A method for constructing a return water temperature prediction model for ground radiation terminal regulation

Abstract

The patent discloses a return water temperature prediction model construction method for ground radiation terminal regulation. By using users with room temperature monitoring, a return water temperature basic prediction model of various users is established. The time variable and indoor temperature are included in the input characteristics, fully considering the influence of personnel behavior on the prediction result, and effectively improving the prediction accuracy of the return water temperature. For users without room temperature monitoring, the users are classified from the perspective of consistent heat load demand, and a two-stage prediction model is designed. The return water temperature basic prediction model is established by using the room temperature monitoring households. Considering the difference of ground radiation heat dissipation performance and the difference of user behavior, the ratio and difference of daily heat consumption of the representative household and the prediction household are introduced into the input characteristics, and a return water temperature difference prediction model of the two is established. The personalized prediction model of the prediction household return water temperature is obtained by adding the basic prediction model and the difference prediction model.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of heating systems, in particular to a return water temperature prediction model construction method for ground radiation terminal regulation. BACKGROUND

[0002] In current heating systems, the core strategy of room temperature regulation theoretically relies on the room temperature feedback regulation mechanism. This mechanism continuously monitors the actual indoor temperature and compares it with the preset temperature to automatically adjust the heating output and control the indoor temperature.

[0003] However, this method has the following two difficulties in practical ground radiation heating applications:

[0004] Currently, only a small number of users have configured room temperature monitoring devices, limiting the popularization and optimization potential of this control strategy.

[0005] The inherent thermal inertia characteristics of ground radiation heating systems result in slow response, and if the room temperature feedback regulation is used, overshoot may occur, resulting in large room temperature fluctuations and affecting the accuracy and stability of the control.

[0006] In modern centralized heating systems, return water temperature is a key indicator for measuring and reflecting user heating quality. User return water temperature data is more easily obtained than room temperature data, and in ground radiation heating, return water temperature feedback control has the advantages of small thermal inertia and fast response compared to room temperature feedback control. Therefore, accurate prediction of return water temperature is not only crucial for grasping the overall heating efficiency, but also directly related to improving the accuracy and efficiency of user-side room temperature regulation. By optimizing the prediction method of return water temperature, the heating strategy can be adjusted more actively and effectively.

[0007] Current return water temperature prediction is mainly at the heat exchange station level and the building unit level, assuming that the return water temperatures of users included in the heat exchange station or building are the same, i.e., based on the return water temperature consistency method for valve regulation of users, the specific method is:

[0008] At the heat exchange station level, a heat exchange station contains several units, each unit supplies heat to one or more buildings, and the return water temperature of each unit is predicted and considered to be consistent with the return water temperatures of all users corresponding to the unit.

[0009] At the building unit level, if the return water temperature of the entire building is predicted, it is assumed that the return water temperatures of all users in the building are consistent; if the building is divided into zones for prediction, it is assumed that the return water temperatures of users in different zones are consistent.

[0010] The return water temperature prediction model based on the consistent return water temperature considers that the return water temperature of each user is the same at the heat exchange station level or the building unit level, but this method cannot guarantee the heat balance between users and cannot achieve the required room temperature of each user. The ground coverage rate, floor material, ground decoration and furniture placement of the ground radiant heating system can affect the heat transfer thermal resistance and the room temperature thermal response delay time. When the indoor coverage rate is low, the large heat dissipation area can ensure that the room temperature is within the required range with low return water temperature; on the contrary, when the indoor coverage rate is high, the small heat dissipation area requires high return water temperature to ensure the room temperature. Therefore, the same return water temperature cannot make the room temperature of all users at the same level, and can also hide the real situation in the room. SUMMARY

[0011] The purpose of the present application is to overcome the above-mentioned deficiencies and provide a return water temperature prediction model construction method for ground radiant terminal regulation. The present application introduces the difference of daily heat consumption of users of the same type as an input variable, establishes a personalized return water temperature prediction model at the user level, and more accurately realizes on-demand heating.

