Indoor temperature determination method fusing physical constraints

Abstract

The application provides a method for determining indoor temperature by fusing physical constraints, which can be applied to the fields of temperature control and machine learning. The method comprises: processing heating system driving data at a historical time by using a temperature prediction model to obtain an indoor evaluation temperature at a target time, wherein the temperature prediction model is obtained by training a neural network model based on a building heat loss value, the building heat loss value is obtained according to a difference between a sample evaluation temperature change rate based on a sample indoor evaluation temperature and a sample simulation temperature change rate based on a simulation algorithm, and the simulation temperature change rate represents a temperature instantaneous change result that meets the energy conservation of a simulation sample building interior at the historical time due to the instantaneous difference between the heat gained and the heat lost. The building heat loss value is used to physically constrain the indoor evaluation temperature of the temperature prediction model to meet the instantaneous heat balance of the energy conservation.

Description

Technical Field

[0001] This application relates to the fields of temperature control and machine learning technology, and specifically to a method for determining indoor temperature that incorporates physical constraints. Background Technology

[0002] Radiant floor heating systems are widely used in building heating due to their high thermal comfort and energy efficiency. However, in related studies, neural network models have shown weak generalization ability in predicting indoor temperature fields in buildings. Summary of the Invention

[0003] In view of the above problems, this application provides a method for determining indoor temperature that incorporates physical constraints.

[0004] According to a first aspect of this application, a method for determining indoor temperature by incorporating physical constraints is provided, comprising: processing heating system driving data at historical time points using a temperature prediction model to obtain an indoor assessment temperature at a target time point, wherein the heating system driving data includes at least one of the inlet water flow temperature, water flow velocity, or outdoor temperature of the target building, the target time point being later than the historical time point, the temperature prediction model being obtained by training a neural network model based on building heat loss values, the building heat loss values ​​being obtained based on the difference between the sample assessment temperature change rate obtained based on the sample indoor assessment temperature and the sample simulation temperature change rate obtained based on the simulation algorithm, the simulation temperature change rate representing the instantaneous temperature change result that satisfies energy conservation at historical time points inside the simulated sample building due to the instantaneous difference between heat gain and heat loss, and the building heat loss values ​​being used to physically constrain the indoor assessment temperature of the temperature prediction model to satisfy the instantaneous thermal balance of energy conservation.

[0005] The second aspect of this application provides an indoor temperature determination device, comprising: an evaluation module for processing heating system driving data at historical times using a temperature prediction model to obtain an indoor evaluation temperature at a target time, wherein the heating system driving data includes at least one of the inlet water flow temperature, water flow velocity, or outdoor temperature of the target building, the target time being later than the historical time, the temperature prediction model being obtained by training a neural network model based on building heat loss values, the building heat loss values ​​being obtained based on the difference between the sample evaluation temperature change rate obtained based on the sample indoor evaluation temperature and the sample simulation temperature change rate obtained based on the simulation algorithm, the simulation temperature change rate representing the instantaneous temperature change result that satisfies energy conservation in the simulated sample building at historical times due to the instantaneous difference between heat gain and heat loss, and the building heat loss values ​​being used to physically constrain the indoor evaluation temperature of the temperature prediction model to satisfy the instantaneous thermal balance of energy conservation.

[0006] A third aspect of this application provides an electronic device comprising: one or more processors; and a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the method described above.

[0007] A fourth aspect of this application also provides a computer-readable storage medium having a computer program or instructions stored thereon, which, when executed by a processor, implement the steps of the above-described method.

[0008] The fifth aspect of this application also provides a computer program product, including a computer program or instructions that, when executed by a processor, implement the steps of the above-described method.

[0009] According to the indoor temperature determination method integrating physical constraints provided in this application, the simulated temperature change rate represents the instantaneous temperature change rate that satisfies energy conservation within the simulated building at historical moments due to the instantaneous difference between heat gain and heat loss. The evaluated temperature change rate represents the instantaneous temperature change rate within the model-predicted building at historical moments due to the instantaneous difference between heat gain and heat loss. By constructing the building thermal loss value of the neural network model based on the degree of difference between the simulated and evaluated temperature change rates, the physical laws of building thermodynamics are explicitly introduced as constraints during the training process of the neural network model. This makes the indoor evaluated temperature predicted by the temperature prediction model in real building scenarios more consistent with the actual heat transfer laws, improving the generalization ability and transferability of the temperature prediction model under different building structures or heating system operation drives. It also improves prediction accuracy and efficiency, providing a reliable basis for the operation, regulation, and optimization of floor radiant heating systems. Attached Figure Description

[0010] The above-mentioned contents, other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0011] Figure 1 A flowchart of an indoor temperature determination method incorporating physical constraints according to an embodiment of this application is shown.

[0012] Figure 2 A scene diagram of a simulated sample building according to an embodiment of this application is shown.

[0013] Figure 3 A schematic diagram of a temperature prediction model according to an embodiment of this application is shown.

[0014] Figure 4 A schematic diagram illustrating the determination of the composite loss value according to an embodiment of this application is shown.

[0015] Figure 5 A schematic diagram of a sample room temperature visualization according to an embodiment of this application is shown.

[0016] Figure 6 A schematic diagram illustrating error visualization according to an embodiment of this application is shown.

[0017] Figure 7 A structural block diagram of an indoor temperature determination device according to an embodiment of this application is shown.

