JP2021071243A - Air conditioning system and air conditioner - Google Patents

Abstract

To provide an air conditioning system capable of controlling a load on a server device.SOLUTION: An air conditioning system comprises an indoor unit, a server device which generates plural types of learning models, and an adaptor to be connected when establishing communication between the indoor unit and the server device. The adaptor transmits, to the server device, identification information to identify the indoor unit, indoor unit operation history data for a predetermined period and learning model type information to identify a type of an object learning model to be used in the indoor unit. The server device has: an identification section which identifies the type of the object learning model from the learning model type information and the indoor unit, which is a transmission destination, from the identification information; a learning section which generates the object learning model using the operation history data to be used for the type of the identified object learning model; and a server side transmission section which transmits the generated object learning model to the adaptor connected to the indoor unit identified as the destination.SELECTED DRAWING: Figure 1

Description

The present invention relates to air conditioning systems and air conditioners.

An air conditioning system that efficiently and automatically controls an indoor unit in an air conditioning machine is known (Patent Document 1). In this air-conditioning system, the storage unit has a learning model, and the learning model is used to reflect the time-series suitable temperature environment according to the user's preference, behavior pattern, etc. in the control by standard specification setting. Can be done.

JP-A-2015-117933

For example, air conditioners are provided with many functions or with the minimum necessary functions according to diversified user preferences. Therefore, regarding the learning model, it is conceivable to provide an air conditioner that stores a plurality of types of learning models or a single learning model depending on the type of air conditioner. On the other hand, various types of learning models are created by connecting to the air conditioner via a communication network and using the operation details of the user in the air-conditioned space collected by the indoor unit in the air conditioner and the operation history data such as the actual environment. A cloud-side server device to be generated may be provided. In this case, the air conditioner acquires various learning models from the server device on the cloud side and stores the acquired various learning models.

When the server device updates the learning model, the server device transmits the updated learning model to each air conditioner. Then, each air conditioner can update the learning model to the latest state.

However, when the server device generates or updates the learning model, if the learning model is uniformly generated or updated regardless of whether or not each indoor unit connected to the server device requires it, the server device does not need it. There was a problem that a heavy load was applied.

In view of such a problem, it is an object of the present invention to provide an air conditioner system and an air conditioner capable of suppressing a load applied to a server device.

The air conditioning system of the present invention includes an indoor unit, a server device that generates a plurality of types of learning models, and an adapter that connects the indoor unit and the server device by communication. The adapter transmits identification information for identifying the indoor unit, operation history data for a predetermined time of the indoor unit, and learning model type information for identifying the type of learning model used in the indoor unit to the server device. It has an adapter side transmitter. The server device has an identification unit, a learning unit, and a server-side transmission unit. The identification unit identifies the type of the learning model from the learning model type information received from the adapter, and identifies the destination indoor unit from the identification information received from the adapter. The learning unit generates the learning model by using the operation history data used for generating the learning model identified by the identification unit. The server-side transmission unit transmits the learning model generated by the learning unit to the adapter connected to the indoor unit of the transmission destination identified by the identification unit.

The air conditioning system and the air conditioner of the present invention can suppress the load applied to the server device.

FIG. 1 is an explanatory diagram showing an example of an air conditioning system of this embodiment. FIG. 2 is an explanatory diagram showing an example of the temperature distribution area of the air-conditioned space measured by the detection unit. FIG. 3 is a block diagram showing an example of the configuration of the adapter. FIG. 4 is a block diagram showing an example of the configuration of the server device. FIG. 5 is an explanatory diagram showing an example of the contents of the operation history data. FIG. 6 is an explanatory diagram showing an example of learning model type information. FIG. 7 is a flowchart showing an example of the processing operation of the adapter related to the transmission processing. FIG. 8 is a flowchart showing an example of the processing operation of the server device involved in the learning process.

Hereinafter, examples of the air conditioning system disclosed in the present application will be described in detail with reference to the drawings. The disclosed technology is not limited by the present embodiment. In addition, each of the examples shown below may be appropriately modified as long as it does not cause a contradiction.

