A method and apparatus for determining a thermal network model

Abstract

The embodiment of the application discloses a kind of determination method and device of thermal network model, it is related to air conditioning technical field.The method comprises: according to the preset indoor heat generation schedule and the air conditioner operating data and meteorological data of target room in first historical time period, first training sample set is constructed;According to first training sample set, the parameter identification of the thermal network model of target room is carried out, and parameter identification result is obtained;Parameter identification result and indoor heat generation schedule are substituted into model, and first thermal network model is obtained;According to the air conditioner operating data and meteorological data of target room in second historical time period, second training sample set is constructed;According to second training sample set and first thermal network model, indoor heat generation schedule is corrected, and the indoor heat generation schedule after correction is obtained;Parameter identification result and the indoor heat generation schedule after correction are substituted into model, and second thermal network model is obtained.

Description

Technical Field

[0001] This application relates to the field of air conditioning technology, and in particular to a method and apparatus for determining a thermal network model. Background Technology

[0002] With the rapid development of my country's economy and society and the gradual improvement of people's living standards, the requirements for thermal comfort and quality of building environments such as office and residential environments are also increasing. The widespread application of air conditioning systems provides a comfortable and healthy indoor environment for buildings, but it also generates a large amount of energy consumption. In today's increasingly tense energy situation, it is quite important to control the indoor temperature by accurately predicting the dynamic indoor air conditioning heat load or cooling load, thereby reducing equipment energy consumption.

[0003] To predict the air conditioning load and indoor temperature of a room, a room thermal network model can be established. However, thermal network models in related technologies often have the following problems: they ignore the influence of internal disturbances (i.e., indoor heat generation) on the air conditioning load and indoor temperature, resulting in low accuracy of the established thermal network model. Summary of the Invention

[0004] This application provides a method and apparatus for determining a thermal network model, which is used to establish a more accurate thermal network model, thereby improving the accuracy of predicting the air conditioning load or indoor temperature in a room.

[0005] In a first aspect, embodiments of this application provide a method for determining a thermal network model. The method includes: constructing a first training sample set based on a preset indoor heat generation timetable and air conditioning operation data and meteorological data of a target room within a first historical time period; the indoor heat generation timetable is used to record the indoor heat generation of the target room at different times; identifying parameters of the thermal network model of the target room based on the first training sample set to obtain parameter identification results; substituting the parameter identification results and the preset indoor heat generation timetable into the thermal network model to obtain a first thermal network model; constructing a second training sample set based on air conditioning operation data and meteorological data of the target room within a second historical time period; correcting the indoor heat generation timetable based on the second training sample set and the first thermal network model to obtain a corrected indoor heat generation timetable; and substituting the parameter identification results and the corrected indoor heat generation timetable into the thermal network model to obtain a second thermal network model.

[0006] The technical solution provided in this application provides at least the following beneficial effects: The thermal network model itself has certain physical meaning and is a relatively simple physical model with low modeling difficulty. After parameter identification of the thermal network model of the target room, a first thermal network model can be constructed based on the parameter identification results and a preset indoor heat generation timetable. The first thermal network model obtained above can predict the indoor temperature or air conditioning load of the room, but its accuracy is not high enough because the indoor heat generation timetable used in the first thermal network model is a preset value, which differs significantly from the actual value, thus the prediction result is not accurate enough. To improve the accuracy of the thermal network model, the indoor heat generation timetable is corrected based on the second training sample set and the first thermal network model, making the value of the indoor heat generation timetable closer to the actual value. In this way, a second thermal network model with higher accuracy can be obtained based on the parameter identification results and the corrected indoor heat generation timetable that is closer to the actual value. Therefore, the second thermal network model can be used to predict the air conditioning load or indoor temperature in the room, and its prediction result has higher accuracy.

[0007] In some embodiments, the above-mentioned parameter identification of the thermal network model of the target room based on the first training sample set to obtain parameter identification results includes: determining the objective function of parameter identification based on the parameters to be predicted of the thermal network model; and solving the objective function of parameter identification using a genetic algorithm based on the first training sample set to obtain parameter identification results.

[0008] It should be understood that genetic algorithms are effective at finding globally optimal solutions. They can be used to search for a set of stable and reliable globally optimal solutions for the parameters to be identified within the boundary range of the parameters in a thermal network model, facilitating subsequent model construction. Furthermore, depending on the parameters to be predicted, the objective function of the genetic algorithm for parameter identification varies, ensuring that the obtained parameter identification results are more applicable to the prediction of the corresponding parameters, thereby improving the accuracy of model predictions.

[0009] In some embodiments, when the parameter to be predicted in the above-described thermal network model is the indoor temperature of the target room, the objective function for parameter identification is:

[0010]

[0011] Where θ is the parameter to be identified, n is the total number of training samples, and t data，i Let t be the actual indoor temperature at time i. pre，i Let be the predicted indoor temperature value at time i. It is the average of the actual indoor temperature values;

[0012] Alternatively, when the parameter to be predicted in the above heat network model is the air conditioning load of the target room, the objective function for parameter identification is:

[0013]

[0014] Where θ is the parameter to be identified, n is the total number of training samples, and q data，f Let q be the actual air conditioning load of the room at time i. pre，i Let be the predicted air conditioning load of the room at time i. It is the average value of the actual room air conditioning load.