[0018] Figure 8 A block diagram of an electronic device suitable for implementing an indoor temperature determination method incorporating physical constraints, according to an embodiment of this application, is shown. Detailed Implementation

[0019] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0020] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0021] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0022] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0023] In the process of implementing this application, it was found that the relevant methods have the following main limitations:

[0024] (1) Although the relevant numerical simulation methods take into account the complex heat transfer process of the indoor temperature field under the floor radiation condition, their calculation scale is large and the time is long, which makes it difficult to meet the needs of the underfloor heating system for minute-level temperature field prediction and real-time control in actual operation, and cannot realize the rapid inference of the overall spatiotemporal distribution of the indoor temperature field.

[0025] (2) Many related studies use the temperature of a single target point or a few fixed points to replace the indoor core temperature layer as the main basis for the operation and regulation of the floor radiation system, which makes it difficult to reflect the continuous spatial distribution characteristics of indoor temperature.

[0026] (3) The relevant data-driven prediction models rely on training data and the prediction object is the system operating parameters (water temperature) rather than the indoor temperature itself. They also do not explicitly introduce physical equations such as energy conservation as constraints, which can easily lead to prediction results deviating from the actual heat transfer law. The model's generalization ability and transferability under different building structures or operating conditions are insufficient, making it difficult to achieve physical consistency and spatial distribution rationality of prediction results under complex operating conditions.

[0027] In view of this, embodiments of this application provide an indoor temperature determination method that integrates physical constraints. The temperature prediction model uses the indoor assessed temperature changes of all spaces within a sample building at continuous target times as the prediction object, which can reflect the continuous spatial distribution characteristics of indoor temperature. In addition, the sample simulated temperature change rate represents the instantaneous temperature change rate that satisfies energy conservation within the simulated sample building at historical times due to the instantaneous difference between heat gain and heat loss. The sample assessed temperature change rate characterizes the instantaneous temperature change rate within the sample building based on the model prediction, which is generated at historical times due to the instantaneous difference between heat gain and heat loss. By constructing the building thermal loss value of the neural network model based on the degree of difference between the sample simulated temperature change rate and the sample assessed temperature change rate, the physical laws of building thermodynamics are explicitly introduced as constraints during the training process of the neural network model. This makes the indoor assessed temperature predicted by the temperature prediction model in real building scenarios more consistent with the actual heat transfer laws, improves the generalization ability and transferability of the temperature prediction model under different building structures or heating system operation drives, and also improves prediction accuracy and efficiency, providing a reliable basis for the operation, regulation and optimization of floor radiant heating systems.

[0028] Figure 1 A flowchart of an indoor temperature determination method incorporating physical constraints according to an embodiment of this application is shown.

[0029] like Figure 1 As shown, the indoor temperature determination method that incorporates physical constraints in this embodiment includes operation S110.

[0030] In operation S110, the heating system drive data from historical moments is processed using a temperature prediction model to obtain the indoor assessment temperature at the target moment. The temperature prediction model is obtained by training a neural network model based on building heat loss values. The building heat loss values ​​are derived from the difference between the sample assessment temperature change rate obtained from the sample indoor assessment temperature and the sample simulated temperature change rate obtained from the simulation algorithm. The simulated temperature change rate characterizes the instantaneous temperature change result that satisfies energy conservation within the simulated sample building at historical moments due to the instantaneous difference between heat gain and heat loss.

[0031] Using simulation software, a three-dimensional building model is constructed based on the variable and invariable parameters of the sample building. At least one variable parameter, such as different inlet water flow temperature, water flow velocity, or outdoor temperature of the sample building, is combined to obtain multiple sample heating system driving data with time-series relationships. The simulation software performs transient simulation on the multiple heating system driving data to obtain the indoor simulation temperature of each of the multiple heating system driving data, and calculates the simulation temperature change rate based on the indoor simulation temperature and the heating system driving data.

[0032] For example, the heating system is an underfloor heating system.

[0033] When the heating system's driving data changes at a certain historical moment, it directly alters the amount of heat dissipated from the underfloor heating coils into the room. Simultaneously, the amount of heat lost from the room to the outside also changes, resulting in an energy difference that causes a change in the simulated indoor temperature.

[0034] The heat gain is the amount of heat dissipated from the underfloor heating coils to the room, and the heat loss is the amount of heat lost from the room to the outside. According to the law of conservation of energy, the instantaneous difference between the heat gain and the heat loss will be converted into the indoor air and internal heat storage of the sample building, thus manifesting as a change in the simulated indoor temperature.

[0035] The temperature change rate of the simulated sample represents the instantaneous temperature change inside the simulated sample building at historical moments, which satisfies the law of energy conservation due to the instantaneous difference between heat gain and heat loss.

[0036] Variable parameters may include outdoor temperature and inlet water temperature and flow velocity of the underfloor heating coils in the heating system.

[0037] Invariable parameters may include the geometric structure of the sample building, the material parameters of the building envelope, the layout of the underfloor heating pipes, the arrangement of indoor heat sources, or the parameters of thermal boundary conditions.

[0038] For example, building envelope material parameters include parameters such as walls, floors, doors and windows, obstructions, wall thickness, and door and window locations; underfloor heating coil layout includes parameters such as coil diameter, arrangement, and pipe spacing; indoor heat source layout can include the distribution of moving people, equipment, and furniture; thermal boundary condition parameters can include convective heat transfer temperatures with different walls.

[0039] Figure 2 A scene diagram of a simulated sample building according to an embodiment of this application is shown.

[0040] like Figure 2 As shown, a sample room in the sample building measures 5.4m × 5.8m × 4m. The south-facing wall is an exterior wall without windows or doors, while the other three sides are interior walls. The north-facing interior wall has a normally closed door that connects to other rooms in the sample building. The wall and ceiling thickness is 150mm, and the floor thickness is 80mm. The underfloor heating pipes are DN20 pipes made of random copolymer polypropylene (PPR), arranged in a fixed serpentine pattern with a fixed pipe spacing of 100mm. The center of the pipe is 50mm from the floor surface. There are no moving personnel, equipment, or other obstructions inside the building.

