A method, system, and storage medium for predicting and controlling heating load based on room temperature feedback.

Abstract

This application relates to a heating load prediction and control method, system, and storage medium based on room temperature feedback, belonging to the field of smart heating systems. It includes establishing a load prediction model; acquiring real room temperature, set room temperature, and weather outside temperature data; calculating a compensation outside temperature based on the real room temperature, set room temperature, and weather outside temperature; replacing the weather outside temperature data with the compensation outside temperature data and substituting it into the load prediction model to obtain the load prediction result. This application provides objective feedback on the actual heating quality based on historical data, corrects and standardizes the prediction results of the load prediction model, effectively improving the prediction accuracy of the load prediction model.

Description

Technical Field

[0001] This application relates to the field of smart heating systems, and in particular to a method, system, and storage medium for predicting and controlling heating load based on room temperature feedback. Background Technology

[0002] Urban centralized heating systems are an important public welfare project in northern China. They are generally heated by a thermal power plant or regional boiler room, and then transported to the community heat exchange station for secondary heat exchange before being delivered to households through the municipal pipeline network.

[0003] With the development of intelligent and digital technologies, load prediction models for heat exchange stations, built from a large amount of data generated during the historical operation of heating systems, have gradually become the main basis for the automated control of many heat exchange systems. The prediction results of the heat load prediction model can provide a reference for the heat load changes during the operation of the centralized heating system, assisting managers in effectively adjusting the operating strategies of the heating system.

[0004] However, existing heat load models are trained entirely on historical data. The higher the quality of heating regulation in historical data, the higher the prediction accuracy of the heat load prediction model. However, there is a lack of objective feedback on the actual heating quality, and heating managers cannot know the actual heating effect. Summary of the Invention

[0005] In order to objectively provide feedback on the actual heating quality based on historical data, thereby correcting and standardizing the prediction results of the load prediction model and effectively improving the prediction accuracy of the load prediction model, this application provides a heating load prediction and control method, system and storage medium based on room temperature feedback.

[0006] In the first aspect, this application provides a heating load prediction and control method based on room temperature feedback, which adopts the following technical solution:

[0007] A heating load prediction and control method based on room temperature feedback includes the following steps:

[0008] Establish a load forecasting model;

[0009] Acquire real room temperature, set room temperature, and weather outside temperature data;

[0010] Calculate the compensation for outside temperature based on the actual room temperature, the set room temperature, and the weather outside temperature;

[0011]

[0012] Where To is the compensated outside temperature, Tair is the ambient outside temperature, Tr is the actual room temperature, and Ts is the set room temperature. The compensation coefficient;

[0013] Replace the weather outside temperature data with compensated outside temperature data and substitute it into the load forecasting model to obtain the load forecasting results.

[0014] By adopting the above technical solution, the temperature difference between the actual room temperature and the set room temperature can reflect the actual effect of indoor temperature on indoor heating in the heating system. This provides objective feedback on the indoor heating quality after adjusting the prediction results of the load prediction model trained entirely on historical data. Furthermore, it compensates for the actual outdoor temperature data based on the temperature difference between the actual and set room temperatures and the corresponding compensation coefficient. The outdoor temperature data is replaced with compensated outdoor temperature data and substituted into the load prediction model, enabling the model to make predictions based on the compensated outdoor temperature data. This reduces the temperature difference between the actual and set room temperatures, achieves the correction of the load prediction model's prediction results, and standardizes the correction process, thereby improving the prediction accuracy of the load prediction model and the heating quality of the heating system.

[0015] In one specific implementation scheme, replacing the weather outside temperature data with compensated outside temperature data and substituting it into the load forecasting model specifically includes:

[0016] Replace the past x hours' weather outside temperature Tairx_ with the past x hours' compensated outside temperature Tox_ and substitute it into the load forecasting model;

[0017] Replace the current weather outside temperature Tair_ with the current compensated outside temperature To_ and substitute it into the load forecasting model;

[0018] Replace the weather outside temperature Tair_y for the next y hours with the compensated outside temperature To_y for the next y hours and substitute it into the load forecasting model;

[0019] Where x and y are both hours.

[0020] By adopting the above technical solution, when constructing the input feature set of the load forecasting model, it is necessary to consider the impact of the current weather temperature as well as the weather temperature over the past and future periods on the load forecasting results. Therefore, when compensating for the weather temperature, it is also necessary to compensate for the weather temperature over the past x hours, the current weather temperature, and the weather temperature over the future y hours. This makes the compensation effect of weather temperature based on room temperature feedback more comprehensive and effectively improves the accuracy of the load forecasting results.

[0021] In one specific implementation, replacing the future y-hour weather outside temperature Tair_y with the future y-hour compensated outside temperature To_y and substituting it into the load forecasting model specifically includes:

[0022]

[0023] Where Tr_ represents the current actual room temperature, and Ts_ represents the current set room temperature. For compensation coefficient, Let be the average hourly rate of change in room temperature during the i-th hour.

