Hot water supply system

Abstract

The hot-water supply system (1) of the present disclosure includes a presumption unit (33) that predicts a total hot-water supply demand amount from time-series data of a first index that indicates the heat of water used by each of a plurality of hot-water supply targets (4).

Description

Technical Field

[0001] This disclosure relates to a hot water supply system. Background Technology

[0002] Patent Document 1 discloses a hot water supply system comprising a hot water generating mechanism such as a fuel cell or a gas engine, and a hot water storage tank utilizing the heat discharged from the hot water generating mechanism. This hot water supply system provides an optimal operating plan for the fuel cell by learning the demand patterns for hot water supply.

[0003] Existing technical documents

[0004] Patent documents

[0005] Patent Document 1: Japanese Patent Publication No. 2006-183947 Summary of the Invention

[0006] -The technical problem the invention aims to solve-

[0007] Patent Document 1 describes a hot water supply system that primarily uses the hot water supply from bathtubs and showers to learn the demand pattern for hot water. Therefore, when supplying hot water to multiple sources, the accuracy of predicting the required heat volume may not be very high.

[0008] The purpose of this disclosure is to improve the accuracy of predicting the total hot water supply demand in a hot water supply system that has multiple hot water supply targets and a water storage tank that supplies hot water to those targets.

[0009] - Technical solutions for solving technical problems -

[0010] The first aspect of this disclosure is a hot water supply system, which includes a hot water supply device 10, multiple supply paths 5, and a diversion unit 33.

[0011] The hot water supply device 10 includes a heating device 20, a water tank 40, and a water circuit 50. The heating device 20 heats the water, the water tank 40 stores the water heated by the heating device 20, and the water in the water tank 40 circulates in the water circuit 50.

[0012] Multiple supply paths 5 are connected to multiple hot water supply objects 4 respectively, and multiple supply paths 5 supply water from the water tank 40.

[0013] The prediction unit 33 predicts the total hot water supply demand based on the time series data of a first indicator, which represents the heat of water used by each of the plurality of hot water supply objects 4.

[0014] In the first aspect, since the total hot water supply demand is inferred based on the time series data of the heat of the water used by each hot water supply object 4, the prediction accuracy of the total hot water supply demand can be improved.

[0015] The second aspect is based on the first aspect.

[0016] The hot water supply system has a first learning unit 32, which learns by associating the time-series data with the total hot water supply demand.

[0017] The prediction unit 33 predicts the total hot water supply demand based on the learning results of the first learning unit 32.

[0018] In the second aspect, the first learning unit 32 can be used to predict the total demand for hot water supply.

[0019] The third aspect is based on the second aspect.

[0020] The first learning unit 32 learns through machine learning.

[0021] In the third aspect, the first learning unit 32 is able to learn through machine learning. Based on the learning results of machine learning, it is able to predict the total demand for hot water supply.

[0022] The fourth aspect is based on the first or third aspect.

[0023] The prediction unit 33 includes a prediction model M1, which is generated by machine learning to predict the total hot water supply demand based on the time series data.

[0024] The prediction unit 33 uses the prediction model M1 to predict the total hot water supply demand.

[0025] In the fourth aspect, the speculative model M1 can be used to predict the total demand for hot water supply.

[0026] The fifth aspect is based on any one of the first to fourth aspects.

[0027] The first indicator is the temperature and amount of water used by each of the plurality of hot water supply objects 4.

[0028] In the fifth aspect, the first index can be determined based on the temperature and quantity of the water used by the hot water supply object 4.

[0029] The sixth aspect is based on any one of the first to fourth aspects.

[0030] The first indicator is the temperature of the water used in each of the plurality of hot water supply objects 4 and the water pressure in each of the supply paths 5.

[0031] In the sixth aspect, the first index can be determined based on the water pressure and temperature of the water flowing in the supply path 5 connected to each hot water supply object 4, and the amount of water flowing out of the water tank 40.

[0032] The seventh aspect is based on any one of the first to sixth aspects.

[0033] The prediction unit 33 predicts the total hot water supply demand based on the time series data, the number of hot water supply objects 4, the type of hot water supply objects 4, and the faucet specifications of the hot water supply objects 4.

[0034] In the seventh aspect, the information used to predict the total hot water supply and demand includes not only time-series data on the heat of the water used by each hot water supply recipient 4, but also information specified by each hot water supply recipient. Thus, because multiple pieces of useful information are available for predicting the total hot water supply and demand, the accuracy of the total hot water supply and demand prediction can be improved compared to predictions based solely on the time-series data on the heat of the water.

[0035] The eighth aspect is based on any one of the first to seventh aspects.

[0036] The prediction unit 33 predicts the total hot water supply demand based on the time series data of the selected hot water supply target 4.

[0037] In the eighth aspect, the calculation of hot water supply objects that have a relatively small impact on the prediction of total hot water supply and demand can be omitted. By omitting hot water supply objects with noisy properties that would reduce prediction accuracy, the prediction accuracy of total hot water supply and demand can be improved.

[0038] The ninth aspect is based on any one of the first to eighth aspects.

[0039] The hot water supply system has a control unit 30, which controls the operation of the heating device 20 according to the total hot water supply demand.

[0040] In the ninth aspect, by predicting the operation control of the heating device 20, the heating device 20 can be operated with high efficiency.

[0041] The tenth aspect is based on the ninth aspect,

[0042] The hot water supply system has a second learning unit 35, which learns by associating the total hot water supply demand with the operating status of the heating device 20.

[0043] The control unit 30 controls the operation of the heating device 20 based on the learning results of the second learning unit 35.

