Acquiring user consumption patterns of domestic hot water and controlling domestic hot water production based on the patterns

Abstract

A computer-implemented method for obtaining a user consumption pattern (UCP) of domestic hot water, comprising the steps of obtaining data representative of an amount of heat (ΣQT1) extracted from a thermal storage tank (20) within a first time period (T1), generating a first history (H1) or a set of data representative of the amount of heat (ΣQ), in particular the cumulative amount of heat, extracted from the thermal storage tank (20) over a number of said first time periods (T1), and obtaining a user consumption pattern of domestic hot water by applying a user consumption pattern determination algorithm to the generated first history (H1) or set of data representative of the amount of heat extracted from the thermal storage tank (20), the user consumption pattern determination algorithm being an algorithm trained on the history (or multiple histories) or set of data (or multiple sets of data) representative of the amount of heat extracted (ΣQ) and defining the user consumption pattern of domestic hot water with one or more machine learning algorithms. Furthermore, the present disclosure relates to a controller (1) and a system (100) for controlling the production and / or distribution of domestic hot water. Furthermore, the present disclosure relates to a computer program and a computer-readable medium having the computer program stored thereon.

Description

[Technical field]

[0001] The present disclosure relates to a computer-implemented method for obtaining a domestic hot water user consumption pattern, a computer-implemented method for generating a domestic hot water consumption forecast, and a computer-implemented method for controlling domestic hot water production and / or distribution. The present disclosure further relates to an associated controller for generating a domestic hot water consumption forecast, a controller for controlling domestic hot water production and / or distribution, and a domestic hot water production and / or distribution system. The present disclosure further relates to a corresponding computer program, and a computer readable medium having said computer program stored thereon. [Background technology]

[0002] In recent years, buildings such as homes and office spaces have been equipped with smart home networks that automatically control devices, appliances, and systems, such as heating, ventilation, and air conditioning (HVAC) systems, lighting systems, alarm systems, home theater and entertainment systems, and other systems. The smart home network may include a control panel through which a user enters settings, preferences, and scheduling information that is used to automatically control various devices, appliances, and systems within the building. For example, this may allow a user to enter a desired temperature and a schedule indicating the times when the user will be away from home. The home automation system uses this information to control the HVAC system to heat or cool the room to the desired temperature when the user is present and to conserve energy by turning off power-consuming components of the HVAC system when the user is away from home, sleeping, etc.

[0003] A similar idea exists in the field of domestic hot water production or distribution. Heating demand in the domestic environment consists of two main sources: sanitary hot water production and space heating. Sanitary hot water is usually produced either on demand (requiring relatively high-powered devices such as gas boilers) or using a buffered approach using hot water tanks that can be heated by various sources such as electricity, solar, gas, heat pumps, etc. Smaller dwellings with improved insulation (e.g. passive housing) have reduced the residential space heating requirement in absolute terms. Meanwhile, domestic hot water demand has remained and even increased slightly. Sanitary hot water production therefore accounts for a larger relative share of the total domestic heating demand.

[0004] For systems with a hot water tank, the minimum tank temperature is usually set constant throughout the day to provide enough hot water for users. This desired minimum temperature is chosen to be on the safe side. To conserve energy, newer systems use a fixed pattern on a weekly schedule, for example lowering the minimum tank temperature at night and raising the tank temperature in the morning when demand is expected to increase.

[0005] To further improve the energy efficiency of domestic hot water systems using a hot water storage tank, it is necessary to provide a domestic hot water production and / or distribution system with a means for more accurately predicting the actual domestic hot water usage. This allows the system to minimize the heat and hot water stored in the thermal storage tank (hot water storage) while maintaining the comfort of the user. For a system to have such a means for more accurately predicting the domestic hot water consumption, it is also necessary for the system to have a means for more accurately estimating the amount of available and usable hot water in the tank. Typically this is done by detecting and measuring the domestic hot water usage. Known systems for this purpose use a monitoring system including a flow meter and a temperature sensor. Based on the output of these units, the system estimates the energy assumed to be taken from the hot water storage tank and estimates the amount of usable hot water remaining in the tank.

[0006] For example, US 2015 / 0226460 A1 describes a retrofit water heater monitoring and prediction system, method, and computer program product for a water heater boiler system including a water heater, a cold water pipe, and a hot water pipe, the system including: an intake temperature sensor configured to measure a water temperature in a cold water intake pipe; a flow meter configured to measure a flow rate of water flowing through the water heater boiler system; a discharge temperature sensor configured to measure a water temperature in a hot water discharge pipe; a processing unit that receives sensor data from the intake temperature sensor, the flow meter, and the discharge temperature sensor, the processing unit configured to calculate an amount of available hot water in the water heater based on the sensor data; and a display panel coupled to the processing unit and configured to display at least one estimated real-time usage value calculated by the processing unit based on the amount of available hot water. Summary of the Invention

[0007] In view of the above, it is desirable to provide a computer-implemented method for obtaining a user's consumption pattern of domestic hot water, a computer-implemented method for generating a domestic hot water consumption forecast, a computer-implemented method for controlling domestic hot water production and / or distribution, a controller for generating a domestic hot water consumption forecast, a controller for controlling domestic hot water production and / or distribution, a domestic hot water production and / or distribution system, a computer program and a computer readable medium having said computer program stored thereon, which allow for a more accurate generation or determination of a forecast or prediction of domestic hot water consumption, thereby improving energy efficiency and reducing the environmental footprint of domestic hot water production and / or distribution. A further object is to enable a smart control of domestic hot water production and / or distribution by adapting the hot water heating process to individual usage conditions with the aim of reducing energy consumption while maintaining user comfort.

[0008] This object may be achieved by a computer implemented method for obtaining a user consumption pattern according to claim 1, a computer implemented method for generating a domestic hot water consumption forecast according to claim 12, a computer implemented method for controlling domestic hot water production and / or distribution according to claim 17, a controller for generating a domestic hot water consumption forecast according to claim 21, a controller for controlling domestic hot water production and / or distribution according to claim 22, a domestic hot water production and / or distribution system according to claim 23, a computer program according to claim 25 and a computer readable medium according to claim 26. Some embodiments are set out in the dependent claims, the following description and the accompanying drawings.

[0009] The present disclosure provides a user consumption pattern determination algorithm trained based on a history or set of data representing the amount of heat extracted from the thermal storage tank and / or a history or set of data representing the amount of heat extracted from the thermal storage tank or each normalized thermal storage tank.

[0010] In this way, by using trained algorithms that can take advantage of data and experience collected from many users over a long period of time, it is possible to more accurately determine user consumption patterns, particularly those of individual users or individual households.

[0011] According to one embodiment of the present disclosure, a computer-implemented method for obtaining a user consumption pattern of domestic hot water is provided, the computer-implemented method comprising: obtaining data representative of the amount of heat, in particular the amount of equivalent energy, extracted from the thermal storage tank, in particular the pressurized tank, preferably within a first predetermined time period; generating a first history or set of data representative of the amount of heat, particularly the cumulative amount of heat, removed from the thermal storage tank, preferably over a predetermined number of first time periods; and acquiring a user consumption pattern of domestic hot water by applying a user consumption pattern determination algorithm to the first history or data set that generates data representing the amount of heat extracted from the heat storage tank; The user consumption pattern determination algorithm is an algorithm, in particular a time series forecasting algorithm, that is trained on a history (or multiple histories) or set (or multiple sets) of data representing the amount of heat extracted from the thermal storage tank or multiple normalized thermal storage tanks, and that uses one or more machine learning algorithms to define a user's domestic hot water consumption pattern.

[0012] In the context of this disclosure, the term "obtain" with respect to obtaining or collecting data representative of the amount of heat removed from the thermal storage tank should be understood to mean that the respective data may be determined by using sensors such as temperature sensors and flow meters, or by accessing or reading data representative of the amount of heat removed from the thermal storage tank that is recorded in a memory. The memory may store periodic temperature and flow measurements of the heat or domestic hot water stored in the thermal storage tank for a period of time.

[0013] Further, in the context of this disclosure, the term "representing" with respect to acquisition of data should be understood to enable a system and / or computer-implemented method to determine or reconstruct from the acquired data a specific amount of heat stored in the thermal storage tank at a particular time and / or for a particular period of time.

[0014] Furthermore, in the context of this disclosure, the term "heat" with respect to removal from the thermal storage tank or with respect to the amount of heat stored in the thermal storage tank (discussed below) is used to define, for example, the energy contained in the water removed from the thermal storage tank; that is, the heat or energy removed from the thermal storage tank upon removal (discharge). Thus, with respect to heat stored in the thermal storage tank, the term "heat" refers, for example, to the remaining equivalent hot water (EHW) at t0.

