Bathroom hot water supply system

Abstract

To provide a bathroom hot water supply system which can appropriately estimate the number of persons per household of a residence in which a water heater is installed.SOLUTION: A server is configured to estimate the number of persons per household of a residence in which a predetermined water heater is installed from data acquired from the predetermined water heater by using a learned model. The data acquired from the predetermined water heater includes first data including at least one of the total use amount of hot and cold water and the total use amount of fuel in a first period, second data including at least one of the total entrance time to a bathroom and the total bathing time to a bathtub in the first period, and third data including at least one of the total entrance frequencies to the bathroom and the total bathing frequencies to the bathtub in the first period, and the learned model is generated by machine learning using explanatory variable data including the first data, the second data, the third data acquired from a plurality of water heaters as explanatory variables and using the number of persons per household preliminarily registered to each of the plurality of water heaters as objective variables.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a bathroom hot water supply system.

Background Art

[0002] There is known a bathroom hot water supply system in which a water heater having a hot water supply function to a bathtub and a server are communicably connected via a network, and various data such as the usage state of the water heater are transmitted to the server. By acquiring various data for each water heater in the server, it is possible to propose maintenance and usage methods according to the usage situation of the user.

[0003] As data related to such a water heater, it is conceivable to acquire the number of household members in the residence where the water heater is installed. However, if the user is made to set and input the number of household members to the water heater, there is a risk that the number of household members will not be updated even if the number of household members changes due to a life event or the like. In addition, it is also assumed that there are users who use the water heater without setting and inputting the number of household members in the first place. For this reason, there is a risk that the server may not be able to appropriately acquire the data on the number of household members.

[0004] The following Patent Document 1 describes a system for estimating the presence or absence of childbirth in a household based on energy consumption performance data indicating the energy consumption performance in the household. In addition, the following Patent Document 2 describes estimating changes in the user's lifestyle based on operation log information including the operation date and time of home appliances. Further, the following Patent Document 3 describes determining the presence or absence of a life event in which the power usage situation of a house changes based on the history of power consumption data in the house.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Patent Document 2

Patent Document 3

Summary of the Invention

Problems to be Solved by the Invention

[0006] In applying the estimation of life events or the like as described above to a hot water supply system, there is room for improvement in Patent Documents 1 to 3 above.

[0007] The present invention has been made to solve the above problems, and an object thereof is to provide a bathroom hot water supply system capable of appropriately estimating the number of household members in a residence where a hot water supply machine is installed.

Means for Solving the Problems

[0008] In order to achieve the above object, a bathroom hot water supply system according to an aspect of the present invention includes a plurality of hot water supply machines, each installed at least one per residence, for supplying hot water to bathtubs installed in bathrooms of a plurality of residences, and a server communicably connected to the plurality of hot water supply machines. The server is configured to estimate the number of household members in the residence where the predetermined hot water supply machine is installed from data acquired from the predetermined hot water supply machine included in the plurality of hot water supply machines using a learned model generated by machine learning. The data acquired from the predetermined hot water supply machine includes first data including at least one of the total usage amount of hot and cold water and the total usage amount of fuel in a first period, second data including at least one of the total entry time into the bathroom and the total bathing time in the bathtub in the first period, and third data including at least one of the total number of entries into the bathroom and the total number of bathing times in the bathtub in the first period. The learned model is a learned model generated by machine learning using, as explanatory variables, the explanatory variable data including the first data, the second data, and the third data acquired from the plurality of hot water supply machines, and, as the objective variable, the number of household members registered in advance for each of the plurality of hot water supply machines.

[0009] According to the above configuration, the number of household members in the residence where the water heater is installed is estimated based on the water heater and the usage status of the bathroom. At this time, the first data regarding the total usage amount of hot water or fuel having a high correlation with the number of household members, the second data regarding the entry time or bath time, and the third data regarding the number of entries or the number of baths are used as explanatory variables, and a learned model generated by machine learning with the number of household members pre-registered in the corresponding water heater 1 as the objective function is used. Therefore, by inputting the first data, the second data, and the third data of the water heater installed in the residence for which the number of household members is to be estimated into the learned model, the number of household members in the target residence can be accurately estimated.

[0010] The server may acquire the explanatory variable data in the plurality of water heaters and the corresponding number of household members, perform machine learning with the explanatory variable data as the explanatory variable and the number of household members as the objective variable, and generate the learned model. Thereby, the transfer of data used for machine learning can be smoothly performed.