[0044] In the tenth aspect, the operation of the heating device 20 can be controlled based on the learning results of the second learning unit 35.

[0045] The eleventh aspect is based on the tenth aspect.

[0046] The second learning unit 35 learns through machine learning.

[0047] In the eleventh aspect, the second learning unit 35 is capable of learning through machine learning. Based on the learning results of machine learning, it is able to predict the operation control of the heating device 20.

[0048] The twelfth aspect is based on the ninth or eleventh aspect.

[0049] The control unit 30 includes an operation prediction model M2, which is generated through machine learning to predict the operation control of the heating device 20 based on the total hot water supply demand.

[0050] The control unit 30 uses the operation prediction model M2 to control the operation of the heating device 20.

[0051] In the twelfth aspect, the operation prediction model M2 can be used to predict the operation control of the heating device 20.

[0052] The thirteenth aspect is based on any one of the first to twelfth aspects.

[0053] The heating device 20 is a heat pump type.

[0054] In the thirteenth aspect, even in heat pumps with relatively low start-up capacity, the risk of running out of hot water can be reduced, and the heating device 20 can be operated with high efficiency. Attached Figure Description

[0055] Figure 1 This is an overall structural diagram of the hot water supply system involved in the implementation method;

[0056] Figure 2 This is a block diagram of a hot water supply system;

[0057] Figure 3 This is a diagram showing the flow of refrigerant during the operation of the heating device;

[0058] Figure 4 This is a flowchart illustrating the operation of the hot water supply system;

[0059] Figure 5 This is a graph showing the relationship between total hot water supply demand and the operation control of heating devices;

[0060] Figure 6 This is a block diagram of the hot water supply system involved in the modified example 1 of the implementation method;

[0061] Figure 7 This is a block diagram of the hot water supply system involved in the modified example 2 of the implementation method;

[0062] Figure 8 This is a flowchart illustrating the operation of the hot water supply system;

[0063] Figure 9 This is a block diagram of the hot water supply system involved in variation 3 of the implementation method;

[0064] Figure 10 This is a block diagram of the hot water supply system involved in variation 4 of the implementation method;

[0065] Figure 11 This is a flowchart illustrating the operation of a hot water supply system. Detailed Implementation

[0066] The embodiments will now be described with reference to the accompanying drawings. It should be noted that the following embodiments are merely preferred examples and are not intended to limit the scope of the invention, its application, or its uses.

[0067] (Implementation Method)

[0068] like Figure 1 As shown, this disclosure discloses a hot water supply system 1. The hot water supply system 1 heats water supplied from a water source and stores the heated water in a water tank 40. The hot water stored in the water tank 40 is supplied to multiple hot water supply objects 4. The water source includes a water supply system. The hot water supply objects 4 include bathtubs, showers, faucets, etc.

[0069] The hot water supply system 1 includes a heating device 20, a water tank 40, a water circuit 50, a supply path 5, a first pipe 6, a pressure sensor 60, temperature sensors 61 and 63, and a control unit 30.

[0070] <Heating device>

[0071] The heating device 20 in this embodiment is a heat pump type. The heating device 20 generates heat energy for heating water. The heating device 20 is a vapor compression type. The heating device 20 has a refrigerant circuit 21. Refrigerant is filled in the refrigerant circuit 21. The refrigerant circuit 21 includes a compressor 22, a heat source heat exchanger 23, an expansion valve 24, and a heat exchanger 25.

[0072] The compressor 22 compresses the refrigerant that has been drawn in and then sprays out the compressed refrigerant.

[0073] The heat source heat exchanger 23 is an air-cooled heat exchanger. The heat source heat exchanger 23 is located outdoors. The heating unit 20 has an outdoor fan 27. The outdoor fan 27 is located near the heat source heat exchanger 23. The heat source heat exchanger 23 facilitates heat exchange between the air supplied by the outdoor fan 27 and the refrigerant.

[0074] Expansion valve 24 is a pressure-reducing mechanism for pressurizing the refrigerant. Expansion valve 24 is located between the liquid end of heat exchanger 25 and the liquid end of heat source heat exchanger 23. The pressure-reducing mechanism is not limited to an expansion valve; it can also be a capillary tube, an expander, etc. The expander recovers the energy of the refrigerant as power.

[0075] The heat exchanger 25 is a liquid-cooled heat exchanger. The heat exchanger 25 has a first flow path 25a and a second flow path 25b. The second flow path 25b is connected to the refrigerant circuit 21. The first flow path 25a is connected to the water circuit 50. The heat exchanger 25 facilitates heat exchange between the water flowing in the first flow path 25a and the refrigerant flowing in the second flow path 25b.

[0076] In the heat exchanger 25, a first flow path 25a is formed along a second flow path 25b. In this embodiment, during the heating operation described in detail below, the direction of the refrigerant flowing in the second flow path 25b is substantially opposite to the direction of the water flowing in the first flow path 25a. In other words, the heat exchanger 25 functions as a counter-current heat exchanger during heating operation.

[0077] <Water Tank and Water Circuit>

[0078] Water tank 40 is a container for storing water. Water tank 40 is formed into a longitudinally elongated cylindrical shape. Water tank 40 has a cylindrical body 41, a bottom 42 that seals the lower end of the body 41, and a top 43 that seals the upper end of the body 41.

[0079] Water in tank 40 circulates in water circuit 50. A first flow path 25a of heat exchanger 25 is connected to water circuit 50. Water circuit 50 includes an upstream flow path 51 and a downstream flow path 52.

[0080] The inflow end of the upstream flow path 51 is connected to the bottom 42 of the water tank 40. The outflow end of the upstream flow path 51 is connected to the inflow end of the first flow path 25a.

