a predetermined time in the air-conditioned space, and controls each device that constitutes the main body 21A in the indoor unit 21 to switch to power-saving operation, which stops the air conditioning operation that is currently being performed.

[0037] <Server device configuration> Fig. 4 is a block diagram showing an example of the configuration of the server device 5. The server device 5 shown in Fig. 4 has, for example, a communication unit 51, a storage unit 52, and a CPU 53. The communication unit 51 in this embodiment is a communication IF that communicatively connects the relay device 6 and the CPU 53. The storage unit 52 has, for example, an HDD (Hard Disk Drive), ROM, RAM, etc., and stores various information such as data and programs. For example, the CPU 53 controls the entire server device 5.

[0038] The storage unit 52 in the server device 5 shown in Fig. 4 includes, for example, a history data memory 52A and a model storage unit 52B. For example, the history data memory 52A stores the operating state quantities received from the communication adapter 3. For example, the model storage unit 52B stores the first learning model and the second learning model generated by the server device 5. The first learning model is, for example, a temperature unevenness prediction model and a sensible temperature prediction model. The second learning model is, for example, a building load prediction model and a presence / absence prediction model.

[0039] The CPU 53 in the server device 5 includes, for example, a receiving unit 53A, an acquiring unit 53B, a generating unit 53C, and a transmitting unit 53D.

[0040] The receiving unit 53A is connected to, for example, the communication adapters 3 of multiple indoor units 21, receives the operation state quantities from the communication adapters 3 via the router 4, the communication network 8, and the relay device 6, and stores the received operation state quantities in the history data memory 52A. The acquiring unit 53B acquires the operation state quantities stored in the history data memory 52A. The generating unit 53C generates, for example, a first learning model and a second learning model based on the operation state quantities acquired by the acquiring unit 53B.

[0041] <Remote control configuration> FIG. 5 is a block diagram showing an example of the configuration of the remote control 23. The remote control 23 shown in FIG. 5 has, for example, a remote control communication unit 61, a touch panel 62, and a remote control control unit 63. The remote control communication unit 61 transmits command signals to the indoor unit 21 by infrared rays, radio waves, or the like. The touch panel 62 has an operation unit 71 and a display unit 72. The display unit 72 is an area for displaying various information. The operation unit 71 is an operation button displayed on the display unit 72, and corresponds to various commands. The user inputs instructions by pressing the operation button. The remote control control unit 63 controls the entire remote control 23.

[0042] Fig. 6 is an explanatory diagram showing an example of the display content of the touch panel 62 of the remote control 23. The touch panel 62 shown in Fig. 6 displays an operation section 71 on an operation screen 72A when an operation mode is set. The operation screen 72A has, as the operation section 71, an AI AUTO button 71A, a HEAT button 71B, an AUTO button 71C, a COOL button 71D, a DEHUMIDIFY button 71E, a BLOW button 71F, a STOP button 71G, an OK button 71H, and a CANCEL button 71J.

[0043] The AI ​​AUTO button 71A corresponds to the activation of a learning model in the inference unit 34D in the communication adapter 23. When a user presses the AI ​​AUTO button 71A, a command signal to execute an AI function is sent to the indoor unit 21. In this embodiment, the AI ​​AUTO button 71A corresponds to the activation of the first learning model 33B1. The HEAT button 71B is, for example, a button that requests the control unit 21D of the indoor unit 21 to start an air conditioning operation (heating operation) that increases the indoor temperature of the air-conditioned space. The AUTO button 71C is, for example, a button that requests the control unit 21D of the indoor unit 21 to start an air conditioning operation (automatic operation) that automatically adjusts the indoor temperature of the air-conditioned space. The COOL button 71D is, for example, a button that requests the control unit 21D of the indoor unit 21 to start an air conditioning operation (cooling operation) that decreases the indoor temperature of the air-conditioned space. The DEHUMIDIFY button 71E is, for example, a button that requests the control unit 21D of the indoor unit 21 to start an air conditioning operation (dehumidification operation) that dehumidifies the air-conditioned space. The blow button 71F is, for example, a button for requesting the control unit 21D of the indoor unit 21 to start air conditioning operation (fan operation) for blowing air toward the air-conditioned space. The stop button 71G is, for example, a button for requesting the control unit 21D of the indoor unit 21 to stop the air conditioning operation that is being performed. The cancel button 71J is, for example, a button for canceling the operation screen 72A that is being displayed. The OK button 71H is, for example, a button for accepting the operation screen 72A that is being displayed.

[0044] FIG. 7 is an explanatory diagram showing an example of transition of display contents on the touch panel 62. When the remote control unit 63 detects button operation of the AI ​​AUTO button 71A on the operation screen 72A shown in FIG. 6, it switches from the operation screen 72A shown in FIG. 6 to a startup screen 72B indicating the start of activation of the AI ​​AUTO mode. The AI ​​AUTO mode is a mode in which the function of a first learning model, which infers a driving method related to the user's comfort, among multiple learning models is exercised. The startup screen 72B is a screen that displays a message indicating that the function of the first learning model, among the multiple learning models, has been activated. After displaying the startup screen 72B, the remote control unit 63 switches from the startup screen 72B to a driving screen 72C indicating that driving is in AI AUTO mode.