[0024] Because the load forecast results are adjusted based on current room temperature feedback, the actual room temperature will become increasingly closer to the set room temperature in the future. This means that calculating the compensated external temperature for the future based on the temperature difference between the current actual and set room temperatures will introduce some deviation. By employing the above technical solution, the cumulative value of room temperature change over the next few hours is calculated based on the average hourly rate of change. This yields the predicted temperature difference between the actual and set room temperatures in the next few hours. Then, the predicted compensated external temperature for the next few hours is calculated based on this difference, making the predicted compensated external temperature data for the next few hours closer to the actual result and improving data accuracy.

[0025] In a specific feasible implementation, the following steps are also included:

[0026] Obtain historical time-series data of actual room temperature and calculate the average hourly rate of change of room temperature R:

[0027]

[0028] in, Let m be the actual room temperature at hour j in the historical time series data of actual room temperature, m be the total number of historical time series data of actual room temperature, and m1 be the actual room temperature in the historical time series data of actual room temperature. The total number, m2 is the actual historical time series data of room temperature. The total number, R1 is the average hourly rate of change of room temperature during the heating process, and R2 is the average hourly rate of change of room temperature during the cooling process;

[0029] The step of replacing the future y-hour weather outside temperature Tair_y with the future y-hour compensated outside temperature To_y and substituting it into the load forecasting model also includes:

[0030] when hour, ;otherwise, .

[0031] By adopting the above technical solution, when heating indoors, the heat transfer methods are different during the heating and cooling processes, resulting in significant differences in the rate of temperature change during these processes. Therefore, the rate of room temperature change in historical time series data is classified, and the average hourly rate of room temperature change during the heating process and the average hourly rate of room temperature change during the cooling process are calculated separately. When calculating the compensated external temperature for future times, different average hourly rates of room temperature change can be selected based on different temperature change trends, making the calculation results of the compensated external temperature for future times more accurate.

[0032] In one specific feasible implementation, the compensation coefficient Determined according to the following formula:

[0033]

[0034] Where Kw is the heat transfer coefficient of the building envelope, Fw is the heat transfer area of ​​the building envelope, G is the instantaneous flow rate of the secondary pipe network, c is the specific heat capacity of water, Tg_2 is the supply water temperature of the secondary pipe network, Th_2 is the return water temperature of the secondary pipe network, Tr is the actual room temperature, and Tpj_2 is the average temperature of the secondary pipe network. ;

[0035] or

[0036]

[0037] Where Tr is the actual room temperature, Tair is the ambient outside temperature, and Tpj_2 is the average temperature of the secondary pipe network. Tg_2 is the water supply temperature of the secondary pipeline network, and Th_2 is the water return temperature of the secondary pipeline network.

[0038] By adopting the above technical solution, the compensation coefficient can be calculated based on the real-time operating data and indoor building design data collected by the heat exchange station, without the need to collect new data or perform complex calculations on the heat exchange process, thus ensuring the practical operability of the load prediction and control process based on room temperature feedback.

[0039] In one specific implementation scheme, the weather temperature data in the input feature set of the load forecasting model includes: weather temperature in the past two hours, weather temperature in the past one hour, current weather temperature, weather temperature in the next one hour, weather temperature in the next two hours, weather temperature in the next three hours, weather temperature in the next four hours, weather temperature in the next five hours, and weather temperature in the next six hours.

[0040] By adopting the above technical solution, and considering the impact of weather outside temperature data at different times on load forecasting results, the data of weather outside temperature in the past two hours, past one hour, current time, future one hour, future two hours, future three hours, future four hours, future five hours, and future six hours are all used as input data for the load forecasting model, thereby improving the accuracy of the instantaneous heat forecasting model.

[0041] In a specific feasible implementation, the establishment of the load forecasting model specifically includes the following steps:

[0042] Acquire historical outdoor meteorological data and unit operation data of the heat exchange station;

[0043] Preprocessing of outdoor meteorological data and unit operation data;

[0044] Preprocessed outdoor meteorological data and unit operation data are used as input data, and heat load and secondary grid heating are used as output prediction results. A neural network method is used to train and establish a load prediction model.

[0045] In a specific feasible implementation, the establishment of the load forecasting model specifically includes the following steps:

[0046] Acquire historical outdoor meteorological data and unit operation data of the heat exchange station;

[0047] Preprocessing of outdoor meteorological data and unit operation data;

[0048] Preprocessed outdoor meteorological data was used as input data, and heat load was used as the output prediction result. A neural network method was used to train and establish a load prediction model.

[0049] Using pre-processed unit operating data and heat load as input data, and secondary network heating as output prediction results, a neural network method is used to train and establish a secondary network heating prediction model.

[0050] Secondly, this application provides a heating load prediction and control system based on room temperature feedback, which adopts the following technical solution:

[0051] A heating load prediction and control system based on room temperature feedback includes:

[0052] Model building module: Builds the load forecasting model;

[0053] Data acquisition module: Acquires real room temperature, set room temperature, and ambient outdoor temperature data;

[0054] Compensated external temperature calculation module: Calculates compensated external temperature based on actual room temperature, set room temperature, and weather external temperature;

[0055]

[0056] Where To is the compensated outside temperature, Tair is the ambient outside temperature, Tr is the actual room temperature, and Ts is the set room temperature. The compensation coefficient;

[0057] Room temperature feedback control module: Replaces the weather outside temperature data with the compensated outside temperature data and substitutes it into the load prediction model to obtain the load prediction results.