[0011] The explanatory variable data may include at least any one of the fourth data including at least one of the standard deviation of the entry time to the bathroom and the standard deviation of the bath time to the bathtub in the first period, and the fifth data including at least one of the standard deviation of the entry time to the bathroom and the standard deviation of the bath time to the bathtub in the first period. By adding at least any one of the fourth data representing the variation in the entry time and / or the bath time, and the fifth data representing the variation in the entry time and / or the bath time as explanatory variables in addition to the first data, the second data, and the third data, machine learning with the number of household members in the residence where the water heater is installed as the objective variable can be performed more accurately.

[0012] The bathroom hot water supply system may include a notification device that notifies the household population estimated by the server. By notifying the user of the estimated household population, an opportunity can be provided for the user to confirm the household population estimated by the server, so that accurate registration of the household population can be realized for many hot water supply machines.

[0013] The bathroom hot water supply system includes a storage device that stores the household population and a notification device. When the household population estimated using the learned model by the server is different from the household population stored in the storage device, the server may cause the notification device to issue a notification. According to this, when the estimated household population is different from the previously registered household population, the user can be prompted to confirm or change the registered content, so that even when there is a change in the household population in the residence where the hot water supply machine is installed, it is possible to prevent the user from forgetting to change the registration of the household population and leaving it unattended.

[0014] The predetermined hot water supply machine includes a hot water supply machine main body and a controller that controls the hot water supply machine main body. The controller may receive the data of the household population estimated by the server and control the hot water supply machine main body based on the household population. By performing hot water supply control according to the household population, it is possible to realize control with higher user satisfaction.

Advantages of the Invention

[0015] The present invention has the configuration described above and has the effect of being able to appropriately estimate the household population of a residence where a hot water supply machine is installed.

Brief Description of the Drawings

[0016]

Figure 1

Figure 2

Figure 3

Figure 4

Embodiment for Carrying Out the Invention

[0017] Hereinafter, preferred embodiments will be described with reference to the drawings. In the following, the same or corresponding elements are denoted by the same reference numerals throughout all the drawings, and the overlapping descriptions thereof are omitted. Also, the present invention is not limited to the following embodiments.

[0018] [Embodiment] FIG. 1 is a block diagram showing a schematic configuration of a bathroom hot water supply system in an embodiment of the present invention. In FIG. 1, among a plurality of hot water supply machines 1, only one hot water supply machine 1 is described in more detail, and the illustration of the other hot water supply machines 1 is omitted, but the other hot water supply machines 1 also have the same configuration.

[0019] The bathroom hot water supply system 100 in the present embodiment includes a plurality of hot water supply machines 1 installed at least one for each dwelling and a server 2 in order to supply hot water to a bathtub 10 installed in a bathroom 6 of a plurality of dwellings. The hot water supply machine 1 includes a hot water supply machine main body 4 and a controller 5. Further, the hot water supply machine 1 includes a remote control for remotely operating the hot water supply machine main body 4. The remote control includes a bathroom remote control 7 installed in the bathroom 6 and a kitchen remote control 8 installed in the kitchen. The kitchen remote control 8 includes a monitor 30 as a notifier for notifying setting information and information transmitted from the server 2. In addition to or instead of the monitor 30, the kitchen remote control 8 may include a predetermined lamp or speaker as a notifier. Also, the bathroom remote control 7 may also include a notifier.

[0020] Server 2 includes a management server 21 and an analysis server 22. The analysis server 22 is communicatively connected to the management server 21 via a communication network 3 such as the Internet. Alternatively, the management server 21 and the analysis server 22 may be directly connected so as to be able to transmit and receive data by a communication cable or the like. In the example of FIG. 1, an example is shown in which the management server 21 and the analysis server 22 are configured as separate servers, but the management server 21 and the analysis server 22 may be configured as a common (single) server.

[0021] The management server 21 is communicatively connected to a plurality of water heaters 1 via a communication network 3 such as the Internet. For this purpose, the water heater 1 includes a relay device 9 dedicated to the water heater, which is communicatively connected to the controller 5. In the present embodiment, the relay device 9 is built into the kitchen remote controller 8. Alternatively, the relay device 9 may be configured as a device separate from the kitchen remote controller 8. The controller 5 of the water heater 1, the remote controllers 7 and 8, and the relay device 9 are configured to be able to communicate with each other by two-wire communication.

