Heating load dynamic prediction method and device and computer equipment

Abstract

The invention provides a heat supply load dynamic prediction method and device and computer equipment, and relates to the technical field of heat supply system regulation and control. The method comprises the following steps: acquiring a reference heat supply load of a target building in a preset historical unit time period; obtaining user behavior data and meteorological data of the target building in at least one historical unit time period; determining a user behavior coefficient of a preset future unit time period according to the user behavior data; determining a meteorological environment coefficient of a preset future unit time period according to the meteorological data; and according to the reference heat supply load, the user behavior coefficient and the meteorological environment coefficient, determining a target predicted heat supply load of the target building in a preset future unit time period. Therefore, the complex characteristics of the environmental factors and the dynamic characteristics of the user behaviors can be balanced, and the actual scene requirement for efficient regulation and control of the heat supply system is met.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of heating system regulation, in particular to a heating load dynamic prediction method and device and computer equipment. BACKGROUND

[0002] As the core technology of heating system regulation, the accuracy of heating load prediction directly determines the balance between energy distribution efficiency and user thermal comfort. If the predicted value is too high, it will lead to excessive energy supply and energy waste, and if it is too low, it will affect the user's heating experience. To solve the problem of the accuracy of heating load prediction, there are three main schemes in the prior art. The first is to construct a linear mapping relationship with outdoor temperature as the core variable, and directly deduce the load adjustment amount through temperature change. The second is to establish a static calculation model based on the physical properties of the building, that is, the load is equal to the product of the building volume, the heat transfer coefficient and the indoor and outdoor temperature difference. The third is to mine the statistical correlation between historical load and meteorological parameters, and use sliding average algorithm for trend extrapolation. The above mainstream schemes cut in from the angles of environmental factors, building properties and data rules, and to some extent, realize the preliminary quantization of load prediction.

[0003] However, due to the multiple influences of environmental factor fluctuations, building property differences and user heating behavior dynamic changes on heating load, the existing prediction methods always fail to balance the complex characteristics of environmental factors and the dynamic characteristics of user behavior, and cannot meet the actual scene needs of efficient regulation of the heating system, thus leading to significant deviation in heating load prediction. SUMMARY

[0004] The purpose of the embodiments of the present application is to provide a heating load dynamic prediction method, device and computer equipment to balance the complex characteristics of environmental factors and the dynamic characteristics of user behavior, and meet the actual scene needs of efficient regulation of the heating system.

[0005] In order to achieve the above purpose, the technical solutions adopted by the embodiments of the present application are as follows: In a first aspect, the embodiments of the present application provide a heating load dynamic prediction method, comprising: obtaining a reference heating load of a target building in a preset historical unit time period; obtaining user behavior data and meteorological data of the target building in at least one historical unit time period; determining a user behavior coefficient of a preset future unit time period according to the user behavior data; determining a meteorological environment coefficient of the preset future unit time period according to the meteorological data; and determining a target predicted heating load of the target building in the preset future unit time period according to the reference heating load, the user behavior coefficient and the meteorological environment coefficient.

[0006] Optionally, the determining the user behavior coefficient of the preset future unit time period according to the user behavior data comprises: determining a behavior coefficient of each of a plurality of preset behavior categories according to behavior data of the plurality of preset behavior categories in the preset user behavior data; and determining the user behavior coefficient according to the behavior coefficients of the plurality of preset behavior categories.

[0007] Optionally, the determining the user behavior coefficient of each of the plurality of preset behavior categories according to the behavior data of the plurality of preset behavior categories in the preset user behavior data comprises: determining actual energy consumption data of a corresponding time period of each of the plurality of preset behavior categories and total energy consumption in a historical unit time period according to historical user behavior data of the plurality of preset behavior categories; and determining a behavior coefficient corresponding to each of the plurality of preset behavior categories according to the actual energy consumption data of the corresponding time period of each of the behavior data and the total energy consumption in the historical unit time period.

[0008] Optionally, the determining the meteorological environment coefficient of the preset future unit time period according to the meteorological data comprises: determining a meteorological coefficient of each of a plurality of preset meteorological categories according to meteorological data of the plurality of preset meteorological categories in the meteorological data; and determining the meteorological environment coefficient according to the meteorological coefficients of the plurality of preset meteorological categories.

[0009] Optionally, the determining the meteorological coefficient of each of the plurality of preset meteorological categories according to the meteorological data of the plurality of preset meteorological categories in the meteorological data comprises: determining a predicted meteorological coefficient of each of the plurality of preset meteorological categories by using a first mapping relationship between the meteorological data of the plurality of preset meteorological categories and the meteorological coefficients corresponding to the meteorological data.

[0010] Optionally, the obtaining the user behavior data and the meteorological data of the target building in at least one historical unit time period comprises: obtaining behavior data of the plurality of preset behavior categories in each of the historical unit time periods input by using a plurality of first standard data templates; and obtaining meteorological data of a plurality of preset meteorological categories in each of the historical unit time periods collected by a plurality of second standard data templates.

[0011] Optionally, the method further comprises: generating a heating load curve of the preset future unit time period according to a target predicted heating load of the preset future unit time period, and marking predicted heating load values of a plurality of key time points in the heating load curve.

[0012] Optionally, the method further comprises: after the end of the preset future unit time period, determining a prediction deviation according to the target predicted heat supply load and an actual heat supply load of a current unit time period; and adjusting a user behavior coefficient and / or a meteorological environment coefficient of a next future unit time period according to the prediction deviation.

[0013] In a second aspect, an embodiment of the present application provides a heat supply load dynamic prediction device, comprising: a first obtaining module, configured to obtain a reference heat supply load of a target building in a preset historical unit time period; a second obtaining module, configured to obtain user behavior data and meteorological data of the target building in at least one historical unit time period; a first determining module, configured to determine a user behavior coefficient of a preset future unit time period according to the user behavior data; a second determining module, configured to determine a meteorological environment coefficient of the preset future unit time period according to the meteorological data; a third determining module, configured to determine a target predicted heat supply load of the target building in the preset future unit time period according to the reference heat supply load, the user behavior coefficient and the meteorological environment coefficient.

[0014] In a third aspect, an embodiment of the present application provides a computer device, comprising a processor and a memory, wherein the memory stores machine executable instructions capable of being executed by the processor, and the processor can execute the machine executable instructions to implement the method of any one of the first aspect.

