Equivalent thermal resistance model construction method and system and building energy consumption calculation method and system

Abstract

The invention provides an equivalent thermal resistance model construction method and system and a building energy consumption calculation method and system. The method comprises the steps of obtaining load data and building feature data of a building of which energy consumption is to be calculated; according to the load data and the building feature data, utilizing a pre-constructed load prediction model to obtain predicted load data of the building; fitting is carried out according to the predicted load data of the building and the temperature difference data of the time period corresponding to the predicted load data, and the vector change relation between the equivalent thermal resistance of the building under the multiple time scales and the temperature difference data is obtained; an equivalent thermal resistance model is obtained according to the vector change relation between the equivalent thermal resistance and the temperature difference data of the building under the multiple time scales; according to the method, the equivalent thermal resistance model is constructed by fusing the data driving and the physical driving, and the historical load data and the building characteristic data are considered in the data driving process, so that the established equivalent thermal resistance model can establish a vector change relation and can also improve the interpretability between the data.

Description

Technical Field

[0001] The present invention relates to the technical field of energy conservation and emission reduction, and in particular to a method and system for constructing an equivalent thermal resistance model and calculating building energy consumption. Background Art

[0002] At present, building energy consumption, transportation energy consumption and industrial energy consumption are listed as the three major "black holes" of energy consumption. The continuous increase in the total amount of buildings and the improvement of living comfort have especially led to the increase in building energy consumption. Among them, building energy consumption is mainly heating and air-conditioning energy consumption, and building energy consumption accounts for about 30% in the energy consumption structure. In the annual energy consumption of public buildings, about 50%-60% is consumed by heating systems and air-conditioning refrigeration. Therefore, the research and development of energy-saving, innovative and low-carbon technologies for public buildings is an important and urgently needed research field. In recent years, due to the continuous increase in winter heating demand with the continuous expansion of urban building area, as well as higher requirements for energy conservation and environmental protection, the development and popularization of centralized heating and clean heating are being vigorously promoted, so that the coverage area of ​​urban centralized heating systems is expanding year by year. With people's pursuit of a higher quality of life, urban heating users also put forward higher requirements for heating quality. The centralized heating system currently still adopts the control method regulated by the heat source side, and the management is relatively extensive, resulting in uneven urban heat use and lagging heat network regulation. For urban centralized heating systems, how to achieve on-demand and precise heating has become the key to improving efficiency and service quality. Therefore, effective and accurate calculation of building energy consumption is of great significance for achieving high-quality heating, and also makes great contributions to the planning of urban energy systems, improving the operating efficiency of heating systems, and energy conservation and emission reduction.

[0003] Existing methods for calculating building energy consumption are mainly driven by physics or data. Taking the heating system as an example, it is a system with huge and complicated mechanisms, and it is difficult to establish a corresponding mathematical expression. It is difficult to apply the equivalent thermal resistance model constructed by a pure mechanism model to the actual project to calculate the building energy consumption. The equivalent thermal resistance model constructed by pure data drive to calculate the building energy consumption relies solely on historical data as the feature and cannot explain the internal relationship, making the model lack of interpretability. Therefore, whether the energy consumption calculation is performed by pure physics drive or pure data drive, it will lead to the problem that the error of building energy consumption calculation deviates greatly from the actual value. Summary of the invention

[0004] In order to solve the problem that the existing equivalent thermal resistance model construction method is difficult to establish mathematical expression if it is driven by pure physics, which makes it difficult to carry out practical application; if it is driven by pure data, it relies on historical data as a feature and has low interpretability, the present invention proposes an equivalent thermal resistance model construction method, including:

[0005] Obtain load data and building characteristic data of the building to be calculated for energy consumption;

[0006] According to the load data and the building characteristic data, using a pre-built load prediction model, obtain predicted load data of the building;

[0007] Fitting is performed based on the predicted load data of the building and the temperature difference data of the time period corresponding to the predicted load data to obtain a vector change relationship between the equivalent thermal resistance of the building and the temperature difference data at multiple time scales;

[0008] According to the vector variation relationship between the equivalent thermal resistance and temperature difference data of the building at multiple time scales, an equivalent thermal resistance model is obtained;

[0009] The load prediction model is obtained by training a long short-term memory neural network using the time series load data of the building in the historical time period and the building characteristic data in the historical time period as training data.

[0010] Optionally, the load forecasting model includes the following training process:

[0011] Using the time series load data of the building in the historical time period and the building characteristic data in the historical time period as the input of the training data;

[0012] Using the actual load data in the historical time period as output of training data;

[0013] Based on the input and output of the training data, the long short-term memory neural network is trained to obtain a load forecasting model.

[0014] Optionally, the long short-term memory neural network is trained based on the input and output of the training data to obtain a load forecasting model, including:

[0015] Based on the input of the training data, a predicted output value corresponding to the input is obtained through a long short-term memory neural network;

[0016] The root mean square error between the actual load data corresponding to the input of the training data and the predicted output value is used as the loss function, and the parameters of the long short-term memory neural network are iteratively updated. When the parameters of the long short-term memory neural network reach the preset convergence conditions, the load forecasting model is obtained.

[0017] Optionally, fitting the predicted load data of the building and the temperature difference data of the time period corresponding to the predicted load data to obtain the vector change relationship between the equivalent thermal resistance and the temperature difference data of the building at multiple time scales includes:

[0018] Dividing the predicted load data of the building and the temperature difference data of the time period corresponding to the predicted load data by time scale to obtain the predicted load data and temperature difference data under different time scales;

[0019] The predicted load data and temperature difference data at different time scales are fitted respectively by a nonlinear regression algorithm to obtain the vector variation relationship between the equivalent thermal resistance and the temperature difference data of the building at multiple time scales.

[0020] Optionally, the expression corresponding to the equivalent thermal resistance model is as follows:

[0021]

[0022] Wherein, R represents the equivalent thermal resistance of the building; ΔT represents the temperature difference data; and Q represents the predicted load data.