[0015] The term "Equivalent Hot Water (EHW)" refers to the maximum hot water volume "V 40 According to EN16147, the maximum amount of mixed water at 40°C in one discharge (from a specific thermal storage tank) shall be determined by calculating the hot water energy at the time of discharge. Hot water flow rate f maxand the temperature of the inflowing cold water θ WC and the temperature of the outflowing hot water θ WH The maximum hot water volume V40 is calculated using the following formula:

number

[0016] Furthermore, in the context of this disclosure, the term "equivalent energy" (also called "available energy") for withdrawal from the thermal storage tank should be understood as the amount of water (liters) at 40°C that has the thermal energy of water relative to 10 degrees. That is, for a volume V (liters) at a temperature T (°C),

number

[0017] Furthermore, the term "user consumption pattern (UCP)" in the present disclosure defines parameters such as the temperature and / or amount of hot water extracted from the thermal storage tank, and indicates the amount of heat extracted from the thermal storage tank, particularly the accumulated amount of heat within a regular time interval or a specific period of time. For example, if a user takes a shower every morning at around 7:00 AM, the consumption pattern of this user will indicate an increased demand for domestic hot water during the time period starting at 6:00 AM and ending at 8:00 AM.

[0018] However, in the context of the present disclosure, the term "predetermined" with respect to a "period" or "number of first periods" should be understood as a particular point in time, such as an initial point in time when the computer-implemented method or the user consumption pattern determination algorithm is initiated (initialization), or an update point in time when the user consumption pattern determination algorithm is updated (algorithm update), and the "period" or "number of first periods" is determined manually or automatically, and in the automatic case, its value is determined by the associated algorithm.

[0019] Thus, at the time when the computer-implemented method for obtaining a user consumption pattern or the algorithm for determining a user consumption pattern is started, the first period may be set to 1 hour and the number of first periods may be set to 12, thereby defining a prediction horizon of 12 hours. In other words, the first history or set of data representing heat quantities spans 12 hours. In a next step, for example after 12 hours or at a time of day (e.g. midnight or noon), based on the first history or set of data generated, the algorithm for determining a user consumption pattern may be updated, whereby for optimization reasons the first period may be changed or set to 4 hours and the number of first periods may be set to 6. This defines a new prediction horizon of 24 hours, which may be combined with the first prediction horizon to form a cumulative prediction horizon of 36 hours.

[0020] Furthermore, in the context of the present disclosure, the term "normalized" with respect to the thermal storage tanks should be understood as making the characteristics of each thermal storage tank correspond to the characteristics of one thermal storage tank using or applying the computer-implemented method of obtaining the domestic hot water user consumption pattern. In other words, the characteristics of the thermal storage tank on which the algorithm is trained should not deviate so much from the characteristics of the actual thermal storage tank that it is expected to adversely affect the obtaining or determining of the domestic hot water user consumption pattern. Therefore, the thermal storage tank used should be similar to the one on which the algorithm is trained. Alternatively, the algorithm may be trained to correct such deviations (e.g., automatically adjust the size of the thermal storage tank used to the size of the tank on which the algorithm is trained, whereas the algorithm may be trained on thermal storage tanks of different sizes and types and the algorithm can be adjusted accordingly).

[0021] According to a further embodiment of the present disclosure, the computer-implemented method further comprises: There may also be a step of generating a second history or set of data representative of the amount of heat removed from the thermal storage tank, preferably over a predetermined number of the first history or set of data, the first history or set of data preferably covering a one day or 24 hour period and the second history or set of data preferably covering a one week or 7 day or 168 hour period.

[0022] As described above, the prediction period is defined by relating the first period and the number of first periods, but the same is true for the second period (the number of first histories). Therefore, the first and second histories can be considered as one history or data collection spanning (the number of first periods x the first periods) x the number of first histories.

[0023] Further, in some embodiments of the present disclosure, the preferred predetermined first time period may span or extend over 1 week, 2 days, 1 day, 12 hours, 8 hours, 6 hours, 4 hours, 1 hour, 30 minutes, 10 minutes or 1 minute, and / or the preferred predetermined number of first time periods in the first history or data collection is 1, 2, 3, 4, 6, 24, 48, 144 or 1440.

[0024] Further, in some embodiments, the preferred predetermined first time period may be determined based on the length / duration of the amount of data in the first history or data set, particularly at the initial and / or update times, i.e., as discussed above, the prediction time period may be changed as more data becomes available representing the amount of heat removed from the thermal storage tank over the prediction time period.

[0025] According to a further embodiment of the present disclosure, at an initial point (the start of the computer-implemented method or algorithm for determining user consumption patterns), if no first history or data set exists or only a short first history or data set of view data is available, a data filler is used to input into the algorithm for determining user consumption patterns.

[0026] Further, in some embodiments, preferably after a predetermined fifth period (followed by the second to third periods), the user consumption pattern determination algorithm may be newly launched / initiated or updated or a new initial point in time is set.

[0027] According to a further embodiment of the present disclosure (general ML model), the consumption pattern determination algorithm is and historical data or set (or sets) of data representing the amount of heat extracted from the thermal storage tank or normalized each thermal storage tank by a number of users or households, preferably over a second predetermined time period (training period); In particular, training may be performed on the time of day for each first period within each history or collection of data representing the amount of heat removed from the reservoir or each normalized reservoir.

[0028] Further, in an alternative embodiment (individual model), the consumption pattern determination algorithm is - historical data or set (or sets) of data representative of the amount of heat extracted from the thermal storage tank by a user, particularly an individual user, or a household, particularly an individual household, preferably over a second predetermined time period; In particular, training may be performed on the time of day within each history or collection of data representing the amount of heat removed from the thermal storage tank, and in particular for each first period of time.

[0029] According to further embodiments of the present disclosure, the preferred second predetermined period of time extends or spans 30 days, 60 days, 90 days, 180 days, one year or two years.

[0030] Further, in some embodiments, the consumption pattern determination algorithm further or alternatively comprises: In particular, the day of the week for each first period and / or for each history or collection of data representing the amount of heat extracted; and / or and / or the date of the year and / or the week of the year and / or the month of the year, in particular for each first period and / or for each historical data or data set; weather conditions, in particular for each first period and / or for each history or collection of data representing the quantity of heat extracted; and / or and / or, each history or collection of data representing the amount of heat extracted, and / or the outside air temperature, in particular for each first period of time; the vacation status of the user or household for each first period and / or for each history or collection of data representing the amount of heat extracted; and / or energy prices for each first period and / or for each history or collection of data representing the amount of heat extracted; and / or and / or, in particular, for each first period and / or for each history or collection of data representing the amount of heat extracted, the green energy availability; the geographic location of each history or collection of data, in particular for each first period and / or representing the quantity of heat extracted; and / or In particular, training may be performed on cultural factors for each first period and / or for each history or collection of data representing the amount of heat extracted.

[0031] In the context of the present invention, the term "green energy availability" in relation to the teaching of the algorithm should be understood as meaning that a disciplined or thrifty user tends to take a bath, for example, on a sunny day when there is preferably a large amount of solar power available locally, especially in winter. The same applies to times when electricity / energy prices are low.

[0032] On the other hand, geographic location and / or cultural factors may also affect users' domestic hot water consumption habits. For example, users living in areas closer to the equator may shower more frequently than those in cooler regions. Furthermore, cultural factors such as religion and wealth may also lead users to bathe more frequently. The same may be true for the amount of hot water used per shower or bath.

[0033] According to a further embodiment of the present disclosure, the user consumption pattern determination algorithm may further or additionally take into account metadata in determining the user consumption pattern (UCP), preferably selected from the group consisting of: number of occupants (residing in each household), age of occupants, average age of occupants, gender of occupants, geographic location, cultural factors which may be automatically determined based on geographic location, annual hot water consumption.

[0034] Additionally, in some embodiments, the user consumption pattern determination algorithm further comprises: assigning users or households to predefined clusters or groups based on at least one metadata; determining a user consumption pattern determination sub-algorithm based on the assigned clusters or groups; The user consumption pattern determination sub-algorithm is preferably trained on data of multiple users or households having at least one identical metadata.

[0035] According to a further embodiment of the present disclosure, after the user consumption pattern determination sub-algorithm has been determined, an individual user consumption pattern determination algorithm may be generated by training a pre-defined user consumption pattern determination sub-algorithm based on the acquired first historical data or data set (or multiple data sets) and / or the second historical data or data set (or multiple data sets) representative of the amount of heat extracted from the thermal storage tank preferably by an individual user or preferably an individual household over a preferred pre-defined third period of time, the preferred pre-defined third period of time preferably spanning 1 day, 2 days, 10 days, 30 days, 60 days, 90 days, 180 days, a year or a continuous period.

[0036] According to further embodiments of the present disclosure, the user consumption pattern determination algorithm may be trained to determine individual user habits including showering in the morning, showering in the evening, bathing in the evening, and using an average heat amount, particularly an equivalent energy amount, for showering or bathing.

[0037] The present disclosure further relates to a computer-implemented method for generating a domestic hot water consumption forecast, comprising: obtaining a domestic hot water user consumption pattern (UCP) using the computer implemented method steps described above for obtaining a domestic hot water user consumption pattern; and applying a domestic hot water consumption prediction algorithm to the obtained user consumption pattern (UCP) to generate a domestic hot water consumption prediction.