[0015] Technical effects: The present application provides a heat supply load dynamic prediction method and device, and a computer device, and relates to the technical field of heat supply system regulation and control. The method comprises the following steps: obtaining a reference heat supply load of a target building in a preset historical unit time period; obtaining user behavior data and meteorological data of the target building in at least one historical unit time period; determining a user behavior coefficient of a preset future unit time period according to the user behavior data; determining a meteorological environment coefficient of the preset future unit time period according to the meteorological data; that is, the user behavior coefficient and the meteorological environment coefficient are both dynamically derived based on historical data, so as to quickly respond to various dynamic changes, avoid prediction failure of the existing historical data fitting method under abnormal working conditions, and enhance the applicability of the method. Then, a target predicted heat supply load of the target building in the preset future unit time period is determined according to the reference heat supply load, the user behavior coefficient and the meteorological environment coefficient. Thus, the present application covers the changes of human heat demand and the changes of environmental heat dissipation conditions through the fusion of the user behavior coefficient and the meteorological environment coefficient, realizes all-round consideration of load influencing factors, improves prediction accuracy and accuracy, and overcomes the limitations of single-factor prediction in the prior art. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required to be used in the embodiments of the present application will be briefly introduced as follows. It should be understood that the following drawings only show some of the embodiments of the present application, and therefore should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0017] Figure 1 A structural schematic diagram of a computer device provided by the embodiments of the present application is shown in FIG. 1. Figure 2 A flowchart of a heat supply load dynamic prediction method provided by the embodiments of the present application is shown in FIG. 2. Figure 1 ; Figure 3 A flowchart of a heat supply load dynamic prediction method provided by the embodiments of the present application is shown in FIG. 3. Figure 2 ; Figure 4 A flowchart of a heat supply load dynamic prediction method provided by the embodiments of the present application is shown in FIG. 4. Figure 3 ; Figure 4 A flowchart of a heat supply load dynamic prediction method provided by the embodiments of the present application is shown in FIG. 5. Figure 5 ; Figure 5 A flowchart of a heat supply load dynamic prediction method provided by the embodiments of the present application is shown in FIG. 6. Figure 6 ; Figure 6 A flowchart of a heat supply load dynamic prediction method provided by the embodiments of the present application is shown in FIG. 7. Figure 7 ; Figure 7 A flowchart of a heat supply load dynamic prediction method provided by the embodiments of the present application is shown in FIG. 8. Figure 8 ; Figure 8 A structural schematic diagram of a heat supply load dynamic prediction device provided by the embodiments of the present application is shown in FIG. 9. DETAILED DESCRIPTION

[0018] The technical solutions of the embodiments of the present application will be described in detail below with reference to the drawings of the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments.

[0019] The heat supply load dynamic prediction method, device and computer device of the embodiments of the present application will be described in detail below with reference to the drawings.

[0020] Figure 9 A structural schematic diagram of a computer device provided by the embodiments of the present application is shown in FIG. 1.Figure 1 As shown, the computer device 100 may include a processor 110 and a memory 120.

[0021] The memory 120 stores machine-executable instructions that can be executed by the processor 110. When the computer device 100 is running, these machine-readable instructions are executed. The processor 110 and the memory 120 communicate via a bus. The processor 110 can execute these machine-executable instructions to implement a dynamic heating load prediction method.

[0022] The memory 120, processor 110, and bus components are electrically connected directly or indirectly to enable data transmission or interaction. For example, these components can be electrically connected to each other via one or more communication buses or signal lines. The mobile storage device includes at least one software function module that can be stored in the memory 120 as software or firmware or embedded in the operating system (OS) of a computer device. The processor 110 is used to execute executable modules stored in the memory 120, such as the software function modules and computer programs included in the dynamic prediction method for heating load of mobile storage media.

[0023] The memory 120 may be, but is not limited to, random access memory (RAM), read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), etc.

[0024] The computer device 100 can be selected according to actual needs; for example, it can be a mobile terminal, a computer, etc. Furthermore, the computer device 100 has software or an application program (APP) capable of executing dynamic heating load prediction methods.

[0025] The dynamic heating load prediction method provided in this application embodiment can be executed by the processor in the computer device 100. The dynamic heating load prediction method provided in this application embodiment will be explained further below. Figure 1 A flowchart illustrating a dynamic heating load prediction method provided in this application embodiment. Figure 2 .like Figure 1 As shown, the method may include: S210, obtaining a reference heating load of the target building in a preset historical unit time period.

[0026] The preset historical unit time period can be selected according to actual conditions. For example, the preset historical unit time period can be selected as 7-15 normal days without special interference, i.e., working days or living days without holidays, equipment maintenance, or extreme weather.

[0027] In a possible implementation manner, the actual heating load data in the preset historical unit time period is called from a heating metering system (such as a heat meter, an energy consumption monitoring platform, etc.) of the target building, and after removing abnormal values (such as sudden increase or decrease in load caused by extreme weather), the average value (unit: kW / MW) is calculated to obtain the reference heating load of the target building in the preset historical unit time period. The reference heating load represents a regular heating level excluding special cases.

[0028] S220, obtaining user behavior data and meteorological data of the target building in at least one historical unit time period.

[0029] In a possible implementation manner, the user behavior data and meteorological data of the target building in at least one historical unit time period are obtained. The user behavior data is used to reflect the influence of personnel activities in the target building on the heating demand; and the meteorological data is used to reflect the influence of the outdoor environment on the heat dissipation or heating demand of the building. By analyzing the correlation between the historical user behavior data and the historical heating load (such as load increase when personnel concentrate in the evening peak), and the correlation between the historical meteorological data and the historical heating load (such as load increase when it is cold and windy), a mapping rule of the historical data and the heating load is established, thereby laying a foundation for predicting the future heating load.

[0030] S230, determining a user behavior coefficient of a preset future unit time period according to the user behavior data.

[0031] In a possible implementation manner, the user behavior coefficient of the preset future unit time period is determined according to the correlation rule between the historical user behavior data and the historical heating load to quantify the deviation degree of the future user behavior from the regular load.

[0032] It should be noted that the time unit of the prediction period corresponds to the historical period. If the historical period is set as a unit time length, the prediction period also uses the same unit time length.

[0033] S240, determining a meteorological environment coefficient of a preset future unit time period according to the meteorological data.

[0034] In one possible approach, the meteorological environmental coefficient for a predetermined future unit time period is determined based on the correlation between historical meteorological data and historical heating load. The degree to which future meteorological conditions deviate from the normal load.

[0035] S250. Based on the baseline heating load, user behavior coefficient, and meteorological environment coefficient, determine the target predicted heating load of the target building within a preset future unit time period.

[0036] In one possible implementation, based on the baseline heating load User behavior coefficient and meteorological environment coefficient The target predicted heating load Q of the target building in the preset future unit time period is determined by the following formula (1).

[0037] Q= × × Formula (1) Among them, the benchmark heating load As a standard benchmark, based on user behavior coefficients The impact of user behavior on load is adjusted; for example, if there is a concentration of people, Kb > 1, and the load increases. This is achieved through the application of meteorological environmental factors. The influence of meteorological conditions on the load is adjusted. For example, if the temperature is low and the wind is strong, then Ke > 1, and the load will increase. Finally, a personalized load forecast value for the future period is obtained.