[0022] The water heater main body 4 is installed at a predetermined location outside or inside the house. The controller 5 is provided inside the housing of the water heater main body 4. The water heater main body 4 is, for example, a combustion heating type water heater device (heat source machine). The water heater main body 4 is connected to a bathtub 10 installed in the bathroom 6 by piping or the like. The water heater main body 4 has a function of supplying hot water to the bathtub 10 and a function of circulating the hot water stored in the bathtub 10 through the piping and reheating the circulating hot water. The water heater main body 4 includes a combustion device for burning fuel gas to heat the water introduced from the outside or the circulating hot water.

[0023] The controller 5 controls the operation of the water heater main body 4. For this purpose, the controller 5 is composed of a storage unit that stores various data, control programs, etc., and a processing circuit that includes a microcontroller having a CPU and memories (such as RAM and ROM) and performs various operations based on the control program. As controls for the water heater main body 4, the controller 5 is configured to be able to execute bathtub automatic control in which water is supplied to the bathtub 10 until the hot water stored in the bathtub 10 reaches the set water level or set hot water volume, and the temperature of the hot water stored in the bathtub 10 is set to the set temperature, and follow-up control for reheating the hot water stored in the bathtub 10.

[0024] The water heater 1 is equipped with various sensors such as a water flow rate sensor 11, a water level sensor 13, a human presence sensor 14, and an incoming water temperature sensor (not shown). The water flow rate sensor 11 detects the flow rate of the water flowing through the water inlet passage supplied from the outside. The water level sensor 13 is provided in the hot water circulation passage in the water heater main body 4 and detects the water level of the hot water stored in the bathtub 10. The human presence sensor 14 detects the presence or absence of a person in the bathroom 6. The human presence sensor 14 is provided, for example, on the bathroom remote controller 7.

[0025] The controller 5 acquires various data detected by these sensors and uses them for controlling the operation of the water heater main body 4. The controller 5 determines the heating amount of the hot water supplied to the bathtub 10 based on the data acquired from the various sensors and the operation content input to the remote controllers 7 and 8. In the storage unit of the controller 5, a conversion value of the fuel gas usage amount with respect to the unit heating amount determined according to the fuel gas type is stored in advance. The controller 5 calculates the fuel gas usage amount from the determined heating amount and the conversion value. Therefore, the fuel gas usage amount is a value associated with the heating amount. Note that instead of the configuration in which the controller 5 calculates the fuel gas usage amount, the water heater 1 may be equipped with a gas flow rate sensor that detects the usage amount of the fuel gas.

[0026] Furthermore, the controller 5 stores in the storage unit the data acquired from various sensors and the calculated fuel gas usage amount together with the time of acquisition, and transmits them to the management server 21 at predetermined timings. The management server 21 includes a storage device 20 and stores the various data related to the water heater 1 sent thereto in association with the water heater ID for identifying the water heater 1. Note that a user ID for identifying the user who owns the water heater 1 may be used as the ID for identifying the water heater 1.

[0027] The bathroom hot water supply system 100 is configured to be able to register the number of household members in the dwelling where the water heater 1 is installed. For example, the water heater 1 may be configured such that the user can operate the kitchen remote controller 8 to register the number of household members. In this case, the controller 5 transmits the data of the registered number of household members to the management server 21.

[0028] Alternatively, when a management app that enables viewing of the operating state of the water heater 1 or remote operation of the water heater 1 is installed on a communication terminal such as the user's smartphone, the number of household members may be registered from the communication terminal. In this case, the communication terminal on which the management app is installed is connected to the management server 21 via the communication network 3 and transmits the data of the number of household members to the management server 21.

[0029] Or, when the user registration information sent by the user by mail or the like includes the number of household members, the administrator who has received the user registration information may perform a registration operation on the management server 21.

[0030] The analysis server 22 is configured to estimate the number of household members in the dwelling where the water heater 1 is installed from the data acquired from the water heater 1 for each water heater ID. For this purpose, the storage device 27 of the analysis server 22 stores a learned model 26 generated by machine learning from various data acquired from a plurality of water heaters 1. The analysis server 22 includes a learning unit 23 that performs machine learning to generate a learned model, which will be described later, and an estimation unit 24 that performs an estimation, which will be described later, using the learned model generated by the learning unit 23. Note that the "number of household members" in this specification and the claims can be defined as the number of people using a single predetermined water heater 1.

[0031] The analysis server 22 acquires explanatory variable data obtained from various data acquired from a plurality of water heaters 1 from the management server 21 and data on the corresponding number of household members registered in advance. The learning unit 23 performs machine learning with the explanatory variable data as the explanatory variable and the number of household members as the target variable. The learning unit 23 generates a household bathing vector in which the explanatory variable data and the data on the number of household members based on various data of the water heater 1 collected during a predetermined first period (number of days, for example, 31 days) are tabulated. The explanatory variable data may include essential explanatory variables, recommended explanatory variables, and other explanatory variables. The data on the number of household members can be paraphrased as target variable data.