[0045] <Indoor unit operation using learning mode> FIG. 8 is a flowchart showing an example of the processing operation of the communication adapter 3 related to the control process. The CPU 34 in the communication adapter 3 detects an AI automatic command from the remote control 23 via the indoor unit 21 in response to button operation of the AI ​​automatic button 71A on the remote control 23 (step S11). The AI ​​automatic command is a command to execute an AI function. In this embodiment, the AI ​​automatic command is a command to cause the communication adapter 3 to start a first learning model among multiple learning models. When the CPU 34 detects the AI ​​automatic command in step S11, it determines whether the communication adapter 3 is in an off state (step S12). If the communication adapter 3 is in an off state (step S12: Yes), the CPU 34 switches the communication adapter 3 on (step S13). When the communication adapter 3 is switched on, the communication adapter 3 is connected to the server device 5 via the router 4. When the communication adapter 3 is connected to the server device 5, the operating state quantities, which are information necessary for learning, are transmitted to the server device 5 via the communication adapter 3. Furthermore, the server device 5 that receives the operating state quantities learns the user's preferences and generates a learned model. Then, the trained model generated on the server device 5 is transmitted from the server device 5 to the communication adapter 3.

[0046] After the communication adapter 3 is switched on, the CPU 34 activates the first learning model 33B1 of the learning models (step S14) and ends the processing operation shown in Figure 8. Note that activating the first learning model 33B1 means bringing the first learning model into a state where it can operate. As a result, the communication adapter 3 will perform automatic driving using the activated first learning model 33B1.

[0047] If the communication adapter 3 is not in the off state, that is, if the communication adapter 3 is in the on state (step S12: No), the CPU 34 returns to the processing of step S14 to activate the first learning model 33B1.

[0048] <Effects of the Example> In the air conditioner 2 of this embodiment, even when both the first learning model 33B1 related to user comfort and the second learning model 33B2 related to user convenience are installed, the first learning model 33B1 related to user comfort can be prioritized according to the user's operation. As a result, user comfort can be ensured while ensuring diversification of AI functions.

[0049] When the inference unit 34D detects the operation of the AI ​​automatic button 71A on the remote control 23 while the communication adapter 3 is in the OFF state, the inference unit 34D switches the communication adapter 3 to the ON state, and the first learning model 33B1 is activated. As a result, the first learning model 33B1 related to the user's comfort can be prioritized in accordance with the user's operation.

[0050] Although the first learning model 33B1 in this embodiment is exemplified by the sensible temperature prediction model 35 and the temperature unevenness prediction model 36, any learning model related to user comfort may be used and may be modified as appropriate. Also, the second learning model 33B2 is exemplified by the building load prediction model 37 and the presence / absence prediction model 38, but any learning model related to user convenience may be used and may be modified as appropriate.

[0051] In the example shown, the communication adapter 3 activates the sensible temperature prediction model 35 and the temperature unevenness prediction model 36 in the first learning model 33B1 when it detects button operation of the AI ​​automatic button 71A on the remote control 23, but it may also activate one or more arbitrary models within the first learning model 33B1, and this can be changed as appropriate.

[0052] In addition, the communication adapter 3 has been exemplified as starting the first learning model 33B1 when it detects button operation of the AI ​​automatic button 71A on the operation screen of the remote control 23, but for example, the terminal device 7 or the indoor unit 21 may be provided with an AI automatic button, and the first learning model 33B1 may be started in response to button operation of the AI ​​automatic button, and changes can be made as appropriate.

[0053] As an example of an operation method related to user convenience inferred by the second learning model 33B2, a case where switching to power-saving operation in which operation is stopped using the presence / absence prediction model 38 has been given, but it may also be an inference method related to maintenance of the air conditioner 2, the start of heating sterilization operation, the start of filter cleaning operation, operation suppression or operation start, and can be changed as appropriate. Note that heating sterilization operation is an operation in which the indoor heat exchanger in the indoor unit 21 is heated to 55 degrees or higher to sterilize it. Filter cleaning operation is an operation in which the filter in the indoor unit 21 is cleaned. [Explanation of symbols]

[0054] 1. Air conditioning system 2. Air conditioners 3 Communication Adapter 5. Server equipment 21 Indoor unit 21B Sensor 21D Control Unit 22 Outdoor unit 23 Remote Control 33B Adapter side learning memory unit 33B1 First Learning Model 33B2 Second Learning Model 34D Reasoning Department 53C Generation part 71A AI Auto Button

Claims

1. A detection unit that detects an operating state quantity of an air conditioner; an operation unit for a user to operate the air conditioner; an inference unit that infers a subsequent driving method based on the detection result of the detection unit, The inference unit a first learning model that infers a driving method related to the user's comfort based on a learning result of learning the user's operation tendency; a second learning model that infers an operating method related to the convenience of the user based on a learning result of learning the tendency of the user's presence or absence in the space in which the air conditioner is installed or a learning result of learning the building load of the space, The operation unit includes: An air conditioner characterized in that it has a start button that starts the inference unit, and when the start button is operated, the first learning model is started preferentially between the first learning model and the second learning model.

2. An air conditioning system having an indoor unit, an outdoor unit, an air conditioner having a communication adapter that communicates with the indoor unit, and a server device that communicates with the communication adapter, The indoor unit is a detection unit that detects an operation state quantity of the air conditioner; an operation unit for a user to operate the air conditioner, The communication adapter an inference unit that infers a subsequent driving method based on the detection result of the detection unit; The inference unit a first learning model that infers a driving method related to the user's comfort based on a learning result of learning the user's operation tendency; a second learning model that infers an operating method related to the convenience of the user based on a learning result of learning the tendency of the user's presence or absence in the space in which the air conditioner is installed or a learning result of learning the building load of the space, The server device a generation unit that generates the first learning model, The operation unit includes: a start button for starting the inference unit; The inference unit An air conditioning system characterized in that, when the start button is operated, the first learning model is started preferentially out of the first learning model and the second learning model.

3. The inference unit The air conditioning system according to claim 2, wherein when the start button is operated while the communication adapter is in a stopped state, the communication adapter is switched to a powered state and the first learning model is started.

Images