[0041] The temperature prediction model is built based on a convolutional neural network (CNN).

[0042] Figure 3 A schematic diagram of a temperature prediction model according to an embodiment of this application is shown.

[0043] like Figure 3 As shown, the temperature prediction model includes a convolutional layer (Conv3D), a max-pooling layer, a fully connected layer, a reshape layer, a fusion layer, and a deconvolutional layer. The sample heating system driving data is input into the convolutional and max-pooling layers to obtain a (4, 4, 4, 256) tensor feature. The sample heating system driving data is then input into the (4, 4, 4, 128) tensor feature obtained after processing by two fully connected layers and the reshape layer. In the fusion layer, the (4, 4, 4, 256) tensor feature is fused with the (4, 4, 4, 128) tensor feature. Finally, the result is input into the three fusion layers and three deconvolutional layers in the decoder to output the (32, 32, 32) sample indoor evaluation temperature.

[0044] The system inputs historical heating system driving data into the temperature prediction model and outputs the target indoor temperature at the target time. The historical time can be the previous time step, and the target time can be any of the next consecutive future time steps adjacent to the previous time step. For example, the historical time is 3:00, and the target time is 3:01.

[0045] The rate of change of the sample assessment temperature is calculated based on the indoor assessment temperature of the sample building. The rate of change of the sample assessment temperature represents the instantaneous temperature change at a historical moment caused by the instantaneous difference between heat gain and heat loss inside the building.

[0046] The building's thermal loss value is obtained by comparing the difference between the assessed temperature change rate and the simulated temperature change rate. If the building's thermal loss value exceeds a preset threshold, the parameters of the temperature prediction model are adjusted until the building's thermal loss value is less than or equal to the preset threshold, at which point training of the temperature prediction model is stopped.

[0047] The building thermal loss value is used to minimize the difference between the sample evaluation temperature change rate and the sample simulation temperature change rate, thereby achieving an instantaneous thermal balance where the indoor evaluation temperature predicted by the physically constrained trained temperature prediction model satisfies energy conservation.

[0048] Responding to historical heating system driving data, a trained temperature prediction model is used to process the historical heating system driving data to obtain the indoor assessment temperature of the target building at multiple consecutive target times in the future, thereby obtaining the three-dimensional indoor assessment temperature field change.

[0049] Since the simulated temperature change rate represents the instantaneous temperature change rate inside the simulated building at a historical moment due to the instantaneous difference between heat gain and heat loss, satisfying the law of energy conservation, while the evaluated temperature change rate represents the instantaneous temperature change rate inside the model-predicted building at a historical moment due to the instantaneous difference between heat gain and heat loss, the building thermal loss value of the neural network model is constructed based on the degree of difference between the simulated and evaluated temperature change rates. This explicitly introduces the physical laws of building thermodynamics as constraints during the training process of the neural network model, making the indoor evaluated temperature predicted by the temperature prediction model in real building scenarios more consistent with the actual heat transfer laws. This improves the generalization ability and transferability of the temperature prediction model under different building structures or heating system operation drives, while also improving prediction accuracy and efficiency, providing a reliable basis for the operation, regulation, and optimization of floor radiant heating systems.

[0050] According to an embodiment of this application, the indoor temperature determination method incorporating physical constraints further includes: adjusting the heating system drive data of the heating system when the indoor assessment temperature does not meet the preset constant temperature conditions; processing the adjusted heating system drive data using a temperature prediction model to obtain the adjusted indoor assessment temperature at the target time; and stopping the adjustment of the heating system when the adjusted indoor assessment temperature meets the preset constant temperature conditions.

[0051] For example, the preset constant temperature condition can be 25℃.

[0052] If the indoor assessment temperature is higher than the preset constant temperature condition, the inlet water temperature or flow velocity of the heating system can be reduced. The reduced inlet water temperature and flow velocity are then processed using a temperature prediction model to predict the target indoor assessment temperature at the desired time. The adjusted indoor assessment temperature is lower than the original indoor assessment temperature.

[0053] If the indoor assessment temperature is lower than the preset constant temperature condition, the inlet water temperature or flow velocity of the heating system can be increased. The increased inlet water temperature and flow velocity are then processed using a temperature prediction model to predict the target indoor assessment temperature at the desired time. The adjusted indoor assessment temperature is higher than the original indoor assessment temperature.

[0054] After iteratively adjusting the inlet water temperature or flow rate of the heating system multiple times until the adjusted indoor assessment temperature meets the preset constant temperature conditions, the adjustment is stopped.

[0055] By using a temperature prediction model to process the heating system drive data from the previous time step, the indoor assessment temperature for future consecutive time steps can be accurately predicted. If the indoor assessment temperature output by the temperature prediction model does not meet the preset constant temperature conditions, the heating system drive data from the previous time step can be adjusted to continuously and dynamically maintain the three-dimensional temperature field in the target building at the preset constant temperature conditions.

[0056] According to the embodiments of this application, the sample heating system driving data of multiple samples are obtained by regulating the sample heating system driving data at the initial regulation time of the sample based on the sample regulation strategy. The sample regulation strategy includes at least one of the following: sample simulation time period parameters, sample time regulation step size parameters, sample outdoor temperature range parameters, sample outdoor temperature regulation step size parameters, sample inlet water flow temperature range parameters, sample inlet water flow temperature regulation step size parameters, sample water flow velocity range parameters, or sample water flow velocity regulation step size parameters.