FIG. 1 is an explanatory diagram showing an example of the air conditioning system 1 of this embodiment. The air conditioning system 1 shown in FIG. 1 includes an indoor unit 2, an adapter 3, an access point 4, a server device 5, a relay device 6, a communication terminal 7, a communication network 8, a remote controller 9, and an outdoor unit. Has 10 and.

The indoor unit 2 is, for example, a part of an air conditioner that is arranged indoors and heats or cools the air in the room. The user of the indoor unit 2 can remotely control the indoor unit 2 by operating the remote controller 9. The indoor unit 2 has a main body 2A, a control unit 2B for controlling the main body 2A, a detection unit 2C, and a storage unit 2D. In the main body 2A, each device such as an indoor fan, an expansion valve, an indoor heat exchanger, and a wind direction plate is housed in a housing. The control unit 2B controls the operation of the indoor fan, the expansion valve, and the wind direction plate described above. The control unit 2B adjusts the opening degree of the expansion valve, flows the amount of refrigerant required for exerting the air conditioning capacity required by the user through the indoor heat exchanger, and drives the indoor fan to drive the indoor heat. The indoor air that has undergone heat exchange with the refrigerant by the exchanger is deflected by controlling the wind direction plate and blown out into the air-conditioned space. As a result, the air-conditioned space is heated, cooled, and dehumidified. The detection unit 2C is, for example, a radiation sensor that measures the radiant temperature of the floor surface of the air-conditioned space in which the indoor unit 2 is installed, a room temperature sensor that measures the indoor temperature of the air-conditioned space, and a humidity that measures the indoor humidity of the air-conditioned space. A sensor, a temperature sensor or the like that measures an indoor heat exchange temperature, which is the temperature of an indoor heat exchanger provided in an indoor heat exchanger (not shown) of the indoor unit 2.

The detection unit 2C as a radiation sensor divides the floor surface of the air-conditioned space into a plurality of areas, for example, detects the radiation temperature of each floor surface of the nine areas A1 to A9 as shown in FIG. In FIG. 2, the area on the front side when viewed from the air outlet of the indoor unit 2 is the area A1 to A3, the area on the back side when viewed from the air outlet of the indoor unit 2 is the area A7 to A9, and the area on the front side and the back. Areas A4 to A6 are intermediate areas between the area on the side and the area on the side. Areas A1, A4, and A7 are areas in the left (L: Left) direction when viewed from the air outlet of the indoor unit 2, and areas A3, A6, and A9 are areas on the right (R: left) when viewed from the air outlet of the indoor unit 2. The area in the Right) direction and the areas A2, A5, and A8 are the areas in the center (C: Center) direction when viewed from the outlet of the indoor unit 2. The storage unit 2D stores learning model identification information that identifies the type of learning model requested from the server device 5 among a plurality of learning models that can be used by the indoor unit 2.

The adapter 3 has a communication function for wirelessly connecting the indoor unit 2 and the access point 4 and a control function for AI-controlling the indoor unit 2. The adapter 3 is provided for each indoor unit 2. The adapter 3 will be described in detail later with reference to FIG. The access point 4 is a device that connects the adapter 3 and the communication network 8 by wireless communication using, for example, a WLAN (Wireless Local Area Network) or the like. The communication network 8 is, for example, a communication network such as the Internet. The server device 5 has a function of generating a learning model used by the adapter 3 for AI control of the indoor unit 2, a database for storing operation history data, and the like. The server device 5 is located in, for example, a data center.

The relay device 6 has a function of being connected to the communication network 8 by communication and also being connected to the server device 5 by communication. The relay device 6 receives the operation history data and the like used for generating the learning model in the server device 5 or updating the learning model from the indoor unit 2 via the adapter 3 and the communication network 8 and transmits the operation history data to the server device 5. Further, the relay device 6 transmits the learning model generated or updated by the server device 5 to the adapter 3 via the communication network 8. The relay device 6 is arranged in, for example, a data center or the like.