[0032] FIG. 2 is a diagram showing an example of a household bathing vector in the present embodiment. The essential explanatory variables include a first data D1, a second data D2, and a third data D3. The first data D1 includes at least one of data D11 on the total usage amount of hot and cold water and data D12 on the total usage amount of fuel during the first period. The data D11 on the total usage amount of hot and cold water is obtained by accumulating data detected by the water flow sensor 11. Also, the data D12 on the total usage amount of fuel is obtained by accumulating data on the fuel gas usage amount calculated by the controller 5. As described above, since the fuel gas usage amount is associated with the heating amount, the total usage amount of fuel can be paraphrased as the total heating amount.

[0033] The second data D2 includes at least one of the data D21 of the total entry time into the bathroom 6 and the data D22 of the total bathing time in the bathtub 10 during the first period. The data D21 of the total entry time into the bathroom 6 is obtained by accumulating the data of the detection period detected by the human presence sensor 14. Also, the data D22 of the total bathing time in the bathtub 10 is obtained by accumulating the data of the detection period detected by the water level sensor 13. The detection period of the water level sensor 13 is the period from the time when the increase amount of the water level per unit time becomes equal to or more than the first reference value to the time when the decrease amount of the water level per unit time becomes equal to or more than the second reference value. The first reference value and the second reference value may be the same value or different values.

[0034] The third data D3 includes at least one of the data D31 of the total number of entries into the bathroom 6 and the data D32 of the total number of baths in the bathtub 10 during the first period. The data D31 of the total number of entries into the bathroom 6 is obtained by accumulating the data of the number of detections detected by the human presence sensor 14. Also, the data D32 of the total number of baths in the bathtub 10 is obtained by accumulating the data of the number of detections detected by the water level sensor 13.

[0035] Also, the recommended explanatory variable includes at least one of the fourth data D4 and the fifth data D5. The fourth data D4 includes at least one of the data D41 of the standard deviation of the entry time into the bathroom 6 and the data D42 of the standard deviation of the bathing time in the bathtub 10 during the first period. The data D41 of the standard deviation of the entry time into the bathroom 6 is calculated from the data of the detection period for each detection detected by the human presence sensor 14 and the data D61 of the average entry time described later. The data D42 of the standard deviation of the bathing time in the bathtub 10 is calculated from the data of the detection period for each detection detected by the water level sensor 13 and the data D62 of the average bathing time described later.

[0036] Further, the fifth data D5 includes at least one of the data D51 of the standard deviation of the bathroom entry time and the data D52 of the standard deviation of the bathtub bathing time in the first period. The data D51 of the standard deviation of the bathroom entry time is calculated from the data of the detection time for each detection by the human presence sensor 14 and the data D63 of the average entry time described later. The data D52 of the standard deviation of the bathtub bathing time is calculated from the data of the detection time for each detection by the water level sensor 13 and the data D64 of the average bathing time described later.

[0037] Other explanatory variables may include, for example, the data D61 of the average entry time, the data D62 of the average bathing time, the data D63 of the average entry time, the data D64 of the average bathing time, the data D65 of the number of bathing days, etc. in the first period. The data D61 of the average entry time is the average entry time per day and is calculated from the data D21 of the total entry time and the first period. The data D62 of the average bathing time is the average bathing time per day and is calculated from the data D22 of the total bathing time and the data D65 of the number of bathing days.

[0038] The data D63 of the average entry time is the average entry time per day and is calculated from the data of the detection time for each detection by the human presence sensor 14 and the first period. The data D64 of the average bathing time is the average bathing time per day and is calculated from the data of the detection time for each detection by the water level sensor 13 and the data D65 of the number of bathing days.

[0039] As shown in FIG. 2, a household bathing vector is generated in which these explanatory variable data and the household number data Do are tabulated for each water heater ID (A, B, C,...). In FIG. 2, for example, each data D11, D12, D21,..., Do with the water heater ID being A is denoted as D11a, D12a, D21a,..., Doa, and each data D11, D12, D21,..., Do with the water heater ID being B is denoted as D11b, D12b, D21b,..., Dob. The same applies to each data D11, D12, D21,..., Do for other water heater IDs.