[0057] Based on the sample control strategy, the sample heating system driving data at the initial control time of the sample are controlled to obtain the sample heating system driving data of multiple samples.

[0058] The sample simulation time period parameter represents the duration of the simulation for each sample. For example, 30 minutes.

[0059] The sample time step size parameter represents the time interval between different moments in the simulation of the sample. Simulation calculations are performed on the variable and invariant parameters of the sample to obtain the indoor simulated temperature and the outlet water temperature of the heating system at multiple consecutive moments. For example, with a sample time step size parameter of 1 minute, the simulation software calculates the indoor simulated temperature of the sample at 2:00 and again at 2:01.

[0060] The outdoor temperature range parameter represents the temperature control range outside the sample building; the outdoor temperature control step size parameter represents the step size of the temperature change outside the sample building each time.

[0061] The sample inlet water flow temperature range parameter characterizes the control range of the inlet water flow temperature of the underfloor heating coil in the heating system; the sample inlet water flow temperature control step size parameter characterizes the step size of each control of the inlet water flow temperature of the underfloor heating coil in the heating system.

[0062] The sample water flow velocity range parameter characterizes the control range of the inlet water flow velocity of the underfloor heating coil in the heating system; the sample water flow velocity control step size parameter characterizes the step size of each control of the inlet water flow velocity of the underfloor heating coil in the heating system.

[0063] The initial control time of the sample is the starting time of the simulation software. The sample heating system drive data at the initial control time is the initialization data.

[0064] Each item in the sample control strategy can be used to control the sample heating system driving data at the initial control time of the sample, thereby obtaining multiple samples and sample heating system driving data of the samples.

[0065] The sample control strategy is determined based on the actual operation and maintenance of the sample buildings and the sampling records of outdoor temperatures. For example, based on the outdoor temperature records of XX City in winter over the past five years, the outdoor temperature range parameter for the sample is determined to be -20℃ to 10℃, and the outdoor temperature control step parameter is 5℃; based on the operating parameter range of the heating system of the sample buildings, the inlet water temperature range parameter is determined to be 35℃ to 50℃, the inlet water temperature control step parameter is 3℃, the water flow velocity range parameter is 0.4m / s to 1.0m / s, and the water flow velocity control step parameter is 0.2m / s (as shown in Table 1).

[0066] Table 1 shows the sample control strategy according to an embodiment of this application.

[0067]

[0068] A total of 168 samples were obtained by combining data from various parameters for simulation. For example, the sample heating system driving data for sample 1 are: simulation time 3:00, outdoor temperature 0℃, inlet water temperature 40℃, and water flow velocity 0.5 m / s.

[0069] The coordinated configuration of the sample simulation time period parameters and the time-based control step parameters enables flexible adaptation of the control time granularity, which can meet the needs of long-term trend prediction and support minute-level real-time response. The sample outdoor temperature range and the sample outdoor temperature control step parameters establish a precise mapping relationship between meteorological conditions and heating load, enabling the system to make forward-looking heat load predictions and step-by-step energy adjustments based on changes in ambient temperature. The combination of the sample inlet water flow temperature range parameters and the sample water flow velocity range parameters forms a comprehensive constraint and adjustment capability on the heat exchange medium. The multi-dimensional refined parameter system enables dynamic optimization of the sample heating system driving data.

[0070] According to an embodiment of this application, the wall includes an inner wall and an outer wall. The sample heat exchange temperature of the outer wall is determined based on the sample outdoor temperature, and the sample heat exchange temperature of the inner wall is determined based on the sample outdoor temperature and a preset linear relationship. The preset linear relationship is determined based on the historical outdoor temperature and the historical heat exchange temperature of the inner wall at a historical time.

[0071] The sample building's walls consist of three interior walls, one exterior wall, and one ceiling. The exterior walls are located between the exterior and interior, while the interior walls are located between the interior and other interior spaces.

[0072] When constructing a 3D building model based on the material parameters of the building envelope, it is necessary to consider setting a preset linear relationship for energy conversion between the two sides of the wall according to the location and material of the wall, so as to accurately simulate the temperature changes in the room transmitted from the outdoor temperature fluctuations through the wall.

[0073] The heat exchange temperature of the sample on the south exterior wall is equal to the outdoor temperature T of the sample. outdoor .

[0074] Based on the historical heat exchange temperatures collected at multiple historical moments during the historical period, corresponding to the absence of heating in the sample building under multiple historical outdoor temperatures, the preset linear relationships of the remaining interior walls and ceilings were constructed (as shown in Table 2).

[0075] Table 2 shows a preset linear relationship table according to an embodiment of this application.

[0076]

[0077] When the simulation software is used, the outdoor temperature parameters of the sample heating system driving data corresponding to the sample are converted according to the preset linear relationship of each wall to obtain the sample heat exchange temperature corresponding to each wall. Then, based on the area of ​​each wall, the sample heat exchange temperature and the convective heat transfer coefficient, the total heat exchange of the overall heat flow inside and outside the sample building is obtained.

[0078] For example, the convective heat transfer coefficient can be 8.72.

[0079] The sample heat exchange temperature characterizes the instantaneous temperature difference between the two sides of the wall. For exterior walls, it can reflect the magnitude of the potential difference in heat transfer from the outdoor climate disturbance to the interior through the wall; for interior walls or ceiling slabs, it can reflect the intensity of energy exchange between the heating system inside the sample building and the walls.