[0058] Thirdly, this application provides a computer-readable storage medium, which adopts the following technical solution:

[0059] A computer-readable storage medium is characterized in that it stores a computer program that can be loaded by a processor and executed as described above in the heating load prediction and control method based on room temperature feedback.

[0060] In summary, this application includes at least one of the following beneficial technical effects:

[0061] Based on the temperature difference between the actual room temperature and the set room temperature that can reflect the heating effect, as well as the compensation coefficient, the compensated outside temperature after compensating for the outside temperature is calculated, and it is used as the replacement data of the outside temperature in the load forecasting process. This realizes the correction and standardization of the forecasting results of the load forecasting model, and improves the forecasting accuracy of the load forecasting model and the heating quality of the heating system.

[0062] By compensating for past, current, and future outdoor temperatures respectively, the compensation effect of outdoor temperature based on room temperature feedback is made more comprehensive, effectively improving the accuracy of load forecasting results.

[0063] The cumulative value of room temperature change in the next few hours is calculated based on the average hourly rate of change of room temperature during the heating and cooling processes, and the temperature difference between the actual room temperature and the set room temperature in the next few hours is obtained, making the calculation results of the compensated external temperature in the future more accurate. Attached Figure Description

[0064] Figure 1 This is a schematic flowchart of the heating load prediction and control method based on room temperature feedback in an embodiment of this application. Detailed Implementation

[0065] The following is in conjunction with the appendix Figure 1 This application will be described in further detail. Example 1:

[0066] This application discloses a heating load prediction and control method based on room temperature feedback.

[0067] Reference Figure 1 The heating load prediction and control method based on room temperature feedback includes the following steps:

[0068] S1: Establish a load forecasting model, specifically including:

[0069] S101: Obtain historical outdoor meteorological data, unit operation data, heat load data of the heat exchange station, and secondary network heating data for the heat exchange station.

[0070] Specifically, the outdoor meteorological data of the heat exchange station includes real-time light intensity, real-time outdoor wind speed, real-time relative humidity, outdoor temperature in the past two hours, outdoor temperature in the past one hour, current outdoor temperature, outdoor temperature in the next one hour, outdoor temperature in the next two hours, outdoor temperature in the next three hours, outdoor temperature in the next four hours, outdoor temperature in the next five hours, and outdoor temperature in the next six hours.

[0071] The unit operating data includes the primary network supply water temperature, primary network return water temperature, primary network supply water pressure, primary network return water pressure, secondary network supply water pressure, secondary network return water pressure, circulating pump frequency, and instantaneous flow rate of the secondary network. For heat exchange station units with distributed pump systems, the data also includes the distributed pump frequency; for heat exchange station units with electrically controlled valve systems, the data also includes the electrically controlled valve opening degree.

[0072] S102: Preprocess outdoor meteorological data, unit operation data, heat exchange station heat load data, and secondary network heating data.

[0073] Specifically, the process involves removing duplicate data from outdoor meteorological data, unit operation data, heat exchange station heat load data, and secondary network heating data; using the upper and lower limits of the heating industry for each data set to eliminate outliers; using three times the standard deviation of each data sequence to eliminate outliers; finally, using unsupervised learning K-means clustering to eliminate outliers; and using interpolation methods to fill in missing values ​​and converting the data into a unified format.

[0074] S103: Using preprocessed outdoor meteorological data and unit operation data as input data, and heat load data and secondary network heating data as output prediction results, a load prediction model is trained and established using a neural network method.

[0075] Specifically, the preprocessed outdoor meteorological data, unit operation data, heat exchange station heat load data, and secondary network heating data are divided into training set, validation set, and test set in a ratio of 7:2:1.

[0076] The input feature set is constructed using the historical time series of preprocessed outdoor meteorological data and unit operation data, and the corresponding historical time series of preprocessed heat load data and secondary network heating data are used as the output prediction results. A BP neural network is used to train and establish a load prediction model with a learning rate of 0.001.

[0077] S2: Obtain real room temperature, set room temperature, and weather outside temperature data.

[0078] Specifically, the historical time series of outdoor temperature data of the heat exchange station is obtained, including the outdoor temperature of the past two hours, the outdoor temperature of the past one hour, the current outdoor temperature, the outdoor temperature of the next one hour, the outdoor temperature of the next two hours, the outdoor temperature of the next three hours, the outdoor temperature of the next four hours, the outdoor temperature of the next five hours, and the outdoor temperature of the next six hours from the outdoor meteorological data of the heat exchange station obtained in step S101 above.

[0079] Furthermore, based on the historical time series of outdoor temperature data at the heat exchange station, the corresponding historical time series of real room temperature data and set room temperature data are obtained.