The relay device 6 has a first relay unit 6A, a second relay unit 6B, a third relay unit 6C, and a relay side learning storage unit 6D. The first relay unit 6A relays various data related to AI control between the adapter 3 and the server device 5. Specifically, the first relay unit 6A transmits the operation history data and the like used for generating or updating the learning model received from the adapter 3 via the communication network 8 to the server device 5, and the server device 5 generates the data. Alternatively, the updated learning model is transmitted to the adapter 3 via the communication network 8. The second relay unit 6B sets the operating conditions (operating mode such as cooling / heating, set temperature, etc.) of the indoor unit 2 set by the user from outside using the communication terminal 7, which will be described later, to the access point 4 and the communication network. Receives via 8, and transmits this to the indoor unit 2 via the communication network 8, the access point 4, and the adapter 3. The third relay unit 6C receives, for example, external data such as a weather forecast via the communication network 8, and transmits the received external data to the server device 5. Further, the third relay unit 6C transmits the received external data to the adapter 3 via the communication network 8 and the access point 4. When the relay side learning storage unit 6D receives the learning model from the server device 5 in the first relay unit 6A, the relay side learning storage unit 6D stores the learning model. When the relay side learning storage unit 6D receives the next learning model from the server device 5, the relay side learning storage unit 6D overwrites the received learning model with the learning model being stored and stores it. The first relay unit 6A transmits the learning model stored in the relay side learning storage unit 6D to the adapter 3 via the communication network 8.

FIG. 3 is a block diagram showing an example of the configuration of the adapter 3. The adapter 3 shown in FIG. 3 includes a first communication unit 11, a second communication unit 12, a storage unit 13, and a CPU (Central Processing Unit) 14. The first communication unit 11 is a communication IF (Interface) such as a UART (Universal Asynchronous Receiver Transmitter) that is connected to the control unit 2B in the indoor unit 2 by communication. The second communication unit 12 is, for example, a communication IF such as a WLAN, which is connected to the access point 4 by communication. The storage unit 13 has, for example, a ROM (Read Only Memory), a RAM (Random Access Memory), and the like, and stores various information such as data and programs. The CPU 14 controls the entire adapter 3.

The storage unit 13 includes an operation history memory 13A that temporarily stores the operation history data acquired from the indoor unit 2, an adapter-side learning storage unit 13B that stores the learning model acquired from the server device 5 via the relay device 6, and an external device. It has an external memory 13C for storing data.

The CPU 14 includes an acquisition unit 14A, an adapter-side transmission unit 14B, a reception unit 14C, a setting unit 14D, a prediction unit 14E, and a control unit 14F. The acquisition unit 14A acquires operation history data from the indoor unit 2 at a predetermined cycle, for example, every 5 minutes. The acquisition unit 14A stores the operation history data acquired every 5 minutes in the operation history memory 13A. The operation history data will be described in detail later with reference to FIG.

The adapter-side transmission unit 14B transmits the operation history data stored in the operation history memory 13A to the access point 4 via the second communication unit 12, and the access point 4 transmits the received operation history data to the communication network 8. The first relay unit 6A of the relay device 6 transmits the received operation history data to the server device 5. In addition to the operation history data, the adapter-side transmission unit 14B transmits the indoor unit ID and learning model type information described later to the access point 4 via the second communication unit 12, and the access point 4 receives the indoor unit. The machine ID and the learning model type information are transmitted to the relay device 6 via the communication network 8. The first relay unit 6A of the relay device 6 transmits the received indoor unit ID and learning model type information to the server device 5. The indoor unit ID is identification information for identifying the indoor unit 2. The learning model type information, which will be described later, is information that identifies the type of learning model requested for the server device 5 among the plurality of learning models that can be used by the indoor unit 2, and is the information that identifies the type of the learning model requested for the server device 5. It is remembered in.