[0038] The heating load dynamic prediction method provided in this application obtains the baseline heating load of the target building within a preset historical time period; obtains user behavior data and meteorological data of the target building within at least one historical time period; determines the user behavior coefficient for a preset future time period based on the user behavior data; and determines the meteorological environment coefficient for a preset future time period based on the meteorological data. In other words, both the user behavior coefficient and the meteorological environment coefficient are dynamically derived from historical data to quickly respond to various dynamic changes, avoiding the prediction failure of existing historical data fitting methods under unconventional operating conditions and enhancing the applicability of this method. Then, based on the baseline heating load, user behavior coefficient, and meteorological environment coefficient, the target predicted heating load of the target building within the preset future time period is determined. Therefore, this application, through the dual-factor fusion of user behavior coefficient and meteorological environment coefficient, covers changes in human heating demand and changes in environmental heat dissipation conditions, achieving a comprehensive consideration of load influencing factors, improving prediction accuracy and precision, and overcoming the limitations of single-factor prediction in existing technologies.

[0039] exist Figure 2 On this basis, Figure 2 A flowchart illustrating a dynamic heating load prediction method provided in this application embodiment.Figure 3 As shown in the above method, the user behavior coefficient of the preset future unit time period is determined according to the user behavior data, comprising: Figure 2 S310, the behavior coefficients of the plurality of preset behavior categories are respectively determined according to the behavior data of the plurality of preset behavior categories in the preset user behavior data.

[0040] Among them, the preset behavior category can be selected according to the actual situation. For example, the preset behavior category can be divided into work and rest regularity category, personnel density category and special event category.

[0041] Among them, the work and rest regularity category: reflects the activity habit of the user in a fixed time period, directly affects the basic heat demand in different time periods, such as personnel concentration and high heat consumption during working hours; personnel dispersion and low heat consumption during lunch break. The activity habit of the fixed time period can record the typical activity period according to the preset area type, such as filling in the class time or lunch time in school; marking the morning peak period or evening peak period in the community. The personnel density category: reflects the number of people actually in the building, which is the core driving factor of heat demand, such as high load when full and low load when empty. The number of people can be counted according to the actual situation, such as calculating the number of people on duty according to the total number of workstations multiplied by the attendance rate in office buildings; calculating the occupancy rate according to the actual number of occupants divided by the total number of beds in dormitories. The special event category: reflects the temporary behavior that deviates from the routine, which leads to the deviation of heat demand from the routine, such as temporary meeting, heating equipment maintenance or holiday.

[0042] It should be noted that each preset behavior category needs to meet the independent influence and data can be collected, that is, the change of behavior of one behavior category will not completely cover the influence of another category, such as work and rest regularity is when there are people, personnel density is how many people, and the influence dimension is different. And the data of each behavior category can be obtained from daily management records, such as attendance from attendance system, number of people from access control system, and special event from administrative declaration.

[0043] In one possible implementation, for the behavior data of the plurality of preset behavior categories in the preset user behavior data, the behavior data of each preset behavior category is collected separately, and then the behavior coefficient of each preset behavior category is determined based on the correlation between the historical behavior data and the historical heating load So that the behavior coefficient of each preset behavior category Quantifies the influence of a single dimension of behavior.

[0044] S320, the user behavior coefficient is determined according to the behavior coefficients of the plurality of preset behavior categories.

[0045] In one possible implementation, the influence of the behavior coefficients of the plurality of preset behavior categories on the heating load is additive and synergistic. For example, during working hours,​ =1.1) + Full ( =1.0) + No events ( =1.0), then the total impact is 1.1×1.0×1.0=1.1 times the conventional heating load. Therefore, the user behavior coefficient needs to be obtained by weighted multiplication of the sub-behavior coefficients and then by the following formula (2). .

[0046] = × × ×...× Formula (2) Where n is a positive integer greater than or equal to 1.

[0047] The heating load dynamic prediction method provided in this application involves determining behavior coefficients for multiple preset behavior categories based on behavior data from preset user behavior data. The method then uses a categorization operator to calculate the behavior coefficients corresponding to each preset behavior category, ensuring that each sub-preset behavior category's behavior coefficient corresponds to only a single behavior dimension, thus avoiding cross-interference and improving the calculation accuracy of user behavior coefficients. Finally, the method determines user behavior coefficients based on the behavior coefficients of multiple preset behavior categories. Therefore, regardless of how many behavior categories change simultaneously, this application only needs to calculate the sub-behavior coefficients of each preset behavior category separately and then fuse them, improving adaptability to complex behavior scenarios and reducing costs.

[0048] Figure 3 A flowchart illustrating a dynamic heating load prediction method provided in this application embodiment. Figure 4 .like Figure 3 As shown, the above method determines the behavior coefficients for multiple preset behavior categories based on the behavior data of multiple preset behavior categories in the preset user behavior data, including: S410. Based on historical user behavior data for multiple preset behavior categories, determine the actual energy consumption data for the corresponding time periods for each preset behavior category, as well as the total energy consumption for the corresponding historical unit time period.

[0049] The historical time unit is used to extract historical periods of undisturbed baseline data, and can be selected according to actual conditions. For example, this historical time unit can be selected as the last 7-15 normal days, during which there is no extreme weather or heating equipment failure, ensuring that energy consumption is only affected by user behavior.

[0050] The behavior data corresponding period refers to a time interval in which a certain type of preset behavior occurs, such as the working period of the work and rest regularity type, which is from 9:00 am to 18:00 pm; the full staff period of the personnel density type, which is from 10:00 am to 12:00 noon, and needs to be located from historical user behavior data, such as obtaining the working time from the attendance system or obtaining the full staff period from the access control data.

[0051] In a possible implementation manner, according to the historical user behavior data of a plurality of preset behavior categories, the specific period in which each type of preset behavior occurs is determined. For example, the historical attendance data (behavior data of the work and rest regularity type) of a certain office building shows that the working period is from 9:00 to 18:00 every day, and the lunch break period is from 12:00 to 13:00, so the working period corresponding to the work and rest regularity type is from 9:00 to 18:00, and the lunch break period corresponding to the work and rest regularity type is from 12:00 to 13:00; the historical access control data (behavior data of the personnel density type) shows that the number of on-site staff is greater than or equal to 90 (full staff) from 10:00 to 12:00 every day, and the number of on-site staff is less than or equal to 50 (half staff) from 14:00 to 16:00 every day, so the full staff period corresponding to the personnel density type is from 10:00 to 12:00, and the half staff period corresponding to the personnel density type is from 14:00 to 16:00.

[0052] Then, the actual energy consumption data of each preset behavior category corresponding period is extracted from the historical energy consumption monitoring system (such as a heat meter, an energy consumption platform) of the target building, and the average energy consumption of the corresponding period is calculated according to the actual energy consumption data of the plurality of preset behavior categories corresponding period, so as to avoid the interference of single-day fluctuations. For example, the energy consumption data of the target office building in the past 7 normal days is extracted: the daily average energy consumption of the working period (8 hours / day) of the work and rest regularity type is 810 kWh, so the total energy consumption of the working period of 7 days is 5670 kWh. Finally, the total energy consumption of the historical unit time period (such as 1 normal day) is calculated as the benchmark reference for subsequent coefficient calculation. For example, the daily average total energy consumption of the target office building in the past 7 normal days is 1200 kWh, so the total energy consumption of 7 days is 8400 kWh.

[0053] S420, according to the actual energy consumption data of each type of behavior data corresponding period and the total energy consumption of the historical unit time period, the behavior coefficient corresponding to each type of preset behavior category is determined.