[0080] The actual room temperature is the average of multiple actual room temperature measurements taken at different locations or by multiple measuring instruments at that moment. The set room temperature is the target value set by the user for the indoor temperature at that moment. If the set room temperature for a certain moment is missing in the historical time series, the median of multiple actual room temperature measurements at that moment is taken as the set room temperature.

[0081] S3: Calculates and compensates for the outside temperature based on the actual room temperature, the set room temperature, and the weather outside temperature.

[0082] Specifically, the compensation for external temperature is determined according to the following formula:

[0083]

[0084] Where To is the compensated outside temperature, Tair is the ambient outside temperature, Tr is the actual room temperature, and Ts is the set room temperature. This is the compensation coefficient.

[0085] Furthermore, the compensation coefficient Determined according to the following formula:

[0086]

[0087] Where Kw is the heat transfer coefficient of the building envelope, Fw is the heat transfer area of ​​the building envelope, G is the instantaneous flow rate of the secondary pipe network, c is the specific heat capacity of water, Tg_2 is the supply water temperature of the secondary pipe network, Th_2 is the return water temperature of the secondary pipe network, Tr is the actual room temperature, and Tpj_2 is the average temperature of the secondary pipe network. ;

[0088] Or, compensation coefficient Determined according to the following formula:

[0089]

[0090] Where Tr is the actual room temperature, Tair is the ambient outside temperature, and Tpj_2 is the average temperature of the secondary pipe network. Tg_2 is the water supply temperature of the secondary pipeline network, and Th_2 is the water return temperature of the secondary pipeline network.

[0091] S4: Replace the weather outside temperature data with the compensated outside temperature data and substitute it into the load forecasting model to obtain the load forecasting results.

[0092] Since the external temperature data in the input features of the load forecasting model includes past, current, and future external temperatures, the external temperature data is replaced with compensated external temperature data and substituted into the load forecasting model. Specifically, this includes:

[0093] Replace the past x hours' weather outside temperature Tairx_ with the past x hours' compensated outside temperature Tox_ and substitute it into the load forecasting model;

[0094] Replace the current weather outside temperature Tair_ with the current compensated outside temperature To_ and substitute it into the load forecasting model;

[0095] Replace the weather outside temperature Tair_y for the next y hours with the compensated outside temperature To_y for the next y hours and substitute it into the load forecasting model;

[0096] Where x and y are both hours.

[0097] In this application, the weather outside temperature data in the input features of the load forecasting model includes the weather outside temperature of the past two hours, the weather outside temperature of the past one hour, the current weather outside temperature, the weather outside temperature of the next one hour, the weather outside temperature of the next two hours, the weather outside temperature of the next three hours, the weather outside temperature of the next four hours, the weather outside temperature of the next five hours, and the weather outside temperature of the next six hours. The weather outside temperature data is replaced with compensated outside temperature data and substituted into the load forecasting model, specifically including:

[0098] Replace the past two hours' weather outside temperature Tair2_ with the past two hours' compensated outside temperature To2_ and substitute it into the load forecasting model;

[0099] Replace the past hour's weather outside temperature Tair1_ with the past hour's compensated outside temperature To1_ and substitute it into the load forecasting model;

[0100] Replace the current weather outside temperature Tair_ with the current compensated outside temperature To_ and substitute it into the load forecasting model;

[0101] Replace the outdoor temperature Tair_1 for the next hour with the compensated outdoor temperature To_1 for the next hour and substitute it into the load forecasting model;

[0102] Replace the outdoor temperature Tair_2 for the next two hours with the compensated outdoor temperature To_2 for the next two hours and substitute it into the load forecasting model;

[0103] Replace the outdoor temperature Tair_3 for the next three hours with the compensated outdoor temperature To_3 for the next three hours and substitute it into the load forecasting model;

[0104] Replace the outdoor temperature Tair_4 for the next four hours with the compensated outdoor temperature To_4 for the next four hours and substitute it into the load forecasting model;

[0105] Replace the outdoor temperature Tair_5 for the next five hours with the compensated outdoor temperature To_5 for the next five hours and substitute it into the load forecasting model;

[0106] Replace the outdoor temperature Tair_6 for the next six hours with the compensated outdoor temperature To_6 for the next six hours and substitute it into the load forecasting model;

[0107] Furthermore, in step S3 above, the compensated external temperature Tox_ for the past x hours is determined according to the following formula:

[0108]

[0109] Where Tairx_ represents the outside temperature over the past x hours, Tr_ represents the current actual room temperature, and Ts_ represents the current set room temperature. This is the compensation coefficient.

[0110] In step S3 above, the current compensated external temperature To is determined according to the following formula:

[0111]

[0112] Where Tair_ represents the current outside temperature, Tr_ represents the current actual room temperature, and Ts_ represents the current set room temperature. This is the compensation coefficient.

[0113] Because of the current adjustment process based on room temperature feedback to the load prediction results, the actual room temperature will become increasingly closer to the set room temperature in the future. This means that calculating the compensated external temperature for future times based on the temperature difference between the current actual room temperature and the set room temperature will result in a certain deviation. In step S3 above, the compensated external temperature To_y for the next y hours is determined according to the following formula:

[0114]

[0115] Where Tr_ represents the current actual room temperature, and Ts_ represents the current set room temperature. For compensation coefficient, Let be the average hourly rate of change in room temperature during the i-th hour.