[0040] The first data D1, the second data D2, and the third data D3 are all data with a higher likelihood that the larger the numerical value, the larger the household size. As a result of intensive research, the inventors of the present invention have found that by combining these three types of data D1, D2, and D3 as essential explanatory variables and inputting them into the learning unit 23, machine learning with the household size of the dwelling where the water heater 1 is installed as the target variable can be performed with high accuracy. The present invention is based on this finding.

[0041] That is, the learning unit 23 constructs a supervised learning model using the explanatory variable data including the first data D1, the second data D2, and the third data as explanatory variables and the corresponding household size as the target variable. In the present embodiment, in order to estimate the household size, a supervised learning model that performs multi-class classification is adopted. For example, when the upper limit at the time of registering the household size is 6 people, the supervised learning model is constructed as a learning model that outputs any one of 1 to 6 as the household size. As the learning model, for example, any one or a combination of known learning models such as neural networks, support vector machines, decision trees, and gradient boosting decision trees can be adopted.

[0042] First, the learning unit 23 performs data cleansing on the explanatory variable data and the household size data in the household bathing vector. In data cleansing, the learning unit 23 checks for outliers outside a predetermined assumed range or the presence or absence of data loss, and if there are outliers or losses, deletes the data or does not adopt all the data related to the water heater 1.

[0043] Next, the learning unit 23 normalizes the explanatory variable data after data cleansing for each of the data D11, D12, D21,.... The content of the normalization may be, for example, normalization that scales the minimum value to 0 and the maximum value to 1, or standard deviation that scales the mean to 0 and the variance to 1. By performing normalization, it is possible to reduce the influence on machine learning due to differences in scales between multiple data types (for example, between the first data D1 and the second data D2).

[0044] FIG. 3 is a diagram showing an example of a neural network that can be adopted as a learning model in the present embodiment. The neural network 25 includes an input layer N1 having m input nodes, an output layer N2 having n output nodes, and an intermediate layer N3 between the input layer N1 and the output layer N2. The intermediate layer N3 is composed of one or a plurality of layers. The number m of input nodes corresponds to the number of data types of the explanatory variable data (at least three). The number n of output nodes corresponds to the upper limit value (6 in the above example) when registering the number of household members.

[0045] In such a learning model, the standardized explanatory variable data is input to each input node of the input layer N1 as an explanatory variable. In the example of FIG. 3, an example is shown in which four essential explanatory variables are input: the data D11 of the total amount of hot and cold water used belonging to the first data D1, the data D21 of the total room entry time and the data D22 of the total bathing time belonging to the second data D2, and the data D32 of the total number of bathing times belonging to the third data D3. That is, in the example of FIG. 3, the number m of input nodes is 4.

[0046] Each output node in the output layer N2 outputs the probability P1 that the number of household members is 1, the probability P2 that the number of household members is 2,..., and the probability P6 that the number of household members is 6 when the explanatory variables in a certain water heater 1 are input to the input layer N1. In supervised learning, learning is performed so that the probability at the output node corresponding to the pre-registered number of household members becomes the highest when the explanatory variables in a certain water heater 1 are input to the input layer N1.

[0047] In such supervised learning, the explanatory variables input to the input nodes of the input layer N1 may include recommended explanatory variables. That is, in addition to the first data D1 (data D11 and / or D12), the second data D2 (data D21 and / or D22), and the third data D3 (data D31 and / or D32), which are essential explanatory variables, at least one of the above data D41, D42, D51, and D52 may be input to the input nodes of the input layer N1.

[0048] As described above, the fourth data D4 and the fifth data D5 are data indicating the standard deviations of the entry time, bathing time, entry time, and bathing time, that is, the degree of variation. It can be said that the greater the numerical values of these data D4 and D5, the greater the degree of variation. The larger the number of household members, the greater the tendency for variations such as the entry time to be larger. Therefore, the fourth data D4 and the fifth data D5 are also data with a higher likelihood that the larger the numerical value, the larger the number of household members.

[0049] Therefore, by inputting at least any one of the fourth data D4 and the fifth data D5 as an explanatory variable into the learning unit 23 in addition to the data D1, D2, and D3 which are essential explanatory variables, machine learning can be performed more accurately with the number of household members of the dwelling where the water heater 1 is installed as the target variable. Furthermore, any of the sixth data D6 which are other explanatory variables may be input as an explanatory variable into the learning unit 23.

[0050] For a plurality of water heaters 1 in which the number of household members is registered in advance, machine learning is repeatedly performed using the explanatory variable data and the registered number of household members, whereby a learning model composed of the neural network 25 is generated as a learned model 26. The generated learned model 26 is stored in the storage device 27 of the analysis server 22.