[0054] In a possible implementation manner, taking the historical unit time period as 1 day as an example, the benchmark energy intensity is equal to the daily average total energy consumption ÷ 24 h, to determine the benchmark energy intensity, that is, the average energy consumption per unit time of the historical unit time period. Then, according to the actual energy consumption data of each type of behavior data corresponding period and the total energy consumption of the historical unit time period, the behavior coefficient corresponding to each type of preset behavior category is calculated respectively . That is, the behavior coefficient corresponding to a certain type of preset behavior category It equals the average energy consumption per unit time of the time period corresponding to the preset behavior divided by the average energy consumption per unit time of the historical time period. That is, the energy consumption intensity of the behavior period is at the benchmark energy consumption intensity, thereby reflecting the degree of deviation of the preset behavior from the benchmark in terms of energy consumption.

[0055] For example, during a regular workday (9 hours / day), the average daily energy consumption is 810 kWh. Therefore, the average energy consumption per unit time during the corresponding workday is 810 kWh ÷ 9h = 90 kWh / h. The baseline energy intensity is 50 kWh / h, so 90 kWh / h ÷ 50 kWh / h = 1.8, which means that the energy consumption per hour during the workday is 1.8 times the average level.

[0056] It should be noted that if the energy consumption for a certain type of preset behavior during a specific time period is lower than the baseline energy consumption intensity, such as the early morning period for a regular work-rest schedule (0:00-6:00 AM) with an average daily energy consumption of 60 kWh and a baseline energy consumption intensity of 50 kWh / h, then the average energy consumption per unit time for that behavior during that time period is 60 kWh ÷ 6 = 10 kWh / h. Therefore, the behavior coefficient corresponding to this type of preset behavior category will be... =10kWh / h ÷ 50kWh / h = 0.2, where 0.2 reflects that this type of preset behavior results in energy consumption below the average level. The behavior coefficient corresponds to each type of preset behavior. Using 1.0 as the dividing line: A value greater than 1.0 indicates that this type of preset behavior increases energy consumption; <1.0 indicates that the preset behavior reduces energy consumption; =1.0 indicates that this type of preset behavior has no effect.

[0057] The heating load dynamic prediction method provided in this application determines the actual energy consumption data for each time period corresponding to multiple preset behavior categories, as well as the total energy consumption within that historical unit time period, based on historical user behavior data for multiple preset behavior categories. Then, based on the actual energy consumption data for each behavior category's corresponding time period and the total energy consumption within that historical unit time period, a behavior coefficient is determined for each preset behavior category. Therefore, the behavior coefficients for each preset behavior category in this application are calculated entirely based on historical actual energy consumption data and historical total energy consumption within a unit time period. Each behavior coefficient can be traced back to specific energy consumption data, ensuring that the behavior coefficients for each preset behavior category are consistent with the actual preset behavior-energy consumption correlation, laying the foundation for subsequent prediction accuracy. Furthermore, by calculating the behavior coefficients for each preset behavior category, the energy consumption contribution of a single behavior category can be accurately located.

[0058] Figure 5 A flowchart illustrating a dynamic heating load prediction method provided in this application embodiment. Figure 4 .like Figure 5As shown, the above method determines the meteorological environmental coefficient for a preset future unit time period based on meteorological data, including: S510. Based on the meteorological data of multiple preset meteorological categories in the meteorological data, determine the meteorological coefficients of multiple preset meteorological categories respectively.

[0059] The preset meteorological categories refer to the types of meteorological factors pre-classified based on their impact mechanisms on the heat dissipation or heating demand of the target building, and must cover all key impact dimensions. These can be selected according to actual conditions. For example, preset meteorological categories can be divided into outdoor temperature, wind force level, and weather conditions.

[0060] Among them, the outdoor temperature category is used to reflect the temperature difference between the inside and outside of the target building. The greater the temperature difference, the faster the building dissipates heat and the higher the heating demand. For example, the outdoor temperature can be recorded multiple times in the morning, noon and evening, and the average value can be taken as the temperature value of the day, denoted as Tout.

[0061] Weather conditions: These reflect the intensity of solar radiation and air humidity at the target location. For example, on sunny days, solar radiation reduces heat loss from the target building, while high humidity on rainy or snowy days increases the perceived coldness, indirectly increasing heating demand. For instance, three categories can be labeled: "Sunny," "Cloudy," and "Rainy / Snowy," corresponding to weather correction coefficients W1, W2, and W3, respectively. For example, on a "Sunny" day, strong solar radiation results in more heat gain for the target building, so the corresponding weather correction coefficient W1 is 1.0; on a "Cloudy" day, lack of sunlight increases heat loss, so the corresponding weather correction coefficient W2 is 1.1; and on a "Rainy / Snowy" day, high humidity and lack of sunlight significantly increase heat loss, so the corresponding weather correction coefficient W3 is 1.2.

[0062] Wind force ratings reflect the intensity of convective heat transfer on the surface of a target building. Stronger winds mean faster airflow, faster heat loss, and higher heating demands. For example, wind force ratings can be categorized as "light," "moderate," and "strong," corresponding to wind force correction factors F1, F2, and F3, respectively. "Strong winds" require increased load due to rapid heat dissipation. For instance, "light" winds (force 1-2) with weak convection have a corresponding wind force correction factor F1 of 1.0; "moderate" winds (force 3-4) with moderate convection have a corresponding wind force correction factor F2 of 1.1; and "strong" winds (force 5 and above) with strong convection have a corresponding wind force correction factor F3 of 1.2.

[0063] It should be noted that each preset meteorological category must meet the requirements of independent impact and clearly defined physical mechanism. For example, outdoor temperature affects temperature difference, wind force level affects convection, and weather conditions affect radiation. The three have different pathways for heat dissipation, and the meteorological data can be obtained directly from preset weather stations or preset meteorological software, or through various sensors.

[0064] In one possible implementation, for each preset meteorological category in the meteorological data, a meteorological sub-coefficient is calculated based on the correlation between historical meteorological data and historical heating load. So that the meteorological sub-coefficients for each preset meteorological category are... Only the intensity of the impact of a single meteorological factor is quantified.

[0065] It should be noted that the meteorological sub-coefficients for each preset meteorological category... The value is based on "1.0", which means that the meteorological conditions of this preset category have no additional impact, that is, the heat dissipation is normal. >1.0 indicates that the meteorological conditions of this preset category enhance heat dissipation, meaning that more heat supply is required; <1.0 indicates that the weather conditions of this preset category reduce heat dissipation, thus reducing the amount of heat supplied.

[0066] S520. Determine the meteorological environment coefficient based on the meteorological coefficients of multiple preset meteorological categories.

[0067] In one possible implementation, the effects of meteorological coefficients from multiple preset weather categories on building heat dissipation are synergistically additive. For example, low temperatures combined with strong winds and rain / snow will lead to a significant increase in heat dissipation, thus increasing the total meteorological environmental coefficient. The meteorological environment coefficient needs to be obtained by weighted multiplication of the meteorological coefficients and then by the following formula (3).