[0116] Furthermore, step S3 above also includes acquiring historical time-series data of the actual room temperature and calculating the average hourly rate of change of room temperature R:

[0117]

[0118] in, Let m be the actual room temperature at hour j in the historical time series data of actual room temperature, m be the total number of historical time series data of actual room temperature, and m1 be the actual room temperature in the historical time series data of actual room temperature. The total number, m2 is the actual historical time series data of room temperature. The total number of R1 and R2 is the average hourly rate of change of room temperature during the heating process and the average hourly rate of change of room temperature during the cooling process.

[0119] In the process of calculating the compensated external temperature To_y for the next y hours, when hour, ;otherwise, .

[0120] In step S4, the compensated external temperature To2_ for the past two hours, To1_ for the past one hour, the current compensated external temperature To_, the compensated external temperature To_ for the next one hour, To_2 for the next two hours, To_3 for the next three hours, To_4 for the next four hours, To_5 for the next five hours, and To_6 for the next six hours, obtained from step S3 above, are all substituted into the load prediction model to obtain the load prediction value and the secondary network heating prediction value based on the temperature difference feedback between the actual room temperature and the set room temperature.

[0121] The implementation principle of a heating load prediction and control method based on room temperature feedback in this application is as follows: the temperature difference between the actual room temperature and the set room temperature reflects the actual effect of indoor temperature on indoor heating in the heating system. The actual weather outside temperature data is compensated based on the temperature difference between the actual room temperature and the set room temperature and the corresponding compensation coefficient. The weather outside temperature data is replaced with compensated outside temperature data and substituted into the load prediction model. The changes in the temperature difference between the actual room temperature and the set room temperature in the future are taken into account, thereby realizing the correction of the prediction results of the load prediction model and the standardization of the correction process, improving the prediction accuracy of the load prediction model and the heating quality of the heating system.

[0122] Figure 1 This is a flowchart illustrating the heating load prediction and control method based on room temperature feedback, as described in an embodiment of this application. It should be understood that, although... Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows; unless explicitly stated otherwise, there is no strict order requirement for the execution of these steps, and they can be executed in other orders; and Figure 1At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0123] This application also discloses a heating load prediction and control system based on room temperature feedback.

[0124] A heating load prediction and control system based on room temperature feedback includes:

[0125] Model building module: Builds load forecasting models.

[0126] Acquire historical outdoor meteorological data, unit operation data, heat load data of the heat exchange station, and secondary network heating data for the heat exchange station.

[0127] Specifically, the outdoor meteorological data of the heat exchange station includes real-time light intensity, real-time outdoor wind speed, real-time relative humidity, outdoor temperature in the past two hours, outdoor temperature in the past one hour, current outdoor temperature, outdoor temperature in the next one hour, outdoor temperature in the next two hours, outdoor temperature in the next three hours, outdoor temperature in the next four hours, outdoor temperature in the next five hours, and outdoor temperature in the next six hours.

[0128] The unit operating data includes the primary network supply water temperature, primary network return water temperature, primary network supply water pressure, primary network return water pressure, secondary network supply water pressure, secondary network return water pressure, circulating pump frequency, and instantaneous flow rate of the secondary network. For heat exchange station units with distributed pump systems, the data also includes the distributed pump frequency; for heat exchange station units with electrically controlled valve systems, the data also includes the electrically controlled valve opening degree.

[0129] Preprocessing is performed on outdoor meteorological data, unit operation data, heat exchange station heat load data, and secondary network heating data.

[0130] Specifically, the process involves removing duplicate data from outdoor meteorological data, unit operation data, heat exchange station heat load data, and secondary network heating data; using the upper and lower limits of the heating industry for each data set to eliminate outliers; using three times the standard deviation of each data sequence to eliminate outliers; finally, using unsupervised learning K-means clustering to eliminate outliers; and using interpolation methods to fill in missing values ​​and converting the data into a unified format.

[0131] The preprocessed outdoor meteorological data and unit operation data are used as input data, and the heat load data and secondary network heating data are used as output prediction results. The neural network method is used to train and establish a load prediction model.

[0132] Specifically, the preprocessed outdoor meteorological data, unit operation data, heat exchange station heat load data, and secondary network heating data are divided into training set, validation set, and test set in a ratio of 7:2:1.

[0133] The input feature set is constructed using the historical time series of preprocessed outdoor meteorological data and unit operation data, and the corresponding historical time series of preprocessed heat load data and secondary network heating data are used as the output prediction results. A BP neural network is used to train and establish a load prediction model with a learning rate of 0.001.

[0134] Data acquisition module: Acquires real room temperature, set room temperature, and weather outside temperature data.

[0135] Specifically, the system acquires historical time series weather temperature data for the heat exchange station, including the weather temperature for the past two hours, the weather temperature for the past one hour, the current weather temperature, the weather temperature for the next one hour, the weather temperature for the next two hours, the weather temperature for the next three hours, the weather temperature for the next four hours, the weather temperature for the next five hours, and the weather temperature for the next six hours from the outdoor meteorological data of the heat exchange station obtained in the model building module.