[0081] The inflow end of the downstream flow path 52 is connected to the outflow end of the first flow path 25a. The outflow end of the downstream flow path 52 is connected to the top 43 of the water tank 40.

[0082] Water circuit 50 has a water pump 53. Water pump 53 circulates water in water circuit 50. Water pump 53 delivers water from water tank 40 to first flow path 25a using heat exchanger 25. Then, water pump 53 delivers water to first flow path 25a and then to water tank 40.

[0083] <First Pipeline and Supply Route>

[0084] The inlet end of the first pipe 6 is connected to the water tank 40. The outlet end of the first pipe 6 is connected to the inlet ends of each of the multiple supply paths 5. The outlet ends of each supply path 5 are connected to each hot water supply object 4.

[0085] <Pressure Sensor>

[0086] Pressure sensors 60 are connected to each supply path 5. Pressure sensors 60 detect the pressure of the water in the supply path 5. In other words, pressure sensors 60 detect the pressure of the water supplied to each hot water supply object 4.

[0087] <Temperature Sensor>

[0088] The hot water supply system 1 has a first temperature sensor 61 and a second temperature sensor 63. The first temperature sensor 61 is installed on each hot water supply object 4. The first temperature sensor 61 detects the temperature of the water used by the hot water supply object 4. The second temperature sensor 63 is installed at the inlet end of the first pipe 6. The second temperature sensor 63 detects the temperature of the water flowing from the water tank 40 into the first pipe 6.

[0089] <Flow Sensor>

[0090] A flow sensor 62 is installed at the inflow end of the first pipe 6. The flow sensor 62 detects the amount of water that has flowed from the water tank 40 into the first pipe 6.

[0091] Control Department

[0092] Figure 2The control unit 30 shown includes a microcomputer and a storage device (specifically, a semiconductor memory) storing software for operating the microcomputer. The control unit 30 is connected via wired or wireless means to various devices and sensors of the hot water supply unit 10. The control unit 30 controls the heating unit 20 and the devices on the water circuit 50. The devices on the water circuit 50 include a water pump 53.

[0093] The control unit 30 includes a storage unit 31, a first learning unit 32, a prediction unit 33, and a second learning unit 35.

[0094] Storage unit 31 stores time-series data of a first indicator. The first indicator represents the heat of the water used by each hot water supply object 4. Specifically, the first indicator is the temperature of the water used by each hot water supply object 4 and the water pressure in each supply path 5. Hereinafter, the time-series data of the first indicator will be referred to as first time-series data. The first time-series data is the time-series data of this disclosure.

[0095] Here, the total hot water supply demand means the amount of heat from the water used by the entire hot water supply device 10 within a specified time. The total hot water supply demand is equivalent to the sum of the heat from the water used by each of the hot water supply objects 4.

[0096] Storage unit 31 stores the actual total hot water supply demand as time-series data. The actual total hot water supply demand is determined by measuring the heat of the water flowing from water tank 40 to first pipe 6 using a detection device. Specifically, the total hot water supply demand is determined based on the values ​​of flow sensor 62 and second temperature sensor 63. The time-series data of the total hot water supply demand stored in storage unit 31 is referred to as second time-series data.

[0097] The first learning unit 32 learns by associating first time-series data stored in the storage unit 31 within a specified period with second time-series data within the same time period as the first time-series data. The first learning unit 32 learns through machine learning.

[0098] The prediction unit 33 predicts the hot water supply demand based on the first time series data. Specifically, the prediction unit 33 predicts the total hot water supply demand based on the learning results of the first learning unit 32. More specifically, the prediction unit 33 uses a learned prediction model M1 to predict the total hot water supply demand. The learned prediction model M1 is obtained by learning through machine learning by associating the first time series data and the second time series data stored in the storage unit 31. The prediction unit 33 predicts the total hot water supply demand within a specified time period, such as the total hot water supply demand for the next day. Here, as... Figure 5As shown, the total hot water supply demand predicted by the prediction unit 33 can also be time series data that varies at regular intervals (e.g., 1 hour).

[0099] The prediction model M1 is included in the prediction unit 33. The prediction model M1 is generated to predict the total hot water supply demand based on first time-series data. The prediction model M1 is constructed as a multi-layer neural network that has acquired predictive capabilities through machine learning. In this embodiment, the prediction model M1 is generated through "supervised learning." The neural network used to generate the prediction model M1 learns using learning data and a recognition function. The learning data is a set of data pairs consisting of input data and training data corresponding to the input data.

[0100] The input data is a first time-series data within a specified period stored in storage unit 31. Specifically, the input data is time-series data of the water pressure in supply path 5 and the water temperature used by the hot water supply object 4 connected to supply path 5 within the specified period. The training data is a second time-series data within the same period as the input data. By having the neural network perform "supervised learning" using the aforementioned learning data, a learned prediction model M1 is generated as the learning result.

[0101] The prediction unit 33 thus uses the learned prediction model M1 to predict the total hot water supply demand. The prediction unit 33 outputs the total hot water supply demand by inputting first-time-series data stored in the storage unit 31 for a specified period (e.g., one week up to the previous day) into the learned prediction model M1. The prediction unit 33 thus predicts the total hot water supply demand.

[0102] The second learning unit 35 learns by associating the total hot water supply demand with the operating status of the heating device 20. The second learning unit 35 learns through machine learning. The control unit 30 controls the operation of the heating device 20 based on the learning results of the second learning unit 35. Specifically, the control unit 30 uses an operation prediction model M2 to control the operation of the heating device 20.