The receiving unit 14C receives, for example, a learning model such as a temperature unevenness prediction model, a bodily sensation prediction model, and a building load prediction model from the server device 5. Specifically, the server device 5 transmits a learning model to the relay device 6, and the first relay unit 6A of the relay device 6 transmits the received learning model to the access point 4 via the communication network 8 for access. At point 4, the received learning model is transmitted to the receiving unit 14C of the adapter 3. The receiving unit 14C stores the learning model received via the second communication unit 12 in the learning storage unit 13B on the adapter side. The setting unit 14D applies the learning model stored in the adapter-side learning storage unit 13B to the prediction unit 14E.

For example, when the indoor temperature becomes stable, the prediction unit 14E transmits the indoor temperature unevenness predicted by using the temperature unevenness prediction model to the control unit 2B of the indoor unit 2, and the control unit 2B receiving this is transmitted. , Control the indoor fan and wind direction plate to eliminate temperature unevenness. The stable state of the indoor temperature means that the temperature difference between the indoor temperature (the temperature of the entire air-conditioned space, not the radiant temperature of each area A1 to A9) and the set temperature of the indoor unit 2 is a predetermined threshold value, for example, ± 0. . This is the case where the temperature within 5 degrees continues for a predetermined time, for example, 5 minutes or more. The prediction unit 14E may directly control each device constituting the main body 2A of the indoor unit 2 based on the learning model, and can be changed as appropriate. The prediction unit 14E may execute the prediction regarding the indoor unit 2 based on the learning model, and can be changed as appropriate. The control unit 14F controls the entire CPU 14.

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 communicates with an adapter 3 provided in a plurality of indoor units 2 via a relay device 6, and includes a communication unit 31, a storage unit 32, and a CPU 33. The communication unit 31 is a communication IF that communicates and connects with the relay device 6. The storage unit 32 is composed of, for example, an HDD (Hard Disk Drive), a ROM, or a RAM, and stores various information such as data and programs. The CPU 33 controls the entire server device 5.

The storage unit 32 in the server device 5 has a data memory 32A and a server-side learning storage unit 32B. The data memory 32A stores the operation history data of each indoor unit 2 received from each adapter 3 for each indoor unit 2. When transmitting the operation history data, the adapter 3 also transmits an indoor unit ID that identifies the indoor unit 2 connected to the adapter 3. Then, the server device 5 that has received the operation history data from each indoor unit 2 stores the operation history data in the data memory 32A for each indoor unit ID included in the received operation history data. The server-side learning storage unit 32B stores the learning model generated or updated by the server device 5. The CPU 33 in the server device 5 includes a learning unit 33A, a receiving unit 33B, an identification unit 33C, and a server-side transmitting unit 33D.

The learning unit 33A receives each operation history data from the adapters 3 of the plurality of indoor units 2 via the access point 4, the communication network 8, the relay device 6, and the receiving unit 33B. Then, the learning unit 33A generates a plurality of learning models by using the operation history data for N weeks (for example, 3 weeks) stored in the data memory 32A among the operation history data received from each adapter 3. Or update. The learning unit 33A generates or updates a learning model for each indoor unit 2 based on the learning result.

Then, the learning unit 33A stores the newly generated learning model or the learning model obtained by updating the existing learning model in the server-side learning storage unit 32B. The identification unit 33C identifies the indoor unit ID, which will be described later, acquired from the adapter 3, and identifies the type of the learning model to be provided from the learning model type information 100, which will be described later, acquired from the adapter 3. The learning model to be provided is, for example, a learning model that can be used in the indoor unit 2. The server-side transmission unit 33D provides a learning model newly generated for each indoor unit 2 by the learning unit 33A or a learning model updated from the existing learning model via the relay device 6, the communication network 8, and the access point 4. It is transmitted to the adapter 3 of the indoor unit 2.

The outdoor unit 10 is connected to the indoor unit 2 by a refrigerant pipe, and is provided with an outdoor fan, a compressor, and the like. For example, the air conditioner 50 is composed of the indoor unit 2, the outdoor unit 10, the adapter 3, and the remote controller 9. Further, the communication terminal 7 is a terminal device such as a smartphone owned by the user.