[0015] In some embodiments, the above-mentioned method of correcting the indoor heating timetable based on the second training sample set and the first thermal network model to obtain the corrected indoor heating timetable includes: determining the objective function of the correction process based on the parameters to be predicted in the thermal network model; and solving the objective function of the correction process using a genetic algorithm based on the second training sample set to obtain the corrected indoor heating timetable.

[0016] It should be understood that genetic algorithms are effective at finding globally optimal solutions. Therefore, genetic methods can be used to search for a set of stable and reliable globally optimal solutions for indoor heat generation within the boundary range of the parameters of the thermal network model, thereby correcting the indoor heat generation timetable and making it more accurate. Furthermore, depending on the parameters to be predicted, the objective function of the genetic algorithm during parameter identification will also differ, ensuring that the obtained indoor heat generation timetable is more applicable to the prediction of the corresponding parameters, thus improving the accuracy of the model's predictions.

[0017] In some embodiments, when the parameter to be predicted in the above thermal network model is the indoor temperature of the target room, the objective function of the correction process is:

[0018]

[0019] Where s is the time-by-time value of the indoor heat generation to be corrected, n is the total number of training samples, and t data，i Let t be the actual indoor temperature at time i. pre，i Let be the predicted indoor temperature value at time i. It is the average of the actual indoor temperature values;

[0020] Alternatively, when the parameter to be predicted in the above thermal network model is the air conditioning load of the target room, the objective function of the correction process is:

[0021]

[0022] Where s is the time-series value of the indoor heat generation to be corrected, n is the total number of training samples, and q data，i Let q be the actual air conditioning load of the room at time i. pre，iLet be the predicted air conditioning load of the room at time i. It is the average value of the actual room air conditioning load.

[0023] In some embodiments, the above-mentioned modification of the indoor calorific value timetable based on the second training sample set and the first thermal network model to obtain the modified indoor calorific value timetable includes: dividing the second training sample set according to the day type to obtain a training sample subset corresponding to each day type; and for each day type, modifying the indoor calorific value timetable based on the training sample subset corresponding to the day type and the first thermal network model to obtain the modified indoor calorific value timetable corresponding to the day type.

[0024] It should be understood that indoor heat generation in rooms generally differs significantly between weekdays and weekends. For example, buildings such as offices and schools tend to have higher indoor heat generation on weekdays and lower indoor heat generation on weekends; similarly, hotels and shopping malls tend to have lower indoor heat generation on weekdays and higher indoor heat generation on weekends. Therefore, dividing the training samples according to day type yields a revised indoor heat generation timetable for different day types, thereby improving the accuracy of subsequent modeling.

[0025] Secondly, embodiments of this application provide a device for determining a thermal network model. The device includes: an acquisition module, configured to construct a first training sample set based on a preset indoor heat generation timetable and air conditioning operation data and meteorological data of a target room within a first historical time period; the indoor heat generation timetable is used to record the indoor heat generation of the target room at different times. A processing module, configured to perform parameter identification on the thermal network model of the target room based on the first training sample set, obtaining parameter identification results; and substituting the parameter identification results and the indoor heat generation timetable into the thermal network model to obtain a first thermal network model. The acquisition module is further configured to construct a second training sample set based on air conditioning operation data and meteorological data of the target room within a second historical time period. The processing module is further configured to correct the indoor heat generation timetable based on the second training sample set and the first thermal network model, obtaining a corrected indoor heat generation timetable; and substituting the parameter identification results and the corrected indoor heat generation timetable into the thermal network model to obtain a second thermal network model.

[0026] In some embodiments, the above processing module is specifically used to: determine the objective function for parameter identification based on the parameters to be predicted in the thermal network model; and solve the objective function for parameter identification using a genetic algorithm based on the first training sample set to obtain the parameter identification result.

[0027] In some embodiments, when the parameter to be predicted in the thermal network model is the indoor temperature of the target room, the objective function for parameter identification is:

[0028]

[0029] Where θ is the parameter to be identified, n is the total number of training samples, and t data，i Let t be the actual indoor temperature at time i. pre，i Let be the predicted indoor temperature value at time i. It is the average of the actual indoor temperature values;

[0030] Alternatively, when the parameter to be predicted in the thermal network model is the air conditioning load of the target room, the objective function for parameter identification is:

[0031]

[0032] Where θ is the parameter to be identified, n is the total number of training samples, and q data，i Let q be the actual air conditioning load of the room at time i. pre，i Let be the predicted air conditioning load of the room at time i. It is the average value of the actual room air conditioning load.

[0033] In some embodiments, the above processing module is specifically used to: determine the objective function of the correction process based on the parameters to be predicted in the thermal network model; and solve the objective function of the correction process using a genetic algorithm based on the second training sample set to obtain the corrected indoor heat generation timetable.

[0034] In some embodiments, when the parameter to be predicted in the thermal network model is the indoor temperature of the target room, the objective function of the correction process is:

[0035]

[0036] Where s is the time-by-time value of the indoor heat generation to be corrected, n is the total number of training samples, and t data，i Let t be the actual indoor temperature at time i. pre，i Let be the predicted indoor temperature value at time i. It is the average of the actual indoor temperature values;

[0037] Alternatively, when the parameter to be predicted in the thermal network model is the air conditioning load of the target room, the objective function of the correction process is:

[0038]

[0039] Where s is the time-series value of the indoor heat generation to be corrected, n is the total number of training samples, and q data，i Let q be the actual air conditioning load of the room at time i. pre，i Let be the predicted air conditioning load of the room at time i. It is the average value of the actual room air conditioning load.