[0103] The control unit 30 includes an operation prediction model M2. The operation prediction model M2 is generated through machine learning to predict the operation control of the heating device 20 based on the total hot water supply demand. The control unit 30 uses this learned operation prediction model M2 to control the operation of the heating device 20.

[0104] The operation prediction model M2 is generated through "reinforcement learning." Specifically, the second learning unit 35 sets the reward to the daily electricity cost and the state variable to the operating state of the heating device 20. Here, the operating state of the heating device 20 refers to, for example, whether the heating device 20 is on or off. The second learning unit 35 inputs second time series data over a specified period as input data into the operation prediction model M2. Thus, the second learning unit 35 learns to minimize the electricity cost required for the heating device 20 to operate for a day. By inputting the total hot water supply demand predicted by the prediction unit 33 into the thus-generated, learned operation prediction model M2, it is possible to perform operation control of the heating device 20 to minimize power consumption.

[0105] <Heating Operation>

[0106] like Figure 3 As shown, the control unit 30 controls the heating device 20 to operate. Specifically, the control unit 30 operates the compressor 22 and the outdoor fan 27. The control unit 30 appropriately adjusts the opening of the expansion valve 24. The control unit 30 operates the water pump 53.

[0107] The refrigerant compressed by compressor 22 flows through a second flow path 25b utilizing heat exchanger 25. In heat exchanger 25, the refrigerant in the second flow path 25b releases heat to the water in the first flow path 25a. After releasing heat in the second flow path 25b, or condensing, the refrigerant, after being depressurized by expansion valve 24, flows through heat source heat exchanger 23. In heat source heat exchanger 23, the refrigerant absorbs heat from the outside air and evaporates. The refrigerant evaporated in heat source heat exchanger 23 is then drawn into compressor 22.

[0108] In water circuit 50, water in water tank 40 flows out to upstream flow path 51. Water in upstream flow path 51 flows through first flow path 25a utilizing heat exchanger 25. Water in first flow path 25a is heated by refrigerant in heating device 20.

[0109] The heated water in the water tank 40 flows through the first pipe 6 and the designated supply path 5. The water flowing through the supply path 5 flows out to the outside from the hot water supply object 4 connected to the supply path 5.

[0110] <Operating status of the hot water supply system>

[0111] Next, refer to Figure 4 An example of the operation of the hot water supply system 1 in this case will be explained.

[0112] In step ST1, the control unit 30 inputs the first time series data of the week up to the previous day into the learned prediction model M1.

[0113] In step ST2, the control unit 30 outputs the total hot water supply demand for the next day (future) from the learned prediction model M1 from step ST1. The total hot water supply demand output here is time series data of the next day, varying every hour.

[0114] In step ST3, the control unit 30 inputs the total hot water supply demand for the next day, which was output in step ST2, into the learned operation prediction model M2.

[0115] In step ST4, the control unit 30 outputs the operation control for the heating device 20 for the following day from the learned operation prediction model M2. The output operation control is, for example, as follows: Figure 5 As shown, the heating device 20 is operated according to a schedule that turns on or off every hour throughout the following day. The control unit 30 controls the heating operation of the heating device 20 according to this schedule. The heating device 20 heats the water in the water tank 40 according to the predicted total hot water supply demand, so as to supply the required amount of hot water to the hot water supply target 4 at each time period.

[0116] In step ST5, the control unit 30 controls the heating device 20 to operate according to the operation control output in step ST4. For example, in Figure 5 In the operational plan, the control unit 30 controls the heating device 20 to heat the water in the water tank 40, so as to supply the required amount of hot water from 13:00 to 14:00 to the hot water supply target 4. On the other hand, the control unit 30 controls the heating device 20 so that it does not operate during 14:00 to 15:00. This is because, in this operational plan, it is determined that not operating during 14:00 to 15:00 is more efficient based on the required amount of hot water supply during 14:00 to 15:00 and the remaining hot water storage in the water tank 40. Then, the control unit 30 controls the heating device 20 to heat the water in the water tank 40 according to the required amount of hot water supply for each time period after 15:00.

[0117] (Effects of the implementation method)

[0118] The hot water supply system 1 of this embodiment includes a prediction unit 33, which predicts the total hot water supply demand based on time series data (first time series data) of a first indicator, where the first indicator represents the heat of water used by each of the multiple hot water supply objects 4. In this way, since the total hot water supply demand is predicted based on the first indicator of each hot water supply object 4, the prediction accuracy of the total hot water supply demand can be improved compared to the case where the prediction is based solely on the hot water supply objects 4 with relatively large hot water supply volumes (e.g., showers and bathtubs).

[0119] Furthermore, even in residential buildings with multiple households, such as multi-unit housing complexes, the accuracy of predicting the total hot water supply demand can be improved. Specifically, in multi-unit housing complexes, the amount of hot water supplied to each household (4) varies. Therefore, if, for example, household A requires more hot water for its shower and bathtub than other households, the predicted hot water supply for household A will affect the overall predicted hot water supply for the entire complex, potentially lowering the accuracy of the predicted hot water supply for other households. However, according to the hot water supply system 1 of this embodiment, the total hot water supply demand for the entire complex is estimated based on time-series data of the first index of each household's hot water supply (4). This improves the accuracy of predicting the total hot water supply demand for the entire complex, and consequently, also improves the accuracy of predicting the total hot water supply demand for each household.

[0120] The hot water supply system 1 of the embodiment includes a first learning unit 32, which learns by associating first time-series data with second time-series data (total hot water supply demand). A prediction unit 33 predicts the total hot water supply demand based on the learning results of the first learning unit 32. In this way, the total hot water supply demand can be predicted based on the learning results of the first learning unit 32.