[0012] To achieve the above purpose, the present application provides the following technical scheme: a return water temperature prediction model construction method for ground radiant terminal regulation, comprising the following steps:

[0013] S1, user classification:

[0014] According to the building thermal performance, house type, position of external envelope structure, total number of floors and floor interval of the user, the user is classified;

[0015] S2, laying monitoring network hardware facilities:

[0016] Laying user water supply temperature sensors, user return water temperature sensors and heat meters in the user's heating network;

[0017] S3, constructing representative household A and prediction household B;

[0018] Selecting users of the same category in step S1 after classification as representative household A with room temperature monitoring and prediction household B without room temperature monitoring, and laying indoor temperature sensors in the representative household A with room temperature monitoring;

[0019] S4, data collection and processing;

[0020] S5, establishment of two-stage return water temperature prediction model for each type of user;

[0021] S6, valve control;

[0022] Based on the predicted user return water temperature value, the valve of the user is controlled.

[0023] Preferably, the specific steps of step S1 are:

[0024] According to the energy-saving level, the building thermal performance is divided into five categories: non-energy-saving, one-step energy-saving, two-step energy-saving, three-step energy-saving and four-step energy-saving.

[0025] According to the energy-saving level, the user group is classified according to the house type, and the same house type user is classified according to the position of the external envelope structure. The basic classification is bottom-middle, bottom-side, middle-middle, middle-side, top-middle and top-side.

[0026] Finally, according to the cold air penetration characteristics, the total number of floors and the floor interval are classified.

[0027] Preferably, the specific operation of step S3 is:

[0028] Data collection: Collect the data of water supply temperature, return water temperature, daily heat consumption, outdoor temperature, wind speed, solar radiation and other data of each user. For representative household A, indoor temperature is also collected.

[0029] Data processing: Use statistical methods to remove and supplement abnormal data, and process the data into historical time data of the required time granularity.

[0030] Preferably, the specific operation of step S5 is:

[0031] First stage return water temperature basic prediction model: For each type of user, a return water temperature basic prediction model is established for each type of user.

[0032] The input variables are the historical time and current time of the representative household's water supply temperature, return water temperature, outdoor temperature, wind speed, solar radiation, room temperature, future time meteorological prediction parameters and indoor temperature setting parameters, and time variables.

[0033] The output variable is the next time return water temperature of representative household A.

[0034] Second stage return water temperature difference prediction model: For each type of user, a return water temperature difference prediction model is established for each type of user.

[0035] The input variables are the historical time and current time of the representative household's water supply temperature, return water temperature, outdoor temperature, wind speed, solar radiation, room temperature, daily heat consumption ratio and difference between the two users, future time meteorological prediction parameters and indoor temperature setting parameters, and time variables.

[0036] The output variable is the difference ΔTh(t+1) between the next time return water temperature of the predicted household B and the representative household A. B-A

[0037] ​The ratio QK of the representative household to the predicted household daily heat consumption and the difference AQ are calculated according to the following formula:

[0038]

[0039] AQ = Q 代表户 - Q 预测对象 ;

[0040] The first-stage return water temperature basic prediction model and the second-stage return water temperature difference prediction model are combined to obtain the predicted return water temperature Th(t+1) of the user B B ;

[0041] The calculation formula is as follows:

[0042] Th(t+1) B = Th(t+1) A + AQ(t+1) B-A .

[0043] Compared with the prior art, the present application has the following advantages:

[0044] The present application proposes a two-stage return water temperature prediction model construction method for the ground radiation heating terminal, and controls the valve through the established user individualized return water temperature prediction model to realize room temperature regulation and control. Firstly, the individualized return water temperature prediction model is developed for the user group that has deployed a room temperature monitoring system, i.e. the representative household, aiming to capture individual differences and realize accurate regulation and control. Secondly, the two-stage prediction model is designed for the users in the same user category that have not configured a room temperature monitoring system. The basic return water temperature prediction model is established by using the representative household, the ratio QK of the representative household to the predicted household daily heat consumption and the difference AQ are introduced as input variables, the difference prediction model of the return water temperature between the representative household and the predicted household is established, and finally the robust and accurate predicted household return water temperature prediction model is established to realize on-demand heating more accurately. The user individualized return water temperature prediction model is established by using a small number of representative households with room temperature monitoring, the prediction accuracy of the model is improved, and the problem that the return water temperature regulation effect of the prior art is not verified by the actual room temperature control effect is solved.