[0040] In some embodiments, the above processing module is specifically used to: divide the second training sample set according to the day type to obtain a training sample subset corresponding to each day type; and for each day type, correct the indoor heat generation timetable according to the training sample subset corresponding to the day type and the first thermal network model to obtain the corrected indoor heat generation timetable corresponding to the day type.

[0041] Thirdly, embodiments of this application provide a controller, including: one or more processors; one or more memories; wherein the one or more memories are used to store computer program code, the computer program code including computer instructions, and when the one or more processors execute the computer instructions, the controller executes the methods provided in the first aspect and possible implementations.

[0042] Fourthly, a computer-readable storage medium is provided, the computer-readable storage medium including computer instructions that, when executed on a computer, cause the computer to perform the methods provided in the first aspect and possible implementations.

[0043] Fifthly, a computer program product containing computer instructions is provided, which, when executed on a computer, causes the computer to perform the methods provided in the first aspect and possible implementations described above.

[0044] It should be noted that the aforementioned computer instructions may be stored, in whole or in part, on a computer-readable storage medium. This computer-readable storage medium may be packaged together with the controller's processor, or it may be packaged separately from the controller's processor; this application does not impose any limitations on this.

[0045] The beneficial effects described in aspects two through five of this application can be referred to the analysis of the beneficial effects of aspect one, and will not be repeated here. Attached Figure Description

[0046] The accompanying drawings are provided to further understand the technical solutions of the present invention and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of the present invention and do not constitute a limitation on the technical solutions of the present invention.

[0047] Figure 1 A schematic diagram of the system structure is shown based on a thermal network model according to some embodiments;

[0048] Figure 2 This is a flowchart illustrating a method for determining a thermal network model according to some embodiments. Figure 1 ;

[0049] Figure 3This is a schematic diagram of the spatial distribution of a typical office room according to some embodiments;

[0050] Figure 4 A schematic diagram of an equivalent thermal network model of a room according to some embodiments;

[0051] Figure 5 A flowchart illustrating another method for determining a thermal network model according to some embodiments. Figure 2 ;

[0052] Figure 6 This is a flowchart illustrating another method for determining a thermal network model according to some embodiments. Figure 3 ;

[0053] Figure 7 This is a flowchart illustrating another method for determining a thermal network model according to some embodiments. Figure 4 ;

[0054] Figure 8 This is a flowchart illustrating another method for determining a thermal network model according to some embodiments. Figure 5 ;

[0055] Figure 9 This is a flowchart illustrating another method for determining a thermal network model according to some embodiments. Figure 6 ;

[0056] Figure 10 This is a flowchart illustrating another method for determining a thermal network model according to some embodiments. Figure 7 ;

[0057] Figure 11 This is a schematic diagram of the air conditioning load prediction results of a heat network model according to some embodiments;

[0058] Figure 12 This is a schematic diagram of the structure of a control device according to some embodiments;

[0059] Figure 13 This is a schematic diagram of the hardware structure of a controller according to some embodiments. Detailed Implementation

[0060] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0061] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0062] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "connected" and "linked" should be interpreted broadly, for example, as a fixed connection, a detachable connection, or an integral connection. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances. Furthermore, when describing pipelines, the terms "connected" and "linked" as used in this application have the meaning of establishing electrical connection. The specific meaning needs to be understood in conjunction with the context.

[0063] Unless the context otherwise requires, throughout the specification and claims, the term "comprise" and its other forms, such as the third-person singular "comprises" and the present participle "comprising," are interpreted as open-ended and encompassing, meaning "including, but not limited to." In the description of the specification, terms such as "one embodiment," "some embodiments," "exemplary embodiments," "example," "specific example," or "some examples," etc., are intended to indicate that a particular feature, structure, material, or characteristic associated with that embodiment or example is included in at least one embodiment or example of this disclosure. The illustrative representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics mentioned may be included in any suitable manner in any one or more embodiments or examples.

[0064] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0065] As described in the background section, with the continuous development of public building scale in recent years, public building energy consumption has become one of the key targets for energy conservation and emission reduction. Generally, people predict indoor air conditioning load or indoor temperature by establishing load or temperature prediction models. However, in related technologies, changes in internal disturbance factors are often ignored when establishing models, resulting in low prediction accuracy.

[0066] Commonly used building load or temperature prediction models include: white box model, black box model, and gray box model.

[0067] (1) White Box Model

[0068] A white box model is a purely physical model, typically used to calculate a building's air conditioning load based on the principles of room heat exchange. Common white box modeling methods include the harmonic response method, the response coefficient method, the cooling load temperature difference method, and the total equivalent time method. This type of white box model requires a large number of known parameters, such as building physical geometry parameters, design parameters, and meteorological parameters, making the calculation process complex and the modeling difficult.

[0069] (2) Black Box Model

[0070] Black-box models rely on data-driven predictions of building loads. Typically, they require long-term actual measurement data for training, applying mathematical and statistical models to reflect the relationship between building air conditioning loads and operational data to achieve predictions. Examples of commonly used black-box modeling methods include regression analysis, time series forecasting, artificial intelligence forecasting, genetic forecasting, and neural network forecasting. However, the physical meaning of these black-box models is not clearly defined, and their reliance on large amounts of historical data for computation makes them inconvenient.