[0038] Further, in some embodiments of the present disclosure, the domestic hot water consumption forecasting algorithm may take into account detected deviations in generating the domestic hot water consumption forecast, preferably comprising user or household vacation status, weather conditions, unexpected events such as showering earlier than average, guests, parties, etc.

[0039] According to a further embodiment of the present disclosure, if a deviation is detected, the domestic hot water consumption forecast may be adjusted automatically and / or a user may be requested to confirm the detected deviation, preferably via a control terminal, preferably a remote control terminal or a voice recognition system, and the domestic hot water consumption forecast may be adjusted in response to the user's confirmation and / or input.

[0040] Further, in some embodiments, the domestic hot water consumption forecast may be determined for a preferred fourth predetermined period (control period) of 10 minutes, 30 minutes, 1 hour, 2 hours, 4 hours, 6 hours, or 12 hours.

[0041] Further in accordance with the present disclosure there is provided a computer-implemented method for controlling the production and / or distribution of domestic hot water, in particular by controlling a system for producing and / or distributing domestic hot water, comprising: generating a domestic hot water consumption forecast using the computer-implemented method for generating a domestic hot water consumption forecast described above; and controlling the production and / or distribution of domestic hot water based on the generated domestic hot water consumption forecast.

[0042] In this disclosure, the term "control" with respect to "control of domestic hot water production and / or distribution" means modifying the production of domestic hot water based on, for example, a detected temperature or detected flow rate, and / or based on an estimate or pattern (e.g., a temperature distribution pattern, a user consumption pattern, or a user consumption forecast), for example by starting or stopping heating of stored hot water by a loaded coil, thereby increasing or decreasing the production of domestic hot water.

[0043] Additionally, in this disclosure, the term "production" in relation to "control of domestic hot water production and / or distribution" defines the production of hot water and means raising the temperature to above about 40°C. Specifically, to produce a given amount of heat or domestic hot water, a hot fluid is supplied to the loading coil, which is heated and supplied, for example, by a heat pump. Heat transfer between the fluid flowing through the loading coil and the hot water stored in the hot water tank heats the hot water stored in the tank, i.e., the temperature of the hot water increases.

[0044] However, in this disclosure the term "distribution" means distributing or making available (supplying) heat or hot water (especially water at a temperature above 40° C.) to each user, for example in the shower or bath.

[0045] Furthermore, in a further embodiment of the present disclosure, the amount of heat, particularly the amount of equivalent energy, to be stored in the thermal storage tank may be determined based on the generated domestic hot water consumption forecast.

[0046] According to a further embodiment of the present disclosure, the amount of heat, in particular the amount of equivalent energy, to be stored in the thermal storage tank may be determined based on the generated domestic hot water consumption forecast.

[0047] Further, in some embodiments, the amount of heat, in particular the amount of equivalent thermal energy, to be stored in the thermal storage tank may be determined by applying a thermal control algorithm to the generated domestic hot water consumption forecast, the thermal control algorithm being preferably a trained algorithm.

[0048] Additionally, in some embodiments of the present disclosure, the thermal control algorithm may be trained based on electrical energy prices (e.g., day / night differences, high availability of green energy such as solar or wind power), availability of local or green energy, weather conditions, weather forecasts, etc.

[0049] According to some embodiments of the present disclosure, depending on electric energy prices, local or green energy availability, weather conditions and / or weather forecasts, the heat control algorithm may activate or begin heat production or increase the amount of heat stored in the thermal storage tank, even if the generated domestic hot water consumption forecast does not require an increase in the thermal storage.

[0050] Further, according to the present disclosure, there is provided a controller for generating a domestic hot water consumption forecast, comprising a control unit and means configured to perform each step of the above-mentioned computer-implemented method for generating a domestic hot water consumption forecast.

[0051] Furthermore, the present disclosure relates to a controller for controlling the production and / or distribution of domestic hot water, in particular by controlling a system for producing and / or distributing domestic hot water, comprising a control unit and means configured to perform the steps of the above-mentioned computer-implemented method for controlling the production and / or distribution of domestic hot water.

[0052] The present disclosure further provides a system for producing and / or distributing domestic hot water, comprising a controller, in particular the above-mentioned controller for generating a domestic hot water consumption forecast, and means configured to implement each of the steps of the above-mentioned method for controlling the production and / or distribution of domestic hot water.

[0053] In some embodiments of the present disclosure, the domestic hot water production and / or distribution system further comprises: A heat storage tank, in particular a hot water storage tank, more particularly a pressurized tank; and means arranged to determine the amount of heat, in particular the amount of equivalent energy, stored in the thermal storage tank and / or to determine the amount of heat, in particular the amount of equivalent energy, removed from the thermal storage tank.

[0054] Since the controller and system are configured to perform the above-mentioned steps of the computer-implemented method, further features disclosed in relation to the computer-implemented method may also be applied to the controller and system, and similarly to the computer-implemented method.

[0055] According to the present disclosure there is further provided a computer program comprising instructions for performing the steps of the above-mentioned computer-implemented method of controlling the production and / or distribution of domestic hot water by the above-mentioned system for producing and / or distributing domestic hot water and / or by the above-mentioned controller for controlling the production of domestic hot water and / or for performing the steps of the above-mentioned computer-implemented method of obtaining a user consumption pattern of domestic hot water by the above-mentioned controller for generating a domestic hot water consumption forecast.

[0056] Further, in accordance with the present disclosure, there is provided a computer readable medium having stored thereon the above-described computer program.

[0057] In this regard, the above-mentioned computer-implemented method can be implemented not only by a controller and a system configured to perform the above-mentioned steps of the computer-implemented method, but also by cloud computing, i.e. by transmitting acquired data representative of the amount of heat to be extracted from a particular thermal storage tank, in particular the data of the thermal storage tank's actual temperature sensor, to a cloud, which is configured to execute the above-mentioned steps of the computer-implemented method for acquiring a user consumption pattern, and sending the acquired user domestic hot water consumption pattern back to a controller which generates a domestic hot water consumption forecast by executing the above-mentioned steps of the method for generating a domestic hot water consumption forecast.

[0058] Furthermore, since the computer program and computer readable medium are associated with the above-mentioned controller and system for generating domestic hot water consumption forecasts and / or controlling domestic hot water production and / or distribution, further features disclosed in relation to the computer-implemented method, controller and system can also be applied to the computer program and computer readable medium, and vice versa.

[0059] According to a first aspect of the present disclosure, there is provided a computer-implemented method for monitoring and / or controlling the production and / or distribution of domestic hot water, in particular by controlling a system for the production and / or distribution of domestic hot water, comprising: - detecting or obtaining at least two real or actual temperatures of a fluid to be stored in the tank, in particular hot water for sanitary purposes, at least at two different positions along the height of the thermal storage tank, in particular a pressurized tank, preferably parallel to the direction of gravity, at least at several points in time, preferably at multiple points in time; and obtaining at least one temperature distribution pattern and / or corresponding heat distribution pattern data of the heat of the fluid stored in the heat storage tank by applying a temperature distribution pattern algorithm to at least two detected or obtained temperatures detected or obtained at least at several points in time, preferably at multiple points in time.

[0060] In this way, the amount of available and / or usable hot water in the hot water tank can be more accurately estimated and monitored, and the energy efficiency of domestic hot water production and / or distribution can be improved. Furthermore, the heat and / or equivalent hot water stored in the hot water tank can be accurately estimated using only a temperature sensor, eliminating the need for a flow detector to detect the amount of hot water removed from the hot water tank. In the present invention, the term "monitoring" with respect to "monitoring domestic hot water production and / or distribution" is used to define, for example, detecting and optionally recording the production of domestic hot water using virtual and real temperature sensors. For example, detecting and / or determining changes or variations in the temperature distribution pattern in the hot water tank as hot water is removed from the hot water tank, and based thereon, determining and / or recording the amount of heat (kWh) remaining in the hot water tank and / or removed from the hot water tank.

[0061] Furthermore, in the present invention, the term "control" with respect to "control of domestic hot water production and / or distribution" means modifying the production of domestic hot water based, for example, on detected temperature or flow rate and / or based on an estimated value or pattern (e.g., temperature distribution pattern), to increase or decrease the production of domestic hot water, for example, by starting or stopping heating of stored hot water by a loaded coil.

[0062] Furthermore, in the present invention, the term "production" in relation to "monitoring and / or control of domestic hot water production" defines the production of hot water and means an increase in temperature to above 40°C. Specifically, to produce a predetermined amount of domestic hot water, a high-temperature fluid is supplied to the loading coil, and this high-temperature fluid is heated and supplied by, for example, a heat pump. The hot water stored in the tank is heated by heat transfer between the fluid flowing through the loading coil and the hot water stored in the hot water tank, i.e., the temperature of the hot water increases.