[0121] In the hot water supply system 1 of the embodiment, the first learning unit 32 learns through machine learning. In this way, the total hot water supply demand can be predicted based on the learning results obtained through machine learning.

[0122] In the hot water supply system 1 of this embodiment, the prediction unit 33 includes a prediction model M1, and uses the prediction model M1 to predict the total hot water supply demand. The prediction model M1 is generated through machine learning to predict the total hot water supply demand based on first time-series data. In this example, the learned prediction model M1 is generated through supervised learning based on first time-series data within a specified period. By using such a learned prediction model M1, the prediction accuracy of the total hot water supply demand can be reliably improved.

[0123] Furthermore, the learned prediction model M1 is updated through sequential learning. Therefore, the more times the hot water supply system 1 is used, the more times the learned prediction model M1 is updated, resulting in improved accuracy of the total hot water supply demand prediction output from the learned prediction model M1.

[0124] In the hot water supply system 1 of the implementation method, the first indicators are the temperature of the water used by each of the multiple hot water supply objects 4 and the water pressure in each supply path 5. By using the time series data of the water temperature used by the hot water supply object 4 and the time series data of the water pressure in the supply path 5 connected to the hot water supply object 4 as input data, and using the second time series data as training data, a learned prediction model M1 can be generated through supervised learning.

[0125] The hot water supply system 1 in this embodiment includes a control unit 30, which controls the operation of the heating device 20 based on the total hot water supply demand. In this way, the heating device 20 can operate according to, for example, the heat required by the hot water supply target 4 during different time periods of the following day.

[0126] The hot water supply system 1 of this embodiment includes a second learning unit 35, and a control unit 30 controls the operation of the heating device 20 based on the learning results of the second learning unit 35. The second learning unit 35 learns by associating the total hot water supply demand with the operating state of the heating device 20. In this way, the operation of the heating device 20 can be controlled based on the learning results of the second learning unit 35.

[0127] In the hot water supply system 1 of the embodiment, the second learning unit 35 learns through machine learning. In this way, the operation control of the heating device 20 can be predicted based on the learning results obtained through machine learning.

[0128] In the hot water supply system 1 of this embodiment, the control unit 30 includes an operation prediction model M2, which is used to control the operation of the heating device 20. The operation prediction model M2 is generated through machine learning to predict the operation control of the heating device 20 based on the total hot water supply demand. In this example, by using the operation prediction model M2 learned through machine learning, the heating device 20 can operate efficiently, and its operation can be controlled to minimize the daily electricity cost compared to the previous day. Because the operation of the heating device 20 can be controlled based on the total hot water supply demand, the possibility of the hot water in the water tank 40 running out during hot water supply can be prevented.

[0129] Furthermore, the learned operation prediction model M2 is updated through sequential learning. Therefore, the more frequently the hot water supply system 1 is used, the more frequently the learned operation prediction model M2 is updated, resulting in the ability to predict energy-saving operation controls. Thus, in this example, the more frequently the hot water supply system 1 is used, the lower the daily electricity cost becomes.

[0130] In the hot water supply system 1 of the embodiment, the heating device 20 is a heat pump. In this way, even in a heat pump with relatively low start-up capacity, the risk of running out of hot water can be reduced, and high-efficiency operation can be achieved.

[0131] (Variation Example 1)

[0132] In the hot water supply system 1 of this example, the prediction unit 33 includes a pre-learned prediction model M1. A structure different from the above embodiment will now be described.

[0133] like Figure 6 As shown, the control unit 30 does not have a first learning unit 32. In this example, the prediction model M1 is pre-generated to predict the total hot water supply demand based on the first time series data before the user uses the hot water supply system 1 (before the hot water supply system 1 leaves the factory).

[0134] The inference model M1 in this example is also generated through "supervised learning." Prescribed learning data stored on a data server is input into the inference model M1. Specifically, the input data includes nationwide user information (number of people in each household, age, gender, and residential area), information on the hot water supply objects 4 owned by each user (number, type, and faucet information of the hot water supply objects), and time-series data on water pressure and water temperature used by the hot water supply objects in the supply path 5 connected to the hot water supply objects 4 of each user's hot water supply system 1. The training data is a second time-series data stored on the data server. The data server stores time-series data from the past few years.

[0135] The neural network is then subjected to "supervised learning" using this learning data. As a result of the learning, a fully learned prediction model M1 is generated for this example. Prediction unit 33 uses the fully learned prediction model M1 to predict the total demand for hot water supply.

[0136] User data is stored in storage unit 31. The user data includes information such as family structure (number of family members, age, gender, etc.) and residential location.

[0137] Thus, in the hot water supply system 1 of this example, the total hot water supply demand is also predicted based on the first-time series data of each hot water supply recipient, thereby improving the prediction accuracy of the total hot water supply demand. In particular, the learned inference model M1 in this example is generated using the first-time series data of users nationwide stored in a data server. Therefore, compared to inputting only the first-time series data, by inputting information such as the user's residential area and family structure along with the first-time series data into the learned inference model M1, a more accurate prediction of the total hot water supply demand corresponding to the user can be obtained.

[0138] Furthermore, since the hot water supply system 1 is equipped with a fully learned prediction model M1 at the time of manufacture, users can use the hot water supply system 1 with high prediction accuracy of total hot water supply demand from the start of use.

[0139] (Variation Example 2)

[0140] In this example, the hot water supply system 1 uses a defined logic formula to predict the total hot water supply demand.

[0141] Here, the hot water supply demand means the amount of heat of water used by each hot water supply object 4 within a specified period. The hot water supply demand is equivalent to the sum of the heat of water used by each hot water supply object 4.