[0051] The estimation unit 24 estimates the number of household members of the dwelling where the water heater 1 is installed from the data acquired from a predetermined water heater 1 using the learned model 26. For example, when the management server 21 receives various data in a water heater 1 where the number of household members is not registered, it stores the data in the storage device 20 and transmits it to the analysis server 22.

[0052] Figure 4 is a diagram showing the schematic configuration of the estimation unit in the present embodiment. The estimation unit 24 of the analysis server 22 includes an input unit 28, a learned model 26, and an output unit 29. The estimation unit 24 of the analysis server 22 generates necessary explanatory variable data based on various data in the received water heater 1. The necessary explanatory variable data is the explanatory variable data used during learning (data D11, D21, D22, D32 in the example of FIG. 2). The input unit 28 of the estimation unit 24 inputs the explanatory variable data D11x, D21x, D22x, D32x in the water heater 1 (water heater ID: X) to the input nodes of the input layer N1 in the learned model 26 stored in the storage device 27.

[0053] When the explanatory variable data in the water heater 1 is input to the input nodes of the learned model 26, values of probabilities P1 to P6, which are the number of household members assigned to each, are output from the output nodes in the output layer N2. The output unit 29 of the estimation unit 24 outputs the number of household members i assigned to the output node Pi (i = 1, 2,..., 6), which is the largest value among the probabilities P1 to P6 output from each output node of the output layer N2, as an estimated value of the number of household members in the dwelling where the water heater 1 is installed.

[0054] The estimated value of the number of household members output from the estimation unit 24 is associated with the water heater ID of the explanatory variable data used for the estimation and is stored in the storage device 20 of the management server 21. The bathroom hot water supply system 100 can execute various processes based on the estimated value of the number of household members estimated by the estimation unit 24 of the analysis server 22.

[0055] For example, the management server 21 transmits the estimated value of the number of household members to the corresponding water heater 1. The controller 5 of the corresponding water heater 1 displays and notifies the estimated number of household members on the monitor 30, which is a notification device. By notifying the user of the estimated number of household members, an opportunity for the user to confirm the number of household members estimated by the server 2 can be provided, so that accurate registration of the number of household members can be realized for many water heaters 1.

[0056] Furthermore, the controller 5 may notify the estimated number of household members and ask the user whether the number of household members is correct. For example, the controller 5 displays a question text on the monitor 30 asking whether the estimated number of household members is correct, and accepts an answer input operation by the kitchen remote controller 8. If the answer input from the user is Yes (an answer indicating that the number of household members is correct), the controller 5 transmits a confirmation signal indicating that the number of household members is correct to the management server 21. If the answer input from the user is No (an answer indicating that the number of household members is incorrect), the controller 5 displays a message on the monitor 30 prompting the user to input an appropriate number of household members.

[0057] If there is an input of an appropriate number of household members from the user, the controller 5 transmits a correction signal for the number of household members to the management server 21. The management server 21 updates and stores the corrected number of household members in the storage device 20 in association with the water heater ID of the water heater 1 that is the transmission source. By prompting the user to correct this when the estimated number of household members is different from the actual number of household members, it is possible to register the correct number of household members for many water heaters 1.

[0058] When a management application that can view the operating state of the water heater 1 or remotely operate the water heater 1 is installed on a communication terminal such as the user's smartphone, the management server 21 may notify the estimated number of household members on the management application. In this case, the notifier is constituted by a monitor or the like of the communication terminal.

[0059] Also, the controller 5 may store the number of household members in a storage device (not shown) of the water heater 1. The controller 5 can control the water heater main body 4 based on the data of the estimated number of household members. For example, the controller 5 may detect the number of bathing times every day, and automatically end the bathtub automatic control when the number of bathing times reaches the number corresponding to the estimated number of household members. By performing hot water supply control according to the number of household members, it is possible to realize control with higher user satisfaction.

[0060] Note that the control of the water heater main body 4 based on the estimated number of household members may be mainly executed by the controller 5, or may be passively executed by the controller 5 based on a control command from the management server 21.

[0061] In addition, the estimation of the number of household members can be performed not only on the water heater 1 for which the number of household members is not registered, but also on the water heater 1 for which the number of household members has already been registered. For example, regardless of whether the number of household members is registered or not, the number of household members may be estimated every predetermined second period. The second period is a period longer than the first period which is the data collection period. The second period can be set to a period that is approximately an integer multiple of the first period, such as 6 months, 1 year, etc.

