Heat supply network management system using artificial intelligence and method for controlling same

Abstract

Disclosed are a heat supply network management system using artificial intelligence and a method for controlling same. The method for controlling a heat supply network management system using artificial intelligence, according to the present invention, comprises the steps of: generating an artificial intelligence model by performing machine learning for predicting heat consumption; collecting sensing data from a sensing device with which a heat consumer is provided; and performing heat consumption prediction by inputting a date, public holiday data, weather data, and the collected sensing data into the generated artificial intelligence model.

Description

Heat supply network management system using artificial intelligence and its control method

[0001] The present invention relates to a heat supply network management system using artificial intelligence and a control method thereof, and more specifically, to a heat supply network management system using artificial intelligence that determines and controls the final heat supply amount by combining the amount of heat consumed by each heat consumer with artificial intelligence and a control method thereof.

[0002] Artificial intelligence technology has been developing rapidly recently, and is being used in various fields, including industry.

[0003] Artificial intelligence is a scientific discipline concerned with building computers and machines that can reason, learn, and act in ways that would generally require human intelligence or involve data on a scale larger than humans can analyze. It is a broad field that encompasses several disciplines, including computer science, data analysis and statistics, hardware and software engineering, linguistics, neuroscience, as well as philosophy and psychology. At the operational level of business, AI refers to a set of technologies, primarily machine learning and deep learning, that perform data analysis, prediction and forecasting, object classification, natural language processing, recommendations, and intelligent data retrieval.

[0004] Meanwhile, heating facilities are needed for hot water for showers, etc., or for heating the floor, and there are three main heating methods: individual heating, central heating, and district heating. Individual heating is a method of supplying heating by installing a boiler in each household, and is a method that is widely used in single-family homes.

[0005] Central heating is a method of supplying heating and hot water from a large boiler room installed inside an apartment complex, while district heating is a method of supplying hot water needed for heating to a local unit, for example, by producing electricity at a combined heat and power plant and supplying the remaining heat as hot water to each residence.

[0006] Among these, district heating methods have recently been widely used for reasons such as efficiency.

[0007] In this way, district heating involves producers / suppliers who produce and supply heat, and consumers who receive and use it. However, in the past, heat (hot water) production was centered on producers / suppliers, resulting in energy waste.

[0008] In other words, the actual heat energy (hot water) is being used by the end-users, but by operating the heat energy (hot water) production facilities with a focus on suppliers without proper prediction of the amount of usage by the end-users, unnecessary energy consumption is occurring.

[0009] Additionally, in the past, each generation did not properly operate to save heating energy, resulting in unnecessary waste of heat energy (hot water), which ultimately led to energy waste by causing heat suppliers to generate unnecessary heat energy.

[0010] Therefore, if the heat consumption of the heat user can be predicted more accurately, the energy saving goal can be achieved by adjusting the supply on the heat supplier side.

[0011] As mentioned earlier, artificial intelligence is demonstrating outstanding performance in data analysis, so it is time to introduce technology that uses artificial intelligence to predict heat consumption of these heat users.

[0012] Accordingly, the present invention has been devised to meet the above-mentioned conventional request, and its purpose is to provide a heat supply network management system using artificial intelligence and a control method thereof for more accurately predicting the heat consumption of a heat consumer using an artificial intelligence module and calculating an appropriate heat supply amount of a heat supplier.

[0013] In order to achieve the above-described purpose, a control method of a heat supply network management system using artificial intelligence according to the present invention may include a step of generating an artificial intelligence model by performing machine learning for heat consumption prediction; a step of collecting sensing data from a sensing device equipped at a heat consumer; and a step of inputting date, holiday data, weather data, and the collected sensing data into the generated artificial intelligence model to perform heat consumption prediction.

[0014] Here, a step of grouping each heat user by consumption tendency based on the sensing data of each heat user collected through the above is further included, and the heat consumption of each group formed through the above is input into an artificial intelligence model to perform heat consumption prediction for each group, and then the heat consumption can be predicted by integrating the data to perform overall heat consumption prediction.