[0142] In this example, the hot water supply system 1 predicts the hot water supply demand of each hot water supply target 4 based on time-series data of the heat of the water used by each hot water supply target 4 within a specified period. Then, the hot water supply system 1 predicts the total hot water supply demand based on the predicted hot water supply demand of each hot water supply target 4. The following describes a structure different from the above-described embodiment and variation 1.

[0143] like Figure 7 As shown, in the hot water supply system 1 of this example, the control unit 30 has a calculation unit 34. The control unit 30 in this example does not have a prediction model M1. The first learning unit 32 does not perform machine learning.

[0144] The calculation unit 34 calculates the time series data of the hot water supply demand of each hot water supply object 4 within a specified period based on the first time series data stored in the storage unit 31.

[0145] The prediction unit 33 predicts the hot water supply demand of each hot water supply target 4 within a specified period, based on the hot water supply demand calculated by the calculation unit 34. The prediction of the hot water supply demand of each hot water supply target 4 is performed using a predetermined logical expression and a first coefficient. The total hot water supply demand predicted by the prediction unit 33 can also be time-series data changing at regular intervals (e.g., 1 hour) for the next day. The prediction unit 33 predicts the total hot water supply demand based on the predicted hot water supply demand of each hot water supply target 4. The prediction of the total hot water supply demand is performed using a predetermined logical expression and a second coefficient. The first and second coefficients are input into the prediction unit 33 by the first learning unit 32.

[0146] The first learning unit 32 adjusts a first coefficient for each hot water supply target 4 to reduce the residual between the hot water supply demand predicted by the prediction unit 33 for a specified period and the actual hot water supply demand used during that specified period. The first learning unit 32 inputs the adjusted first coefficient into the prediction unit 33. The first learning unit 32 adjusts a second coefficient to reduce the residual between the total hot water supply demand predicted by the prediction unit 33 for a specified period and the actual total hot water supply demand used during that specified period. The first learning unit 32 inputs the second coefficient into the prediction unit 33.

[0147] <Operating status of the hot water supply system>

[0148] Next, refer to Figure 8 An example of the operation of the hot water supply system 1 in this case will be described. Steps ST24 to ST26 are the same as steps ST3 to ST5 in the above embodiment, so the description is omitted.

[0149] In step ST21, the control unit 30 calculates the time series data of the hot water supply demand of each hot water supply object 4 during the specified period (e.g., a week up to the previous day) based on the first time series data stored in the storage unit 31 during the specified period.

[0150] In step ST22, the control unit 30 uses a prescribed logic formula and a first coefficient to predict the hot water supply demand of each hot water supply target 4 for the next day based on the time series data of hot water supply demand obtained in step ST21.

[0151] In step ST23, the control unit 30 predicts the total hot water supply demand for the next day based on the hot water supply demand of each hot water supply target 4 predicted in step ST22 for the next day.

[0152] In step ST27, the control unit 30 adjusts a first coefficient to reduce the prediction error based on the residual between the predicted hot water supply demand of each hot water supply target 4 in step ST22 and the actual hot water supply demand of each hot water supply target 4. The control unit 30 then inputs the adjusted first coefficient into the prediction unit 33.

[0153] In step ST28, the control unit 30 adjusts a second coefficient based on the residual between the total hot water supply demand predicted in step ST23 and the actual total hot water supply demand used, in order to reduce the prediction error. The control unit 30 then inputs the adjusted second coefficient into the prediction unit 33.

[0154] In this example, in the prediction of hot water supply demand for each hot water supply target 4, by repeatedly adjusting the input of the first coefficient, the residual between the predicted hot water supply demand of each hot water supply target 4 and the actual hot water supply demand of each hot water supply target 4 can be reduced. Furthermore, by repeatedly adjusting the input of the second coefficient, the residual between the predicted hot water supply demand and the actual hot water supply demand can be reduced. As a result, the prediction accuracy of the total hot water supply demand can be improved.

[0155] (Variation Example 3)

[0156] In the hot water supply system 1 of this example, the control unit 30 includes a pre-learned operation prediction model M2. The following describes a structure different from the above-described embodiment and its variations.

[0157] like Figure 9 As shown, the control unit 30 in this example does not have a second learning unit 35. The operation prediction model M2 in this example is pre-generated to predict operation control based on the total hot water supply demand before the user uses the hot water supply system 1 (before the hot water supply system 1 leaves the factory).

[0158] In this example, the operation prediction model M2 is generated through reinforcement learning. The input data stored in the data server in the above-described implementation (nationwide user information (number of people in a household, age, gender, and residential area), information on the hot water supply devices 4 owned by each user (number, type, and faucet information of the hot water supply devices), and the total hot water supply demand of each user, etc.) is input into the operation prediction model M2. By setting the reward as a day's electricity cost and the state variable as the operating state of the heating device 20, a learned operation prediction model M2 is generated. This learned model M2 learns to minimize the electricity cost required for the heating device 20 to operate for a day. Thus, by inputting the total hot water supply demand into the learned operation prediction model M2, operation control of the heating device 20 is output to minimize power consumption.

[0159] Thus, in this example, using the operation prediction model M2 learned through machine learning, it is possible to control the heating device 20 to minimize the daily electricity cost compared to the previous day. Specifically, the learned operation prediction model M2 in this example is generated using the total hot water supply demand of users nationwide, stored on a data server. Therefore, compared to simply inputting the total hot water supply demand, by inputting information such as the user's residential area and family structure, along with the total hot water supply demand, into the learned operation prediction model M2, it is possible to predict the operation of the heating device 20 to reduce electricity costs based on user forecasts.