[0068] = × × ×...× Formula (3) Where n is a positive integer greater than or equal to 1.

[0069] The heating load dynamic prediction method provided in this application determines meteorological coefficients for multiple preset meteorological categories based on meteorological data from multiple preset meteorological categories. Then, based on these meteorological coefficients, a meteorological environment coefficient is determined, avoiding the accuracy loss inherent in single-factor models and improving prediction accuracy. Therefore, the meteorological environment coefficient of this application can be obtained by superimposing sub-meteorological coefficients, without relying on historical data, thus avoiding prediction failures in extreme scenarios. Simultaneously, this meteorological environment coefficient reflects the impact of meteorological conditions on heating load, working in conjunction with user behavior coefficients to improve prediction accuracy and provide a reliable basis for the dynamic control of the heating system.

[0070] Optionally, in the above method, meteorological coefficients for multiple preset meteorological categories are determined based on meteorological data for multiple preset meteorological categories, including: Based on meteorological data for multiple preset meteorological categories, the predicted meteorological coefficients for multiple preset meteorological categories are determined by using the first mapping relationship between the meteorological data and meteorological coefficients corresponding to the multiple preset meteorological categories.

[0071] The first mapping relationship is for future forecast scenarios. Based on the historical static mapping, a dynamic correction factor for the forecast scenario is added to adapt to the uncertainty, seasonal variation, and regional specificity of future weather forecasts.

[0072] In one possible implementation, based on meteorological data for multiple preset meteorological categories, a first mapping relationship between the corresponding meteorological data and meteorological coefficients for each of the multiple preset meteorological categories is used to obtain the predicted meteorological coefficients for multiple preset meteorological categories in a future preset time period. .

[0073] The heating load dynamic prediction method provided in this application determines the predicted meteorological coefficients for multiple preset meteorological categories based on meteorological data for each category, using a first mapping relationship between the corresponding meteorological data and meteorological coefficients. Therefore, the predicted meteorological coefficients for multiple preset meteorological categories in this application not only improve the accuracy of the meteorological coefficients but also allow the entire prediction to flexibly adapt to future scenarios in different regions and seasons, significantly enhancing the practicality of engineering projects.

[0074] Figure 6 A flowchart illustrating a dynamic heating load prediction method provided in this application embodiment. Figure 5 .like Figure 6 As shown, the above method obtains user behavior data and meteorological data of the target building within at least one historical time period, including: S610. Obtain behavioral data of multiple preset behavioral categories within each historical unit time period using multiple first standard data templates.

[0075] The first standard data template refers to a structured data entry table or form designed for each preset behavior category (such as work-rest patterns, population density, special events, etc.). Each preset behavior category corresponds to one template. The fields of this first standard data template are strongly linked to the main influencing factors of each preset behavior to avoid data loss or formatting issues. Furthermore, the fields of each first standard data template must allow for the direct extraction of key information for calculating behavior coefficients, avoiding redundancy.

[0076] In one possible implementation, a first standard data template is used to separately enter data for different preset behavior categories using the corresponding standard data template. For example, work schedule data is entered into the work schedule record sheet; personnel data is entered into the personnel statistics daily report; this avoids mixing different categories of data, which would prevent subsequent segmentation. Then, for each historical normal day (e.g., the last 7 workdays without special interference), the three types of behavior data for that day must be completely entered using the three standard data templates. Simultaneously, it is ensured that the collected data is continuous and periodic historical behavior data, providing sufficient samples for subsequent calculations of the behavior category-energy consumption correlation.

[0077] S620: Acquire meteorological data for multiple preset meteorological categories within each historical unit time period collected by multiple types of meteorological sensors using multiple second standard data templates.

[0078] The second standard data template refers to a structured data acquisition template designed for each preset meteorological category (such as outdoor temperature, wind force level, weather conditions, etc.). Each preset meteorological category corresponds to one template; that is, multiple second standard data templates = temperature acquisition template + wind force acquisition template + weather acquisition template. The fields of the second standard data template directly match the output data of multiple preset meteorological sensors. For example, the temperature sensor outputs time minus the temperature value; the corresponding field in the second standard data template is the acquisition time plus the temperature value, eliminating the need for manual data format conversion. Furthermore, the fields of each second standard data template must allow for the direct extraction of key information for calculating meteorological coefficients and must be consistent with the preset sensor data format.

[0079] In one possible implementation, a second standard data template is used. Data for each preset meteorological category is collected by a dedicated preset sensor and directly interfaced with the corresponding standard data template. For example, temperature data comes only from a temperature sensor, and wind data comes only from a wind speed sensor, avoiding inaccurate data caused by using temperature sensors to extrapolate wind speed. Then, for each historical normal day, the three types of meteorological data for that day must be completely collected using the three standard data templates. At the same time, it is ensured that the historical time unit of the meteorological data and the behavioral data is consistent (e.g., both are in daily units). This allows for accurate matching of the daily behavioral data, daily meteorological data, and daily energy consumption data when calculating the meteorological-energy consumption correlation later.

[0080] The heating load dynamic prediction method provided in this application involves acquiring behavioral data for multiple preset behavioral categories within each historical time unit using multiple first standard data templates; and acquiring meteorological data for multiple preset meteorological categories within each historical time unit using multiple types of meteorological sensors collected using multiple second standard data templates. Thus, this application standardizes data acquisition sources using first and second standard data templates to achieve the goals of data standardization, efficient acquisition, accurate matching, and low-barrier operation. Simultaneously, standardized information acquisition templates are designed for user behavioral data and meteorological data respectively, ensuring the effective acquisition of both types of factor data, providing a data foundation for dual-factor fusion, and lowering the barrier to entry.

[0081] Optionally, the above-mentioned method for dynamic prediction of heating load also includes: Based on the target heating load of a preset future unit time period, a heating load curve for the preset future unit time period is generated, and the predicted heating load values ​​at multiple key time points are marked on the heating load curve.

[0082] The preset future unit time period is usually the core cycle for heating system scheduling and can be selected according to actual conditions. For example, it can be divided into 24 consecutive future hour-level unit time periods based on 1 hour / interval to ensure that the time granularity meets the needs of time-sharing control.

[0083] The target predicted heating load is derived from the target predicted heating load for each future hourly unit time period. (t=1~24), and then the target predicted heating load is calculated according to the above formula (1). .

[0084] In one possible implementation, the coordinate axes and format of the heating load curve are first selected: the time axis is the X-axis, and continuous intervals of the preset future unit time period are marked, ensuring that each hourly interval corresponds to one coordinate point; the heating load value is the Y-axis, with units of kilowatts (kW) or megawatts (MW), and the vertical axis range must cover all target predicted heating loads. The maximum and minimum values ​​should be determined to avoid load values ​​exceeding the coordinate axes; then, the type of heating load curve should be determined, such as a line graph, to clearly show hourly changes, and the line should be smoothed to avoid predicting heating load based on a single target. Even minute errors can cause sudden spikes and drops in the heating load curve. Then, the target predicted heating load for 24 future hourly time periods is further analyzed. Data points are used in chronological order and connected sequentially to form a continuous curve. After connecting the data points, the curve will show a typical fluctuation trend, such as low load in the early morning → morning peak (7-9 am) → midday trough (12-1 pm) → evening peak (6-8 pm) → nighttime decline, which is suitable for the heating patterns of residential or office buildings.