[0071] (3) Gray box model

[0072] Gray-box models combine the advantages of white-box and black-box models. Generally, gray-box models establish simple physical models, use less training data for modeling, have relatively clear physical meaning, and lower modeling costs. In related technologies, gray-box models use resistance capacitance (RC) models, also known as thermal network models. However, thermal network models in related technologies often have the following problems: they ignore the influence of internal disturbances (i.e., indoor heat generation) on the air conditioning load and indoor temperature in the room, and only identify the structural thermal characteristics of the room based on external disturbances of the thermal network model, such as outdoor temperature and solar radiation, to obtain the final prediction model; thus, the prediction model does not fully consider the influence of internal disturbances, resulting in decreased accuracy. Alternatively, they preset the indoor heat generation and identify the structural thermal characteristics of the room based on the external disturbances of the thermal network model (e.g., outdoor temperature) and the preset indoor heat generation to obtain the final prediction model; however, because the preset indoor heat generation is not accurate enough, the accuracy of the trained prediction model is also relatively low.

[0073] To address this, this application provides a method for determining a thermal network model. This method establishes a first thermal network model of the room to be tested based on a preset indoor heat generation timetable and a training sample set within a first historical time period. Then, the preset indoor heat generation timetable in the first thermal network model is corrected to make its values ​​closer to actual values. Thus, based on the parameter identification results and the corrected indoor heat generation timetable, a more accurate second thermal network model can be obtained, thereby improving the accuracy of predicting the air conditioning load or indoor temperature in the room.

[0074] To further describe the scheme of this application, Figure 1 This is a schematic diagram of a thermal network model determination system provided in this application embodiment. (Refer to...) Figure 1 The thermal network model determination system 100 includes a server 101, an air conditioner 102, and an indoor temperature sensor 103. In some embodiments, the thermal network model determination system 100 further includes an outdoor temperature sensor 104.

[0075] In some embodiments, server 101 is used to execute the thermal network model determination method described in the embodiments of this application. Server 101 may be a single server, or it may be a server cluster consisting of multiple servers. In some embodiments, the server cluster may also be a distributed cluster. This application does not limit the specific form of server 101. Optionally, server 101 can obtain meteorological data, such as outdoor temperature values ​​and solar radiation, through a network. Optionally, server 101 may pre-store a trained thermal network model to predict the air conditioning load or indoor temperature value of a target room.

[0076] In some embodiments, the air conditioner 102 establishes a communication connection with the server 101 via wired or wireless means for heating or cooling a target room. Optionally, the air conditioner 102 can provide the server 101 with historical and real-time air conditioning supply (i.e., air conditioning load). Optionally, the air conditioner 101 includes a controller (not shown in the figure), which can be used to execute the thermal network model determination method described in the embodiments of this application.

[0077] In some embodiments, the indoor temperature sensor 103 establishes a communication connection with the server 101 via wired or wireless means to detect the indoor temperature. Optionally, the indoor temperature sensor 103 can be integrated into the indoor unit of the air conditioner 102. Optionally, the indoor temperature sensor 103 can be a sensor independent of the air conditioner 102, located in the target room to be predicted.

[0078] In some embodiments, the outdoor temperature sensor 104 establishes a communication connection with the server 101 via wired or wireless means to detect the outdoor temperature. Optionally, the outdoor temperature sensor 104 can be integrated into the outdoor unit of the air conditioner 102. Optionally, the outdoor temperature sensor 104 can be a sensor independent of the air conditioner 102, located outside the target room to be predicted.

[0079] The embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0080] This application provides a method for determining a thermal network model, applied to a thermal network model determining apparatus. For example, the thermal network model determining apparatus may be... Figure 1 The system shown may include server 101 or air conditioner 102; this is not limited. For example... Figure 2 As shown, the method may include the following steps:

[0081] S101. The thermal network model determination device constructs a first training sample set based on a preset indoor heat generation timetable and the air conditioning operation data and meteorological data of the target room within the first historical time period.

[0082] The target room is the room whose air conditioning load or indoor temperature is to be predicted.

[0083] An indoor heat generation time table is used to record the indoor heat generation of a target room at different times. Indoor heat generation is the indoor heating power per unit area, also known as indoor heating power density.

[0084] In some embodiments, the air conditioning operation data includes: indoor temperature value and air conditioning supply (i.e., air conditioning load). The air conditioning operation data may also include other parameters related to air conditioning operation, which are not limited in this application embodiment.

[0085] In some embodiments, meteorological data includes outdoor temperature and solar radiation. Meteorological data may also include other climate or weather-related parameters (e.g., humidity), which are not limited in this application.

[0086] In some embodiments, the first training sample set includes several training samples, and each training sample includes the indoor heat generation of the target room at a certain time, various air conditioning operation data, and various meteorological data.

[0087] In some embodiments, the thermal network model determination device can obtain historical indoor temperature values ​​through temperature sensors built into the indoor unit of the air conditioning system, or through other temperature sensors installed in the room. Air conditioning supply includes both cooling and heating capacity. The thermal network model determination device can obtain historical air conditioning supply data from a server, the air conditioner controller, or the air conditioner's storage.