[0080] The heat transfer temperature of the exterior wall samples responds to outdoor climate disturbances, while the heat transfer temperature of the interior wall samples is converted to the outdoor temperature of the samples through a preset linear relationship coupled with historical heat transfer effects. This allows the simulation calculation of the total heat transfer of the sample building to take into account both the characteristics of instantaneous meteorological shocks and the heat storage delay of the building envelope. As a result, the room temperature fluctuation trend can be predicted in advance in the heating system simulation. The differentiated modeling of the heat flow of the interior and exterior walls and the linear correction mechanism driven by historical data enable accurate dynamic prediction of the heat load of the sample building.

[0081] According to an embodiment of this application, the building heat loss value is obtained based on the following operations: For any sample among multiple samples with a sampling time sequence relationship, the sample heating system driving data corresponding to the sample is processed using a simulation algorithm to obtain sample simulation temperature data, which includes the sample indoor simulation temperature; the sample heating system driving data corresponding to the sample is processed using a neural network model to obtain the sample indoor evaluation temperature; the sample simulation temperature change rate is determined based on the sample heating system driving data and the sample simulation temperature data; the sample evaluation temperature change rate is determined based on the sample indoor evaluation temperature, the sample indoor simulation temperature, and the sampling time interval; and the building heat loss value is obtained based on the degree of difference between the sample simulation temperature change rate and the sample evaluation temperature change rate corresponding to each of the multiple samples.

[0082] The sample heating system drive data corresponding to multiple samples with sampling time sequence relationships are processed into a one-dimensional array in list format, denoted as (T outdoor , Tw in, v w ), where T outdoor The outdoor temperature of the sample is Tw_in, the inlet water temperature of the sample is Tw_in, and v_in is v_in. w The velocity of the sample water flow.

[0083] A variable parameter database was constructed based on the driving data of the heating systems corresponding to multiple samples. A 3D building model was built using simulation software based on the invariable parameters of the sample buildings. Simulations were performed for each sample in the variable parameter database to obtain the simulated indoor temperature and the simulated outlet water flow temperature. The simulated temperature data includes the simulated indoor temperature Tm. true And the simulated temperature of the water flow at the sample outlet (Tw out).

[0084] The indoor simulated temperature of the sample is the indoor temperature calculated based on simulation software, and the outlet water flow simulated temperature of the sample is the coil outlet water flow temperature of the heating system calculated based on simulation software. The outlet water flow simulated temperature of the sample is processed into a single-value format as a calculation parameter for the sample simulated temperature change rate.

[0085] The training samples for the neural network model include 5040 samples, each with its own sample heating system driving data and sample label. The sample label for each sample is the simulated indoor temperature.

[0086] The sample heating system driving data is processed using a neural network model to obtain the sample indoor assessment temperature. The sample indoor assessment temperature is based on the indoor temperature predicted by the model.

[0087] Energy conservation analysis was performed on the sample building. The heating capacity of the heating system and the heat loss from heat exchange between the indoor and outdoor environments were calculated based on the driving data of the sample heating system and the sample simulation temperature data. Then, the sample simulation temperature change rate was determined based on the heating capacity and heat loss of the heating system.

[0088] The sample simulation temperature change rate represents the rate of change of the real temperature over time based on the simulation software.

[0089] In one embodiment, the sample is evaluated for the rate of temperature change. As shown in formula (1):

[0090] (1).

[0091] in, The indoor temperature of the sample chamber characterizes the sample. The simulated temperature inside the sample chamber characterizes the sample. The sampling time interval characterizing the temperature prediction, e.g., 1 minute, is used to assess the rate of temperature change in the samples. The unit is K / s, which represents the rate of temperature change over time predicted based on a neural network model.

[0092] The simulated temperature change rate and the evaluated temperature change rate of each sample are compared to obtain the degree of difference. Then, multiple degrees of difference are evaluated to obtain the building heat loss value.

[0093] The simulated temperature change rate is the result of the simulation software simulating the instantaneous temperature change inside the sample building while satisfying energy conservation. The evaluated temperature change rate is the result of the neural network model predicting the instantaneous temperature change inside the sample building. By constructing the building thermal loss value of the neural network model based on the difference between the simulated temperature change rate and the evaluated temperature change rate, the physical laws of building thermodynamics are explicitly introduced as constraints in the training process of the neural network model. This improves the generalization ability and transferability of the model under different building structures or heating system operation, enabling the trained neural network model to meet the actual needs of floor radiant heating systems for high-precision and complex indoor temperature prediction.

[0094] According to embodiments of this application, the sample simulation temperature data further includes the sample outlet water flow simulation temperature of the sample heating system, and the sample heating system driving data further includes the sample heat exchange temperature of different walls of the sample building, where the sample heat exchange temperature is the temperature difference between the high-temperature side and the low-temperature side of the wall. Based on the sample heating system driving data and the sample simulation temperature data, the sample simulation temperature change rate is determined, including: determining the sample heating system heating value based on the sample inlet water flow temperature, sample water flow velocity, and sample outlet water flow simulation temperature; determining the sample heat source heating value based on the number, heat dissipation, and heat dissipation area of ​​the sample heat sources within the sample building; determining the sample environment heat loss value based on the sample indoor simulation temperature and sample heat exchange temperature; and determining the sample simulation temperature change rate based on the sample heating system heating value, sample heat source heating value, sample environment heat loss value, and sample air heat capacity.

[0095] In one embodiment, the sample heating system heating value As shown in formula (2):

[0096] (2).

[0097] in, Density of water, expressed in kg / m³ 3 , The specific heat capacity of water, expressed in J / (kg·K). Characterized by the cross-sectional area of ​​the heating system coils, in meters (m²). 2 , Characterizing the water flow velocity in the sample, in m / s. Characterized by the temperature of the inlet water flow of the sample, in K. Characterizes the simulated temperature of the water flow at the sample outlet, in K.