[0085] After determining the heating load curve, the predicted heating load values ​​for multiple key time points are marked based on a combination of user behavior data and meteorological data. It should be noted that marking the predicted heating load values ​​for multiple key time points must ensure coverage of high-load scheduling points, low-load adjustment points, and abnormal risk points. For example, behavior-driven peak points are periods when user heating demand is concentrated; behavior-driven trough points are periods when user heating demand is low; meteorological-driven extreme points are periods when weather conditions cause a sudden increase or decrease in load; and scheduling transition points are critical periods for heating system shift changes or parameter adjustments.

[0086] The labeling method can combine numerical labels with visual emphasis to avoid the labels being buried in the heating load curve. For example, the target predicted heating load can be directly labeled next to the heating load curve data points at key time points. The specific values ​​(e.g., 2592kW for 7:00-8:00) are specified; and the data points and numerical labels corresponding to key time points are differentiated. For example, peak points are marked with solid red dots and bold red font; trough points are marked with hollow blue dots and blue font; and extreme points are marked with yellow triangles and yellow font, so that maintenance personnel can identify the core information at a glance.

[0087] It should be noted that the above-mentioned target predicts heating load. The predicted heating load values ​​at multiple key time points can also be marked according to the actual situation. For example, the predicted value of the morning peak, the value of the midday trough, and the value of the evening peak can be displayed intuitively using a bar chart.

[0088] The heating load dynamic prediction method provided in this application involves predicting the heating load based on a preset target for a future unit time period, generating a heating load curve for that unit time period, and annotating the heating load curve hourly to clearly present the dynamic changes in the heating load, thereby revealing future energy demand fluctuations. Furthermore, the predicted heating load values ​​at multiple key time points are marked on the heating load curve to provide quantitative targets for future heating load scheduling. Based on the annotated information, the causes can be traced, and countermeasures can be taken in advance to avoid heating system failures or energy interruptions. Therefore, this application enables the implementation of test results → scheduling execution → operation and maintenance support, improves the accuracy of two-factor fusion prediction, and reduces the failure rate.

[0089] Figure 7 A flowchart illustrating a dynamic heating load prediction method provided in this application embodiment. Figure 6 .like Figure 7 As shown, the above-mentioned dynamic prediction method for heating load also includes: S710. After the preset future unit time period ends, determine the prediction deviation based on the target predicted heating load and the actual heating load of the current unit time period.

[0090] S720. Adjust the user behavior coefficient and / or meteorological environment coefficient for the next future unit time period based on the prediction deviation.

[0091] Among them, the preset future unit time period refers to the previous scheduling cycle in which prediction and energy supply have been completed. After the scheduling cycle ends, the deviation calculation and coefficient adjustment stage begins. The target predicted heating load Q is the set of hourly predicted heating loads in the previous period, which is the benchmark reference value for coefficient adjustment; the actual heating load Q' in the current unit time period is the set of hourly actual heating loads collected by the heating metering system (such as heat meters, energy consumption monitoring platforms, etc.) in the previous period, which is the true basis for calculating the deviation. The next future time unit is the next scheduling cycle to be predicted, and the adjusted user behavior coefficient. and / or meteorological environment coefficient This will be used for load forecasting during this period.

[0092] In one possible implementation, after the preset future unit time period ends, the prediction deviation E is determined by the following formula (4) based on the target predicted heating load Q and the actual heating load Q' of the current unit time period.

[0093] E= Formula (4) The smaller the prediction deviation E, the more accurate the overall prediction of the previous period.

[0094] Based on the magnitude of the prediction deviation E, it can be divided into four categories: no deviation (E≤5%), slight deviation (5%<E≤8%), moderate deviation (8%<E≤15%), and severe deviation (E>15%). Different levels correspond to different adjustment intensities; the larger the prediction deviation E, the larger the adjustment ratio, to avoid insufficient or excessive adjustment in a single instance.

[0095] It should be noted that if the prediction deviation E exceeds the preset deviation threshold, the user behavior coefficient needs to be adjusted. and meteorological environment coefficient If the prediction deviation E is less than or equal to the preset deviation threshold, then no adjustment of the coefficients is required, and the current user behavior coefficients can continue to be used. and meteorological environment coefficient .

[0096] For example, when the prediction deviation E > a preset deviation threshold, and when the target predicted heating load Q is greater than the actual heating load Q' in the current unit time period, the user behavior coefficient is reduced by a first preset ratio. and meteorological environment coefficient When the target predicted heating load Q is less than the actual heating load Q' in the current unit time period, the user behavior coefficient is increased by the second preset ratio. and meteorological environment coefficient .

[0097] Among them, user behavior coefficient and meteorological environment coefficient The adjustment is based on the following: If the prediction deviation E mainly stems from behavioral data (such as discrepancies between actual number of users, special events, and predictions), then only the user behavior coefficient will be adjusted. If the forecast deviation E mainly stems from meteorological data (such as discrepancies between actual temperature and wind speed and forecasts), then only the meteorological environment coefficient should be adjusted. If both types of prediction bias E are significant (e.g., prediction bias E > 15%, and the biases are large during key periods), then the user behavior coefficients need to be adjusted simultaneously. and meteorological environment coefficient .

[0098] The first preset ratio and the second preset ratio can be different or the same. For example, if the first preset ratio and the second preset ratio are the same, the preset ratio can be based on the ratio λ of the actual heating load Q' of the previous cycle and the target predicted heating load Q, which can be expressed by the following formula (5).

[0099] λ=[ ] / n formula (5) Where n is a positive integer greater than or equal to 1, representing the number of data collection cycles; N is the sum of multiple cycles. λ > 1 indicates that the predicted load is underestimated, and the coefficient needs to be adjusted upwards; λ < 1 indicates that the predicted load is overestimated, and the coefficient needs to be adjusted downwards.

[0100] According to the above formula (5), the ratio λ ensures that the adjustment magnitude is positively correlated with the prediction deviation E. For example, if only the user behavior coefficient is adjusted... Then the new user behavior coefficient =Original User Behavior Coefficient ×λ; If only the meteorological environment coefficient is adjusted The new meteorological environment coefficient =Original meteorological environment coefficient ×λ; If user behavior coefficients are adjusted synchronously and meteorological environment coefficient Then the new user behavior coefficient =Original User Behavior Coefficient ×λ, New meteorological environment coefficient =Original meteorological environment coefficient ×λ.

[0101] It should be noted that the adjusted user behavior coefficient and / or meteorological environment coefficient Small-scale validation is needed to avoid over-correction. The first key time period of the next cycle could be selected, such as tomorrow's morning rush hour from 7:00 to 8:00, using the new user behavior coefficient. and / or new meteorological environment coefficient Calculate the new target predicted heating load Q; if the preset deviation E between the new actual heating load Q' and the new target predicted heating load Q during this period is ≤5%, then the new user behavior coefficient is calculated. and / or new meteorological environment coefficient This will officially take effect and be used for full-time forecasting in the next cycle; if the preset deviation E is still greater than 5%, the above adjustment steps will be repeated (i.e., recalculate the ratio λ and fine-tune the user behavior coefficient). and / or meteorological environment coefficient (until the preset deviation E is met).