[0015] Here, heat consumption predictions for each individual group can be performed using artificial intelligence models generated for each individual group.

[0016] Here, if the predicted heat consumption and the actual heat consumption are compared and match within a preset range, a step of determining the input element that has the greatest influence on the prediction and accumulating and storing the same may be further included; and a step of updating the artificial intelligence model by re-performing machine learning for heat consumption prediction using existing data after setting weights for each input element using the results accumulated and stored by the above for a preset period of time.

[0017] Here, the method may further include a step of creating a visualization image by constructing a 3D model of the location and shape of the heat supply pipe and each heat receiver using map information and facility information; and a step of displaying the location of the determination on the visualization image when an abnormal condition of the sensing device or heat supply facility provided in each heat receiver is determined.

[0018] In addition, in order to achieve the above-described purpose, the heat supply network management system using artificial intelligence according to the present invention may include an artificial intelligence model generation unit that generates an artificial intelligence model by performing machine learning for predicting heat consumption; a sensing data collection unit that collects sensing data from a sensing device equipped at a heat consumer; and a heat consumption prediction unit that inputs date, holiday data, weather data, and the sensing data collected by the sensing data collection unit into the artificial intelligence model generated by the artificial intelligence model generation unit to predict heat consumption.

[0019] Here, the system further includes a grouping unit that groups each heat consumer by consumption tendency based on the sensing data of each heat consumer collected through the sensing data collection unit, and the heat consumption prediction unit inputs each group formed through the grouping unit into an artificial intelligence model to perform heat consumption prediction for each group, and then integrates the data to perform overall heat consumption prediction.

[0020] Here, the heat consumption prediction unit can predict the heat consumption of each individual group using an artificial intelligence model generated for each individual group.

[0021] Here, if the heat consumption predicted by the heat consumption prediction unit and the actual heat consumption are compared and match within a preset range, a result influence determination unit determines and accumulates the input element that has the greatest influence on the prediction of the heat consumption prediction unit; and the artificial intelligence model generation unit can update the artificial intelligence model by re-performing machine learning for heat consumption prediction using existing data after setting weights for each input element using the results accumulated and stored by the result influence determination unit for a preset period of time.

[0022] Here, a visualization processing unit is further included to generate a visualization image constructed by 3D modeling the location and shape of the heat supply pipe and each heat receiver using map information and facility information, and the visualization processing unit can display the corresponding judgment location on the visualization image when an abnormal state of the sensing device or heat supply facility equipped in each heat receiver is determined.

[0023] Figure 1 is a schematic diagram of the entire system including a heat supply network management system using artificial intelligence according to one embodiment of the present invention.

[0024] Figure 2 is a functional block diagram of the management server of Figure 1,

[0025] Figure 3 is an overall control flow diagram of the heat supply network management system of Figure 1.

[0026] Hereinafter, the present invention will be described in detail with reference to the attached drawings.

[0027] The following embodiments of the present invention are merely examples intended to aid understanding of the present invention, and the present invention is not limited to these embodiments. In particular, the present invention may be configured by a combination of at least one of the individual components, individual functions, or individual steps included in each embodiment.

[0028] In particular, for convenience, some claims in the claims include alphabets such as '(a)', but these alphabets do not specify the order of each step.

[0029] The schematic configuration of the entire system including the heat supply network management system (1) according to one embodiment of the present invention is as shown in Fig. 1.

[0030] First, the heat production / supply facility (300) in FIG. 1 is a facility that produces and supplies heat, and may be, for example, a facility that supplies medium-temperature water of 115 degrees or higher.

[0031] For reference, in the case of district heating, if the heat production / supply facility (300) supplies high-temperature medium-temperature water, for example, tap water can be heated and supplied to each household through a heat exchanger in the mechanical room of an apartment complex. Since such technology and structure correspond to known technology, their description is omitted in this embodiment, and it will be explained that when the heat production / supply facility (300) supplies heat, it is received and used by the heat consumer (200).