[0136] Furthermore, based on the historical time series of outdoor temperature data at the heat exchange station, the corresponding historical time series of real room temperature data and set room temperature data are obtained.

[0137] The actual room temperature is the average of multiple actual room temperature measurements taken at different locations or by multiple measuring instruments at that moment. The set room temperature is the target value set by the user for the indoor temperature at that moment. If the set room temperature for a certain moment is missing in the historical time series, the median of multiple actual room temperature measurements at that moment is taken as the set room temperature.

[0138] Compensation for external temperature calculation module: Calculates compensation for external temperature based on the actual room temperature, the set room temperature, and the weather external temperature.

[0139] Specifically, the compensation for external temperature is determined according to the following formula:

[0140]

[0141] Where To is the compensated outside temperature, Tair is the ambient outside temperature, Tr is the actual room temperature, and Ts is the set room temperature. This is the compensation coefficient.

[0142] Furthermore, the compensation coefficient Determined according to the following formula:

[0143]

[0144] Where Kw is the heat transfer coefficient of the building envelope, Fw is the heat transfer area of ​​the building envelope, G is the instantaneous flow rate of the secondary pipe network, c is the specific heat capacity of water, Tg_2 is the supply water temperature of the secondary pipe network, Th_2 is the return water temperature of the secondary pipe network, Tr is the actual room temperature, and Tpj_2 is the average temperature of the secondary pipe network. ;

[0145] Or, compensation coefficient Determined according to the following formula:

[0146]

[0147] Where Tr is the actual room temperature, Tair is the ambient outside temperature, and Tpj_2 is the average temperature of the secondary pipe network. Tg_2 is the water supply temperature of the secondary pipeline network, and Th_2 is the water return temperature of the secondary pipeline network.

[0148] Room temperature feedback control module: Replaces the weather outside temperature data with the compensated outside temperature data and substitutes it into the load prediction model to obtain the load prediction results.

[0149] Since the external temperature data in the input features of the load forecasting model includes past, current, and future external temperatures, the external temperature data is replaced with compensated external temperature data and substituted into the load forecasting model. Specifically, this includes:

[0150] Replace the past x hours' weather outside temperature Tairx_ with the past x hours' compensated outside temperature Tox_ and substitute it into the load forecasting model;

[0151] Replace the current weather outside temperature Tair_ with the current compensated outside temperature To_ and substitute it into the load forecasting model;

[0152] Replace the weather outside temperature Tair_y for the next y hours with the compensated outside temperature To_y for the next y hours and substitute it into the load forecasting model;

[0153] Where x and y are both hours.

[0154] In this application, the weather outside temperature data in the input features of the load forecasting model includes the weather outside temperature of the past two hours, the weather outside temperature of the past one hour, the current weather outside temperature, the weather outside temperature of the next one hour, the weather outside temperature of the next two hours, the weather outside temperature of the next three hours, the weather outside temperature of the next four hours, the weather outside temperature of the next five hours, and the weather outside temperature of the next six hours. The weather outside temperature data is replaced with compensated outside temperature data and substituted into the load forecasting model, specifically including:

[0155] Replace the past two hours' weather outside temperature Tair2_ with the past two hours' compensated outside temperature To2_ and substitute it into the load forecasting model;

[0156] Replace the past hour's weather outside temperature Tair1_ with the past hour's compensated outside temperature To1_ and substitute it into the load forecasting model;

[0157] Replace the current weather outside temperature Tair_ with the current compensated outside temperature To_ and substitute it into the load forecasting model;

[0158] Replace the outdoor temperature Tair_1 for the next hour with the compensated outdoor temperature To_1 for the next hour and substitute it into the load forecasting model;

[0159] Replace the outdoor temperature Tair_2 for the next two hours with the compensated outdoor temperature To_2 for the next two hours and substitute it into the load forecasting model;

[0160] Replace the outdoor temperature Tair_3 for the next three hours with the compensated outdoor temperature To_3 for the next three hours and substitute it into the load forecasting model;

[0161] Replace the outdoor temperature Tair_4 for the next four hours with the compensated outdoor temperature To_4 for the next four hours and substitute it into the load forecasting model;

[0162] Replace the outdoor temperature Tair_5 for the next five hours with the compensated outdoor temperature To_5 for the next five hours and substitute it into the load forecasting model;

[0163] Replace the outdoor temperature Tair_6 for the next six hours with the compensated outdoor temperature To_6 for the next six hours and substitute it into the load forecasting model;

[0164] Furthermore, in the external temperature compensation calculation module, the external temperature Tox_ for the past x hours is determined according to the following formula:

[0165]

[0166] Where Tairx_ represents the outside temperature over the past x hours, Tr_ represents the current actual room temperature, and Ts_ represents the current set room temperature. This is the compensation coefficient.