[0045] The patent considers the influence of heat dissipation terminal thermal performance, user behavior and user thermal characteristic difference on user return water temperature by integrating multiple key parameters, including outdoor weather conditions, indoor temperature, supply and return water temperature, daily heat consumption, etc., and constructs a user personalized return water temperature prediction model applicable to the ground radiation heating system with thermal inertia. Compared with the prior art, the difference of user heat load demand is fully considered by user classification, the difference of ground radiation heat dissipation performance is considered by introducing the difference index of daily heat consumption of the same type of user in the input variable, the influence of user behavior mode is considered by introducing the time variable, so as to enhance the accuracy and practicability of the model prediction, promote the delicacy of heating management and the efficiency of energy use, and improve the overall heating experience and satisfaction of users. BRIEF DESCRIPTION OF DRAWINGS

[0046] Fig. 1 It is a schematic diagram of the user ground radiation system in the application.

[0047] Fig. 2 It is a user return water temperature regulation flow chart in the application.

[0048] In the figure: 1 - user indoor temperature sensor, 2 - user water supply temperature sensor, 3 - regulating valve, 4 - heat meter, 5 - user return water temperature sensor, 6 - floor heating sub-collector, 7 - representative household A with room temperature monitoring, 8 - predicted household B without room temperature monitoring. DETAILED DESCRIPTION

[0049] The technical solutions in the embodiments of the application will be described clearly and completely in combination with the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, not all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the application.

[0050] Please refer to Figs. 1-2 The application provides a technical solution: the patent device includes four parts of detection elements, a data acquisition system, a central data processor and an execution mechanism for controlling valves. The detection elements include a user heat meter, temperature sensors installed on the user water supply and return water main, indoor temperature sensors of representative households, etc. The data acquisition system is responsible for collecting the signals output by the detection elements and converting them into digital signals for use by the central data processor. The central data processor can process, analyze and make corresponding decisions, use the return water temperature prediction model to predict the user return water temperature, and perform real-time data transmission to send signals to the execution mechanism for controlling the valves of the user. The execution mechanism for controlling the valves automatically adjusts the valve opening degree according to the instruction signals from the central data processor, thereby controlling the hot water flow through the valves.

[0051] The patent proposes a two-stage return water temperature prediction model construction method for ground radiant heating terminals. The valve is controlled through the established user individual return water temperature prediction model to realize room temperature regulation. First, for the user group that has deployed a room temperature monitoring system, i.e., the representative household, a personalized return water temperature prediction model is developed to capture individual differences and achieve precise regulation. Second, for users in the same user category who have not configured room temperature monitoring, a two-stage prediction model is designed. The representative household establishes a basic return water temperature prediction model, introduces the representative household and the predicted household daily heat consumption ratio QK and difference AQ as input variables, establishes a difference prediction model between the return water temperatures of the representative household and the predicted household, and finally establishes a robust and accurate predicted household return water temperature prediction model.

[0052] The patent includes the following steps:

[0053] S1, user classification:

[0054] S2, laying monitoring network hardware facilities:

[0055] S3, constructing representative household A and predicted household B;

[0056] S4, data collection and processing;

[0057] S5, establishment of two-stage return water temperature prediction models for various users;

[0058] S6, valve control;

[0059] The specific steps are as follows: user classification. According to the user's building thermal performance, house type, external envelope location, total number of floors, and floor interval, the user is classified. First, for the same unit building group, the building thermal performance is divided according to the energy saving grade, and the building group is divided into five categories: non-energy saving, one-step energy saving, two-step energy saving, three-step energy saving, and four-step energy saving. Then, according to the energy saving grade classification, the user group is classified according to the house type; for the same house type user, it is classified according to the external envelope location, and the basic classification is bottom middle, bottom edge, middle middle, middle edge, top middle, and top edge; finally, according to the cold air penetration characteristics, the total number of floors and the floor interval are classified.