[0032] As described above, the heat receiver (200) receives and uses the heat energy supplied by the heat production / supply facility (300), and may correspond to each household in an apartment, for example.

[0033] These heat receivers (200) may be equipped with sensing data related to heat use, for example, a calorimeter and a flow meter may be equipped.

[0034] Here, the calorimeter measures the accumulated heat, and can measure the heat energy consumed in the household by using, for example, the difference in the temperature of hot water at the inlet and outlet of the hot water supply pipe supplied to each heat consumer (200).

[0035] Additionally, the flow meter detects the flow of a given fluid, and can measure, for example, the flow rate flowing in through the supply pipe of hot water supplied to each heat receiver (200).

[0036] Meanwhile, the management server (100) performs a function of predicting the amount of heat supply using sensing data and predetermined additional information received from a sensing device (210) equipped in a heat receiver (200). Here, a calorimeter and a flow meter may be equipped as the sensing device.

[0037] Here, the calorimeter measures the accumulated heat, and can measure the heat energy consumed in the household by using, for example, the difference in the temperature of hot water at the inlet and outlet of the hot water supply pipe supplied to each heat consumer (200).

[0038] Additionally, the flow meter detects the flow of a given fluid, and can measure, for example, the flow rate flowing in through the supply pipe of hot water supplied to each heat receiver (200).

[0039] Specifically, the management server (100) performs a function of predicting heat consumption based on the current date, holiday data, weather data, and sensing data of the sensing device (210) and controlling the heat supply accordingly.

[0040] In particular, the management server (100) predicts heat consumption using an artificial intelligence module generated by machine learning. At this time, heat consumption is predicted for each heat user (200) or by region, and then the total heat consumption is predicted by adding them up.

[0041] An example of a specific functional block of such a management server (100) is as shown in FIG. 2.

[0042] As shown in the drawing, the management server (100) may be configured to include an artificial intelligence model generation unit (110), a sensing data collection unit (120), a heat consumption prediction unit (130), a grouping execution unit (140), a grouping execution unit (140), a result impact judgment unit (150), a visualization processing unit (160), and a heat supply control unit (170).

[0043] Here, the artificial intelligence model generation unit (110) performs the function of generating an artificial intelligence model by performing machine learning for predicting heat consumption.

[0044] The process of machine learning and the resulting creation of artificial intelligence models are well-known technologies.

[0045] For example, in the case of deep learning that includes multiple layers, the parameter values ​​of each layer are determined while performing machine learning based on a given data, and an artificial intelligence model can be created by determining the parameter values ​​of each layer in this way. Since this process is merely a known technology, a more detailed explanation is omitted.

[0046] However, in this embodiment, the data for machine learning may be the same type of data input to predict heat consumption as described below.

[0047] For example, for machine learning, sensing data, dates, holiday data, weather data, etc. detected by the sensing device equipped in each heat receiver (200) can be input, and in the case of supervised learning, machine learning can be performed by using the actual heat consumption as the output labeling.

[0048] A more detailed description of the data input into the artificial intelligence module will be provided later.

[0049] In addition, the artificial intelligence model generation unit (110) does not only perform machine learning before operation, but can also update the artificial intelligence model by additionally performing machine learning when data satisfying certain conditions is generated. This is described in the supplementary explanation below.

[0050] The sensing data collection unit (120) performs the function of collecting sensing data from the sensing device equipped in each heat receiver (200).

[0051] The sensing data collection unit (120) can communicate directly with the sensing device (210), or, if a gateway (not shown) for external communication is provided between the management server (100) and the sensing device (210), the sensing data collection unit (120) can collect the necessary data by communicating with the sensing device (210) via the communication gateway and receiving the necessary sensing data.

[0052] The sensing data collected by the sensing data collection unit (120) may include various data generated by the sensing device (210), such as the accumulated heat value detected by the calorimeter, the flow rate value detected by the flow meter, etc.