[0102] The heating load dynamic prediction method provided in this application involves determining the prediction deviation based on the target predicted heating load and the actual heating load of the current unit time period after the preset future unit time period ends; and adjusting the user behavior coefficient and / or meteorological environment coefficient for the next future unit time period based on the prediction deviation. Thus, this application accurately locates the source of deviation and makes targeted adjustments through deviation analysis of key time periods, adapting to complex scenario changes.

[0103] To facilitate understanding of the above-described dynamic heating load prediction method, this application also provides a flowchart example of the dynamic heating load prediction method, which will be further described below with reference to the accompanying drawings. Figure 8 A flowchart illustrating a dynamic heating load prediction method provided in this application embodiment. Figure 7 .like Figure 8 As shown in the illustration, the embodiments provided in this application provide... Figure 7 It may include: S810 acquires user behavior data and meteorological data.

[0104] Specifically, user behavior data collection mainly involves the following three categories, all obtained through daily management records: **Daily Routine Data:** Recording typical activity periods by area type, such as "class time" and "lunch break time" for schools; and "morning peak hours" and "evening peak hours" for residential areas. **Personnel Density Data:** Recording the actual number of people each day, such as calculating "number of people on duty = total workstations × attendance rate" for office buildings; and calculating "occupancy rate = actual number of occupants / total number of beds" for dormitories. **Special Events Data:** Marking temporary situations affecting heating, such as "equipment maintenance days" and "Spring Festival holidays."

[0105] Meteorological data acquisition mainly involves inputting the following three categories of basic meteorological data, which can be acquired through preset sensors: Outdoor temperature: Recorded multiple times in the morning, noon, and evening, and the average value is taken as the daily temperature value, denoted as Tout. Weather conditions: Labeled as "Sunny," "Cloudy," and "Rain / Snow," corresponding to weather correction coefficients W1, W2, and W3, respectively. Wind force level: Divided into "Light Breeze," "Moderate Wind," and "Gale," corresponding to wind force correction coefficients F1, F2, and F3, respectively. Increased load is required in strong winds due to rapid heat dissipation.

[0106] S820: Determine the target predicted heating load based on user behavior data and meteorological data.

[0107] Specifically, the baseline heating load is first determined by a weighted calculation based on both user behavior data and meteorological data. The average load of the most recent normal days (without special events) is taken as the baseline value.

[0108] Then, based on the collected user behavior data, the values ​​are dynamically adjusted to obtain the user behavior coefficient. For example, when students are not in their dormitories during class time or when there are fewer people on holidays, the user behavior coefficient... The value is relatively low; there are more people in the community during the evening rush hour, resulting in a lower user behavior coefficient. The value is relatively high. Then, based on meteorological data, the meteorological environment coefficient is calculated and determined. Ke = f(Tout, W, F), where W is the weather correction factor and F is the wind correction factor. This meteorological environment coefficient... The coefficients of the function are determined based on different temperatures and weather conditions. The final target, the predicted heating load Q, can be obtained according to the above formula (1).

[0109] S830: Output the results based on the predicted heating load of the target.

[0110] Specifically, the output of the target predicted heating load Q can include two parts: first, generating a 24-hour heating load curve and marking three key values, such as the predicted value of the morning peak, the value of the midday trough, and the value of the evening peak, which are displayed visually using a bar chart; second, calculating the preset deviation E between the target predicted heating load Q and the actual heating load Q' at the end of each day, which can be obtained through the above formula (4).

[0111] If the preset deviation E is large, the user behavior coefficient for the next day will be automatically adjusted. and meteorological environment coefficient .

[0112] It should be noted that, in terms of prediction accuracy, the synergistic effect of two factors—user behavior coefficients and... Quantification reflects user behavior characteristics, and also utilizes meteorological environmental coefficients. Taking meteorological conditions into account comprehensively, the prediction error for key periods such as morning and evening rush hours is significantly reduced. For example, in school dormitories during evening self-study hours, traditional models only predict based on historical averages, while this application uses user behavior coefficients... The value is higher when people gather after class, and the meteorological environment coefficient is also higher. Adjustments are made based on the prevailing meteorological conditions, and the combined effect of the two factors accurately captures peak loads. In terms of adaptability, the two-factor fusion approach allows the model to better handle various complex situations. When user behavior or meteorological conditions change, the model can adjust the user behavior coefficient accordingly. and meteorological environment coefficient It adapts to these changes without requiring manual parameter resetting. Moreover, it's simple to operate; heating technicians can complete the forecast by filling in relevant information, making it suitable for widespread application in daily life.

[0113] The heating load dynamic prediction method provided in this application acquires user behavior data and meteorological data; determines the target predicted heating load based on the user behavior data and meteorological data; and outputs the results based on the target predicted heating load. Thus, this application organically integrates two key factors, user behavior data and meteorological data, and achieves accurate prediction of heating load through dynamic weighting of the user behavior coefficient Kb and the meteorological environment coefficient Ke, overcoming the limitations of single-factor prediction in existing technologies. Simultaneously, using a benchmark heating load as a reference, the target predicted heating load is obtained, allowing user behavior data and meteorological data factors to work together in the prediction, improving prediction accuracy. Finally, based on daily actual data, a preset deviation is calculated, adjusting not only the user behavior coefficient but also simultaneously optimizing the meteorological environment coefficient, ensuring that the two factors remain synergistically adapted during dynamic adjustment, achieving continuous model optimization without professional intervention.

[0114] Based on the same inventive concept, this application also provides a dynamic heating load prediction device. Since the principle of the device in this application is similar to the dynamic heating load prediction method described above in this application, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be described again.

[0115] Figure 9 This is a schematic diagram of a dynamic heating load prediction device provided in an embodiment of this application. Figure 9 As shown, the heating load dynamic prediction device 90 may include: a first acquisition module 91, a second acquisition module 92, a first determination module 93, a second determination module 94, and a third determination module 95.

[0116] The first acquisition module 91 is used to acquire the baseline heating load of the target building in a preset historical unit time period; The second acquisition module 92 is used to acquire user behavior data and meteorological data of the target building within at least one historical unit time period. The first determining module 93 is used to determine the user behavior coefficient for a preset future unit time period based on user behavior data. The second determining module 94 is used to determine the meteorological environment coefficient for a preset future unit time period based on meteorological data; The third determination module 95 is used to determine the target predicted heating load of the target building within a preset future unit time period based on the baseline heating load, user behavior coefficient, and meteorological environment coefficient.

[0117] In one optional implementation, the first determining module 93 is specifically used to: determine the behavior coefficients of multiple preset behavior categories based on the behavior data of multiple preset behavior categories in the preset user behavior data; and determine the user behavior coefficients based on the behavior coefficients of the multiple preset behavior categories.