[0063] However, in the present invention, the term "distribution" in relation to "monitoring and / or control of domestic hot water distribution" means distributing, i.e. making available (supplying) hot water (especially water with a temperature above 40°C) to each user, for example in the shower or bath.

[0064] The computer-implemented method further comprises: applying a virtual temperature sensor algorithm to at least two actual or real temperatures detected at least at several time points, preferably at multiple time points, to obtain or simulate a plurality of virtual temperatures, preferably at least five virtual temperatures, more preferably more than 10 virtual temperatures, even more preferably more than 20 virtual temperatures, of the fluid stored in the thermal storage tank at different positions along the height of the thermal storage tank; and obtaining or simulating a temperature distribution pattern and / or corresponding heat distribution pattern data of the heat stored in the heat storage tank by applying a temperature distribution pattern algorithm to the at least two detected actual or real temperatures and the obtained or simulated plurality of virtual temperatures; The fictive temperature is preferably obtained or simulated using a neural network.

[0065] The computer-implemented method further comprises: determining the quantity of heat, in particular the quantity of equivalent hot water (EHW,V40), stored in the thermal storage tank by applying a thermal estimation algorithm to the obtained temperature distribution pattern and / or to the at least two detected real or actual temperatures and the obtained virtual temperatures; and / or obtaining at least two temperature distribution patterns and / or corresponding thermal pattern data by applying a temperature distribution pattern algorithm to the plurality of sets of detected and / or obtained temperatures, preferably comprising at least two sets of obtained temperatures, preferably detected actual temperatures and / or obtained virtual temperatures, detected at at least two different time points; The method may further comprise the step of determining the amount of heat, in particular the amount of equivalent hot water, extracted from the thermal storage tank by applying an extraction estimation algorithm to the at least two temperature distribution patterns (indirect extraction estimation).

[0066] According to a further aspect, the temperature distribution pattern obtaining step of the computer-implemented method, in particular the temperature distribution pattern algorithm, comprises: The method comprises a step of determining a temperature distribution pattern of heat stored in the thermal storage tank by processing at least two actual or real temperatures detected at least at several points in time, preferably at multiple points in time, in particular at least two detected actual or real temperatures and a plurality of acquired or simulated virtual temperatures, using a regression algorithm, the regression algorithm being preferably trained on temperature data defining the temperature distribution pattern of heat stored in the thermal storage tank using one or more machine learning algorithms.

[0067] Further, the regression algorithm of the computer-implemented method comprises: Temperatures and / or temperature data detected by a plurality of temperature sensors, including a plurality of temperature sensors arranged at different positions along the height of the thermal storage tank and adapted to detect at least two temperatures, preferably two temperature sensors; and / or The input and / or output temperatures of the heat coil, particularly when heating / heating the fluid stored in the thermal storage tank, and / or the flow rate at the inlet and / or outlet of the fluid to / from the thermal storage tank; and / or Training may be based on the flow rate of the fluid (heated fluid) flowing through the heat coil.

[0068] Furthermore, the computer-implemented method may comprise the steps of obtaining the flow rate and / or amount of fluid, in particular hot water, extracted from the thermal storage tank using at least one flow sensor, preferably positioned at the fluid outlet from the thermal storage tank, and / or determining the amount of heat, in particular equivalent hot water (EHW,V40), extracted from the thermal storage tank by applying an indirect extraction estimation algorithm to the at least two temperature distribution patterns and the flow rate of fluid through the heat coil (indirect extraction estimation).

[0069] Further, the computer-implemented method further comprises: at least two temperature sensors, in particular real temperature sensors; The virtual temperatures may be obtained and / or determined using a plurality of virtual temperature sensors, preferably at least five virtual temperature sensors, in particular more than 10 virtual temperature sensors, in particular more than 20 virtual temperature sensors, which are used to obtain a plurality of virtual temperatures, preferably at least five virtual temperatures; The virtual temperature sensor is preferably provided and / or simulated by an (artificial) neural network.

[0070] In the present invention, the term "real" in "real temperature" and "real temperature sensor" is used to define a temperature sensor that is (actually) physically located in a system for monitoring and / or controlling domestic hot water production and therefore actually measures a (live) actual temperature. In other words, the real temperature sensor is actually physically located in the thermal storage tank and actually measures the temperature of the fluid stored in the thermal storage tank.

[0071] In contrast, in the present invention, the term "virtual" in "virtual temperature" and "virtual temperature sensor" is used to define a temperature sensor that is not physically located in the system for monitoring and / or controlling domestic hot water production. Instead of being physically located, the virtual temperature sensor is simulated to some extent by a neural network, as described in more detail below. The temperature value of the virtual sensor is determined by a neural network trained based on the input of the real temperature sensor, and thus the obtained or simulated temperature is referred to as a "virtual temperature."

[0072] The computer-implemented method further comprises: a) The acquired temperature distribution pattern (several temperature distribution patterns) of the heat stored in the heat storage tank, and / or b) the determined amount of heat or equivalent hot water to be stored in the thermal storage tank; and / or c) the amount of heat or equivalent hot water extracted from the thermal storage tank, as determined using an indirect extraction estimation algorithm; and / or d) obtaining a user consumption pattern by applying a user consumption algorithm to the amount of fluid or hot water withdrawn from the thermal storage tank, determined by using at least one flow sensor.

[0073] The computer-implemented method may further comprise determining a heating pattern and / or a hot water production control pattern for the fluid stored in the thermal storage tank by applying a heating pattern algorithm to the obtained user consumption pattern, wherein the user consumption pattern and / or the heating pattern and / or the hot water production control pattern are bounded into time increments of 1 day, 12 hours, 6 hours, 1 hour, 30 minutes, 10 minutes, and / or 1 minute.

[0074] This means that the user consumption pattern is, for example, a temperature distribution pattern of the heat stored in the heat storage tank and / or a set of heat quantities or equivalent hot water quantities stored in the heat storage tank determined at multiple points in time, for example 10 times per hour. Based on these 10 data sets, an average value is calculated, which defines and / or characterizes an increment (in this case an increment per hour). Based on the increments obtained, the user consumption pattern can be determined. The same applies to the heating pattern and / or the hot water production control pattern.

[0075] For example, if a user consumption pattern indicates that a particular user always has a high demand for domestic hot water at a particular time, such as in the morning between 6am and 8am (for taking showers), the heating pattern can be altered accordingly, i.e. making more hot water available than usual between 6am and 8am.

[0076] Further, in the computer-implemented method, prior to determining the temperature distribution pattern of the thermal storage tank, at least 10, preferably at least 20, and more preferably at least 30 temperatures may be obtained at at least 10 time points, preferably at least 20 time points, and more preferably at least 30 time points.

[0077] In this way, the accuracy of determining the temperature distribution pattern can be improved. In particular, before determining the temperature distribution pattern, multiple temperature sets of at least two temperatures are obtained at several time points, and the temperature distribution pattern is determined based on the multiple temperature sets (history) using an (artificial) neural network.

[0078] Furthermore, the present invention provides a computer-implemented method for monitoring and / or controlling the production and / or distribution of domestic hot water, in particular by monitoring and / or controlling a system for monitoring and / or controlling the production and / or distribution of domestic hot water, - detecting at least two actual temperatures of a fluid to be stored in the thermal storage tank, in particular hot water for sanitary purposes, at least at different positions along the height of the thermal storage tank, in particular a pressurized tank, at least at different times; obtaining an amount of fluid withdrawn from the thermal storage tank by applying a fluid withdrawal estimation algorithm to at least two actual temperatures detected at at least several time points; and obtaining the amount of heat or equivalent hot water removed from the thermal storage tank by applying a direct removal estimation algorithm to the amount of fluid removed from the thermal storage tank and the top layer temperature of the thermal storage tank.

[0079] Furthermore, in the computer-implemented method, the top layer temperature may be detected by a temperature sensor, in particular a real temperature sensor, provided near the outlet of the heat storage tank and / or may be obtained by a top layer real temperature sensor or a virtual temperature sensor in the above-mentioned computer-implemented method.

[0080] The present invention further provides a controller for monitoring and / or controlling a domestic hot water production and / or distribution system comprising a control unit and means configured to perform the above-mentioned steps of the computer-implemented method.

[0081] The invention further provides a system for monitoring and / or controlling the production and / or distribution of domestic hot water comprising a controller, in particular the controller as described above, and means adapted to perform each of the above-mentioned steps of the computer-implemented method.

[0082] The system further A heat storage tank, in particular a hot water storage tank, more particularly a pressurized hot water storage tank; and at least two temperature sensors arranged at two different positions along the height of the hot water tank and configured to detect the temperature of the fluid stored in the hot water tank, in particular the hot water for the sanitary equipment.

[0083] Further, in the system, the number of the at least two temperature sensors is at most five, preferably at most four, and more preferably at most three, and one of the at least two temperature sensors is preferably located in the lower half of the thermal storage tank, and more preferably in the lower third of the thermal storage tank.