[0023] Optionally, the building characteristic data includes one or more of the following: meteorological data, personnel work and rest data, and load characteristic data.

[0024] Based on the same inventive concept, the present invention also provides an equivalent thermal resistance model construction system, comprising:

[0025] A data acquisition module, used to acquire load data and building characteristic data of the building whose energy consumption is to be calculated;

[0026] A data driving module, for obtaining predicted load data of the building according to the load data and the building characteristic data using a pre-built load prediction model;

[0027] A physical driving module, used for fitting the predicted load data of the building and the temperature difference data of the time period corresponding to the predicted load data, to obtain the vector change relationship between the equivalent thermal resistance of the building and the temperature difference data at multiple time scales;

[0028] A thermal resistance determination module, used to obtain an equivalent thermal resistance model according to a vector variation relationship between the equivalent thermal resistance and temperature difference data of the building at multiple time scales;

[0029] The load prediction model is obtained by training a long short-term memory neural network using the time series load data of the building in the historical time period and the building characteristic data in the historical time period as training data.

[0030] Optionally, the equivalent thermal resistance model building system further includes: a model training module, which is used to:

[0031] Using the time series load data of the building in the historical time period and the building characteristic data in the historical time period as the input of the training data;

[0032] Using the actual load data in the historical time period as output of training data;

[0033] Based on the input and output of the training data, the long short-term memory neural network is trained to obtain a load forecasting model.

[0034] Optionally, the model training module includes:

[0035] A prediction output submodule, used to obtain a prediction output value corresponding to the input through a long short-term memory neural network based on the input of the training data;

[0036] The parameter updating submodule is used to iteratively update the parameters of the long short-term memory neural network by taking the root mean square error between the actual load data corresponding to the input of the training data and the predicted output value as the loss function, and obtain the load forecasting model when the parameters of the long short-term memory neural network reach the preset convergence conditions.

[0037] Optionally, the physical driving module includes:

[0038] A scale division submodule is used to divide the predicted load data of the building and the temperature difference data of the time period corresponding to the predicted load data into time scales to obtain the predicted load data and temperature difference data under different time scales;

[0039] The data fitting submodule is used to fit the predicted load data and temperature difference data at different time scales respectively through a nonlinear regression algorithm to obtain the vector change relationship between the equivalent thermal resistance and temperature difference data of the building at multiple time scales.

[0040] Optionally, the expression corresponding to the equivalent thermal resistance model is as follows:

[0041]

[0042] Wherein, R represents the equivalent thermal resistance of the building; ΔT represents the temperature difference data; and Q represents the predicted load data.

[0043] Optionally, the building characteristic data includes one or more of the following: meteorological data, personnel work and rest data, and load characteristic data.

[0044] Based on the same inventive concept, the present invention also provides a method for calculating building energy consumption, comprising:

[0045] Obtain the indoor and outdoor temperature difference data of the building whose energy consumption is to be calculated;

[0046] According to the indoor and outdoor temperature difference data of the building, using an equivalent thermal resistance model, the building load demand of the building is obtained;

[0047] Calculating the energy consumption of the building according to the building load demand;

[0048] The equivalent thermal resistance model is constructed using the equivalent thermal resistance model construction method as described above.

[0049] Optionally, obtaining the building load demand of the building according to the indoor and outdoor temperature difference data of the building by using an equivalent thermal resistance model includes:

[0050] Outputting the equivalent thermal resistance of the building using an equivalent thermal resistance model according to the indoor and outdoor temperature difference data of the building;

[0051] The building load demand of the building is obtained according to the indoor and outdoor temperature difference data of the building and the ratio of the equivalent thermal resistance.

[0052] Optionally, calculating the energy consumption of the building according to the building load demand includes:

[0053] The energy consumption of the building is calculated according to the building load demand and the obtained efficiency conversion weight of the building.

[0054] Optionally, the calculation formula corresponding to the energy consumption of the building is as follows:

[0055]

[0056] Wherein, E represents the energy consumption of the building; D represents the building load demand; and cop represents the efficiency conversion weight corresponding to the building.

[0057] Based on the same inventive concept, the present invention also provides a building energy consumption calculation system, comprising:

[0058] A data acquisition module is used to obtain the indoor and outdoor temperature difference data of the building whose energy consumption is to be calculated;

[0059] A load demand output module, used to obtain the building load demand of the building according to the indoor and outdoor temperature difference data of the building using an equivalent thermal resistance model;

[0060] An energy consumption calculation module, used to calculate the energy consumption of the building according to the building load demand;

[0061] The equivalent thermal resistance model is constructed using the equivalent thermal resistance model construction method as described above.

[0062] Optionally, the load demand calculation module includes:

[0063] A thermal resistance output submodule, for outputting the equivalent thermal resistance of the building according to the indoor and outdoor temperature difference data of the building using an equivalent thermal resistance model;

[0064] The demand calculation submodule is used to obtain the building load demand of the building according to the indoor and outdoor temperature difference data of the building and the ratio of the equivalent thermal resistance.

[0065] Optionally, the energy consumption calculation module is used to:

[0066] The energy consumption of the building is calculated according to the building load demand and the obtained efficiency conversion weight of the building.

[0067] Optionally, the calculation formula corresponding to the energy consumption of the building is as follows:

[0068]

[0069] Wherein, E represents the energy consumption of the building; D represents the building load demand; and cop represents the efficiency conversion weight corresponding to the building.

[0070] In another aspect, the present invention further provides an electronic device, comprising: at least one processor and a memory; the memory and the processor are connected via a bus;

[0071] The memory is used to store one or more programs;

[0072] When the one or more programs are executed by the at least one processor, an equivalent thermal resistance model construction method or a building energy consumption calculation method as described above is implemented.

[0073] On the other hand, the present invention also provides a computer-readable storage medium having an execution program stored thereon, and when the execution program is executed, an equivalent thermal resistance model construction method or a building energy consumption calculation method as described above is implemented.