[0062] When the estimated number of household members is different from the number of household members already stored in the storage device 20 in the corresponding water heater 1, the management server 21 causes the monitor 30 of the corresponding water heater 1 or the management application to notify that fact.

[0063] For example, the number of household members stored in the storage device 20 is displayed on the monitor 30 or the like, and a message is displayed to prompt the user to confirm whether it matches the current number of household members and, if necessary, correct the number of household members. At this time, the estimated number of household members may be displayed on the monitor 30 or the like. Alternatively, only the estimated number of household members may be displayed on the monitor 30 or the like, and a display may be made to prompt the user to confirm whether it is appropriate. According to this, when the estimated number of household members is different from the previously registered number of household members, the user can be prompted to confirm or change the registration content, so that even if there is a change in the number of household members in the residence where the water heater 1 is installed, it is possible to prevent the user from forgetting to change the registration of the number of household members and leaving it unattended.

[0064] Alternatively, instead of displaying a message prompting the user to correct the number of household members, the management server 21 may update and store the estimated number of household members in the storage device 20 regardless of whether it is the same as the number of household members stored in the storage device 20.

[0065] In addition, the management server 21 may display, on the monitor 30 or the like, setting changes according to the registered household size data and guidance messages regarding services. For example, for the water heater 1 with a large household size, a guidance message prompting a setting change to an eco-mode that is more effective as the amount of hot water used increases may be displayed. Also, for example, for the water heater 1 with a household size of one person, a guidance message prompting the setting of a monitoring function based on the detection of entry by the human presence sensor 14 and the detection of bathing by the water level sensor 13 may be displayed.

[0066] As described above, according to the present embodiment, the household size of the residence where the water heater 1 is installed is estimated based on the usage status of the water heater 1 and the bathroom 6. At this time, the first data D1 regarding the total usage amount of hot water or fuel having a high correlation with the household size, the second data D2 regarding the entry time or bathing time, and the third data D3 regarding the number of entries or the number of baths are used as explanatory variables, and the learned model 26 generated by machine learning with the household size registered in advance in the corresponding water heater 1 as the objective function is used. Therefore, by inputting the first data D1, the second data D2, and the third data D3 of the water heater 1 installed in the residence for which the household size is to be estimated into the learned model 26, the household size of the target residence can be accurately estimated.

[0067] Also, according to the present embodiment, since the learning unit 23 of the server 2 performs machine learning, the transfer of data used for machine learning can be performed smoothly.

[0068] From the above description, many improvements and other embodiments of the present invention will be apparent to those skilled in the art. Therefore, the above description should be construed as illustrative only and is provided for the purpose of teaching those skilled in the art the best mode of carrying out the present invention. Without departing from the spirit of the present invention, the details of its structure and / or function can be substantially changed.

[0069] [Other Embodiments] For example, in the above-described embodiment, the analysis server 22 is illustrated as including the learning unit 23. However, the learning unit 23 may not be provided in the bathroom hot water supply system 100. That is, machine learning may be performed in a learning device separate from the bathroom hot water supply system 100, and it is sufficient that the learned model 26 generated as a result is stored in the server 2. In this way, the learned model 26 may be generated only from the data of the explanatory variables and the objective variables acquired during the past learning period. Therefore, for machine learning to generate the learned model 26, some or all of the plurality of water heaters that have acquired various data do not have to be included in the plurality of water heaters 1 connected to the server 2 in the current bathroom hot water supply system 100.

[0070] On the other hand, when the server 2 includes the learning unit 23, the learning unit 23 may perform machine learning using the data of the explanatory variables used in the estimation in the estimation unit 24 and the data of the household size obtained as the estimation result or the data of the household size corrected and input thereto, and update the learned model 26.

[0071] [Summary of the present disclosure] Each of the following items discloses a preferred embodiment of the present disclosure.

[0072] [Item 1] A bathroom hot water supply system includes a plurality of hot water supply devices, with at least one installed for each dwelling, for supplying hot water to bathtubs installed in bathrooms of a plurality of dwellings, and a server communicably connected to the plurality of hot water supply devices. The server is configured to estimate the number of household members in the dwelling where the predetermined hot water supply device is installed from data acquired from the predetermined hot water supply device included in the plurality of hot water supply devices, using a learned model generated by machine learning. The data acquired from the predetermined hot water supply device includes first data including at least one of the total usage amount of hot and cold water and the total usage amount of fuel in a first period, second data including at least one of the total entry time into the bathroom and the total bathing time in the bathtub in the first period, and third data including at least one of the total number of entries into the bathroom and the total number of bathing times in the bathtub in the first period. The learned model is a learned model generated by machine learning that uses the explanatory variable data including the first data, the second data, and the third data acquired from the plurality of hot water supply devices as explanatory variables and the pre-registered number of household members for each of the plurality of hot water supply devices as the target variable.