[0084] Since the controller and system are configured to perform the above-mentioned steps of the computer-implemented method, further features disclosed in relation to the computer-implemented method may also be applied to the controller and system, and similarly to the computer-implemented method.

[0085] The present invention further provides a computer program comprising instructions for causing the above-mentioned controller for a domestic hot water production and / or distribution system and / or a system for domestic hot water production and / or distribution to perform each of the above-mentioned steps of the computer-implemented method for monitoring and / or controlling domestic hot water production and / or distribution.

[0086] The present invention further provides a computer readable medium having stored thereon the above-mentioned computer program for monitoring and / or controlling domestic hot water production and / or distribution.

[0087] In this regard, the above-mentioned computer-implemented method can be implemented not only by a controller and a system configured to perform the above-mentioned steps of the computer-implemented method, but also by cloud computing, i.e., sending the data of the actual temperature sensor of a particular thermal storage tank to the cloud, which is configured to perform the above-mentioned steps of the computer-implemented method, and sending the obtained data (temperature distribution pattern, equivalent hot water stored in the thermal storage tank, heat and / or equivalent hot water extracted from the thermal storage tank, consumer pattern, etc.) back to the controller and / or user of each system for monitoring and / or controlling domestic hot water.

[0088] As the computer program and computer readable medium are associated with the above-mentioned controller and system for the production and / or distribution of domestic hot water, further features disclosed in relation to the computer-implemented method, controller and system may also be applied to the computer program and computer readable medium, and vice versa. [Brief description of the drawings]

[0089] A more complete understanding of the present disclosure and many of the advantages thereof will be readily obtained as the same becomes better understood by reference to the following detailed description, when taken in conjunction with the accompanying drawings, in which: [Figure 1] FIG. 1 is a schematic diagram illustrating a conventional prediction system for a water heating system. [Diagram 2] FIG. 1 is a schematic diagram showing the installation stages of a system for monitoring and / or controlling the production and / or distribution of domestic hot water according to a first embodiment of the present disclosure; [Diagram 3] FIG. 2 is a schematic diagram illustrating the training phase of a system for monitoring and / or controlling the production and / or distribution of domestic hot water according to a second embodiment of the present disclosure; [Figure 4] FIG. 3 is a block diagram illustrating an exemplary signal processing hardware configuration of the system of FIG. 2 according to a further embodiment of the present disclosure. [Diagram 5] FIG. 1 is a flow diagram illustrating a process for generating a domestic hot water consumption forecast according to one embodiment of the present disclosure. [Figure 6] FIG. 1 is a schematic diagram of a system for controlling domestic hot water production and / or distribution according to a further embodiment of the present disclosure; [Figure 7] Schematic diagram showing three different models for generating domestic hot water consumption forecasts and using them for smart control of domestic hot water production and / or distribution [Figure 8] FIG. 3 is a flow diagram showing a process in which the system of FIG. 2 obtains a temperature distribution pattern TDP1 of heat stored in a heat storage tank and determines the amount of stored heat or equivalent hot water according to an embodiment of the present disclosure. [Figure 9]A flow diagram showing the offline data collection process for the training process of the system during the training phase shown in FIG. [Figure 10] FIG. 3 is a block diagram illustrating a process by which the system of FIG. 2 determines the amount of heat (EHW,V40) stored in the thermal storage tank according to one embodiment of the present disclosure. [Figure 11] FIG. 3 is a block diagram illustrating a process by which the system of FIG. 2 determines the amount of heat to be removed from the thermal storage tank according to a further embodiment of the present disclosure. [Figure 12] FIG. 12 is a flow diagram illustrating a training process for the temperature distribution estimator of FIGS. 10 and 11 according to one embodiment of the present disclosure. [Figure 13] FIG. 11 is a flow diagram illustrating a training process for the direct draw estimator of FIG. 10 according to a further embodiment of the present disclosure. [Figure 14] Schematic diagram of a neural network with artificial neurons in the input, hidden and output layers. [Figure 15] 1 is a schematic diagram of a neural network of a user consumption pattern determination algorithm according to an embodiment of the present disclosure; [Figure 16] FIG. 3 illustrates a user consumption pattern obtained by the system for monitoring and / or controlling domestic hot water production and / or distribution illustrated in FIG. 2 . [Figure 17] Schematic diagram showing a heat storage tank with a coil omitted DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0090] <Detailed Description> Some embodiments of the present invention are described below with reference to the drawings. It will be apparent to those skilled in the art of domestic hot water production and / or distribution that the following description of the embodiments is provided for illustrative purposes only and is not intended to limit the present disclosure, which is defined by the appended claims.

[0091] FIG. 1 is a schematic diagram showing a conventional prediction system for a water heater system. In FIG. 1, a hot water heater (boiler) system 400 is extended with a real-time monitoring and prediction system. A typical water heater system includes a water heater 460, a cold water intake pipe 480 that takes in cold water from an external source, and a hot water outlet pipe 470 that leads from the water heater (usually located near the top of the water heater where the hottest hot water is) to a domestic system with multiple pipes for distribution throughout the house (e.g., bathroom, kitchen, laundry room, etc.). The system collects data from three sensors attached to the water heater, including a first temperature sensor (C) 450 that measures the water temperature in the intake line 480, a second temperature sensor (H) 440 that measures the water temperature in the hot water outlet pipe leaving the water heater, and a flow meter (X) 430 that measures the flow rate of water entering the water heater through the cold water intake line. If the water heater system is a closed system, a flow sensor can be installed in either the intake or discharge line. The system analyzes the input from the sensor, calculates the amount of hot water available, and displays a real-time estimate of the available water on a display unit 410, which is preferably located in a bathing area such as a shower or bath.

[0092] 2 is a schematic diagram showing an installation stage of a system for monitoring and / or controlling the production and / or distribution of domestic hot water according to a first embodiment of the present disclosure. The illustrated system includes a controller 1 for monitoring and controlling the production and distribution of domestic hot water, a thermal storage tank 20 for storing hot water (particularly water heated to 40° C. or higher), and five temperature sensors 10A-10B for detecting the temperature of the hot water stored in the thermal storage tank 20.

[0093] Additionally, the thermal storage tank 20, preferably a pressurized tank, includes a coil or loading coil located in the lower half of the tank, as shown in Figure 2. Alternatively, the tank may be configured without a coil, in which case the heat exchanger / source may be external to or part of the tank. In this case, hot water is pumped in and circulated as shown in Figure 17.

[0094] The temperature distribution of the hot water stored in the heat storage tank 20 or hot water tank is stratified, so that hot water can be extracted from the tank even if the water near the bottom is cold (below 40°C). If the heat source for heating the hot water stored in the heat storage tank 20 is, for example, a heat pump, the temperature of the supplied hot water only needs to be slightly (ΔT, about 3°C) higher than the temperature of the water in the lower half of the tank, so the heat pump can operate with a better COP in the initial stage of heating the tank.

[0095] Furthermore, because the hot water stored in heat storage tank 20 is layered, the temperature rises continuously from the bottom to the top of the tank, resulting in a characteristic temperature distribution pattern. Because the temperature rises from the bottom to the top of the tank, temperature sensors 10A-10E provided at different positions along the height direction of hot water storage tank 20 measure different temperatures depending on the position and height of each sensor.

[0096] The hot water tank 20 shown has a cold water inlet / inlet 22A for introducing cold water from an external source and a hot water outlet / outlet 22B for removing hot water from the hot water tank 20. The inlet 22A is located in the bottom third of the tank and the outlet 22B is located near the top of the tank where the hottest temperatures are. The hot water from outlet 22B can be distributed to a home, for example, by multiple pipes for distribution throughout the home.

[0097] Additionally, the illustrated system 100 further includes a pair of temperature sensors 15, 16 for detecting the inlet and outlet temperatures of the fluid (heating fluid) flowing through the loading coil.

[0098] This system uses the (actual) temperature sensors 10A to 10E to measure the (actual) temperatures T 1R_t0 ~T 5R_t0 The actual temperature T 1R_t0 ~T 5R_t0Based on this, the system further obtains a temperature distribution pattern TDP1 of the heat stored in the hot water tank 20 and corresponding heat distribution pattern data. To determine the temperature distribution pattern TDP1 and the corresponding heat distribution pattern data, the controller 1 applies a temperature distribution pattern algorithm, which will be described in more detail below. Then, based on the obtained heat distribution pattern algorithm, the controller 1 determines the amount of heat, in particular the amount of equivalent hot water, to be stored in the hot water tank 20. This may be done by applying a heat estimation algorithm to the obtained temperature distribution pattern (TDP1).

[0099] By repeating the above-described process over time, especially after a certain amount of hot water is taken out of the tub 20 or after the temperature of the hot water stored in the tub 20 drops due to heat loss to the surrounding environment, the system will generate a number of temperature distribution patterns TDP1, TDP2 to TDP n The acquired temperature distribution patterns TDP1, TDP2, TDP n Based on the obtained temperature distribution patterns TDP1, TDP2 to TDP n By applying an indirect extraction estimation algorithm to the above, the amount of heat remaining in the hot water tank 20 and the amount of heat, specifically the amount of equivalent hot water, that is to be extracted from the tank can be determined.