[0074] Compared with the prior art, the present invention has the following beneficial effects:

[0075] 1. The present invention provides a method and system for constructing an equivalent thermal resistance model, including: obtaining load data and building characteristic data of a building whose energy consumption is to be calculated; obtaining predicted load data of the building by using a pre-constructed load prediction model according to the load data and the building characteristic data; fitting the predicted load data of the building and the temperature difference data of the corresponding period of the predicted load data to obtain the vector change relationship between the equivalent thermal resistance and the temperature difference data of the building at multiple time scales; obtaining an equivalent thermal resistance model according to the vector change relationship between the equivalent thermal resistance and the temperature difference data of the building at multiple time scales; wherein the load prediction model is obtained by training a long short-term memory neural network with the time series load data of the building in the historical time period and the building characteristic data in the historical time period as training data; the present application integrates data-driven and physical-driven to construct an equivalent thermal resistance model, and by introducing building characteristic data in the data-driven process, it is beneficial to increase the interpretability between data and the accuracy of load prediction, and by establishing the vector change relationship between the equivalent thermal resistance and the temperature difference data in the physical-driven process, it is beneficial to formulate a more effective energy management strategy and improve the energy efficiency of the building;

[0076] 2. The present invention also provides a method and system for calculating building energy consumption, including: obtaining indoor and outdoor temperature difference data of a building whose energy consumption is to be calculated; obtaining the building load demand of the building according to the indoor and outdoor temperature difference data of the building using an equivalent thermal resistance model; calculating the energy consumption of the building according to the building load demand; the present application applies the equivalent thermal resistance model constructed by integrating data-driven and physical-driven methods to the energy consumption calculation of the building, which can reflect the changes in energy consumption of the building under different temperature difference conditions, and is conducive to accurately evaluating the load demand of the building and increasing the accuracy of the building energy consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 A schematic diagram of a process for constructing an equivalent thermal resistance model provided by the present invention;

[0078] Figure 2 A schematic diagram of the training process of a load prediction model in a method for constructing an equivalent thermal resistance model provided by the present invention;

[0079] Figure 3 A schematic diagram of the overall process framework of a method for constructing an equivalent thermal resistance model provided by the present invention;

[0080] Figure 4 A schematic diagram of a flow chart of a building energy consumption calculation method provided by the present invention;

[0081] Figure 5 A schematic diagram showing the change of the indoor and outdoor temperature difference at 25°C with the time of day in a specific embodiment of a method for calculating building energy consumption provided by the present invention;

[0082] Figure 6 A schematic diagram showing the variation of the heating load of each floor at 25°C with the time of the day in a specific embodiment of a method for calculating building energy consumption provided by the present invention;

[0083] Figure 7 A schematic diagram of comparing the predicted values ​​and the actual values ​​of a multi-feature daily model in a specific embodiment of a building energy consumption calculation method provided by the present invention;

[0084] Figure 8 A schematic diagram showing the change of equivalent thermal resistance of each floor with cooling at 25°C over a day in a specific embodiment of a method for calculating building energy consumption provided by the present invention;

[0085] Fig. 9 A schematic diagram of the structural composition of an equivalent thermal resistance model building system provided by the present invention;

[0086] Fig.10 A schematic diagram of the structure of a building energy consumption calculation system provided by the present invention;

[0087] Fig.11 A schematic diagram of an electronic device. DETAILED DESCRIPTION

[0088] The present invention proposes an equivalent thermal resistance model construction, a building energy consumption calculation method, a system, a device and a medium. The specific implementation methods of the present invention are further described in detail below with reference to the accompanying drawings.

[0089] Embodiment 1:

[0090] The present invention provides a method for constructing an equivalent thermal resistance model, the flow chart of which is as follows: Figure 1 As shown, including:

[0091] Step 1: Obtain the load data and building characteristic data of the building whose energy consumption is to be calculated;

[0092] Step 2: Based on the load data and building characteristic data, the predicted load data of the building is obtained using the pre-built load prediction model;

[0093] Step 3: Fit the predicted load data of the building and the temperature difference data of the corresponding period of the predicted load data to obtain the vector change relationship between the equivalent thermal resistance of the building and the temperature difference data at multiple time scales;

[0094] Step 4: Obtain an equivalent thermal resistance model based on the vector variation relationship between the equivalent thermal resistance and temperature difference data of the building at multiple time scales;

[0095] The load forecasting model is obtained by training the Long Short-Term Memory (LSTM) neural network using the time series load data of the building in the historical time period and the building characteristic data in the historical time period as training data.

[0096] Generally, when a building is subjected to load forecasting, the model is trained by relying on the historical load data of the building's own energy consumption. However, if the load forecasting model is constructed only by using the historical load data, it will be overly dependent on the historical load data, resulting in a lag in the forecasting. To solve this problem, when the load forecasting model is constructed, in addition to considering the historical time series load data, the present invention also introduces the building characteristic data including the factors affecting the building's heating energy consumption into the construction process of the load forecasting model, thereby producing a correction effect on the model constructed by only considering the historical time series load data. For example, the above-mentioned building characteristic data may include one or more of the following: meteorological data ( For example, temperature, humidity, radiation, etc.), social development level (such as GDP and energy prices), date type information (such as holidays and working days), personnel schedule data and load characteristic data (such as building type, building purpose, building structure, building energy efficiency, etc.). For the time series load data of the building, it can be collected through the smart meter or building automation system installed inside the building; for the building characteristic data, it can be obtained through public websites, building drawings, construction records, meteorological station data, etc.; according to the time series load data and building characteristic data obtained in the historical time period, the load prediction model of the building can be trained, specifically:

[0097] In one implementation, the load forecasting model may include the following training process:

[0098] Using the time series load data of the building in the historical time period and the building characteristic data in the historical time period as the input of the training data;

[0099] The actual load data in the historical time period is used as the output of training data;