[0060] Data collection. Collect the data of each user's water supply temperature, return water temperature, daily heat consumption, outdoor temperature, wind speed, solar radiation, etc. For room temperature monitoring users (representative households), indoor temperature data is also collected.

[0061] Data processing. Use statistical methods to remove and supplement abnormal data, and process the data into the required time granularity of historical time data.

[0062] Establishment of two-stage return water temperature prediction models for various users:

[0063] The first stage of the return water temperature basic prediction model: for each type of user with room temperature monitoring user (referred to as representative household A) to establish the return water temperature basic prediction model of each type of user, the input variables are the supply water temperature, return water temperature, outdoor temperature, wind speed, solar radiation, room temperature, future time meteorological prediction parameters and indoor temperature setting parameters and time variables of the representative household at the historical time and the current time, and the output variable is the next time return water temperature of the representative household A. Among them, the time variable includes weekdays (Monday to Friday) and weekends (Saturday and Sunday), and three stages of each day (night sleep period, daytime indoor active period, and indoor idle period due to going out to work).

[0064] The second stage of the return water temperature difference prediction model: for the non-room temperature monitoring prediction user B, a difference prediction model of the return water temperature between the prediction user B and the representative household A is established. The input variables are the supply water temperature, return water temperature, outdoor temperature, wind speed, solar radiation, room temperature, daily heat consumption ratio and difference between the two users, future time meteorological prediction parameters and indoor temperature setting parameters, and time variables of the representative household at the historical time and the current time, and the output variable is the difference ΔTh(t+1) between the next time return water temperatures of the prediction user B and the representative household A. B-A .

[0065] The daily heat consumption ratio QK and the difference ΔQ between the representative household and the prediction household. The following formula is used for calculation:

[0066]

[0067] ΔQ=Q 代表户 -Q 预测对象

[0068] The return water temperature Th(t+1) of the prediction user B is obtained by combining the first stage of the return water temperature basic prediction model and the second stage of the return water temperature difference prediction model. B The following formula is used for calculation:

[0069] Th(t+1) B =Th(t+1) A +ΔTh(t+1) B-A

[0070] Valve control. Based on the predicted return water temperature value of the user, the valve of the user is controlled to realize accurate control of the indoor temperature.

[0071] The return water temperature basic prediction model of each type of user is established by using the user with room temperature monitoring. The time variable and indoor temperature are included in the input features, the influence of personnel behavior on the prediction result is fully considered, and the prediction accuracy of the return water temperature is effectively improved.

[0072] For users without room temperature monitoring, a two-stage prediction model is designed for users from the perspective of consistent heat load demand. A return water temperature basic prediction model is established for users with room temperature monitoring. Considering the difference in ground radiation heat dissipation performance and the difference in user behavior, the ratio and difference of daily heat consumption between the representative user and the prediction user are introduced as input features, and a difference prediction model of return water temperature between the two is established. The personalized prediction model of return water temperature of the prediction user is obtained by adding the basic prediction model and the difference prediction model.

[0073] Compared with the prior art, the beneficial effects of the present application are:

[0074] The present patent proposes a two-stage return water temperature prediction model construction method for ground radiation heating terminals, and controls the valve through the established user personalized return water temperature prediction model to realize room temperature regulation. First, a personalized return water temperature prediction model is developed for the user group that has deployed a room temperature monitoring system, i.e. the representative user, aiming to capture individual differences and achieve precise regulation. Second, a two-stage prediction model is designed for users in the same user category without room temperature monitoring. A basic return water temperature prediction model is established for the representative user, and the ratio QK and difference AQ of daily heat consumption between the representative user and the prediction user are introduced as input variables to establish a difference prediction model of return water temperature between the representative user and the prediction user. Finally, a robust and accurate prediction model of return water temperature of the prediction user is established to achieve on-demand heating more accurately. By establishing a user personalized return water temperature prediction model through a small number of representative users with room temperature monitoring, the prediction accuracy of the model is improved, and the problem that the regulation effect of the return water temperature in the prior art is not verified by the actual control effect of the room temperature is solved.