[0100] Figure 3 is a schematic diagram of a training phase of a system for monitoring and / or controlling the production and / or distribution of domestic hot water according to a second embodiment of the present disclosure, in particular as shown in Figure 2. The system shown comprises all the components described above with respect to the system shown in Figure 2. Furthermore, for training purposes, the system comprises 20 additional (real) temperature sensors arranged at different positions along the height of the hot water tank 20, a flow sensor 30 measuring the flow rate of hot water withdrawn from the hot water tank 20, and a flow sensor 31 measuring the flow rate of fluid (heating fluid) through the coil.

[0101] As previously described, the system 100 is used to train the system to obtain, determine, or simulate the temperature distribution pattern of the heat or hot water stored in the thermal storage tank 20. The training of the system, and in particular the temperature distribution pattern algorithm, the heat estimation algorithm, the indirect withdrawal estimation algorithm, and the regression algorithm, are described in further detail below.

[0102] Fig. 4 is a block diagram showing an exemplary signal processing hardware configuration of the system of Fig. 2 according to a further embodiment of the present disclosure, which may be configured to function as the controller 1 of Fig. 2. The programmable signal processing hardware 200 includes a communication interface (I / F) 210, receives (actual) temperature data from the (actual) temperature sensors 10A-10E described above, generates instructions for the system 100 for monitoring and / or controlling domestic hot water production and / or distribution to perform temperature measurements of the hot water storage tank 20, receives measurement data from the (actual) temperature sensors 10A-10E, determines a temperature distribution pattern TDP of heat stored in the thermal storage tank 20 and corresponding heat distribution pattern data, optionally determines the amount of heat stored in the thermal storage tank and / or determines the amount of heat extracted from the thermal storage tank, and outputs a display control signal to control the display device 215 to display the heat distribution pattern, the amount of heat stored in the thermal storage tank and / or the amount of heat extracted from the thermal storage tank. The signal processing device 200 further comprises a processor (e.g. a central processing unit (CPU) or a graphics processing unit (GPU)) 220 which is a control unit 2, a working memory 230 (e.g. a random access memory), an instruction store 240 which stores computer programs comprising computer readable instructions which, when executed by the processor 220, cause the processor 220 to perform various functions including functions of the system 100 for monitoring and / or controlling domestic hot water production and / or distribution. The instruction store 240 may comprise a ROM (e.g. in the form of an Electrically Erasable Programmable Read Only Memory (EEPROM) or Flash memory) which is pre-written with the computer readable instructions. Alternatively, the instruction storage unit 240 may comprise a RAM or a similar type of memory, and the computer readable instructions for the computer program may be input to the instruction storage unit 240 from a computer program product, such as a non-transitory computer readable storage medium 250 in the form of a CD-ROM or DVD-ROM, or from a computer readable signal 260 carrying the computer readable instructions.In any case, when executed by a processor, the computer program causes the processor to perform at least one of the described methods for monitoring and / or controlling domestic hot water production and / or distribution. Note that the controller 1 can also be implemented in non-programmable hardware, such as an application specific integrated circuit (ASIC).

[0103] With respect to this embodiment of the disclosure, the combination of hardware components 270 shown in Figure 4, consisting of a processor 220, a working memory 230, and an instruction storage device 240 configured to perform the functions of the system 100 for monitoring and / or controlling domestic hot water production and / or distribution, is described in detail below. In embodiments such as this embodiment of the invention in which the system 100 includes a display control signal generator, the functionality of this optional component is also provided by the combination of hardware components 270 together with the communication I / F 210.

[0104] As can be seen from the following description of the operations performed by the controller 1 and / or system 100 of this embodiment, the controller 1 and / or system 100 automatically processes the temperatures and / or temperature data, and optionally the flow rates and / or flow rate data, obtained by the respective sensors in order to determine a very accurate thermal distribution pattern TDP of the heat or equivalent hot water stored in the thermal storage tank or hot water storage tank.

[0105] FIG. 5 is a flow diagram illustrating a process for generating a domestic hot water consumption forecast according to one embodiment of the present disclosure.

[0106] FIG. 6 is a schematic diagram of a system for controlling the production and / or distribution of domestic hot water according to a further embodiment of the present disclosure.

[0107] FIG. 7 is a schematic diagram illustrating three different models for generating domestic hot water consumption forecasts and using them for smart control of domestic hot water production and / or distribution.

[0108] Figure 8 is a flow diagram showing a process by which the system 100 of Figure 2 obtains the temperature distribution pattern TDP1 of heat stored in the thermal storage tank 20 and determines the amount of heat or equivalent hot water stored in the tank 20 and optionally removed from the thermal storage tank 20. Figure 5 also shows an alternative process for obtaining the amount of heat or equivalent hot water removed from the thermal storage tank 20.

[0109] In process S10 of FIG. 8, the controller 1, particularly the control unit 2, receives temperature data, particularly at least two actual temperature data. Each of the two actual temperature data includes a plurality of actual temperature measurements (T 1R_t0 ,T 2R_t0 ;T 1R_t0-1 ,T 2R_t0-2 ;T 1R_t-n ,T 2R_tn ) are included.

[0110] In process S15A of FIG. 8, the controller 1, particularly the control unit 2, applies a virtual temperature sensor algorithm to at least two actual temperatures detected at least at several points in time to calculate a plurality of virtual temperatures (T 1V_t0 ,T 2V_t0 , T NV_t0 ) to get the

[0111] In the process S20A of FIG. 8, the controller 1, in particular the control unit 2, acquires at least two actual temperatures (T 1R_t0 ,T 2R_t0 ;T 1R_t0-1 ,T 2R_t0-2 ;T 1R_t-n ,T 2R_tn ) and the obtained multiple fictive temperatures (T 1V_t0 ,T 2V_t0 , T NV_t0) to obtain a (first) temperature distribution pattern TDP1 of hot water stored in the hot water tank 20. The machine learning temperature distribution pattern algorithm used is an algorithm that has already been trained using the above-described system 100 for training / simulation described with reference to Figure 3. The training / machine learning of the temperature distribution pattern algorithm is described in more detail below with reference to Figures 9, 10 and 11.

[0112] Further, in process S30A of FIG. 8, the controller 1, particularly the control unit 2, determines the amount of heat, particularly the amount of equivalent hot water, to be stored in the heat storage tank or hot water storage tank 20 by applying a heat estimation algorithm to the acquired temperature distribution pattern TDP1.

[0113] Further, in an optional process shown in FIG. 8 (shown in dashed lines), the process S20 is repeated at least once in S40A, thereby applying the temperature distribution pattern algorithm described above to at least two sets of acquired temperatures (T 1R_t0 ,T 2R_t0 ,T 1V_t0 ,T 2V_t0 , T NV_t0 ;T 1R_t0-1 ,T 2R_t0-1 ,T 1V_t0-1 ,T 2V_t0-1 , T NV_t0-1 ) to determine at least two temperature distribution patterns TDP1 and TDP2.

[0114] As a further optional process (shown by dashed lines), in process S50A shown in FIG. 8, the amount of heat and / or equivalent hot water to be extracted from the heat storage tank or hot water tank 20 is determined by applying an indirect extraction estimation algorithm to the two acquired temperature distribution patterns TDP1, TDP2.

[0115] As described above with reference to FIG. 3, in the training phase of the system 100 (training system), not only at least two (real) temperature sensors but, for example, 25 (real) temperature sensors are arranged in the system 100. Thus, in the training phase (machine learning phase) of the system 100, the controller 1 receives 25 pieces of temperature data T 1_t0 , T 2_t0 …T 25_t0 Receive.

[0116] FIG. 9 is a flow diagram illustrating an offline data collection process for the system, particularly the training process of the system's artificial neural network, during the training phase shown in FIG. 3. "Offline data collection" means collecting the data required for training the neural network and the corresponding algorithm before the training is actually performed. That is, all the required data is first collected during a specified period of time, such as a day, a week, or a few months. After collecting the required data, the data is pre-processed and input into the neural network for training.

[0117] In process S100 of FIG. 9, the controller 1, particularly the control unit 2 (processor), uses 25 actual temperature sensors 10A to 10XY arranged along the height direction of the hot water tank 20 to obtain 25 pieces of actual temperature data T 1_t0 ~T 25_t0 Receive.

[0118] In process S110 of FIG. 9, the controller, particularly the control unit 2, detects the acquired temperature T 1_t0 ,T 2_t0 …T 1_t25-0 By processing this using the above-mentioned temperature distribution pattern algorithm, a temperature distribution pattern TDP1 of the hot water stored in the hot water tank 20 is determined.