[0100] Based on the input and output of the training data, the long short-term memory neural network is trained to obtain a load forecasting model;

[0101] In this implementation, in order to make the acquired data more accurate and take into account the energy demand and characteristics of different users, the time series load data in the above historical time period can also be obtained by simulating the construction of the building's indoor exterior envelope structure (such as wall materials, window-to-wall ratio design, building area and number of floors and other detailed parameters) and indoor heat source design parameters (such as equipment, number of personnel, personnel work and rest, etc.) through 3D dynamic simulation software (such as TRNSYS building energy system simulation software or DEST building energy consumption simulation software). The time series load data obtained in this way has high time resolution and is more accurate and reliable; and in the process of constructing the load prediction model, by combining a variety of building feature data (such as meteorological conditions, social development status, date type, personnel work and rest and load characteristics), the LSTM neural network is used to simulate and train the historical time series load data and predict it, which can more comprehensively capture the various factors affecting the load change, so that the constructed model can adapt to the load prediction needs of different building types, different regions and different climatic conditions, thereby significantly improving the accuracy of the prediction.

[0102] In this implementation, the process of training the long short-term memory neural network based on the input and output of the training data to obtain the load forecasting model may include:

[0103] Based on the input of training data, the predicted output value corresponding to the input is obtained through the long short-term memory neural network;

[0104] The root mean square error between the actual load data corresponding to the input of the training data and the predicted output value is used as the loss function, and the parameters of the long short-term memory neural network are iteratively updated. When the parameters of the long short-term memory neural network reach the preset convergence conditions, the load forecasting model is obtained; in this implementation, if Figure 2 As shown in the figure, when the neural network prediction algorithm (i.e., LSTM neural network prediction algorithm) is applied to train and predict the load prediction model, the historical time series load data and building characteristic data are used as input parameters for feature mining and data processing. There is no necessary rule for adjusting the network parameters (such as the number of hidden layers, the number of hidden layer nodes, the number of iterations, and the batch size) during the training process. When analyzing the influence of one of the parameters on the prediction results, the model evaluation index Root Mean Square Error (RMSE) or Mean Absolute Error (MAE) is the smallest, and the model prediction accuracy (ACC) value is the highest. At this time, the network parameter value is the best. Through repeated experimental comparison, a set of parameter combinations with the best model effect is obtained, and the loss function first decreases and then tends to be stable, indicating that the model has successfully converged, which ensures the stability and reliability of the training process.

[0105] For example, when training the load forecasting model, the present invention selects to use the root mean square error RMSE or the mean absolute error MAE as the loss function, and uses the prediction accuracy ACC as the evaluation function to evaluate the performance and effect of the forecasting model. The specific calculation method is as follows:

[0106] The expression of root mean square error RMSE can be as follows:

[0107]

[0108] Among them, RMSE represents the root mean square error value of the LSTM neural network; represents the ith actual load data; y i Represents the i-th predicted load data; i=1…N; N represents the total number of data.

[0109] The expression of mean absolute error MAE is as follows:

[0110]

[0111] Among them, MAE represents the mean absolute error value of the LSTM neural network; represents the ith actual load data; y i Represents the i-th predicted load data; i=1…N; N represents the total number of data.

[0112] The calculation formula corresponding to the prediction accuracy ACC is as follows:

[0113]

[0114] Among them, ACC represents the prediction accuracy value of the LSTM neural network; represents the ith actual load data; y i Represents the i-th predicted load data; i=1…N; N represents the total number of data; the model prediction accuracy can be used to represent the degree of closeness between the model prediction value and the true value, thereby evaluating the quality of the model.

[0115] In this implementation, the use of LSTM neural network is conducive to capturing long-term dependencies in time series data, which is crucial for time series forecasting tasks such as load forecasting. In addition, by optimizing network parameters (such as the number of hidden layers, the number of hidden layer nodes, the number of iterations, and the batch size) during model training, the prediction accuracy, model convergence, and generalization ability can be further improved.

[0116] In the actual application process, the load prediction data of the building can be obtained by inputting the actual characteristic data of the building into the load prediction model obtained by the above training. Combined with the indoor and outdoor temperature difference data of the building in the same period, the vector change relationship between the equivalent thermal resistance and the indoor and outdoor temperature difference can be fitted through physical driving. Specifically:

[0117] In one implementation, the process of fitting the predicted load data of the building and the temperature difference data of the time period corresponding to the predicted load data to obtain the vector change relationship between the equivalent thermal resistance of the building and the temperature difference data at multiple time scales may include:

[0118] Divide the predicted load data of the building and the temperature difference data of the corresponding period of the predicted load data into time scales to obtain the predicted load data and temperature difference data at different time scales;

[0119] Through the nonlinear regression algorithm, the predicted load data and temperature difference data at different time scales are fitted respectively, and the vector change relationship between the equivalent thermal resistance and temperature difference data of the building at multiple time scales is obtained;

[0120] In this implementation, based on the predicted load data output by the predicted load model and the temperature difference data of the indoor and outdoor temperatures in the same period, the nonlinear regression algorithm can identify and fit the vector variation relationship between the nonlinear equivalent thermal resistance and the indoor and outdoor temperature difference under multiple time scales (such as hourly, daily, and yearly). The nonlinear regression algorithm can handle the complex relationship between variables, especially when these relationships are not linear. In the relationship between the thermal resistance and the temperature difference of the building, there are often multiple nonlinear factors, such as the thermal inertia of the building materials, the change of the external environment, etc. The nonlinear regression algorithm can fit these complex nonlinear relationships, so as to more accurately describe the vector variation relationship between the equivalent thermal resistance and the temperature difference data of the building at different time scales, which is conducive to formulating more effective energy management strategies, improving the energy utilization efficiency of the building, and reducing energy consumption and carbon emissions. For example, in the period of large temperature difference, the operating parameters of the heating or cooling system can be adjusted to optimize energy use and reduce energy consumption. In addition, this implementation can reveal the heat exchange characteristics of the building under different climatic conditions, thereby enhancing the adaptability of the building to different climatic conditions, which is conducive to maintaining a stable indoor environment under extreme weather conditions and improving the comfort of living and working.