[0098] The heating value of the sample heating system is the amount of heat the heating system provides to the sample building.

[0099] In one embodiment, the sample heat source heating value of the sample As shown in formula (3):

[0100] (3).

[0101] in, Characterizes the heat dissipation per unit area of ​​the i-th sample heat source. Characterizes the heat dissipation area of ​​the i-th sample heat source. Characterizes the number of heat sources in the sample.

[0102] The sample heat source heating value characterizes the heat supplied to the sample building by multiple sample heat sources. The sample heat source represents the equipment that generates and transfers heat, such as gas appliances.

[0103] In one embodiment, the sample's environmental heat loss value As shown in formula (4):

[0104] (4).

[0105] in, The overall convective heat transfer coefficient of a wall that exchanges heat with the outside environment, expressed in W / (m²). 2 ·K), Characterizes the heat transfer temperature of the sample.

[0106] The sample environment heat loss value represents the heat loss due to heat exchange between the indoor and outdoor environments of the sample building.

[0107] In one embodiment, the sample's simulated temperature change rate As shown in formula (5):

[0108] (5).

[0109] in, Characterizes the air thermal capacity of the sample.

[0110] In the energy conservation mechanism, the energy difference between entering and leaving the room will cause instantaneous changes in the indoor temperature. Therefore, when conducting energy conservation analysis on the sample building, the heat transfer between the heating value of the sample heating system, the heating value of the sample heat source, and the heat loss value of the sample environment should be considered to establish a transient thermal balance model of the sample building, thereby providing a physical basis for calculating the simulated temperature change rate of the sample.

[0111] According to an embodiment of this application, a building heat loss value is obtained based on the simulated temperature change rate and the evaluated temperature change rate corresponding to each of the multiple samples, including: obtaining the building heat loss value based on the ratio between a first quantity and a total quantity, wherein the first quantity is the number of mismatches between the simulated temperature change rate and the evaluated temperature change rate corresponding to the sample, and the total quantity is the number of samples.

[0112] For each of the multiple samples, the simulated temperature change rate and the evaluated temperature change rate of the sample are compared point by point for sign consistency and marked accordingly. If the signs are consistent, it is marked as 0, otherwise it is marked as 1, thus obtaining the comparison matrix of multiple samples.

[0113] A value of 0 indicates that the simulated temperature change rate and the evaluated temperature change rate are the same, while a value of 1 indicates that the simulated temperature change rate and the evaluated temperature change rate are different.

[0114] The first quantity is obtained by comparing the number of samples marked as 1 in the comparison matrix. The building heat loss value is obtained by comparing the ratio between the first quantity and the total number of samples.

[0115] The smaller the building heat loss value, the fewer the number of samples whose simulated temperature change rate and the evaluated temperature change rate are different.

[0116] The proportion of mismatches between the simulated temperature change rate and the evaluated temperature change rate is used as the physical constraint value for building thermal loss, thereby training the model-determined evaluated temperature change rate to approximate the simulated temperature change rate.

[0117] According to an embodiment of this application, the temperature prediction model is obtained by training a neural network model based on building thermal loss values ​​and other loss function values, including error loss values. The error loss values ​​are obtained based on the error loss function, according to the sample indoor evaluation temperature and sample indoor simulation temperature corresponding to each of the multiple samples.

[0118] The error loss value can be the mean squared error (MSE) function.

[0119] The error loss value is obtained by calculating the loss value between the evaluation temperature in the sample room and the simulated temperature in the sample room using the error loss function.

[0120] A composite loss value is obtained based on the error loss value and the building thermal loss value. A neural network model is trained based on the composite loss value to obtain a temperature prediction model.

[0121] In one embodiment, the composite loss value As shown in formula (6):

[0122] (6).

[0123] in, Characterization error loss value, Characterizing the building's heat loss value, Characterization error loss value.

[0124] Figure 4 A schematic diagram illustrating the determination of the composite loss value according to an embodiment of this application is shown.

[0125] like Figure 4 As shown, the sample heating system driving data is input into a neural network model to obtain the sample indoor evaluation temperature; the sample evaluation temperature change rate is determined based on the sample indoor evaluation temperature and the sample indoor simulation temperature; the sample heating system driving data and invariant parameters are processed using a simulation algorithm to obtain the sample simulation temperature data; the sample simulation temperature change rate is determined based on the sample heating system driving data and the sample simulation temperature data; the building thermal loss value is obtained based on the degree of difference between the sample simulation temperature change rate and the sample evaluation temperature change rate corresponding to each of the multiple samples; and the composite loss value is obtained based on the error loss value and the building thermal loss value.

[0126] Set the weighting factor for the error loss value to 1, and set the weighting factor for the building thermal loss value to... This allows us to construct a composite loss value, which is then used as the core monitoring metric for the training process. By simultaneously optimizing the model parameters and the satisfaction of physical constraints through gradient backpropagation, we achieve gradient descent and embed physical constraints into the training process of the temperature prediction model.

[0127] The training samples were divided into training and validation sets in a 7:3 ratio for training and validation of the neural network model, resulting in a temperature prediction model. The training batch size was set to 32, and the initial learning rate was 10%. -3 The batch ratio for validation was 0.25, the tolerance (ReduceLROnPlateau) was set to 5, and the adjustment factor was 0.1.

[0128] A sample is selected with an outdoor temperature of -1℃, an inlet water temperature of 36℃, and a water flow velocity of 0.5m / s. The indoor temperature change of the sample is simulated in the simulation software over a period of 30 minutes. Then, the temperature prediction model is used to predict the indoor temperature of the sample.