[0160] Furthermore, since the hot water supply system 1 has a pre-learned operation prediction model M2 at the factory, users can use the hot water supply system 1 from the start of use, which can reduce daily electricity bills and control the operation of the heating device 20.

[0161] (Variation Example 4)

[0162] In this example, the hot water supply system 1 is based on predetermined control, predicting the operation control of the heating device 20 according to the total hot water supply demand. Below, a structure different from the above-described embodiment and its variations will be described.

[0163] like Figure 10 As shown, in the hot water supply system 1 of this example, the control unit 30 has a hot water depletion judgment unit 36. The control unit 30 does not have an operation prediction model M2.

[0164] If the flow sensor 62 detects zero when the heating device 20 is in the on state, the hot water depletion judgment unit 36 ​​will make a judgment that there is no water in the water tank 40 (hot water depletion).

[0165] The second learning unit 35 learns the operation control of the heating device 20 to reduce the risk of running out of hot water based on the total hot water supply demand. Specifically, if the result of controlling the operation of the heating device 20 based on the total hot water supply demand predicted by the prediction unit 33 is that there is a time when hot water runs out, the second learning unit 35 corrects the operation control of the heating device 20 performed that day. Through this feedback, the second learning unit 35 corrects the operation control of the heating device 20 and predicts the operation control of the heating device 20 for the next day.

[0166] <Operating status of the hot water supply system>

[0167] Next, refer to Figure 11 An example of the operation of the hot water supply system 1 in this case will be described. Steps ST41 to ST42 are the same as steps ST1 to ST2 in the above embodiment, so the description is omitted.

[0168] In step ST43, the control unit 30 predicts the operation control of the heating device 20 based on the total hot water supply demand predicted by the prediction unit 33.

[0169] In step ST44, the control unit 30 controls the heating device 20 in a manner that executes the operation control predicted in step ST43.

[0170] In step ST45, the control unit 30 determines whether hot water has run out when supplying hot water to the hot water supply object 4. If hot water has run out (yes in step ST45), the control unit 30 controls the heating device 20 to heat the water supplied to the water tank 40 (step ST44). If hot water has not run out (no in step ST45), step ST46 is executed.

[0171] In step ST46, the control unit 30 determines whether the operation of the heating device 20 for one day has ended. If the operation has ended (yes in step ST46), step ST47 is executed. If the operation has not ended (no in step ST46), step ST44 is executed again.

[0172] In step ST47, the control unit 30 determines whether the heating device 20 has run out of hot water during a day's operation. If the hot water has run out (yes in step ST47), step ST48 is executed.

[0173] In step ST48, the control unit 30 adjusts the operation control based on the time when the hot water is used up, the amount of hot water heated, the heating temperature, etc.

[0174] In this example, whenever hot water runs out, the control unit 30 adjusts the operation control of the heating device 20. By predicting the operation control based on such adjustments, the risk of running out of hot water the following day can be reduced.

[0175] (Other implementation methods)

[0176] The above-described embodiments and variations can also adopt the following structure.

[0177] The first indicator could also be the temperature of the water used by the hot water supply object 4 and the amount of water in the supply path 5 connected to the hot water supply object 4. In this case, a flow sensor (not shown) for measuring the flow rate of water is installed on each supply path 5. The flow sensor is connected to the control unit 30 via wired or wireless means.

[0178] The first indicator can also be the heat of the water used by each hot water supply object 4. The time series data of the heat of the water used by each hot water supply object 4 can also be obtained from the time series data of the water temperature used by the hot water supply object 4 and the time series data of the water pressure in the supply path 5 connected to the hot water supply object 4. The time series data of the heat of the water used by each hot water supply object 4 can also be obtained from the time series data of the water temperature used by the hot water supply object 4 and the time series data of the water volume in the supply path 5 connected to the hot water supply object 4. The time series data of the heat of the water used by each hot water supply object 4 is input into the prediction model M1.

[0179] In the reinforcement learning of the operation prediction model M2, the operating state of the heating device 20, which is set as a state variable, can also be the speed of the compressor 22, the condensing temperature of the heat exchanger 25, and the opening degree of the expansion valve 24. For example, the control unit 30 can heat the water in the water tank 40 by increasing the speed of the compressor 22, or suppress the rise in the condensing temperature of the heat exchanger 25 by decreasing the opening degree of the expansion valve 24, thereby suppressing the heating of the water in the water tank 40.

[0180] The hot water supply system 1 may also have a specified sensor (not shown) that directly detects the heat of the water used by each hot water supply object 4. In this case, the time series data of the heat of the water used by each hot water supply object 4 detected by the sensor is input into the prediction model M1 without relying on the calculation unit 34.

[0181] The input data to the prediction model M1 may include at least one of the following: the number of hot water supply objects 4, the type of hot water supply object 4, and the faucet specifications of the hot water supply object 4. Thus, by using the information of each hot water supply object 4 as input data based on the first time series data, the total hot water supply demand can be predicted based on the determined types of hot water supply objects 4 and their first time series data. For example, if the hot water supply objects 4 are bathtubs and showers, the bathtubs (filling with bath water) and showers are frequently used simultaneously or in relatively close time periods (before bathing and during bathing). Therefore, the first learning unit 32 can learn the total hot water supply demand by considering the hot water supply demand of the shower based on the first time series data of the bathtubs, and similarly, it can learn the total hot water supply demand by considering the hot water supply demand of the bathtub based on the first time series data of the shower. By using the learned prediction model M1 in this way, the risk of running out of hot water can be reduced.

[0182] The total hot water supply demand that the prediction unit 33 wants to forecast can also be the heat of the water used by the hot water supply device 10 only within a specified time period (e.g., 1 hour from a certain time the next day).