[0121] For example, the expression corresponding to the above equivalent thermal resistance model can be as follows:

[0122]

[0123] Among them, R represents the equivalent thermal resistance of the building; Q represents the predicted load data; ΔT represents the temperature difference data, which can be expressed as: ΔT = T h -Tc , T h Indicates the indoor temperature, T c Represents outdoor temperature; by calculating the equivalent thermal resistance of the building through this example, we can intuitively understand the heat exchange capacity of the building under specific conditions, which is conducive to evaluating the thermal insulation performance and thermal isolation effect of the building, and provide a basis for building design and improvement; and by combining the predicted load data Q with the temperature difference data ΔT, it can reflect the heat exchange characteristics of the building under different climatic conditions, thereby enhancing the adaptability of the building to different climatic conditions and accurately predicting the load changes of the building under different temperature difference conditions, which is conducive to formulating energy management strategies in advance, optimizing energy use, and reducing unnecessary energy consumption.

[0124] In summary, if the existing equivalent thermal resistance model construction method only adopts pure physical drive, it is difficult to establish mathematical expression, which makes it difficult to carry out practical application; if it adopts pure data drive, it relies on historical data as features and has low interpretability. In order to improve the practicality and interpretability of the equivalent thermal resistance model, the present invention uses the historical load data and building feature data of the building as training data on the basis of 3D simulation building analysis, constructs a load prediction model in a data-driven manner, and uses physical model information in combination with indoor and outdoor temperature difference data to empower the load prediction model, thereby establishing Figure 3 The equivalent thermal resistance model that integrates data-driven and physical-driven models shown in the figure will have more reasonable, more explanatory, more accurate and more reliable calculation results than the equivalent thermal resistance model constructed by pure data-driven or pure physical-driven models, which will help provide data support for the optimization of building energy supply system operation.

[0125] Embodiment 2:

[0126] The present invention based on the same inventive concept also provides a method for calculating building energy consumption, the flow chart of which is as follows: Figure 4 As shown, including:

[0127] Step S1: Acquire indoor and outdoor temperature difference data of the building whose energy consumption is to be calculated;

[0128] Step S2: According to the indoor and outdoor temperature difference data of the building, the building load demand of the building is obtained by using the equivalent thermal resistance model;

[0129] Step S3: Calculate the energy consumption of the building according to the building load demand;

[0130] The equivalent thermal resistance model is constructed using an equivalent thermal resistance model construction method as described above.

[0131] In one implementation, the process of obtaining the building load demand of the building based on the indoor and outdoor temperature difference data of the building using the equivalent thermal resistance model may include:

[0132] According to the indoor and outdoor temperature difference data of the building, the equivalent thermal resistance of the building is output using the equivalent thermal resistance model;

[0133] According to the indoor and outdoor temperature difference data of the building and the ratio of the equivalent thermal resistance, the building load demand of the building is obtained;

[0134] In this implementation, by using the equivalent thermal resistance model and combining the indoor and outdoor temperature difference data, the load demand and energy consumption of the building can be evaluated more accurately. This method takes into account the heat exchange characteristics of the building and can reflect the changes in the energy consumption of the building under different temperature difference conditions, providing a scientific basis for building energy consumption management. Accurate building load demand and energy consumption assessment is conducive to optimizing energy use. Building managers can adjust the operating parameters of heating, cooling and other systems according to changes in load demand, avoid energy waste, and improve energy efficiency.

[0135] In one implementation, the process of calculating the energy consumption of a building according to the building load demand may include:

[0136] Calculate the energy consumption of the building based on the building load demand and the efficiency conversion weight of the building;

[0137] In this implementation, the selection of the comprehensive efficiency conversion weight of the heating and cooling system in the building can refer to the standard value of the comprehensive efficiency conversion weight of the heating and cooling system in the "Green Performance Calculation Standard for Civil Buildings" as shown in Table 1:

[0138] Table 1 Conversion weights of comprehensive efficiency of heating and cooling systems

[0139]

[0140]

[0141] The efficiency of heating and cooling systems in different climate zones and building types (residential buildings, public buildings) varies. By referring to the standard values ​​in Table 1, more accurate parameters can be provided for energy consumption calculation, thereby improving the accuracy of calculation. In this implementation, the efficiency conversion weight is introduced to take into account the efficiency differences of heating and cooling systems in different climate zones, which is conducive to more accurately reflecting the actual energy consumption of building heating and cooling systems, thereby promoting regional energy management.

[0142] For example, the calculation formula corresponding to the energy consumption of the above building can be as follows:

[0143]

[0144] Among them, E represents the energy consumption of the building (i.e. the comprehensive energy consumption of the building's heating and cooling system); D represents the building load demand; cop represents the efficiency conversion weight corresponding to the building; since the efficiency conversion weights of different buildings, different systems and different climatic conditions are different, the introduction of the efficiency conversion weight cop in the process of building energy consumption calculation can more accurately reflect the actual energy efficiency of the building's heating and cooling system, making the energy consumption calculation results closer to the actual situation, so it can objectively evaluate the energy efficiency level of the building and provide a basis for energy efficiency improvement and comparison.

[0145] In summary, when calculating the actual energy consumption of a typical building, the present invention applies the equivalent thermal resistance model constructed by integrating data-driven and physical-driven methods to actual projects, obtains the building load demand under multiple time scales, and further calculates the building energy consumption value based on the obtained building load demand and comprehensive efficiency conversion weight. Compared with the building energy consumption calculation under pure data-driven or pure physical-driven, the method of the present invention can avoid the problem of excessive reliance on historical load data, which is conducive to calculating more reasonable and accurate building energy consumption.