[0129] Figure 5 A schematic diagram of a sample room temperature visualization according to an embodiment of this application is shown.

[0130] like Figure 5 As shown, the predicted temperature field in the 3D building model at the 20th minute is visualized. Visualization is performed on three planes: XY (horizontal), XZ (vertical), and YZ (horizontal and vertical), displaying the contour lines of the sample room's assessed temperature distribution. The values ​​on the number axes represent the sample room's assessed temperature at that location.

[0131] Figure 6 A schematic diagram illustrating error visualization according to an embodiment of this application is shown.

[0132] like Figure 6 As shown, the simulated temperature field in the 3D building model at the 20th minute is visualized. The error between the simulated indoor temperature in the sample room and the evaluated indoor temperature in the predicted temperature field is visualized from three planes: XY (horizontal), XZ (vertical), and YZ (horizontal and vertical). The value on the number axis represents the error between the simulated indoor temperature and the evaluated indoor temperature at that position.

[0133] For each time point, the mean absolute error (MAE), mean squared error (MSE), root mean squared error (RMSE), mean absolute percentage error (MAPE), and coefficient of determination (R-squared, R²) were calculated for the simulated temperature and the evaluated temperature in the sample room for multiple samples at that time. The evaluation indicators are shown in Table 3, showing that the MAE is below 0.1 and the R² is above 0.98.

[0134] Table 3 shows the evaluation index table according to the embodiments of this application.

[0135]

[0136] (Continued from Table 3)

[0137]

[0138] Figure 7 A structural block diagram of an indoor temperature determination device according to an embodiment of this application is shown.

[0139] like Figure 7 As shown, the indoor temperature determination device 700 of this embodiment includes an evaluation module 710.

[0140] The evaluation module 710 is used to process the heating system driving data at historical time using a temperature prediction model to obtain the indoor evaluation temperature at the target time. The heating system driving data includes at least one of the inlet water flow temperature, water flow velocity, or outdoor temperature of the target building. The target time is later than the historical time. The temperature prediction model is obtained by training a neural network model based on the building heat loss value. The building heat loss value is obtained based on the difference between the sample evaluation temperature change rate obtained based on the sample indoor evaluation temperature and the sample simulation temperature change rate obtained based on the simulation algorithm. The simulation temperature change rate characterizes the instantaneous temperature change result that satisfies energy conservation in the simulated sample building at historical time due to the instantaneous difference between heat gain and heat loss. The building heat loss value is used to physically constrain the indoor evaluation temperature of the temperature prediction model to satisfy the instantaneous thermal balance of energy conservation.

[0141] Figure 8 A block diagram of an electronic device suitable for implementing an indoor temperature determination method incorporating physical constraints, according to an embodiment of this application, is shown.

[0142] Figure 8 The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0143] like Figure 8 As shown, a computer electronic device 800 according to an embodiment of this application includes a processor 801, which can perform various appropriate actions and processes according to a program stored in a ROM 802 (read-only memory) or a program loaded from a storage portion 808 into a RAM 803 (random access memory). The processor 801 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 801 may also include onboard memory for caching purposes. The processor 801 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this application.

[0144] RAM 803 stores various programs and data required for the operation of electronic device 800. Processor 801, ROM 802, and RAM 803 are interconnected via bus 804. Processor 801 executes various operations of the method flow according to embodiments of this application by executing programs in ROM 802 and / or RAM 803. It should be noted that programs may also be stored in one or more memories other than ROM 802 and RAM 803. Processor 801 may also execute various operations of the method flow according to embodiments of this application by executing programs stored in one or more memories.

[0145] According to embodiments of this application, the electronic device 800 may further include an input / output (I / O) interface 805, which is also connected to a bus 804. The electronic device 800 may also include one or more of the following components connected to the input / output (I / O) interface 805: an input section 806 including a keyboard, mouse, etc.; an output section 807 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 808 including a hard disk, etc.; and a communication section 809 including a network interface card such as a LAN card, modem, etc. The communication section 809 performs communication processing via a network such as the Internet. A drive 810 is also connected to the input / output (I / O) interface 805 as needed. A removable medium 811, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 810 as needed so that computer programs read from it can be installed into the storage section 808 as needed.

[0146] According to embodiments of this application, the method flow according to embodiments of this application can be implemented as a computer software program. For example, embodiments of this application include a computer program product comprising a computer program carried on a computer-readable storage medium, the computer program containing program code for performing the method shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network via communication section 809, and / or installed from removable medium 811. When the computer program is executed by processor 801, it performs the functions defined in the system of embodiments of this application. According to embodiments of this application, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0147] This application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the indoor temperature determination method incorporating physical constraints according to the embodiments of this application.

[0148] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium. Examples include, but are not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0149] For example, according to embodiments of this application, a computer-readable storage medium may include the ROM 802 and / or RAM 803 described above and / or one or more memories other than ROM 802 and RAM 803.

[0150] Embodiments of this application also include a computer program product comprising a computer program containing program code for performing the methods provided in the embodiments of this application. When the computer program product is run on an electronic device, the program code is used to enable the electronic device to implement the indoor temperature determination method incorporating physical constraints provided in the embodiments of this application.

[0151] When the computer program is executed by the processor 801, it performs the functions defined in the system / apparatus of this application embodiment. According to the embodiments of this application, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0152] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and may be downloaded and installed via the communication section 809, and / or installed from a removable medium 811. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0153] According to embodiments of this application, program code for executing the computer programs provided in the embodiments of this application can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0154] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions. Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations are not explicitly described in this application. In particular, without departing from the spirit and teachings of this application, the features described in the various embodiments of this application can be combined and / or combined in various ways. All such combinations and / or combinations fall within the scope of this application.