[0088] In some embodiments, the thermal network model determination device can obtain historical outdoor temperature values ​​from a meteorological monitoring department or a temperature sensor built into the outdoor unit; in addition, the thermal network model determination device can obtain future outdoor temperature values, historical solar radiation, and future solar radiation from a meteorological monitoring department.

[0089] In some embodiments, indoor heat generation includes: heat generation power per unit area of ​​people, lighting power per unit area of ​​lighting, and equipment power per unit area of ​​equipment. The indoor heat generation timetable can be preset according to relevant departmental standards or specifications.

[0090] For example, the preset indoor heat generation timetable can be obtained with reference to the "Energy Conservation Design Standard for Public Buildings". For instance, as shown in Tables 1 to 6:

[0091] Table 1

[0092]

[0093] Table 1 shows the lighting on / off schedules for some public buildings as specified in the "Energy Conservation Design Standard for Public Buildings". The operating periods include 24 periods, each corresponding to one hour. Lighting on / off times are percentages, referring to the proportion of lighting on time within each operating period. The building categories in Table 1 include office buildings, hotel buildings, and shopping mall buildings. In some embodiments, building categories may also include residential buildings, hospital buildings, and campus buildings (not shown in Table 1), which will not be elaborated upon further below.

[0094] Table 2

[0095]

[0096] Table 2 shows the hourly occupancy rate of some public buildings as specified in the "Energy Conservation Design Standard for Public Buildings". The operating period includes 24 operating periods, each corresponding to one hour; the hourly occupancy rate is a percentage value, referring to the proportion of time people spend indoors within each operating period.

[0097] Table 3

[0098]

[0099] Table 3 shows the hourly equipment utilization rates of some public buildings as specified in the "Energy Conservation Design Standard for Public Buildings". The operating period includes 24 operating periods, each corresponding to one hour; the equipment utilization rate is a percentage value, referring to the proportion of equipment usage time within each operating period.

[0100] Table 4

[0101] Architecture <![CDATA[Floor area per capita (m 2 / person)]]> office buildings 10 Hotel Building 25 Shopping mall building 8

[0102] Table 4 shows the per capita building area of ​​some different types of buildings as shown in the "Energy Conservation Design Standard for Public Buildings".

[0103] Table 5

[0104] Architecture <![CDATA[Device power (W / m 2 )]]> office buildings 15 Hotel Building 15 Shopping mall building 13

[0105] Table 5 shows the power density of equipment in some different types of buildings as indicated in the "Energy Conservation Design Standard for Public Buildings".

[0106] Table 6

[0107] Architecture <![CDATA[Illuminance power density (W / m 2 )]]> office buildings 9 Hotel Building 7 Shopping mall building 10

[0108] Table 6 shows the lighting power density of some different types of buildings as indicated in the "Energy Conservation Design Standard for Public Buildings".

[0109] Taking office buildings as an example, when the operating period is the 10th time slot on a weekday, according to Table 4, the average area per person in an office building is 10m². 2 / person, assuming the heat output of a single person is aW / person, then when people are constantly indoors, the heat output per unit area is: Considering that people do not stay in the building permanently, we can refer to Table 2 for the hourly occupancy rate of the rooms. According to Table 2, the occupancy rate during the 10th time period on a weekday is 95%. Therefore, the heat output per unit area of ​​people during the 10th time period on a weekday should be: That is, 0.095*a W / m 2 .

[0110] In addition, according to Table 5, the heat dissipation power of equipment in office buildings per unit area is generally 15W / m². 2 However, this equipment power density value is based on the equipment's power density under constant operating conditions (also known as equipment power per unit area). Therefore, we can refer to Table 3 for the hourly equipment utilization rate. According to Table 3, the equipment utilization rate during the 10th time period on a weekday is 95%. Therefore, the equipment power per unit area during the 10th time period on a weekday is 95% * 15 W / m. 2 That is, 14.25W / m 2 .

[0111] Furthermore, according to Table 6, the lighting power density (also known as lighting power per unit area) of office buildings is generally 9 W / m². 2 However, this lighting power density value is based on the lighting power when the lighting equipment is constantly on. Therefore, we can refer to the lighting switch schedule shown in Table 1. According to Table 1, the lighting switch time accounts for 95% of the total time during the 10th time period on weekdays. Therefore, the lighting power per unit area during the 10th time period on weekdays can be considered as 95% * 9 W / m. 2 That is, 8.55W / m 2 .

[0112] Furthermore, by adding the personnel heat generation power density per unit area, the equipment power per unit area, and the lighting power per unit area during the 10th time period of the workday, the indoor heat generation per unit area during the 10th time period of the workday can be obtained as the indoor heat generation power per unit area during the 10th time period of the workday. Similarly, the indoor heat generation of the office building during other time periods can be obtained according to Tables 1 to 6 above. In this way, the indoor heat generation of the aforementioned office building during each time period can be obtained, and the indoor heat generation for each time period can be organized in tables as a preset heat generation timetable.

[0113] Based on this, in the early stages of establishing the thermal network model, the indoor heat generation timetable can be preset according to the standards or specifications of relevant departments to obtain a heat generation timetable that is closer to the actual operating conditions, thereby improving the accuracy of subsequent training of the first thermal network model.

[0114] In other embodiments, the preset indoor heat generation timetable may be a timetable obtained from actual surveys.