[0167] In the external temperature compensation calculation module, the current external temperature To is determined according to the following formula:

[0168]

[0169] Where Tair_ represents the current outside temperature, Tr_ represents the current actual room temperature, and Ts_ represents the current set room temperature. This is the compensation coefficient.

[0170] Because of the current adjustment process based on room temperature feedback to the load forecast results, the actual room temperature will become increasingly closer to the set room temperature in the future. This means that calculating the compensated external temperature for future moments based on the temperature difference between the current actual room temperature and the set room temperature will introduce some deviation. In the compensated external temperature calculation module, the compensated external temperature To_y for the next y hours is determined according to the following formula:

[0171]

[0172] Where Tr_ represents the current actual room temperature, and Ts_ represents the current set room temperature. For compensation coefficient, Let be the average hourly rate of change in room temperature during the i-th hour.

[0173] Furthermore, the external temperature compensation calculation module also includes acquiring historical time-series data of the actual room temperature and calculating the average hourly room temperature change rate R:

[0174]

[0175] in, Let m be the actual room temperature at hour j in the historical time series data of actual room temperature, m be the total number of historical time series data of actual room temperature, and m1 be the actual room temperature in the historical time series data of actual room temperature. The total number, m2 is the actual historical time series data of room temperature. The total number of R1 and R2 is the average hourly rate of change of room temperature during the heating process and the average hourly rate of change of room temperature during the cooling process.

[0176] In the process of calculating the compensated external temperature To_y for the next y hours, when hour, ;otherwise, .

[0177] In the room temperature feedback control module, the compensated external temperature To2_ of the past two hours, the compensated external temperature To1_ of the past one hour, the current compensated external temperature To_, the compensated external temperature To_1 of the next one hour, the compensated external temperature To_2 of the next two hours, the compensated external temperature To_3 of the next three hours, the compensated external temperature To_4 of the next four hours, the compensated external temperature To_5 of the next five hours, and the compensated external temperature To_6 of the next six hours, obtained from the compensated external temperature calculation module, are all substituted into the load prediction model. The resulting load prediction value and the secondary network heating prediction value are then adjusted based on the temperature difference feedback between the actual room temperature and the set room temperature.

[0178] This application also discloses a computer-readable storage medium.

[0179] Specifically, the computer-readable storage medium stores a computer program that can be loaded by a processor and executed, such as the heating load prediction and control method based on room temperature feedback described above. The computer-readable storage medium includes, for example, 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. Example 2:

[0180] The difference between this embodiment and Embodiment 1 is that the heating load prediction and control method based on room temperature feedback includes step S1: establishing a load prediction model, specifically including:

[0181] S101: Obtain historical outdoor meteorological data, unit operation data, heat load data of the heat exchange station, and secondary network heating data for the heat exchange station.

[0182] S102: Preprocess outdoor meteorological data, unit operation data, heat exchange station heat load data, and secondary network heating data.

[0183] S103: Using preprocessed outdoor meteorological data as input data and heat load as output prediction result, a load prediction model is trained and established using a neural network method.

[0184] Specifically, the outdoor meteorological data of the heat exchange station includes real-time light intensity, real-time outdoor wind speed, real-time relative humidity, outdoor temperature in the past two hours, outdoor temperature in the past one hour, current outdoor temperature, outdoor temperature in the next one hour, outdoor temperature in the next two hours, outdoor temperature in the next three hours, outdoor temperature in the next four hours, outdoor temperature in the next five hours, and outdoor temperature in the next six hours.

[0185] The preprocessed outdoor meteorological data and heat exchange station heat load data were divided into training, validation, and test sets in a 7:2:1 ratio. An input feature set was constructed using the historical time series of the preprocessed outdoor meteorological data, and the corresponding preprocessed heat load data was used as the output prediction result. A BP neural network was used to train and establish a load prediction model with a learning rate of 0.001.

[0186] S104: Using pre-processed unit operating data and heat load as input data, and secondary network heating as output prediction results, a neural network method is used to train and establish a secondary network heating prediction model.

[0187] Specifically, the unit operating data includes the primary network supply water temperature, the primary network return water temperature, the primary network supply water pressure, the primary network return water pressure, the secondary network supply water pressure, the secondary network return water pressure, the circulating pump frequency, and the instantaneous flow rate of the secondary network. For heat exchange station units with distributed pump systems, this also includes the distributed pump frequency; for heat exchange station units with electrically controlled valve systems, this also includes the electrically controlled valve opening degree.

[0188] The preprocessed unit operation data and secondary network heating data were divided into training, validation, and test sets in a 7:2:1 ratio. An input feature set was constructed using the historical time series of the preprocessed unit operation data, and the corresponding historical time series of the preprocessed secondary network heating data was used as the output prediction result. A BP neural network was used to train and establish a load prediction model with a learning rate of 0.001.

[0189] This specific embodiment is merely an explanation of the present invention and is not intended to limit the invention. After reading this specification, those skilled in the art can make modifications to this embodiment without contributing any inventive step, but such modifications are protected by patent law as long as they are within the scope of the claims of the present invention.