[0053] The grouping execution unit (140) performs the function of grouping each heat consumer (200) by consumption tendency based on the sensing data of each heat consumer (200) collected through the sensing data collection unit (120).

[0054] Even if the heat consumers (200) belong to a preset area, the heat consumption tendency may differ depending on the situation of each heat consumer (200). The grouping performing unit (140) can group heat consumers (200) that show similar patterns into the same group based on data collected from a sensing device equipped in each heat consumer (200).

[0055] The heat consumption prediction unit (130) performs a function of predicting heat consumption by inputting the date, holiday data, weather data, and the sensing data collected by the sensing data collection unit (120) into the artificial intelligence model generated by the artificial intelligence model generation unit (110).

[0056] Here, the date includes the month and day, public holiday data includes information on whether it is a public holiday or a long weekend, and weather data may include information on temperature, humidity, wind direction, wind speed, daily temperature range, etc.

[0057] At this time, the prediction of heat consumption may be made for each heat user (200) and then added up later, or it may be predicted by preset regional unit and then added up later.

[0058] In particular, weather data may be current weather data, forecast weather data, or even weather change data such as daily temperature range. Since the technology for obtaining weather data is well known, a more detailed description will be omitted.

[0059] In this case, when an artificial intelligence module is equipped, the heat consumption prediction unit (130) can predict heat consumption for the next day or the following week, for example, by inputting the above-described sensing data, date, shared date data, and weather data.

[0060] In particular, the heat consumption prediction unit (130) can input the heat consumption prediction for each group formed through the grouping execution unit (140) into the artificial intelligence model and then perform a heat consumption prediction for each group, and then integrate the data to perform a total heat consumption prediction.

[0061] Here, the heat consumption prediction unit (130) can perform heat consumption prediction for each individual group using an artificial intelligence model generated for each individual group.

[0062] For example, if the artificial intelligence model generation unit (110) generates a first artificial intelligence model corresponding to the first group and a second artificial intelligence model corresponding to the second group, the heat consumption prediction unit (130) can predict the heat usage of a heat consumer (200) belonging to the first group using the first artificial intelligence model, and can predict the heat usage of a heat consumer (200) belonging to the second group using the second artificial intelligence model.

[0063] In this way, when an abnormal state of a sensing device or heat supply facility of a specific heat user (200) is determined, the heat consumption prediction unit (120) described above can predict heat consumption after correcting the data received from the sensing device (210) of the heat user (200) determined to be in the abnormal state to a standard value.

[0064] Here, correcting to a standard value may correspond to a method such as changing the sensing data of the sensing device (210) of the corresponding heat receiver (200) to a value before the previous abnormality was detected, or changing it to an average value of the sensing data of the sensing device (210) of another heat receiver (200), and further, it may mean correcting the sensing data of the sensing device (210) of the corresponding heat receiver (200) using a value obtained by subtracting / increasing / decreasing the amount of change in the sensing data of the sensing device (210) of another heat receiver (200) from the value before the previous abnormality was detected.

[0065] Meanwhile, the result influence judgment unit (150) compares the heat consumption predicted by the heat consumption prediction unit (130) with the actual heat consumption, and if the results match within a preset range, it performs the function of judging the input element that has the greatest influence on the prediction of the heat consumption prediction unit (130) and accumulating and storing the results.

[0066] For example, the result influence judgment unit (150) compares the actual heat consumption that occurred today with the predicted heat consumption based on data up to yesterday, and if the predicted value is the same or similar, the predicted value is correct. Therefore, it judges which of the data used at that time contributed most to the predicted value being correct (for example, at least one of the machine learning inputs, such as season, usage, and temperature determined by date).

[0067] For this judgment, the result influence judgment unit (150) can arbitrarily change each input value to produce a virtual predicted value (i.e., predicted heat consumption value), and then determine the element with the largest difference from the actual heat consumption as the element that contributes the most to the predicted value.