[0115] For example, we will still take an office building as an example for actual research. Based on the actual planning or usage of the office, we will obtain the per capita usable area. With lighting and equipment constantly on, we will measure the heat generation power per unit area of ​​the room to be predicted, and thus obtain the lighting power density and equipment heat generation power density of the room when people are frequently present and lighting and equipment are constantly on. This is shown in Table 7 below:

[0116] Table 7

[0117] <![CDATA[Per capita usable area (m 2 / person)]]> <![CDATA[Illuminance power density (W / m 2 )]]> <![CDATA[Device power density (W / m 2 )]]> 3.5 10.5 21

[0118] Assuming the heat output of a single person is 35W / person, according to Table 7, the average floor space per person in an office building is 3.5m². 2 If the number of people is 10W / m², then when people remain indoors, the heat output per unit area is 10W / m². 2 Furthermore, according to Table 7, the heat dissipation power of equipment in office buildings per unit area is generally 21W / m². 2 The lighting power per unit area is generally 10.5W / m². 2 Under ideal conditions, where people are constantly in the room and lighting and equipment are always on, the ideal indoor heat generation is (10 + 10.5 + 21) W / m². 2 That is, 41.5W / m 2 .

[0119] Considering that people are not indoors all the time, and that lighting and other equipment are used at different times, it is sometimes necessary to conduct a survey of the actual indoor heat generation. Table 8 shows the time-sharing table of indoor heat generation percentages based on actual surveys:

[0120] Table 8

[0121] Running time period 1 2 3 4 5 6 7 8 9 10 11 12 Indoor heat generation percentage (%) 0 0 0 0 0 0 0 21 58 67 67 45 Running time period 13 14 15 16 17 18 19 20 21 22 23 24 Indoor heat generation percentage (%) 56 67 67 67 67 48 18 6.4 1.1 0 0 0

[0122] The operating period consists of 24 segments, each lasting one hour. The heat generation percentage is a value representing the ratio of the actual indoor heat generation to the ideal indoor heat generation within each operating segment. Taking the 10th operating segment as an example, as shown in Table 7 above, the ideal indoor heat generation is 41.5 W / m². 2 According to Table 8, the indoor heat generation during the 10th operating period accounts for 67%, therefore, the indoor heat generation during the 10th operating period is (41.5 * 67%) W / m². 2 That is, 27.805W / m 2Similarly, the indoor heat generation of the office building during other time periods can be obtained from Tables 7 and 8 above. In this way, the indoor heat generation of the aforementioned office building at various time periods can be obtained and organized into tables as a preset heat generation timetable.

[0123] Based on this, in the early stages of establishing the thermal network model, the indoor heat generation timetable can be preset according to the actual survey to obtain a heat generation timetable that is closer to the actual operation, so as to improve the accuracy of subsequent training of the first thermal network model.

[0124] S102. Based on the first training sample set, the thermal network model determination device identifies the parameters of the thermal network model of the target room and obtains the parameter identification results.

[0125] The target room is the room whose air conditioning load or indoor temperature is to be predicted. The thermal network model of the target room is a capacitive-resistive model established based on the target room.

[0126] It should be understood that the idea of ​​thermoelectric analogy can be applied to equate the thermal network model of the target room to an electrical network model: the heat flow between different nodes in the target room is analogous to current; the temperature difference is analogous to potential difference; the heat storage and release capacity of the building is regarded as heat capacity, analogous to capacitance in a circuit; the thermal resistance of the building is regarded as thermal resistance, analogous to resistance in a circuit; and the heat flow of solar radiation, air conditioning cooling capacity, and indoor heat generation as input nodes is analogous to the current source in a circuit.

[0127] Taking a typical office room as an example, assuming the target room has a volume of 5×5×3m 3 The spatial distribution diagram of the target room is as follows: Figure 3 The rectangular space shown.

[0128] Furthermore, such as Figure 4 As shown, Figure 3 The diagram shows the equivalent heat network model of the room.

[0129] Reference Figure 4 Where C1 and C2 are the equivalent heat capacities of the outer envelope, in J / K (joules per Kelvin); C Z C1 represents the equivalent heat capacity of indoor air, in J / K; C3 and C4 represent the equivalent heat capacity of indoor heat storage bodies, in J / K; T0 represents the equivalent temperature node of outdoor air, in K; T1 represents the equivalent temperature node of the outer surface of the external envelope, an intermediate parameter, in K (Kelvin); T2 represents the equivalent temperature node of the inner surface of the external envelope, an intermediate parameter, in K; T ZT1 represents the indoor air equivalent temperature node, i.e., the indoor temperature, in K; T3 and T4 represent the indoor heat storage body equivalent temperature nodes, which are intermediate parameters, in K; R1 represents the outdoor air equivalent thermal resistance, in K / W (Kelvin / Watt); R2 represents the external envelope equivalent thermal resistance, in K / W; R Z R1 represents the equivalent thermal resistance of the indoor air, in K / W; R3 and R4 represent the equivalent thermal resistance of the indoor heat storage body, in K / W; R win Q represents the equivalent thermal resistance of the exterior window structure, expressed in K / W. HVAC The air conditioning system supply, i.e., the room's air conditioning load, is Q when the air conditioning is providing heating. HVAC When the air conditioner is cooling, Q is a positive value. HVAC It is a negative value, and the unit is W; Q int Indoor heat generation, unit: W / m² 2 Q s Solar radiation, measured in W / m² 2 α1, α2, β1, γ1, and γ2 are correction factors for indoor heat generation.