[0146] Embodiment 3:

[0147] Taking the rooms on each floor of a certain information port building as the research object, an energy consumption calculation method provided by the present invention is described through a specific embodiment. By investigating the energy demand and characteristics of the building, and referring to the detailed indoor and outdoor and enclosure structure parameters in the "Energy-saving Design Standard for Public Buildings", the simulation software TRNSYS is used to simulate and model the building. When the room is maintained under a stable working condition, the cooling and heating load data of the room are obtained respectively. The specific indoor and outdoor parameters are shown in Tables 2 and 3:

[0148] Table 2 Interior design parameters

[0149]

[0150]

[0151] Table 3 Enclosure structure parameters

[0152]

[0153] Taking the room cooling temperature of 25℃ as an example, the relationship between the load of each floor (including the bottom floor Q1, the middle floor Q3 and the top floor Q9) and the equivalent thermal resistance identified by the change of indoor and outdoor temperature difference over the time of the day is as follows: Figure 5 As shown, the horizontal axis represents time T (in hours: h), which is represented by the symbol ■ in the figure; the vertical axis represents the equivalent thermal resistance ▲T (in degrees Celsius: ℃).

[0154] The historical building load data obtained by modeling based on the simulation software TRNSYS is as follows Figure 6 As shown in the figure, the horizontal axis represents different times t (in hours: h) in a day, and the vertical axis represents the heating load data Q (in kilowatts: kW) at different times. The LSTM neural network deep learning algorithm is used to take the historical time series load data as input parameters for feature mining and data processing in combination with building feature data (such as meteorological data, personnel schedules, load characteristics, etc.), thereby establishing a more reasonable and more explanatory building energy consumption vector load prediction model. When constructing the load prediction model, during the training process of the model's network parameters (number of hidden layers, number of hidden layer nodes, number of iterations, batch size), a set of parameter combinations with the best model effect are obtained through repeated experimental comparisons. The corresponding network parameters are shown in Table 4, and the evaluation indicators are shown in Table 5. The corresponding predicted load is compared with the actual load. Figure 7 As shown:

[0155] Table 4 Network parameters of multi-feature daily model

[0156]

[0157] Table 5 Multi-feature daily model evaluation indicators

[0158]

[0159] Based on the obtained predicted load data and the temperature difference data of indoor and outdoor temperatures, the relationship between the nonlinear equivalent thermal resistance R (in Kelvin / kilowatt: K / kW) of each floor (bottom floor R1, middle floor R3 and top floor R9) and the time t of the day can be identified and fitted under multiple time scales, such as Figure 8 shown.

[0160] Based on the above-mentioned relationship between the change of the equivalent thermal resistance identified by the indoor and outdoor temperature difference and the load over time, when it is actually applied, the building load demand at the corresponding time within a day can be obtained through the equivalent thermal resistance model according to the indoor and outdoor temperature difference measured by the actual temperature sensor, and the corresponding building energy consumption value can be calculated based on the building load demand and the comprehensive efficiency conversion weight of the heating and cooling system in Table 1.

[0161] Through this specific embodiment, it can be explained that a building energy consumption calculation method provided by the present invention applies the equivalent thermal resistance model constructed by integrating data-driven and physical-driven methods to the energy consumption calculation of the building. The data-driven method learns the energy consumption pattern of the building based on a large amount of historical load data and building characteristic data, while the physical-driven method performs accurate calculations based on physical laws and building characteristics. The combination of the two can more accurately reflect the actual energy consumption of the building and improve the accuracy of energy consumption calculation. In summary, applying the equivalent thermal resistance model integrating data-driven and physical-driven to the energy consumption calculation of the building can bring about multiple technical effects such as improving the accuracy of energy consumption calculation, enhancing the adaptability of the model, optimizing energy management strategies, supporting smart buildings and Internet of Things technologies, promoting green buildings and sustainable development, and enhancing the credibility of building energy efficiency labels. These effects are of great significance for improving the energy utilization efficiency of buildings, reducing energy consumption and carbon emissions, and promoting green buildings and sustainable development.

[0162] Embodiment 4:

[0163] The present invention based on the same inventive concept also provides an equivalent thermal resistance model construction system, the structural composition diagram is as follows Fig. 9 As shown, including:

[0164] A data acquisition module, used to acquire load data and building characteristic data of the building whose energy consumption is to be calculated;

[0165] A data driven module is used to obtain predicted load data of a building based on load data and building characteristic data using a pre-built load prediction model;

[0166] The physical driving module is used to fit the predicted load data of the building and the temperature difference data of the corresponding period of the predicted load data to obtain the vector change relationship between the equivalent thermal resistance of the building and the temperature difference data at multiple time scales;

[0167] The thermal resistance determination module is used to obtain an equivalent thermal resistance model based on the vector variation relationship between the equivalent thermal resistance and temperature difference data of the building at multiple time scales;

[0168] Among them, the load prediction model is obtained by training the long short-term memory neural network with the time series load data of the building in the historical time period and the building characteristic data in the historical time period as training data.

[0169] For example, the above-mentioned building characteristic data may include one or more of the following: meteorological data, personnel work and rest data, and load characteristic data.

[0170] In one implementation, the above-mentioned equivalent thermal resistance model construction system may further include: a model training module, specifically used for:

[0171] Using the time series load data of the building in the historical time period and the building characteristic data in the historical time period as the input of the training data;

[0172] The actual load data in the historical time period is used as the output of training data;

[0173] Based on the input and output of the training data, the long short-term memory neural network is trained to obtain the load forecasting model.

[0174] In one implementation, the above-mentioned model training module may include:

[0175] The prediction output submodule is used to obtain the predicted output value corresponding to the input through the long short-term memory neural network based on the input of training data;

[0176] The parameter updating submodule is used to iteratively update the parameters of the long short-term memory neural network using the root mean square error between the actual load data corresponding to the input of the training data and the predicted output value as the loss function. When the parameters of the long short-term memory neural network reach the preset convergence conditions, the load forecasting model is obtained.