[0119] 9, a quantity of heat (kWh) and / or a quantity of equivalent hot water (l) is removed from the hot water tank using flow sensor 30 located at the outlet of the hot water tank and tank top temperature sensor 10XY (the topmost sensor of the real and virtual sensors). Additionally, optionally, stored hot water is heated (kWh) through coil 21 while the flow rate of the fluid through coil 21 is measured by coil flow sensor 31, and the inlet and outlet temperatures of the fluid are measured by inlet and outlet temperature sensors 15 and 16 of coil 21.

[0120] Furthermore, in process S130 of FIG. 9, the new temperature T 1_t1 ~T 25_t1 In process S140, similarly to process S110, the temperature distribution pattern TDP2 of the hot water stored in the hot water tank 20 is calculated based on the obtained new temperature T 1_t1 ~T 25_t1 Determined using:

[0121] The above process is continuously repeated until sufficient data is acquired and / or collected to train the neural network. In process S150 of Figure 9, the acquired and / or collected data is used to train the artificial neural network. Training is described in more detail below with reference to Figures 12 and 13.

[0122] Fig. 10 is a block diagram showing a process according to one embodiment of the present invention in which the system 100 of Fig. 2 determines the amount of heat (residual equivalent hot water (EHW,V40) at t0) stored in the heat storage tank. As shown in Fig. 10, the actual temperature is detected by at least two actual temperature sensors 10A, 10B (sensor 1, sensor 2 and sensor 3 in this embodiment) and received by the controller 1, in particular the pre-processor of the controller 1 or the control unit 2.

[0123] The preprocessor takes the best subset for each thermal storage tank 20, preprocesses the actual temperature and / or temperature data received from sensors 1-3 and calculates new features. Calculating new features means that the preprocessor uses history (e.g. actual temperatures previously measured by sensors 1-3) to provide a data package consisting of e.g. 28 data.

[0124] In a next step, a scaler scales down the features in preparation for a neural network model. The scaled features are input to an (artificial) neural network (ANN) that has been trained as described above and will be explained in more detail below with reference to Figures 12 and 13. This (artificial) neural network, consisting of two hidden layers of e.g. 40 nodes, estimates the remaining 22 virtual sensors (of the training system described above) based on the received features. An unscaler then scales the features back to the original range and a joiner combines the data of the three real sensors 1-3 with the data of the 22 virtual sensors.

[0125] The data from the coupler is input to a temperature distribution estimator to determine the temperature distribution pattern TDP of the hot water tank.

[0126] The combiner data is further sent to an interpolator, which increases the number of virtual sensors used to determine the temperature distribution pattern TDP in order to eliminate artefacts in the subsequently converted or calculated heat / equivalent hot water (EHW,V40).

[0127] Furthermore, after the interpolator, the determined data is sent to a hot water converter (EHW, V40) and then optionally processed by a filter to further smooth the output (EHW, V40) if the interpolator is not able to remove all artifacts.

[0128] In a final optional step, the coil flow (l / min) detected by a flow sensor configured to detect the flow of fluid through the coil is used to estimate the extraction by an indirect extraction estimator. This allows for the heat (kwh) and / or equivalent hot water (EHW,V40) extracted from the hot water tank 20 to be estimated. In estimating the heat (kwh) and / or equivalent hot water (l) extracted from the hot water tank 20, the indirect extraction estimator may remove or correct for heat losses due to heat transfer to the surrounding environment and may remove or correct for heat added to the hot water tank 20 by heating via the heat coil 21.

[0129] Figure 11 is a block diagram showing a process according to a further embodiment of the present invention in which the system 100 shown in Figure 2 directly determines the amount of heat and / or equivalent hot water (EHW,V40) extracted from the thermal storage tank. The process or system (control unit) shown in Figure 11 comprises all the features / steps or components of the process or system shown in Figure 10 except the indirect extraction estimator.

[0130] Additionally, the illustrated process includes a second (parallel) process line for directly determining the amount of heat and / or equivalent hot water (EHW, V40) extracted from the thermal storage tank. As shown, the three actual temperatures detected by sensors 1-3 are input to a second pre-processor, which takes the best subset for each thermal storage tank 20 and pre-processes the actual temperatures to calculate new features. The features include the newly input actual temperatures and previously input actual temperatures (history).

[0131] The second scaler scales down the features in preparation for a second model in a second (artificial) neural network (ANN_2), trained as described below with particular reference to FIG. 13 and consisting of two distinct layers of, for example, 40 nodes, to estimate the amount of hot water to be withdrawn from the bath. Here, only the amount of water is estimated without indicating the energy stored in the withdrawn water. The unscaler then scales the features back to the original range and provides an estimate of the withdrawn hot water to the direct withdrawal estimator.

[0132] The direct extraction estimator estimates the heat and / or equivalent hot water extracted from the hot water tank using the estimated hot water extraction volume (provided by the second neural network) and the top layer temperature (considered to be the actual temperature of the hot water extracted from the hot water tank) detected by the temperature sensor at the top of the 25 sensors (22 virtual sensors + 3 real sensors). When estimating the heat or equivalent hot water extracted from the hot water tank 20, the direct extraction estimator may remove or correct for heat loss due to heat transfer to the surrounding environment. Here, it is also possible to use an actually installed temperature sensor instead of one of the 22 virtual sensors of the first neural network. This makes the first neural network unnecessary when determining the heat or equivalent hot water to be extracted.

[0133] Figure 12 is a flow diagram showing a training process of the temperature distribution estimator of Figures 10 and 11 according to one embodiment of the present invention. As shown, in the first step, the numbers in the input layer (number of real sensors x history (number of available data; multiple time points)), the numbers in the hidden layer, and the numbers in the output layer (number of virtual sensors) are initialized. In the next step, an artificial neural network (ANN) is generated, and the weights of the ANN are initially set to random values.

[0134] In the next step, the output values ​​of each layer are calculated for the training input values ​​(data collected in the offline data collection process) and the error of the output layer is calculated based on the estimated value (temperature) and the actual value (temperature).

[0135] Based on the calculated error, new values ​​(updates) of the weights of the output and hidden layers of the ANN are calculated and set. Then, with the updated weights, the calculation of the output of each layer using the training input is repeated. This process is continued until the calculated error falls below a required threshold. Once the threshold is reached, the training of the artificial neural network can be terminated.

[0136] During the above process, the number and positions of real and virtual sensors, the history (number of temperature setpoints at several points in time), the optimal number of layers and the optimal weights can be optimized. That is, out of, say, 25 sensors used during training of the neural network, at least two sensors are selected as real sensors, but the two sensors that give the best overall results in terms of accuracy of estimating the temperature distribution pattern when compared with the actually measured temperature distribution pattern are selected as real sensors. The same applies to the number of real and virtual sensors, the number of previous data sets (history) to be considered, the number of layers and the layer size of the artificial neural network.

[0137] Fig. 13 is a flow diagram showing a training process of the direct withdrawal estimator of Fig. 11 according to a further embodiment of the present invention. This process is essentially the same as the machine learning process shown in Fig. 12, except that the amount of hot water to be withdrawn from the hot water tank is estimated and / or trained instead of the temperature value (temperature distribution pattern). Thus, in the step of calculating the error in the output layer, the estimated amount of heat and / or equivalent hot water withdrawn from the hot water tank is compared with the actual value measured by the flow sensor and by the temperature sensor, if any. This training process trains the second neural network of Fig. 11.

[0138] The above-mentioned regression algorithm may be a neural network as in the present embodiment. The neural network automatically generates a discrimination characteristic by processing input data (for example, temperature data detected by the temperature sensors 10A to 10XY, heat coil input temperature data and / or output temperature data detected by the heat coil temperature sensors 15 and 16, and flow rate data detected by the flow rate sensors 30 and 31) without prior knowledge.

[0139] As shown in Figure 14, a neural network generally consists of an input layer, an output layer, and multiple hidden layers. Each layer consists of multiple artificial neurons (labeled A through F in Figure 14), and each layer may perform various types of transformations on the input. Each artificial neuron may be connected to multiple artificial neurons in adjacent layers. The output of each artificial neuron is calculated by a nonlinear function on the sum of its inputs. The multiple artificial neurons and the connections between them are usually given respective weights (WAD, WAE, etc. in Figure 14) that determine the strength of the signal at a particular connection. These weights are adjusted as the learning progresses, thereby adjusting the output of the neural network. The signal travels from the first layer (input layer) to the last layer (output layer), and may pass through each layer multiple times.

[0140] FIG. 15 is a schematic diagram of a neural network of a user consumption pattern determination algorithm according to one embodiment of the present disclosure.

[0141] FIG. 16 is a diagram showing a user cumulative consumption pattern obtained by the system for monitoring and / or controlling domestic hot water production and / or distribution shown in FIG. 2. FIG. 16 shows the amount of equivalent hot water consumed or withdrawn (m 3 16. As can be seen from FIG. 16, the obtained consumption patterns (over time) vary not only during the day, but also between, for example, weekdays and weekends.