[0130] Based on the principle of thermal equilibrium, we can list Figure 4 The relationships in the heat network model shown are as follows:

[0131]

[0132]

[0133]

[0134]

[0135]

[0136] Furthermore, when identifying the parameters of the thermal network model of the target room, the parameters that need to be identified are shown in Table 9 below. As can be seen from Table 9, the parameters to be identified include: resistive parameters, that is, the heat capacity parameters of the thermal network model, including... Figure 4 The equivalent heat network model diagram shown includes R1, R2, and R... Z R3, R4, and R win Capacitive parameters, also known as heat capacity parameters of the thermal network model, include Figure 4 The equivalent heat network model diagram shown includes C1, C2, and C... Z C3, C4; in addition, the parameters to be identified also include: correction coefficients α1, α2, β1, γ1, γ2.

[0137] Table 9

[0138] Parameter Classification Parameters to be identified resistive parameters <![CDATA[R1、R2、R Z 、R3、R4、R win ]]> Capacitive parameters <![CDATA[C1、C2、C Z 、C3、C4]]> Correction coefficient <![CDATA[α1、α2、β1、γ1、γ2]]>

[0139] In some embodiments, such as Figure 5 As shown above, based on the first training sample set, the parameters of the thermal network model of the target room are identified to obtain the parameter identification results. The specific implementation is as follows:

[0140] S1021. Based on the parameters to be predicted in the thermal network model, the thermal network model determination device determines the objective function for parameter identification.

[0141] In some examples, when the parameter to be predicted is the indoor temperature of the room, the objective function for parameter identification is:

[0142]

[0143] Where θ is the parameter to be identified, n is the total number of training samples, and t data，i Let t be the actual indoor temperature at time i. pre，i Let be the predicted indoor temperature value at time i. It is the average of the actual indoor temperature.

[0144] In other examples, when the parameter to be predicted is the room's air conditioning load, the objective function for parameter identification is:

[0145]

[0146] Where θ is the parameter to be identified, n is the total number of training samples, and q data，i Let q be the actual air conditioning load of the room at time i, i.e., the actual air conditioning supply. pre，i Let be the predicted air conditioning load for the room at time i, which is also the predicted air conditioning supply. It is the average value of the actual air conditioning load in the room.

[0147] S1022. Based on the first training sample set, the thermal network model determination device uses a genetic algorithm to solve the objective function of parameter identification and obtain the parameter identification result.

[0148] Genetic algorithms are search algorithms used in computational mathematics to solve optimization problems; they are a type of evolutionary algorithm. Evolutionary algorithms were initially developed by drawing inspiration from phenomena in evolutionary biology, including heredity, mutation, natural selection, and hybridization. Genetic algorithms are typically implemented through computer simulation. For an optimization problem, a population of abstract representations (called chromosomes) of a certain number of candidate solutions (called individuals) evolves towards better solutions.

[0149] In some embodiments, still with Figure 3 For example, in the room shown, Figure 6 As shown, step S1022 above is specifically implemented as follows:

[0150] Sa1, Determine the genetic strategy.

[0151] In some examples, the above-mentioned determination of genetic strategies includes determining the selection operator, crossover operator, and mutation operator in the genetic process.

[0152] Sa2. Set the generation counter t1 = 0, set the maximum generation N1, and randomly generate M1 individuals as the initial population P1(0) for the parameters to be identified.

[0153] Sa3, calculate the fitness of each individual in the population.

[0154] In some examples, the fitness of each individual in the above-mentioned population is specifically implemented as follows: the parameters to be predicted are obtained according to the above formulas (1)-(5), and the fitness of each individual is calculated according to the predicted parameters and the objective function. For example, when the parameter to be predicted is the indoor temperature of the room and the objective function is formula (6), the smaller the absolute value of the objective function, the higher the fitness of the corresponding individual. When the parameter to be predicted is the air conditioning load of the room and the objective function is formula (7), the smaller the absolute value of the objective function, the higher the fitness of the corresponding individual.

[0155] Sa4: Apply the selection operator to the population.

[0156] It should be understood that the purpose of selection is to directly pass on optimized individuals to the next generation or to generate new individuals through crossover and then pass them on to the next generation. Selection is based on an assessment of the fitness of individuals within the population. For example, individuals with higher fitness are more likely to be selected for inheritance.

[0157] Sa5: Apply the crossover operator to the population.

[0158] It should be understood that crossover generates new individuals that are then passed on to the next generation, with a probability of producing individuals with better fitness. Crossover is the core step of the genetic algorithm.

[0159] Sa6: Apply the mutation operator to the population.

[0160] It should be understood that applying the mutation operator to the population means changing the gene values ​​at certain loci in the individual strings of the population. Based on this, there is a probability of generating individuals with better fitness. Generally, the mutation probability in genetic algorithms is less than the crossover probability.

[0161] After selection, crossover, and mutation operations, population Sa7 and population P1(t1) are used to obtain the next generation population P1(t1+1). Each time a genetic cycle is completed, the generation counter t1 is incremented by one.

[0162] Sa8. If t1 = N1, then the individual with the highest fitness obtained during the evolutionary process is output as the optimal solution, and the calculation is terminated.