Claims

1. A method for predicting and controlling heating load based on room temperature feedback, characterized in that, Includes the following steps: Establish a load forecasting model; Acquire real room temperature, set room temperature, and weather outside temperature data; Calculate the compensation for outside temperature based on the actual room temperature, the set room temperature, and the weather outside temperature; ， Where To is the compensated outside temperature, Tair is the ambient outside temperature, Tr is the actual room temperature, and Ts is the set room temperature. This is the compensation coefficient; Replace the weather outside temperature data with the compensated outside temperature data and substitute it into the load forecasting model to obtain the load forecasting results; The process of replacing weather outside temperature data with compensated outside temperature data and substituting it into the load forecasting model specifically includes: Replace the past x hours' weather outside temperature Tairx_ with the past x hours' compensated outside temperature Tox_ and substitute it into the load forecasting model; Replace the current weather outside temperature Tair_ with the current compensated outside temperature To_ and substitute it into the load forecasting model; Replace the weather outside temperature Tair_y for the next y hours with the compensated outside temperature To_y for the next y hours and substitute it into the load forecasting model; Where x and y are both hours; The step of replacing the future y-hour weather outside temperature Tair_y with the future y-hour compensated outside temperature To_y and substituting it into the load forecasting model specifically includes: ， Where Tr_ represents the current actual room temperature, and Ts_ represents the current set room temperature. For compensation coefficient, Let i be the average hourly rate of change of room temperature in the i-th hour; The establishment of the load forecasting model specifically includes the following steps: Acquire historical outdoor meteorological data and unit operation data of the heat exchange station; Preprocessing of outdoor meteorological data and unit operation data; The preprocessed outdoor meteorological data and unit operation data are used as input data, and the heat load and secondary network heating are used as output prediction results. The neural network method is used to train and establish a load prediction model. Alternatively, the establishment of the load forecasting model specifically includes the following steps: Acquire historical outdoor meteorological data and unit operation data of the heat exchange station; Preprocessing of outdoor meteorological data and unit operation data; Preprocessed outdoor meteorological data was used as input data, and heat load was used as the output prediction result. A neural network method was used to train and establish a load prediction model. Using pre-processed unit operating data and heat load as input data, and secondary network heating as output prediction results, a neural network method is used to train and establish a secondary network heating prediction model.

2. The heating load prediction and control method based on room temperature feedback according to claim 1, characterized in that, It also includes the following steps: Obtain historical time-series data of actual room temperature and calculate the average hourly rate of change of room temperature R: ， in, Let m be the actual room temperature at hour j in the historical time series data of actual room temperature, m be the total number of historical time series data of actual room temperature, and m1 be the actual room temperature in the historical time series data of actual room temperature. The total number, m2 is the actual historical time series data of room temperature. The total number, R1 is the average hourly rate of change of room temperature during the heating process, and R2 is the average hourly rate of change of room temperature during the cooling process; The step of replacing the future y-hour weather outside temperature Tair_y with the future y-hour compensated outside temperature To_y and substituting it into the load forecasting model also includes: when hour, ;otherwise, .

3. The heating load prediction and control method based on room temperature feedback according to claim 1, characterized in that, The compensation coefficient Determined according to the following formula: ， Where Kw is the heat transfer coefficient of the building envelope, Fw is the heat transfer area of ​​the building envelope, G is the instantaneous flow rate of the secondary pipe network, c is the specific heat capacity of water, Tg_2 is the supply water temperature of the secondary pipe network, Th_2 is the return water temperature of the secondary pipe network, Tr is the actual room temperature, and Tpj_2 is the average temperature of the secondary pipe network. ; or ， Where Tr is the actual room temperature, Tair is the ambient outside temperature, and Tpj_2 is the average temperature of the secondary pipe network. Tg_2 is the water supply temperature of the secondary pipeline network, and Th_2 is the water return temperature of the secondary pipeline network.

4. The heating load prediction and control method based on room temperature feedback according to claim 1, characterized in that, The external temperature data in the input feature set of the load forecasting model includes: external temperature of the past two hours, external temperature of the past one hour, current external temperature, external temperature of the next one hour, external temperature of the next two hours, external temperature of the next three hours, external temperature of the next four hours, external temperature of the next five hours, and external temperature of the next six hours.

5. A heating load prediction and control system based on room temperature feedback, characterized in that, The method for performing heating load prediction and control based on room temperature feedback as described in any one of claims 1 to 4 includes: Model building module: Builds the load forecasting model; Data acquisition module: Acquires real room temperature, set room temperature, and ambient outdoor temperature data; Compensated external temperature calculation module: Calculates compensated external temperature based on actual room temperature, set room temperature, and weather external temperature; ， Where To is the compensated outside temperature, Tair is the ambient outside temperature, Tr is the actual room temperature, and Ts is the set room temperature. This is the compensation coefficient; Room temperature feedback control module: Replaces the weather outside temperature data with the compensated outside temperature data and substitutes it into the load prediction model to obtain the load prediction results.

6. A computer-readable storage medium, characterized in that, The computer program is stored and can be loaded by a processor and executed as described in any one of claims 1 to 4, which is a method for predicting and controlling heating load based on room temperature feedback.

Images