FIG. 5 is an explanatory diagram showing an example of the contents of the operation history data. The operation history data includes, for example, an operation state, an operation mode, a set temperature, an indoor temperature, an indoor humidity, an air volume, an air direction, a motion sensor, a temperature distribution, an indoor heat exchange temperature, an outdoor temperature and a time.

The operating state is information indicating whether the operation of the indoor unit 2 is ON or OFF. The operation mode is information indicating which operation mode such as cooling or heating is selected when the indoor unit 2 is operated. The set temperature is a control target value of the temperature in the room where the indoor unit 2 is used. The indoor temperature is a measurement temperature of a room temperature sensor (not shown) provided in the indoor unit 2, for example, the temperature of the indoor air at the suction port. The indoor humidity is the indoor humidity of a humidity sensor (not shown) provided in the indoor unit 2 in which the indoor unit 2 is used. The air volume is the air volume of the conditioned air blown out from the indoor unit 2. The air volume is represented by, for example, the rotation speed of the indoor fan of the indoor unit 2. The wind direction is the direction of the conditioned air blown out from the indoor unit 2. The wind direction is a combination of, for example, a left direction, a center direction, and a right direction, as seen from the air outlet side of the indoor unit 2, and, for example, a horizontal direction, an oblique downward direction, and a downward direction. The motion sensor is a detection result of the presence or absence of a person in the room and the amount of activity. The temperature distribution is, for example, the temperature distribution of the floor or wall in the room detected by the radiation sensor of the indoor unit 2. The temperature distribution is, for example, the distribution of the radiant temperature of the floor surface for each of the nine areas "A1" to "A9", and the space corresponding to each of the areas A1 to A9 (each area A1 to A9). It is a temperature distribution that is regarded as the temperature distribution of the space above). The indoor heat exchange temperature is the temperature of the indoor heat exchanger detected by a temperature sensor provided in the indoor heat exchanger (not shown) of the indoor unit 2. The outdoor temperature is the outside air temperature detected by the temperature sensor of the outdoor unit 10. The adapter 3 collects the outdoor temperature detected by the temperature sensor of the outdoor unit 10 via the indoor unit 2. It should be noted that these operation history data may be recorded together with a time stamp which is the year, month, day, hour, minute, and second at the time of data acquisition. Further, although not shown, the operation history data may include, for example, an indoor unit ID, an hourly cloud amount collected from weather forecast data acquired by the adapter 3 via the communication network 8 and the access point 4, and the like.

Next, the learning model will be described. The learning model of this embodiment is a building load prediction model, a bodily sensation prediction model, and a temperature unevenness prediction model. The building load prediction model is a learning model that predicts the air conditioning load of the room in which the indoor unit 2 is installed. For example, the building load model predicts the air conditioning load of the room in which the indoor unit 2 is installed by using the air conditioning capacity currently exhibited by the indoor unit 2 in addition to the information shown in FIG. The control unit 2B of the indoor unit 2 acquires the air conditioning load predicted by the building load prediction model, adds the acquired air conditioning load, and configures the main body 2A so that the indoor temperature becomes the set temperature at a predetermined time. Control each device.

The sensible prediction model predicts the sensible temperature of the user 5 minutes after the current time according to the operating condition of the air conditioner in each household by using a plurality of indoor temperatures, indoor humiditys, and outdoor temperatures acquired in chronological order. It is a learning model. The indoor unit 2 control unit 2B changes the set temperature based on the user's sensible temperature after 5 minutes predicted by the sensible prediction model, and configures the main body 2A so that the indoor temperature becomes the changed set temperature. Control each device.

The temperature unevenness prediction model uses the measured values of the temperature distributions of multiple times among the radiant temperatures of each area A1 to A9 measured every 5 minutes (hereinafter referred to as the temperature distribution), and finally measures the temperature distribution. It is a learning model that predicts temperature unevenness after a predetermined time, for example, 10 minutes after the time.