[0068] For example, by applying values ​​that are 10% lower and 10% higher than the calorimeter value detected the previous day, and values ​​that are 10% lower and 10% higher than the temperature detected the previous day as input values ​​to the artificial intelligence model, you can identify the factor that causes the greatest difference from the output value of the artificial intelligence model (i.e., the predicted heat consumption value) (for example, either the previous day's heat consumption or the temperature).

[0069] Here, the result influence judgment unit (150) does not meet the above-described conditions, that is, if the result of comparing the heat consumption predicted by the heat consumption prediction unit (130) with the actual heat consumption is outside the preset range, the input element that has the greatest influence on the prediction of the heat consumption prediction unit (130) is judged or not accumulated and stored.

[0070] Meanwhile, the artificial intelligence model generation unit (110) mentioned above can update the artificial intelligence model by re-performing machine learning for heat consumption prediction using existing data after setting weights for each input element using the results accumulated and stored by the result influence judgment unit (150) for a preset period of time.

[0071] For example, if the accumulated stored results by the result influence judgment unit (150) are 2000 times for temperature and 1000 times for heat consumption of the heat receiver (200), a weight of 2 is applied to the temperature element and a weight of 1 is applied to the heat consumption element obtained by the calorimeter.

[0072] This weighting can be done on all past data that has been used to perform machine learning, which allows for the creation of a completely new and updated AI model. This updated AI model can then perform machine learning using the weighted data to provide more accurate heat consumption predictions.

[0073] Meanwhile, the visualization processing unit (160) performs a function of generating a visualization image constructed by 3D modeling the location and shape of the heat supply pipe and each heat receiver (200) using map information and facility information.

[0074] For example, the location and housing type of each heat consumer (200) can be drawn on a map of an area where heat is supplied, and furthermore, the connection structure of the heat supply pipe connected from the heat production / supply facility (300) to each heat consumer (200) can also be drawn on the map.

[0075] In particular, when an abnormal state of a sensing device or heat supply facility equipped in each heat receiver (200) is determined, the visualization processing unit (160) can display the determined location on a visualization image (i.e., on a map).

[0076] Here, the abnormal state of the sensing device or heat supply facility can be determined by comparing the sensing data received from various sensing devices (210) provided in the heat receiver (200), for example.

[0077] For example, based on the result of comparing the cumulative heat value per unit period received from the heat meter equipped in the first heat receiver (200) with the cumulative heat value per unit period received from the heat meter equipped in the second heat receiver (200), a function is performed to determine whether there is an abnormality in the sensing device (210) or heat supply facility equipped in each heat receiver (200).

[0078] For example, if the heat consumption (value determined by the calorimeter) of the first heat receiver (200) shows a different pattern from the heat consumption of other heat receivers (200) despite the significant change in temperature, the abnormality determination unit (130) can determine that an abnormality has occurred in at least one of the sensing device (210) of the first heat receiver (200) or the heat supply facility provided to the first heat receiver (200).

[0079] Additionally, the management server (100) of FIG. 1 can determine a saving pattern for energy saving and provide guidance to each heat user (200).

[0080] That is, the management server (100) can perform a function of determining a heating saving pattern by weather. Specifically, as mentioned above, when the sensing data received from the heat meter and flow meter equipped in each heat user (200) are compared by heat user (200), the management server (100) can determine a heating saving pattern by weather using the comparison result and weather data.

[0081] For example, each heat receiver (200) is equipped with a calorimeter and a flow meter, and the values ​​measured by the calorimeter and the flow meter may be different from each other depending on the temperature controller settings of the users of each heat receiver (200).

[0082] For example, at the same temperature, the first heat receiver (200) may use the away mode, or the second heat receiver (200) may use the heating mode. Furthermore, even in the away mode or the heating mode, the operation may differ depending on the detailed time setting or temperature setting. In this way, the influence of the operation of the temperature controller may affect the sensing data of the calorimeter and the flow meter.

[0083] For example, in away mode, heating water is used intermittently, so the amount of hot water supplied through the flow meter is bound to be relatively small, and in heating mode, especially in heating mode with the temperature raised, the amount of hot water is bound to be relatively high.