[0183] In the above embodiment, the second time series data can also be obtained based on the change in the hot water storage temperature within the water tank 40. In this case, a hot water storage temperature sensor (not shown) is installed inside the water tank 40. The hot water storage temperature sensor is connected to the control unit 30 via wired or wireless means. For example, if the temperature value of the hot water storage temperature sensor decreases, it indicates that hot water has been supplied from the water tank 40 and water has been supplied to the water tank 40. Based on how much the temperature of the hot water storage temperature sensor has decreased, the amount of heat in the water flowing out of the water tank 40 can be determined.

[0184] The prediction unit 33 can also predict the total hot water supply demand based on the hot water supply consumption of each of the selected hot water supply targets 4. In this way, by eliminating hot water supply targets with noisy properties that would reduce prediction accuracy, the prediction accuracy of the total hot water supply demand can be improved.

[0185] The prediction model M1 can also be generated through "unsupervised learning." In this case, the neural network repeatedly performs a learning action that uses a clustering algorithm to group multiple input data into multiple categories, ensuring that similar input data (total hot water supply demand) are grouped into the same category. In this way, the learned prediction model M1 can be generated without using training data. The prediction model M1 can also be generated through "reinforcement learning."

[0186] The storage unit 31 may not be located in the control unit 30. Alternatively, the storage unit 31 may be stored in a designated server that can communicate with the control unit 30.

[0187] In the second embodiment and variations described above, the operational prediction model M2 can also be generated through "supervised learning" or "unsupervised learning".

[0188] The first learning unit 32 can also learn using learning methods other than those described in the above embodiments and variations. The second learning unit 35 can also learn using learning methods other than those described in the above embodiments and variations.

[0189] The embodiments and variations have been described above. However, it should be understood that various changes can be made to the configuration and specific details without departing from the spirit and scope of the claims. As long as the function of the object of this disclosure is not affected, the above embodiments and variations can be appropriately combined and substituted. The terms "first" and "second" used above are only used to distinguish statements containing the above terms and are not intended to limit the number or order of the statements.

[0190] -Industry Applicability-

[0191] In conclusion, this disclosure is very useful for hot water supply systems.

[0192] - Symbol Explanation -

[0193] M1 Speculation Model

[0194] M2 Operation Prediction Model

[0195] 4. Hot water supply recipients

[0196] 5. Supply Path

[0197] 10 Hot water supply device

[0198] 20 Heating device

[0199] 30 Control Department

[0200] 32 First Study Department

[0201] 33. Speculation Department

[0202] 35 Second Study Department

[0203] 40 water tank

[0204] 50 Water Circuit

Claims

1. A hot water supply system, characterized in that: The hot water supply system includes a hot water supply device (10), multiple supply paths (5), and a prediction unit (33). The hot water supply device (10) includes a heating device (20), a water tank (40), and a water circuit (50). The heating device (20) heats the water, and the water tank (40) stores the water heated by the heating device (20). The water in the water tank (40) circulates in the water circuit (50). Multiple supply paths (5) are connected to multiple hot water supply objects (4) respectively, and multiple supply paths (5) supply water from the water tank (40). The prediction unit (33) predicts the total hot water supply demand based on time series data of a first indicator, whereby the first indicator represents the heat of water used by each of the plurality of hot water supply objects (4), and the time series data exists for each of the plurality of hot water supply objects (4). The prediction unit (33) predicts the total hot water supply demand based on the time series data and at least one of the number of hot water supply objects (4), the type of hot water supply objects (4), and the faucet specifications of the hot water supply objects (4).

2. The hot water supply system according to claim 1, characterized in that: The hot water supply system has a first learning unit (32), which learns by associating the time series data with the total hot water supply demand. The prediction unit (33) predicts the total hot water supply demand based on the learning results of the first learning unit (32).

3. The hot water supply system according to claim 2, characterized in that: The first learning unit (32) learns through machine learning.

4. The hot water supply system according to claim 1 or 3, characterized in that: The prediction unit (33) includes a prediction model (M1), which is generated by machine learning to predict the total hot water supply demand based on the time series data. The prediction unit (33) uses the prediction model (M1) to predict the total hot water supply demand.

5. The hot water supply system according to any one of claims 1 to 3, characterized in that: The first indicator is the temperature and amount of water used by each of the plurality of hot water supply objects (4).

6. The hot water supply system according to any one of claims 1 to 3, characterized in that: The first indicator is the temperature of the water used in each of the plurality of hot water supply objects (4) and the water pressure in each of the supply paths (5).

7. The hot water supply system according to any one of claims 1 to 3, characterized in that: The prediction unit (33) predicts the total hot water supply demand based on the time series data of the selected hot water supply targets (4).

8. The hot water supply system according to any one of claims 1 to 3, characterized in that: The hot water supply system has a control unit (30) that controls the operation of the heating device (20) according to the total hot water supply demand.

9. The hot water supply system according to claim 8, characterized in that: The hot water supply system has a second learning unit (35) that learns by associating the total hot water supply demand with the operating status of the heating device (20). The control unit (30) controls the operation of the heating device (20) based on the learning results of the second learning unit (35).

10. The hot water supply system according to claim 9, characterized in that: The second learning unit (35) learns through machine learning.

11. The hot water supply system according to claim 8, characterized in that: The control unit (30) includes an operation prediction model (M2), which is generated by machine learning to predict the operation control of the heating device (20) based on the total hot water supply demand. The control unit (30) uses the operation prediction model (M2) to control the operation of the heating device (20).

12. The hot water supply system according to any one of claims 1 to 3, characterized in that: The heating device (20) is a heat pump type.

Images