FIG. 6 is an explanatory diagram showing the learning model type information of the present embodiment. The learning model type information 100 shown in FIG. 6 is composed of a 16-bit bit string, and a binary value for identifying the applicability of each learning model is stored in the storage unit 2D of the indoor unit 2 for each bit. .. For example, when the bit is "1", an applicable learning model is shown, and when the bit is "0", an inapplicable learning model is shown. As described above, the learning model type information 100 is stored in the storage unit 2D of the indoor unit 2. Therefore, by storing the learning model type information in the indoor unit 2, the adapter 3 is shared between the air conditioner that stores a plurality of types of learning models and the air conditioner that stores a single learning model. It becomes easy to change. Whether or not the learning model can be applied is determined for each bit of the 1st to 16th bits (the 1st bit on the rightmost side and the 16th bit on the leftmost side in FIG. 6) in the bit string in the learning model type information shown in FIG. assign. The adapter 3 and the server device 5 recognize the type of the learning model assigned for each bit position in the learning model type information. In the learning model type information shown in FIG. 6, for example, "0" indicating whether or not the "building load prediction model" in the first bit, the "experience prediction model" in the second bit, and the "temperature unevenness prediction model" in the third bit can be applied. "Or" 1 "is assigned. Further, as the learning model type information, for example, "0" indicating "inapplicable" is assigned to the 4th to 16th bits, and these inapplicable bits can be set as applicability of other learning models. is there.

When the adapter-side transmitter 14B in the adapter 3 requests the server device 5 to provide two learning models, for example, a "temperature unevenness prediction model" and a "building load prediction model", for example, an indoor unit having a model name: X001. The bits of the bit positions "1" and "3" in the learning model type information of 2 are set to "1", the remaining bits are set to "0", and the learning model type information 100 after setting is transmitted to the server device 5. .. The adapter-side transmission unit 14B transmits the indoor unit ID, the learning model type information 100, and the operation history data acquired from the indoor unit 2 to the server device 5. The server device 5 receives the learning model type information 100 and the indoor unit ID from the adapter 3. In the server device 5, since the bits of the bit positions "1" and "3" in the learning model type information 100 are "1", the adapter 3 connected to the indoor unit 2 is the "temperature unevenness prediction model" and the "building load prediction". It can be identified that it requests the provision of two learning models of "model".

That is, the adapter-side transmission unit 14B transmits the learning model type information 100 to the server device 5. The server device 5 can also identify the learning model applicable to the indoor unit 2 requested by the adapter 3 side based on the bit string in the learning model type information 100 and the indoor unit ID. Further, when the learning model that can be applied is also changed with the program update of the indoor unit 2, the server device 5 generates the changed learning model. As a result, the server device 5 can generate only the changed learning model.

Next, the operation of the air conditioning system 1 of this embodiment will be described. FIG. 7 is a flowchart showing a processing operation when the adapter 3 in the present embodiment transmits the learning model type information 100. The CPU 14 of the adapter 3 starts the transmission process when the current time is the transmission timing of the operation history data. In FIG. 7, the adapter-side transmission unit 14B in the adapter 3 determines whether or not the current timing is the operation history data transmission timing (step S11).

When the current timing is the transmission timing of the operation history data (step S11: Yes), the adapter side transmission unit 14B acquires the operation history data for a predetermined time acquired from the indoor unit 2, for example, 5 minutes (step). S12). The adapter-side transmission unit 14B acquires an indoor unit ID that identifies the connected indoor unit 2 (step S13).

The adapter-side transmission unit 14B acquires the learning model type information 100 in which the bit indicating the applicability of the learning model to be provided is set to “1” from the indoor unit 2 (step S14). The learning model to be provided is a learning model applicable to the indoor unit 2 and is a learning model that requests the server device 5 to provide the learning model. The adapter-side transmission unit 14B transmits the learning model type information 100 including the operation history data and the indoor unit ID to the server device 5 (step S15), and ends the processing operation shown in FIG. 7. If the current timing does not detect the transmission timing of the operation history data (step S11: No), the adapter-side transmission unit 14B ends the processing operation shown in FIG. 7.