[0084] However, using the away mode on a cold day does not necessarily mean that the accumulated heat measured by the calorimeter is small, so the actual heating cost is greatly affected depending on the mode and temperature set.

[0085] Accordingly, the management server (100) can compare the sensing data of the heat meter and the flow meter by heat receiver (200) to determine which mode of operation under the same temperature is helpful in saving heating costs.

[0086] If you can directly know the mode and temperature set on the actual thermostat, you can use the data from that thermostat. However, since most thermostats are purchased and installed individually and may not have an external communication function, you can indirectly determine the operating status of the thermostat using the sensing data from the flow meter, and compare this with the accumulated heat detected by the heat meter to determine which mode or temperature is necessary to save on heating costs.

[0087] Even if the current temperature set in the temperature controller for at least a specific heat user (200) is unknown, it can be determined by comparing the sensing data patterns of other heat users (200) that setting the temperature higher or lower than the current set temperature is desirable for saving heating costs.

[0088] The management server (100) can provide the heating saving pattern determined through the above-described process to at least one heat user (200) subscriber.

[0089] For example, if the subscriber phone number of each heat user (200) is registered in advance, the management server (100) can provide information on a heating saving pattern (e.g., information that the temperature should be set higher than the current temperature) to the corresponding phone number.

[0090] The heat supply control unit (170) performs a function of controlling the heat supply based on the prediction result of the heat consumption prediction unit (120).

[0091] For this purpose, the heat supply control unit (170) must be connected to each heat production / supply facility (300).

[0092] In the above-described embodiment, it is explained on the premise that the function of the heat supply network management system (1) according to one embodiment of the present invention is implemented in the management server (100). However, it is of course possible for the heat supply network management system (1) according to the present invention to include various sensing devices (210) provided in the heat consumer (200).

[0093] Hereinafter, the overall control flow of a heat supply network management system (1) using artificial intelligence according to one embodiment of the present invention will be described with reference to FIG. 3.

[0094] First, the management server (100) creates an artificial intelligence model through a machine learning process.

[0095] Afterwards, the management server (100) collects data necessary for predicting heat consumption.

[0096] For example, the management server (100) receives and collects sensing data from sensing devices equipped in each heat receiver (200), and similarly collects date, holiday data, weather data, etc. from various external servers.

[0097] Next, the management server (100) uses each collected data as an input value for the artificial intelligence module to predict heat consumption.

[0098] After some time, if the predicted values ​​and actual values ​​are compared and have the same similar range, the management server (100) determines and stores the factor that contributed most to the predicted value, i.e. the result influence factor.

[0099] After this process is repeated for several days or weeks, the management server (100) performs an update process on the artificial intelligence model using the accumulated and stored result influence factors. Meanwhile, it goes without saying that the process of performing each of the above-described embodiments can be performed by a program or application stored on a predetermined recording medium (e.g., computer-readable). Here, the recording medium includes all types of electronic recording media such as RAM (Random Access Memory), magnetic recording media such as hard disks, and optical recording media such as CDs (Compact Disks).

[0100] At this time, the program stored in the recording medium can be executed on hardware such as a computer or smartphone to perform each of the embodiments described above. In particular, at least one of the functional blocks of the management server according to the present invention described above can be implemented by such a program or application.

[0101] Furthermore, the present invention is not limited to the specific embodiments described above, and various modifications and variations can be made without departing from the spirit and scope of the invention. It will be apparent that such modifications and variations are included within the scope of the appended claims.

[0102] As described above, according to the present invention, by identifying heat consumption from the perspective of a heat user using an artificial intelligence module, an appropriate amount of heat supply can be determined and applied, thereby preventing unnecessary energy waste.

[0103] In particular, by regularly updating the artificial intelligence module using data that satisfies certain conditions, more accurate heat consumption prediction is possible.