[0075] The present patent integrates multiple key parameters, including outdoor weather conditions, indoor temperature, supply and return water temperature, daily heat consumption, etc., considers the influence of heat dissipation terminal thermal performance, user behavior and user-specific thermal characteristics difference on user return water temperature, and constructs a set of user personalized return water temperature prediction model suitable for ground radiation heating systems with thermal inertia. Compared with the prior art, the difference of user heat load demand is fully considered through user classification, the difference of daily heat consumption of the same type of user is considered through the introduction of the difference index in the input variable, the influence of user behavior mode is considered by introducing the time variable, thereby enhancing the accuracy and practicality of the model prediction, promoting the fineness of heating management and the efficiency of energy use, and improving the overall heating experience and satisfaction of users.

[0076] Although embodiments of the present application have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for constructing a return water temperature prediction model for floor radiant end regulation, characterized by, The method comprises the following steps: S1, user classification: Classify users according to building thermal performance, house type, external envelope location, total number of floors and floor interval; S2, laying monitoring network hardware facilities: Lay user water supply temperature sensors, user return water temperature sensors and heat meters in the user's heating network; S3, constructing representative households A and prediction households B; Select users in the same category classified in step S1 as representative households A with room temperature monitoring and prediction households B without room temperature monitoring, and lay indoor temperature sensors in the representative households A with room temperature monitoring; S4, data collection and processing; S5, establishment of two-stage return water temperature prediction models for each type of user; S6, valve control: Based on the predicted return water temperature value of the user, the valve of the user is controlled; The specific operation of step S5 is: First-stage return water temperature basic prediction model: for each type of user, a return water temperature basic prediction model for each type of user is established in the representative household A; The input variables are the supply water temperature, return water temperature, outdoor temperature, wind speed, solar radiation, room temperature, future time meteorological prediction parameters and indoor temperature setting parameters, and time variables of the representative household at historical time and current time; The output variable is the next time return water temperature of the representative household A; Second-stage return water temperature difference prediction model: for each type of user, a difference prediction model between the return water temperature of the representative household A and the prediction household B is established; The input variables are the supply water temperature, return water temperature, outdoor temperature, wind speed, solar radiation, room temperature, daily heat consumption ratio and difference between the two users, future time meteorological prediction parameters and indoor temperature setting parameters, and time variables of the representative household at historical time and current time; The output variable is the predicted difference ΔTh(t+1) between the return water temperature of house B and the return water temperature of house A at the next time instant B-A ; The representative household and the prediction household daily heat consumption ratio QK and difference ΔQ are calculated as follows: The first-stage return water temperature basic prediction model and the second-stage return water temperature difference prediction model are combined to obtain a predicted return water temperature Th(t+1) of the house B B ; The specific steps of step S1 are: 。 2. The method of claim 1, wherein the method is characterized by: According to the building thermal performance, the building group is divided into five categories: non-energy saving, one-step energy saving, two-step energy saving, three-step energy saving and four-step energy saving; According to the classification of each type of user group according to the house type, the same house type user is classified according to the external envelope location, and the basic classification is bottom middle, bottom edge, middle middle, middle edge, top middle and top edge six categories; Finally, according to the cold air penetration characteristics, the total number of floors and the floor interval are classified. The specific operation of step S3 is:

3. The method of claim 1, wherein the method is characterized by: Data collection: collect the supply water temperature, return water temperature, daily heat consumption, outdoor temperature, wind speed and solar radiation data of each user, and collect the indoor temperature of the representative household A; Data processing: use statistical methods to remove and supplement abnormal data, and process the data into historical time data with the required time granularity. ​

Images