[0177] In one implementation, the physical driver module may include:

[0178] The scale division submodule is used to divide the predicted load data of the building and the temperature difference data of the corresponding period of the predicted load data into time scales to obtain the predicted load data and temperature difference data under different time scales;

[0179] The data fitting submodule is used to fit the predicted load data and temperature difference data at different time scales through a nonlinear regression algorithm to obtain the vector change relationship between the equivalent thermal resistance and temperature difference data of the building at multiple time scales.

[0180] For example, the expression corresponding to the above equivalent thermal resistance model can be as follows:

[0181]

[0182] Among them, R represents the equivalent thermal resistance of the building; ΔT represents the temperature difference data; Q represents the predicted load data.

[0183] Embodiment 5:

[0184] The present invention based on the same inventive concept also provides a building energy consumption calculation system, the structural composition diagram is as follows Fig.10 As shown, including:

[0185] A data acquisition module is used to obtain the indoor and outdoor temperature difference data of the building whose energy consumption is to be calculated;

[0186] The load demand output module is used to obtain the building load demand of the building based on the indoor and outdoor temperature difference data of the building using the equivalent thermal resistance model;

[0187] Energy consumption calculation module, used to calculate the energy consumption of the building according to the building load demand;

[0188] The equivalent thermal resistance model is constructed using an equivalent thermal resistance model construction method as described above.

[0189] In one implementation, the load demand calculation module may include:

[0190] The thermal resistance output submodule is used to output the equivalent thermal resistance of the building based on the indoor and outdoor temperature difference data of the building using the equivalent thermal resistance model;

[0191] The demand calculation submodule is used to obtain the building load demand of the building based on the indoor and outdoor temperature difference data of the building and the ratio of the equivalent thermal resistance.

[0192] In one implementation, the energy consumption calculation module can be used to:

[0193] The energy consumption of the building is calculated based on the building load demand and the obtained building efficiency conversion weight.

[0194] For example, the calculation formula corresponding to the energy consumption of the above building can be as follows:

[0195]

[0196] Among them, E represents the energy consumption of the building; D represents the building load demand; cop represents the efficiency conversion weight corresponding to the building.

[0197] Embodiment 6:

[0198] like Fig.11 As shown, the present invention also provides an electronic device, which may be a computer device, a single-chip device, an intelligent mobile device, etc. The electronic device in this embodiment may include a processor, a memory, a transceiver component, etc. The memory, the processor, and the transceiver component are connected via a bus; the memory may be used to store an execution program, and an exemplary execution program may include instructions; the processor is used to execute the instructions stored in the memory. The memory may also be used to store data, which may be called and / or modified when the instructions are executed.

[0199] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in a storage medium to implement corresponding method flows or corresponding functions, so as to implement the steps of an equivalent thermal resistance model construction method or a building energy consumption calculation method in the above-mentioned embodiment.

[0200] Embodiment 7:

[0201] Based on the same inventive concept, the present invention also provides a readable storage medium, specifically an electronic device readable storage medium (Memory), which is a memory device in an electronic device for storing programs and data. It can be understood that the storage medium here can include both built-in storage media in the electronic device and, of course, extended storage media supported by the electronic device. The storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space, and these instructions can be one or more execution programs (including program codes). It should be noted that the storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. The processor loads and executes one or more instructions stored in the storage medium, which can implement the steps of an equivalent thermal resistance model construction method or a building energy consumption calculation method in the above-mentioned embodiment.

[0202] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0203] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0204] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0205] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0206] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit its protection scope. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that after reading the present invention, those skilled in the art can still make various changes, modifications or equivalent substitutions to the specific implementation methods of the application, but these changes, modifications or equivalent substitutions are all within the protection scope of the claims to be approved.

Claims

1. A method for constructing an equivalent thermal resistance model, characterized in that: include: Obtain load data and building characteristic data of the building to be calculated for energy consumption; According to the load data and the building characteristic data, using a pre-built load prediction model, obtain predicted load data of the building; Fitting is performed based on the predicted load data of the building and the temperature difference data of the time period corresponding to the predicted load data to obtain a vector change relationship between the equivalent thermal resistance of the building and the temperature difference data at multiple time scales; According to the vector variation relationship between the equivalent thermal resistance and temperature difference data of the building at multiple time scales, an equivalent thermal resistance model is obtained; The load prediction model is obtained by training a long short-term memory neural network using the time series load data of the building in the historical time period and the building characteristic data in the historical time period as training data.

2. The method according to claim 1, characterized in that The load forecasting model includes the following training process: Using the time series load data of the building in the historical time period and the building characteristic data in the historical time period as the input of the training data; Using the actual load data in the historical time period as output of training data; Based on the input and output of the training data, the long short-term memory neural network is trained to obtain a load forecasting model.

3. The method according to claim 2, characterized in that The long short-term memory neural network is trained based on the input and output of the training data to obtain a load forecasting model, including: Based on the input of the training data, a predicted output value corresponding to the input is obtained through a long short-term memory neural network; The root mean square error between the actual load data corresponding to the input of the training data and the predicted output value is used as the loss function, and the parameters of the long short-term memory neural network are iteratively updated. When the parameters of the long short-term memory neural network reach the preset convergence conditions, the load forecasting model is obtained.

4. The method according to claim 1, characterized in that The fitting is performed based on the predicted load data of the building and the temperature difference data of the time period corresponding to the predicted load data to obtain the vector change relationship between the equivalent thermal resistance of the building and the temperature difference data at multiple time scales, including: Dividing the predicted load data of the building and the temperature difference data of the time period corresponding to the predicted load data by time scale to obtain the predicted load data and temperature difference data under different time scales; The predicted load data and temperature difference data at different time scales are fitted respectively by a nonlinear regression algorithm to obtain the vector variation relationship between the equivalent thermal resistance and the temperature difference data of the building at multiple time scales.