Claims

1. A control method for a heat supply network management system using artificial intelligence, (a) a step of creating an artificial intelligence model by performing machine learning for predicting heat consumption; (b) a step of collecting sensing data from a sensing device equipped in a heat receiving unit; (c) A control method for a heat supply network management system using artificial intelligence, characterized by including a step of inputting date, public holiday data, weather data and the sensing data collected in step (b) into the artificial intelligence model generated in step (a) to perform heat consumption prediction.

2. In paragraph 1, (d) further includes a step of grouping each heat user by consumption tendency based on the sensing data of each heat user collected through the step (b). A control method for a heat supply network management system using artificial intelligence, characterized in that in the step (c) above, heat consumption prediction for each group formed through the step (d) is performed by inputting the data into an artificial intelligence model, and then heat consumption prediction for each group is performed by integrating the data.

3. In paragraph 2, A control method for a heat supply network management system using artificial intelligence, characterized in that in the step (c) above, heat consumption prediction for each individual group is performed using an artificial intelligence model generated for each individual group.

4. In paragraph 1, (e) a step of determining the input element that has the greatest influence on the prediction of step (c) and accumulating and storing the result of comparing the predicted heat consumption and the actual heat consumption in step (c) and if they match within a preset range; (f) A control method for a heat supply network management system using artificial intelligence, characterized in that it further includes a step of updating an artificial intelligence model by re-performing machine learning for heat consumption prediction using existing data after setting weights for each input element using the results accumulated and stored by step (e) for a preset period of time.

5. In paragraph 1, (g) a step of creating a visual image by building a 3D model of the location and shape of the heat supply pipe and each heat receiver using map information and facility information; (h) A control method of a heat supply network management system using artificial intelligence, characterized in that it further includes a step of displaying the location of the judgment on the visualization image when an abnormal state of a sensing device or heat supply facility equipped in each heat receiver is determined.

6. A computer-readable recording medium recording a program for executing any one of the methods of clauses 1 to 5.

7. An application program stored in a computer-readable recording medium for executing any one of the methods of claims 1 to 5 in combination with hardware.

8. In a heat supply network management system using artificial intelligence, An artificial intelligence model generation unit that generates an artificial intelligence model by performing machine learning for predicting heat consumption; A sensing data collection unit that collects sensing data from a sensing device equipped in a heat receiving unit; A heat supply network management system using artificial intelligence, characterized by including a heat consumption prediction unit that performs heat consumption prediction by inputting date, public holiday data, weather data, and the sensing data collected by the sensing data collection unit into an artificial intelligence model generated by the artificial intelligence model generation unit.

9. In paragraph 8, It further includes a grouping performing unit that groups each heat user by consumption tendency based on the sensing data of each heat user collected through the sensing data collection unit. A heat supply network management system using artificial intelligence, characterized in that the heat consumption prediction unit performs heat consumption prediction for each group formed through the grouping execution unit by inputting the data into an artificial intelligence model, and then performs heat consumption prediction for each group by integrating the data.

10. In paragraph 9, The above heat consumption prediction unit is a heat supply network management system using artificial intelligence, characterized in that it performs heat consumption prediction for each individual group using an artificial intelligence model generated for each individual group.

11. In paragraph 8, A result influence judgment unit for determining the input element that has the greatest influence on the prediction of the heat consumption prediction unit and accumulating and storing the result influence judgment unit when the predicted heat consumption and actual heat consumption in the heat consumption prediction unit are compared and match within a preset range; A heat supply network management system using artificial intelligence, characterized in that the artificial intelligence model generation unit sets weights for each input element using the results accumulated and stored by the result influence judgment unit for a preset period of time, and then re-performs machine learning for heat consumption prediction using existing data to update the artificial intelligence model.

12. In paragraph 8, It further includes a visualization processing unit that creates a visualization image by constructing a 3D model of the location and shape of the heat supply pipe and each heat user using map information and facility information. The above visualization processing unit is a heat supply network management system using artificial intelligence, characterized in that when an abnormal state of a sensing device or heat supply facility equipped in each heat user is determined, the corresponding determination location is displayed on the visualization image.

Images