As described above, when the current timing is the transmission timing of the operation history data, the adapter 3 is a bit indicating that the operation history data for a predetermined time, for example, 48 hours, the indoor unit ID, and the learning model can be applied. The learning model type information 100 in which is set to "1" is acquired from the indoor unit 2. The adapter 3 transmits the operation history data for a predetermined time, the indoor unit ID, and the learning model type information 100 to the server device 5 via the relay device 6. As a result, the adapter 3 can request the server device 5 for a learning model applicable to the indoor unit 2 to which the adapter 3 is connected.

FIG. 8 is a flowchart showing a processing operation when the server device 5 performs a learning process (creating a new learning model and updating an existing learning model) in the present embodiment. The CPU 33 of the server device 5 starts the learning process when the operation history data, the indoor unit ID, and the learning model type information 100 are received from the adapter 3. In FIG. 8, the receiving unit 33B in the server device 5 determines whether or not the operation history data and the learning model type information 100 have been received from the adapter 3 (step S21). When the receiving unit 33B receives the operation history data and the learning model type information 100 from the adapter 3 (step S21: Yes), the receiving unit 33B stores the operation history data in the data memory 32A in association with the indoor unit ID (step S22). The identification unit 33C in the server device 5 identifies the learning model to be provided, that is, the learning model whose bit is “1” from the learning model type information 100 (step S23).

The learning unit 33A in the server device 5 identifies the learning model to be provided based on the identification result of the learning model to be provided, and uses the operation history data in memory used for the learning model to be provided. A learning model to be provided is generated or updated (step S24). The learning unit 33A stores the generated or updated learning model to be provided in the server-side learning storage unit 32B in association with the indoor unit ID (step S25). The server-side transmission unit 33D in the server device 5 stores the generated or updated learning model of the provision target, and then is connected to the indoor unit 2 corresponding to the indoor unit ID of the learning model of the provision target. It is transmitted to the adapter 3 (step S26), and the processing operation shown in FIG. 8 is terminated. When the receiving unit 33B does not acquire various information from the adapter 3 (step S21: No), the receiving unit 33B ends the processing operation shown in FIG.

As described above, the server device 5 refers to the bit positions 1 to 16 in the learning model type information 100 acquired from the adapter 3, and bits depending on whether either “0” or “1” is stored. It identifies whether or not the learning model corresponding to the position is the type of the learning model to be provided requested by the adapter 3. Further, the server device 5 generates or updates a learning model to be provided by using the operation history data acquired from the adapter 3.

The server device 5 refers to the bit position corresponding to the learning model in the learning model type information 100, and generates or updates only the learning model whose stored binary value is “1”. As a result, the server device 5 can reduce the processing when generating or updating the extra learning model. Moreover, the user of the indoor unit 2 can also reduce the extra load by reducing the processing required for generating or updating the extra learning model on the server device 5 side, and also eliminates the burden of the pay-as-you-go charge of the server device 5. be able to.

The adapter 3 of this embodiment includes an indoor unit ID that identifies the indoor unit 2, operation history data for a predetermined time of the indoor unit 2 connected to the adapter 3, and a type of learning model to be used in the indoor unit 2. The learning model type information for identifying the above is transmitted to the server device 5. The server device 5 identifies the type of learning model applicable to the indoor unit 2 from the learning model type information and the adapter 3 connected to the destination indoor unit 2 from the indoor unit ID. Further, the server device 5 generates the learning model using the operation history data used for the identified learning model type, and transmits the generated learning model to the adapter 3 connected to the indoor unit 2 of the transmission destination. To do. As a result, the server device 5 can reduce the processing load when generating or updating an extra learning model. Moreover, the user of the indoor unit 2 can also reduce the extra load by reducing the processing required for generating or updating the extra learning model on the server device 5 side, and can also reduce the pay-as-you-go charge of the server device 5.

When the learning model is changed from the adapter 3, the server device 5 generates the changed learning model. As a result, the server device 5 can generate only the changed learning model.