[0073] [Item 2] The server according to item 1 of the bathroom hot water supply system described above acquires the explanatory variable data and the corresponding number of household members in the plurality of hot water supply devices, performs machine learning using the explanatory variable data as explanatory variables and the number of household members as the target variable, and generates the learned model.

[0074] [Item 3] The explanatory variable data according to item 1 or 2 of the bathroom hot water supply system described above includes at least one of fourth data including at least one of the standard deviation of the entry time into the bathroom and the standard deviation of the bathing time in the bathtub in the first period, and at least one of fifth data including at least one of the standard deviation of the entry time into the bathroom and the standard deviation of the bathing time in the bathtub in the first period.

[0075] [Item 4] The hot water supply system for a bathroom according to any one of Items 1 to 3, comprising a notification device that notifies the household size estimated by the server.

[0076] [Item 5] The hot water supply system for a bathroom according to any one of Items 1 to 4, comprising a storage device that stores the household size and a notification device, wherein when the household size estimated using the learned model by the server is different from the household size stored in the storage device, the server causes the notification device to issue a notification.

[0077] [Item 6] The hot water supply system for a bathroom according to any one of Items 1 to 5, wherein the predetermined hot water supply machine includes a hot water supply machine main body and a controller that controls the hot water supply machine main body, and the controller receives data on the household size estimated by the server and controls the hot water supply machine main body based on the household size.

Industrial Applicability

[0078] The present invention is useful for appropriately estimating the household size of a residence in which a hot water supply machine is installed in a hot water supply system for a bathroom.

Explanation of Signs

[0079] 1... Hot water supply machine, 2... Server, 4... Hot water supply machine main body, 5... Controller, 20... Storage device, 26... Learned model, 30... Monitor (notification device), 100... Hot water supply system for bathroom

Claims

1. A bathroom hot water supply system comprising: a plurality of hot water supply devices, each dwelling having at least one hot water supply device installed therein for supplying hot water to a bathtub installed in the bathroom of the plurality of dwellings; and a server communicably connected to the plurality of hot water supply devices, wherein the server is configured to estimate the number of household members of the dwelling in which the predetermined hot water supply device is installed from data acquired from the predetermined hot water supply device included in the plurality of hot water supply devices using a learned model generated by machine learning, the data acquired from the predetermined hot water supply device includes first data including at least one of the total amount of hot and cold water used and the total amount of fuel used during a first period, second data including at least one of the total time of entering the bathroom and the total bathing time in the bathtub during the first period, and third data including at least one of the total number of times of entering the bathroom and the total number of times of bathing in the bathtub during the first period, and the learned model is a learned model generated by machine learning using, as explanatory variable data, the first data, the second data, and the third data including the explanatory variable data acquired from the plurality of hot water supply devices, and using, as the objective variable, the number of household members pre-registered for each of the plurality of hot water supply devices. A bathroom hot water supply system.

2. The server acquires the explanatory variable data in the plurality of hot water supply devices and the corresponding number of household members, performs machine learning using the explanatory variable data as an explanatory variable and the number of household members as an objective variable, and generates the learned model. The bathroom hot water supply system according to claim 1.

3. The explanatory variable data includes fourth data including at least one of the standard deviation of the time of entering the bathroom and the standard deviation of the bathing time in the bathtub during the first period, and at least any one of fifth data including at least one of the standard deviation of the time of entering the bathroom and the standard deviation of the bathing time in the bathtub during the first period. The bathroom hot water supply system according to claim 1 or 2.

4. The bathroom hot water supply system according to claim 1 or 2, further comprising a notification device for notifying the number of household members estimated by the server.

5. A storage device for storing the number of household members; a notification device; and when the number of household members estimated using the learned model by the server is different from the number of household members stored in the storage device, the server causes the notification device to notify. The bathroom hot water supply system according to claim 1 or 2.

6. The predetermined water heater includes a water heater main body and a controller for controlling the water heater main body. The controller receives the data of the estimated household size estimated by the server and controls the water heater main body based on the household size. The bathroom water heating system according to claim 1 or 2.

Images