[0142] FIG. 17 is a schematic diagram showing a heat storage tank without a coil. [Explanation of symbols]

[0143] 1 Controller 2. Control unit 10A (1st) Actual Temperature Sensor 10B (2nd) Actual temperature sensor 10E (5th) Actual temperature sensor 15 Inlet temperature sensor coil 16 Outlet temperature sensor coil 20 Heat storage tank 21 Heat Coil 22A Inlet / cold water intake 22B Outlet / hot water outlet 30 Discharge hot water flow sensor 31 Heated Fluid Flow Sensor [Prior art documents] [Patent documents]

[0144] [Patent Document 1] US 2015 / 0226460 A1 [Non-patent literature]

[0145] [Non-Patent Document 1] EN16147

Claims

1. 1. A computer-implemented method for obtaining a domestic hot water user consumption pattern (UCP), comprising: The amount of heat (ΣQ Ｔ１ ), in particular obtaining data representative of the equivalent energy amount (S10); A first history (H) of data representative of the amount of heat (ΣQ), particularly the cumulative amount of heat, extracted from the heat storage tank (20) over a number of first periods (T1). １ ) or a data set generation step (S20); The generated first history (H １ and applying a user consumption pattern determination algorithm to the data set (S30) to obtain a user consumption pattern (UCP) of domestic hot water, said user consumption pattern determination algorithm being an algorithm, in particular a time series forecasting algorithm, trained on a history (or a number of histories) or a set (or a number of sets) of data representing the quantity of heat (ΣQ) extracted (from said thermal storage tank (20) or from a number of equivalent thermal storage tanks) and defining a user consumption pattern of domestic hot water by one or more machine learning algorithms, The user consumption pattern determination algorithm comprises: Historical data or a set of data (a plurality of sets of data) representing the amount of heat (ΣQ) extracted by a plurality of users or households over a second time period (T2); and the time of day, particularly for each first period (T1), within each history or collection of data representing in particular the quantity of heat extracted (ΣQ); Trained in Computer-implemented method.

2. A large number of first histories (H １ ) or a second history of data (H ２ ) or a data set (S40), The first history (H １ ) or data set, preferably over a one day or 24 hour period, ２ ) or the data collection preferably spans a period of one week or seven days or 168 hours; 10. The computer-implemented method of claim 1.

3. said first period (T1) lasting for 1 week, 2 days, 1 day, 12 hours, 8 hours, 6 hours, 4 hours, 1 hour, 30 minutes, 10 minutes or 1 minute; and / or The first history (H １ ) or the (predetermined) number of said first periods (T1) of data sets is 1, 2, 3, 4, 6, 24, 48, 144, or 1440; 3. A computer-implemented method according to claim 1 or 2.

4. The (predetermined) second period (T2) lasts for 30 days, 60 days, 90 days, 180 days, 1 year or 2 years; 10. The computer-implemented method of claim 1.

5. The user consumption pattern determination algorithm further comprises: the day of week of each first period (T1) and / or of each history or collection of data representing the quantity of heat extracted (ΣQ), and / or and / or the date of the year and / or the week of the year and / or the month of the year, in particular for each first period and / or for each historical data or data set. the weather conditions, in particular for each first period (T1) and / or for each history or collection of data representative of the quantity of heat extracted (ΣQ), and / or the outside air temperature, in particular for each first period (T1) and / or for each history or collection of data representing the quantity of heat extracted (ΣQ), and / or the vacation status of said user or household, in particular for each first period (T1) and / or for each history or collection of data representative of the quantity of heat extracted (ΣQ), and / or energy prices, in particular for each first period (T1) and / or for each history or collection of data representing the quantity of heat extracted (ΣQ), and / or green energy availability, in particular for each first period (T1) and / or for each history or collection of data representing the amount of extracted heat (ΣQ); and / or the geographical location of each history or collection of data, in particular of each first period (T1) and / or of the quantity of heat extracted (ΣQ), and / or cultural factors, in particular for each first period (T1) and / or for each history or collection of data representing the quantity of heat extracted (ΣQ), Trained in 10. The computer-implemented method of claim 1.

6. the user consumption pattern determination algorithm further considers metadata in determining the user consumption pattern (UCP), the metadata being selected from the group consisting of: number of occupants, age of the occupant(s), average age of the occupants, gender of the occupants, geographic location, cultural factors, annual hot water consumption.

3. A computer-implemented method according to claim 1 or 2.

7. The user consumption pattern determination algorithm further comprises: assigning said users or households to predefined clusters or groups based on at least one of said metadata; determining a user consumption pattern determination sub-algorithm based on the assigned clusters or groups; Equipped with Preferably, said user consumption pattern determination sub-algorithm is trained on data of a plurality of users or households having at least one identical said metadata; 7. The computer-implemented method of claim 6.

8. After the user consumption pattern determination sub-algorithm is determined, the first history (H １ ) data or set (or sets) of data and / or a second history (H ２ ) generating an individual user consumption pattern determination algorithm by training said pre-defined sub-algorithm based on the data or data set (or sets of data); The third period (T3) preferably lasts for 1 day, 2 days, 10 days, 30 days, 60 days, 90 days, 180 days, 1 year or a continuous period; 8. The computer-implemented method of claim 7.

9. The user consumption pattern determination algorithm is trained to determine individual user habits including taking a shower in the morning, taking a shower in the evening, taking a bath in the evening, using the average heat quantity (ΣQ), in particular the equivalent energy quantity, for a shower or a bath; 3. A computer-implemented method according to claim 1 or 2.

10. 1. A computer-implemented method for generating a domestic hot water consumption forecast, comprising: obtaining a user consumption pattern (UCP) of domestic hot water using the computer-implemented method of claim 1 or 2; generating a domestic hot water consumption forecast by applying a domestic hot water consumption forecasting algorithm to the obtained user consumption pattern (UCP); Equipped with Computer-implemented method.

11. the domestic hot water consumption forecasting algorithm taking into account the detected deviations when generating the domestic hot water consumption forecast; The deviations preferably comprise vacation status of the user or household, weather conditions, unexpected events such as earlier than average showers, guests, parties, etc.

11. The computer-implemented method of claim 10.

12. automatically adjusting said domestic hot water consumption forecast if a deviation is detected and / or requesting confirmation of said detected deviation from said user, preferably via a control terminal, and adjusting said domestic hot water consumption forecast in response to said user's confirmation and / or input.

12. The computer-implemented method of claim 11.

13. The domestic hot water consumption forecast is determined for a fourth time period (T4) of 10 minutes, 30 minutes, 1 hour, 2 hours, 4 hours, 6 hours, or 12 hours.

11. The computer-implemented method of claim 10.

14. events of the user or household such as holidays, visitors, parties etc. are determined by the user's input, preferably via a (remote) control terminal, by accessing online data such as a calendar, 11. The computer-implemented method of claim 10.

15. A computer-implemented method for controlling the production and / or distribution of domestic hot water, in particular by controlling a system (100) for producing and / or distributing domestic hot water, comprising: Generating a domestic hot water consumption forecast using the computer-implemented method of claim 10; controlling domestic hot water production and / or distribution based on the generated domestic hot water consumption forecast; Equipped with Computer-implemented method.

16. The amount of heat to be stored in the heat storage tank (20) over a control period T_period is determined based on the generated domestic hot water consumption prediction.

16. The computer-implemented method of claim 15.

17. The amount of heat to be stored in the heat storage tank (20) over a control period T_period is determined by applying a heat control algorithm to the generated domestic hot water consumption forecast; The thermal control algorithm is preferably a trained algorithm.

16. The computer-implemented method of claim 15.

18. The heat control algorithm is trained based on electrical energy prices (dynamic prices, time-of-use rates, etc.), local or green energy availability, weather conditions, energy carbon footprint, weather forecasts, etc.

16. The computer-implemented method of claim 15.

19. A controller (1) for generating a domestic hot water consumption forecast, comprising: A control unit (2); Means arranged to carry out the steps of the method according to claim 10; having Controller (1).

20. A controller (1) for controlling the production and / or distribution of domestic hot water, in particular by controlling a system (100) for the production and / or distribution of domestic hot water, comprising: A control unit (2); Means arranged to carry out the steps of the method according to claim 15; having Controller (1).

21. A domestic hot water production and / or distribution system (100), comprising: A controller, in particular a controller (1) according to claim 19, Means arranged to carry out the steps of the method according to claim 15; Equipped with system.

22. A heat storage tank (20), in particular a hot water storage tank (20), more particularly a pressurized tank; - means arranged to determine the amount of heat, in particular the amount of equivalent energy, stored in the thermal storage tank (20) and / or to determine the amount of heat, in particular the amount of equivalent energy, removed from the thermal storage tank (20); Further comprising:

22. The system (100) of claim 21.

23. 20. The method of claim 19, further comprising instructions for causing the controller to perform the steps of the method of claim 10. Computer program.

24. A computer program according to claim 23, Computer-readable medium.

Images