The adapter 3 executes a prediction regarding the indoor unit 2 based on the learning model acquired from the server device 5. As a result, the user of the indoor unit 2 can realize the optimum control regarding the indoor unit 2 by using the acquired learning model.

The learning model type information 100 is information that manages a plurality of learning models in association with binary values indicating whether or not the learning model can be applied to a specific indoor unit 2. As a result, the server device 5 can reduce the burden required for generating or updating an extra learning model by referring to the bit position of each learning model in the learning model type information 100 learning model type from the adapter 3.

Since the indoor unit 2 includes a storage unit 2D for storing learning model type information for identifying the type of the learning model used in the indoor unit 2, the adapter 3 connected to the indoor unit 2 can be shared.

Although the adapter 3 of this embodiment illustrates the case where the operation history data of the indoor unit 2 is transmitted to the server device 5 via the relay device 6, the operation history data is transmitted as it is without passing through the relay device 6. , May be transmitted to the server device 5, and can be changed as appropriate.

In addition, although the process of determining whether or not the temperature difference between the current room temperature and the set temperature is within ± 0.5 ° C is illustrated, the threshold value of the temperature difference is not limited to ± 0.5 ° C. It can be changed as appropriate.

Further, each component of each of the illustrated parts does not necessarily have to be physically configured as shown in the figure. That is, the specific form of distribution / integration of each part is not limited to the one shown in the figure, and all or part of them are functionally or physically distributed / integrated in arbitrary units according to various loads and usage conditions. Can be configured.

Further, various processing functions performed by each device are performed on the CPU (Central Processing Unit) (or a microcomputer such as an MPU (Micro Processing Unit) or an MCU (Micro Controller Unit)) in whole or in any part thereof. You may try to do it. Further, various processing functions may be executed in whole or in any part on a program analyzed and executed by a CPU (or a microcomputer such as an MPU or MCU) or on hardware by wired logic. Needless to say.

1 Air conditioner system 2 Indoor unit 2D storage unit 3 Adapter 5 Server device 6 Relay device 6D Relay side learning storage unit 10 Outdoor unit 13B Adapter side learning storage unit 14A Acquisition unit 14B Adapter side transmission unit 14E Prediction unit 32B Server side learning storage unit 33A Learning unit 33C Identification unit 33D Server side transmitter 50 Air conditioner

Claims (5)

An air conditioning system having an indoor unit, a server device that generates a plurality of types of learning models, and an adapter that connects the indoor unit and the server device by communication.
The adapter
The adapter side that transmits the identification information that identifies the indoor unit, the operation history data for a predetermined time of the indoor unit, and the learning model type information that identifies the type of the target learning model used in the indoor unit to the server device. Has a transmitter and
The server device
An identification unit that identifies the type of the learning model from the learning model type information received from the adapter, and an indoor unit to be transmitted from the identification information received from the adapter.
A learning unit that generates the learning model using the operation history data used to generate the learning model of the target identified by the identification unit, and a learning unit that generates the learning model.
An air conditioning system including a server-side transmission unit that transmits the learning model generated by the learning unit to the adapter that connects to the indoor unit of the transmission destination identified by the identification unit. The learning unit
The first aspect of the present invention is to generate the changed learning model in the learning model type information when the target learning model in the learning model type information from the adapter is changed. Described air conditioning system. The adapter
The air conditioning system according to claim 1, further comprising a prediction unit that executes a prediction regarding the indoor unit based on the learning model acquired from the server device. The learning model type information is
Each of the learning identification information that identifies the type of the target learning model is information that is managed by associating a binary value indicating whether or not the learning model can be applied to a specific indoor unit. The air conditioning system according to claim 1. An air conditioner that can be controlled by using a learning model generated by an external server device based on operation history information, in order to communicate with the indoor unit, outdoor unit, and external server device. With an adapter,
The indoor unit is an air conditioner including a storage unit that stores learning model type information for identifying the type of learning model used in the indoor unit.

Images