[0118] In one optional implementation, the first determining module 93 is specifically used to: determine the actual energy consumption data of the time period corresponding to the multiple preset behavior categories and the total energy consumption of the historical unit time period based on the historical user behavior data of the multiple preset behavior categories; and determine the behavior coefficient corresponding to each preset behavior category based on the actual energy consumption data of the time period corresponding to each type of behavior data and the total energy consumption of the historical unit time period.

[0119] In one optional implementation, the second determining module 94 is specifically used to: determine the meteorological coefficients of multiple preset meteorological categories based on the meteorological data of multiple preset meteorological categories in the meteorological data; and determine the meteorological environment coefficients based on the meteorological coefficients of multiple preset meteorological categories.

[0120] In one optional implementation, the second determining module 94 is specifically used to: determine the predicted meteorological coefficients of multiple preset meteorological categories based on meteorological data of multiple preset meteorological categories and by adopting a first mapping relationship between meteorological data and meteorological coefficients corresponding to multiple preset meteorological categories.

[0121] In one optional implementation, the second acquisition module 92 is specifically used to: acquire behavioral data of multiple preset behavioral categories within each historical unit time period using multiple first standard data templates; and acquire meteorological data of multiple preset meteorological categories within each historical unit time period collected by multiple types of meteorological sensors using multiple second standard data templates.

[0122] In one optional embodiment, the heating load dynamic prediction device 90 is further configured to: predict the heating load based on a preset target for a future unit time period, generate a heating load curve for a preset future unit time period, and mark the predicted heating load values ​​at multiple key time points in the heating load curve.

[0123] In one optional embodiment, the heating load dynamic prediction device 90 is further configured to: determine the prediction deviation based on the target predicted heating load and the actual heating load of the current unit time period after the preset future unit time period ends; and adjust the user behavior coefficient and / or meteorological environment coefficient for the next future unit time period based on the prediction deviation.

[0124] It should be noted that for details not disclosed in the heating load dynamic prediction device of this application embodiment, please refer to the details disclosed in the heating load dynamic prediction method of this application embodiment, which will not be repeated here.

[0125] These modules can be one or more integrated circuits configured to implement the above methods, such as one or more Application Specific Integrated Circuits (ASICs), one or more microprocessors, or one or more Field Programmable Gate Arrays (FPGAs). Alternatively, when a module is implemented using processing element scheduler code, the processing element can be a general-purpose processor, such as a Central Processing Unit (CPU) or other processor capable of calling program code. Furthermore, these modules can be integrated together as a system-on-a-chip (SOC).

[0126] Optionally, embodiments of this application also provide a computer-readable storage medium storing a computer program. When the computer program is run by a processor, the processor executes the steps of the dynamic heating load prediction method for mobile storage media described in the above embodiments. The specific implementation and technical effects are similar and will not be repeated here.

[0127] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The integrated unit described above can be implemented in hardware or in the form of hardware plus software functional units.

[0128] The integrated units implemented as software functional units described above can be stored in a computer-readable storage medium. These software functional units, stored in a storage medium, include several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute some steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0129] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for dynamic prediction of heating load, characterized in that, include: Obtain the baseline heating load of the target building within a preset historical time period; Obtain user behavior data and meteorological data for the target building within at least one historical time period; Based on the user behavior data, determine the user behavior coefficient for a preset future unit time period; Based on the meteorological data, determine the meteorological environment coefficient for the preset future unit time period; Based on the baseline heating load, the user behavior coefficient, and the meteorological environment coefficient, the target predicted heating load of the target building within the preset future unit time period is determined.

2. The method for dynamic prediction of heating load according to claim 1, characterized in that, The step of determining the user behavior coefficient for the preset future unit time period based on the user behavior data includes: Based on the behavior data of multiple preset behavior categories in the preset user behavior data, the behavior coefficients of the multiple preset behavior categories are determined respectively; The user behavior coefficient is determined based on the behavior coefficients of the multiple preset behavior categories.

3. The method for dynamic prediction of heating load according to claim 2, characterized in that, The step of determining the behavior coefficients for the multiple preset behavior categories based on the behavior data of multiple preset behavior categories in the preset user behavior data includes: Based on the historical user behavior data of the multiple preset behavior categories, the actual energy consumption data of the corresponding time periods of the multiple preset behavior categories, as well as the total energy consumption of the corresponding historical unit time period, are determined respectively. Based on the actual energy consumption data of the time period corresponding to each type of behavior data and the total energy consumption of the historical unit time period, the behavior coefficient corresponding to each of the preset behavior categories is determined.

4. The method for dynamic prediction of heating load according to claim 1, characterized in that, The step of determining the meteorological environmental coefficient for the preset future unit time period based on the meteorological data includes: Based on the meteorological data of multiple preset meteorological categories in the meteorological data, the meteorological coefficients of the multiple preset meteorological categories are determined respectively; The meteorological environment coefficient is determined based on the meteorological coefficients of the multiple preset meteorological categories.

5. The method for dynamic prediction of heating load according to claim 4, characterized in that, The step of determining the meteorological coefficients for each of the multiple preset meteorological categories based on meteorological data from the meteorological data includes: Based on the meteorological data of the multiple preset meteorological categories, the predicted meteorological coefficients of the multiple preset meteorological categories are determined by using the first mapping relationship between the meteorological data and meteorological coefficients corresponding to the multiple preset meteorological categories.

6. The method for dynamic prediction of heating load according to claim 2, characterized in that, The acquisition of user behavior data and meteorological data of the target building within at least one historical time period includes: Obtain behavioral data for the multiple preset behavioral categories within each historical time unit, input using multiple first standard data templates; Meteorological data for each historical unit time period is acquired from multiple preset meteorological categories collected by various meteorological sensors using multiple second standard data templates.

7. The method for dynamic prediction of heating load according to claim 1, characterized in that, The method further includes: Based on the target predicted heating load for the preset future unit time period, a heating load curve for the preset future unit time period is generated, and the predicted heating load values ​​at multiple key time points are marked on the heating load curve.

8. The method for dynamic prediction of heating load according to claim 1, characterized in that, The method further includes: After the preset future unit time period ends, the prediction deviation is determined based on the target predicted heating load and the actual heating load of the current unit time period; Based on the prediction deviation, adjust the user behavior coefficient and / or meteorological environment coefficient for the next future unit time period.

9. A dynamic heating load prediction device, characterized in that, include: The first acquisition module is used to acquire the baseline heating load of the target building in a preset historical time period. The second acquisition module is used to acquire user behavior data and meteorological data of the target building within at least one historical unit time period. The first determining module is used to determine the user behavior coefficient for a preset future unit time period based on the user behavior data. The second determining module is used to determine the meteorological environment coefficient for the preset future unit time period based on the meteorological data; The third determining module is used to determine the target predicted heating load of the target building within the preset future unit time period based on the benchmark heating load, the user behavior coefficient, and the meteorological environment coefficient.

10. A computer device, characterized in that, The method includes a processor and a memory, the memory storing machine-executable instructions that can be executed by the processor to implement the method of any one of claims 1-8.

Images