5. The method according to claim 1, characterized in that The expression corresponding to the equivalent thermal resistance model is as follows: Wherein, R represents the equivalent thermal resistance of the building; ΔT represents the temperature difference data; and Q represents the predicted load data.

6. The method according to claim 1, characterized in that The building characteristic data includes one or more of the following: meteorological data, personnel work and rest data, and load characteristic data.

7. An equivalent thermal resistance model construction system, characterized in that: include: A data acquisition module, used to acquire load data and building characteristic data of the building whose energy consumption is to be calculated; A data driving module, for obtaining predicted load data of the building according to the load data and the building characteristic data using a pre-built load prediction model; A physical driving module, used for fitting the predicted load data of the building and the temperature difference data of the time period corresponding to the predicted load data, to obtain the vector change relationship between the equivalent thermal resistance of the building and the temperature difference data at multiple time scales; A thermal resistance determination module, used to obtain an equivalent thermal resistance model according to a vector variation relationship between the equivalent thermal resistance and temperature difference data of the building at multiple time scales; The load prediction model is obtained by training a long short-term memory neural network using the time series load data of the building in the historical time period and the building characteristic data in the historical time period as training data.

8. The system according to claim 7, characterized in that Also includes: Model training module, used to: Using the time series load data of the building in the historical time period and the building characteristic data in the historical time period as the input of the training data; Using the actual load data in the historical time period as output of training data; Based on the input and output of the training data, the long short-term memory neural network is trained to obtain a load forecasting model.

9. The system according to claim 8, characterized in that The model training module includes: A prediction output submodule, used to obtain a prediction output value corresponding to the input through a long short-term memory neural network based on the input of the training data; The parameter updating submodule is used to iteratively update the parameters of the long short-term memory neural network by taking the root mean square error between the actual load data corresponding to the input of the training data and the predicted output value as the loss function, and obtain the load forecasting model when the parameters of the long short-term memory neural network reach the preset convergence conditions.

10. The system according to claim 7, characterized in that The physical driving module includes: A scale division submodule is used to divide the predicted load data of the building and the temperature difference data of the time period corresponding to the predicted load data into time scales to obtain the predicted load data and temperature difference data under different time scales; The data fitting submodule is used to fit the predicted load data and temperature difference data at different time scales respectively through a nonlinear regression algorithm to obtain the vector change relationship between the equivalent thermal resistance and temperature difference data of the building at multiple time scales.

11. The system according to claim 7, characterized in that The expression corresponding to the equivalent thermal resistance model is as follows: Wherein, R represents the equivalent thermal resistance of the building; ΔT represents the temperature difference data; and Q represents the predicted load data.

12. The system according to claim 7, characterized in that The building characteristic data includes one or more of the following: meteorological data, personnel work and rest data, and load characteristic data.

13. A method for calculating building energy consumption, characterized in that: include: Obtain the indoor and outdoor temperature difference data of the building whose energy consumption is to be calculated; According to the indoor and outdoor temperature difference data of the building, using an equivalent thermal resistance model, the building load demand of the building is obtained; Calculating the energy consumption of the building according to the building load demand; Wherein, the equivalent thermal resistance model is constructed using an equivalent thermal resistance model construction method described in any one of claims 1-6.

14. The method according to claim 13, characterized in that The method of obtaining the building load demand of the building by using an equivalent thermal resistance model based on the indoor and outdoor temperature difference data of the building includes: Outputting the equivalent thermal resistance of the building using an equivalent thermal resistance model according to the indoor and outdoor temperature difference data of the building; The building load demand of the building is obtained according to the indoor and outdoor temperature difference data of the building and the ratio of the equivalent thermal resistance.

15. The method according to claim 13, characterized in that Calculating the energy consumption of the building according to the building load demand includes: The energy consumption of the building is calculated according to the building load demand and the obtained efficiency conversion weight of the building.

16. The method according to claim 13, characterized in that The energy consumption of the building is calculated as follows: Wherein, E represents the energy consumption of the building; D represents the building load demand; and cop represents the efficiency conversion weight corresponding to the building.

17. A building energy consumption calculation system, characterized in that: include: A data acquisition module is used to obtain the indoor and outdoor temperature difference data of the building whose energy consumption is to be calculated; A load demand output module, used to obtain the building load demand of the building according to the indoor and outdoor temperature difference data of the building using an equivalent thermal resistance model; An energy consumption calculation module, used to calculate the energy consumption of the building according to the building load demand; Wherein, the equivalent thermal resistance model is constructed using an equivalent thermal resistance model construction method described in any one of claims 1-6.

18. The system of claim 17, wherein: The load demand calculation module comprises: A thermal resistance output submodule, for outputting the equivalent thermal resistance of the building according to the indoor and outdoor temperature difference data of the building using an equivalent thermal resistance model; The demand calculation submodule is used to obtain the building load demand of the building according to the indoor and outdoor temperature difference data of the building and the ratio of the equivalent thermal resistance.

19. The system of claim 17, wherein: The energy consumption calculation module is used to: The energy consumption of the building is calculated according to the building load demand and the obtained efficiency conversion weight of the building.

20. The system of claim 17, wherein: The energy consumption of the building is calculated as follows: Wherein, E represents the energy consumption of the building; D represents the building load demand; and cop represents the efficiency conversion weight corresponding to the building.

21. An electronic device, characterized in that: include: at least one processor and memory; The memory and the processor are connected via a bus; The memory is used to store one or more programs; When the one or more programs are executed by the one or more processors, an equivalent thermal resistance model construction method as described in any one of claims 1 to 6 or a building energy consumption calculation method as described in any one of claims 13 to 16 is implemented.

22. A computing device readable storage medium, characterized in that: An execution program is stored thereon, and when the execution program is executed, an equivalent thermal resistance model construction method as described in any one of claims 1 to 6 or a building energy consumption calculation method as described in any one of claims 13 to 16 is implemented.

Images