[0155] The embodiments of this application have been described above. However, these embodiments are merely illustrative and not intended to limit the scope of this application. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. Without departing from the scope of this application, those skilled in the art can make various substitutions and modifications, all of which should fall within the scope of this application.

Claims

1. A method for determining indoor temperature that incorporates physical constraints, characterized in that, The method includes: The indoor assessment temperature at the target time is obtained by processing historical heating system drive data using a temperature prediction model. The heating system driving data includes at least one of the following: the inlet water temperature of the heating system, the water flow velocity, or the outdoor temperature of the target building. The target time is later than the historical time. The temperature prediction model is obtained by training a neural network model based on the building heat loss value. The building heat loss value is obtained based on the difference between the sample assessed temperature change rate obtained based on the sample indoor assessed temperature and the sample simulated temperature change rate obtained based on the simulation algorithm. The sample assessed temperature change rate represents the instantaneous temperature change rate of the sample building interior based on the model prediction at a historical time due to the instantaneous difference between heat gain and heat loss. The sample simulated temperature change rate represents the instantaneous temperature change rate of the simulated sample building interior at a historical time due to the instantaneous difference between heat gain and heat loss, which satisfies the energy conservation principle. The building heat loss value is used to physically constrain the indoor assessed temperature of the temperature prediction model to satisfy the instantaneous thermal balance of energy conservation. The building heat loss value is obtained based on the following operation: For any sample among multiple samples that have a sampling time sequence relationship, The sample heating system drive data corresponding to the sample is processed by the simulation algorithm to obtain the sample simulation temperature data, which includes the sample indoor simulation temperature. The neural network model is used to process the sample heating system driving data corresponding to the sample to obtain the sample indoor evaluation temperature. The sample simulation temperature change rate is determined based on the sample heating system drive data and the sample simulation temperature data. The rate of change of the sample evaluation temperature is determined based on the indoor evaluation temperature, the indoor simulated temperature, and the sampling time interval. The building thermal loss value is obtained based on the degree of difference between the simulated temperature change rate corresponding to each of the multiple samples and the assessed temperature change rate of the samples. If the building's heat loss value is greater than a preset threshold, adjust the parameters of the neural network model until the building's heat loss value is less than or equal to the preset threshold, then stop training to obtain the temperature prediction model.

2. The method according to claim 1, characterized in that, The building thermal loss value is obtained based on the simulated temperature change rate and the evaluated temperature change rate corresponding to each of the multiple samples, including: The building heat loss value is obtained based on the ratio between the first quantity and the total quantity, wherein the first quantity is the number of mismatches between the simulated temperature change rate and the evaluated temperature change rate of the sample corresponding to the sample, and the total quantity is the number of samples.

3. The method according to claim 2, characterized in that, The sample simulation temperature data also includes the sample outlet water flow simulation temperature of the sample heating system. The sample heating system drive data also includes the sample heat exchange temperature of different walls of the sample building. The sample heat exchange temperature is the temperature difference between the high-temperature side and the low-temperature side of the wall. Based on the sample heating system drive data and sample simulation temperature data, determine the sample simulation temperature change rate, including: The heating value of the sample heating system is determined based on the sample inlet water temperature, sample water velocity, and sample outlet water simulated temperature. The heating value of the heat source is determined based on the number, heat dissipation, and heat dissipation area of ​​the heat source within the sample building. The heat loss value of the sample environment is determined based on the simulated temperature inside the sample chamber and the sample heat exchange temperature. The sample simulation temperature change rate is determined based on the heating value of the sample heating system, the heating value of the sample heat source, the heat loss value of the sample environment, and the heat capacity of the sample air.

4. The method according to claim 3, characterized in that, The wall includes an inner wall and an outer wall. The sample heat exchange temperature of the outer wall is determined based on the sample outdoor temperature. The sample heat exchange temperature of the inner wall is determined based on the sample outdoor temperature and a preset linear relationship. The preset linear relationship is determined based on the historical outdoor temperature and the historical heat exchange temperature of the inner wall at historical times.

5. The method according to claim 2, characterized in that, The temperature prediction model is obtained by training a neural network model based on the building thermal loss value and other loss function values, including the error loss value; the error loss value is obtained based on the error loss function, according to the sample indoor evaluation temperature and the sample indoor simulation temperature corresponding to each of the multiple samples.

6. The method according to claim 5, characterized in that, The sample heating system driving data of each of the multiple samples is obtained by regulating the sample heating system driving data at the initial regulation time of the sample based on the sample regulation strategy. The sample regulation strategy includes at least one of the following: sample simulation time period parameters, sample time regulation step size parameters, sample outdoor temperature range parameters, sample outdoor temperature regulation step size parameters, sample inlet water flow temperature range parameters, sample inlet water flow temperature regulation step size parameters, sample water flow velocity range parameters, or sample water flow velocity regulation step size parameters.

7. The method according to any one of claims 1 to 2, characterized in that, The method further includes: If the indoor temperature does not meet the preset constant temperature conditions, adjust the heating system drive data of the heating system. The temperature prediction model is used to process the controlled heating system drive data to obtain the controlled indoor assessment temperature at the target time. The control of the heating system is stopped when the controlled indoor assessment temperature meets the preset constant temperature conditions.

8. An electronic device, comprising: One or more processors; Memory, used to store one or more computer programs. The characteristic feature is that the one or more processors invoke the one or more computer programs to implement the steps of the method according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 7.

Images