[0163] It should be noted that the genetic algorithm steps provided in the embodiments of this application are merely examples. In some embodiments, the above-described genetic algorithm may also incorporate strategies such as cluster extinction, elimination, and population isolation to optimize the genetic algorithm. The embodiments of this application do not impose any limitations on this.

[0164] It should be understood that genetic algorithms are effective at finding global optima. Therefore, genetic methods can be used to search for a set of stable and reliable global optima within the parameter boundaries of a thermal network model, which can then be used for subsequent model construction. Furthermore, depending on the parameters to be predicted, the objective function of the genetic algorithm for parameter identification will vary, ensuring that the obtained parameter identification results are more applicable to the prediction of the corresponding parameters, thereby improving the accuracy of model predictions.

[0165] S103. The thermal network model determination device substitutes the parameter identification results and the indoor heat generation timetable into the above thermal network model to obtain the first thermal network model.

[0166] It should be understood that by substituting the parameter identification results and the indoor heat generation timetable into the aforementioned heat network model to obtain the first heat network model, the first heat network model itself has the ability to make simple predictions about the air conditioning load or indoor temperature in the room. However, considering that the indoor heat generation timetable of the first heat network model is a preset indoor heat generation timetable, which may differ significantly from the actual indoor heat generation timetable, it is advisable to consider modifying the indoor heat generation timetable based on the first heat network model to improve the accuracy of indoor heat generation, thereby improving the accuracy of model predictions.

[0167] S104. The thermal network model determination device constructs a second training sample set based on the air conditioning operation data and meteorological data of the target room during the second historical time period.

[0168] The second historical time period differs from the first historical time period. Indoor heat generation is the indoor heating power per unit area, i.e., indoor heating power density. Air conditioning operation data includes: indoor temperature value and air conditioning supply. Meteorological data includes: outdoor temperature value and solar radiation. The target room is the room whose air conditioning load or indoor temperature is to be predicted. The second training sample set includes several training samples. Each training sample includes various air conditioning operation data and meteorological data for the target room at a given time.

[0169] In some embodiments, the thermal network model determination device can obtain historical indoor temperature values ​​through temperature sensors built into the indoor unit of the air conditioning system, or through other temperature sensors installed in the room. Air conditioning supply, also known as air conditioning load, includes both cooling and heating capacity supplied by the air conditioning system. The thermal network model determination device can obtain historical air conditioning supply data from a server, the air conditioning controller, or the air conditioning storage device.

[0170] In some embodiments, the thermal network model determination device can obtain historical outdoor temperature values ​​from a meteorological monitoring department or a temperature sensor built into the outdoor unit; in addition, the thermal network model determination device can obtain future outdoor temperature values, historical solar radiation, and future solar radiation from a meteorological monitoring department.

[0171] In some embodiments, the data values ​​in the second training sample set can be hourly data values; or data values ​​every 10 minutes; or, depending on the actual application, a sample set at other time intervals. This application embodiment does not limit this.

[0172] Optionally, when the indoor heat generation timetable to be corrected is the indoor heat generation timetable of the target room every 10 minutes in the second time period, the second training sample set is the air conditioning operation data and meteorological data of the target room every 10 minutes in the second time period.

[0173] Optionally, when the indoor heat generation timetable to be corrected is the hourly indoor heat generation timetable of the target room within the second time period, the second training sample set consists of the hourly air conditioning operation data and meteorological data of the target room within the second time period.

[0174] It should be understood that when the corrected indoor heating timetable is a 10-minute value, the corresponding sample set should also be a 10-minute sample set; when the corrected indoor heating timetable is an hourly value, the corresponding sample set should also be an hourly sample set. In this way, the dimensions of the sample set correspond to the indoor heating timetable to be corrected, improving the accuracy of the indoor heating timetable correction.

[0175] In some embodiments, the second training sample set further includes a preset indoor heat generation timetable. In this case, the second sample set includes several training samples, each training sample including the indoor heat generation of the target room at a given time, various air conditioning operation data, and various meteorological data. The indoor heat generation is obtained from the preset indoor heat generation timetable.

[0176] Optionally, when the indoor heat generation timetable to be corrected is a 10-minute indoor heat generation timetable of the target room within the second time period, such as... Figure 7 As shown, the thermal network model determination device performs the following steps:

[0177] Sb1. Determine the preset 10-minute indoor heat generation timetable.

[0178] For example, an hourly indoor heating timetable can be preset using relevant departmental standards or specifications, or a preset hourly indoor heating timetable can be obtained through research. The preset method can be referred to the description in step S101 above, and will not be repeated here. The preset hourly indoor heating timetable is linearly processed to determine the preset 10-minute indoor heating timetable. For example, for an indoor heating timetable with 24 operating time periods, when the preset hourly indoor heating timetable shows that the indoor heating of the 20th operating time period is 20W / m²... 2 Furthermore, the indoor heat generation during the 21st operating period was 26 W / m². 2 At that time, the indoor heat generation during the 20th operating period is 20W / m². 2 As the indoor heat generation at 20:00, the indoor heat generation during the 21st operating period is 26W / m². 2 The indoor heat output at 21:00 is considered as follows. The change in indoor heat output from 20:00 to 21:00 is assumed to be a uniform linear change, i.e., the indoor heat output at 20:00, 20:10, 20:20, 20:30, 20:40, 20:50, and 21:00 is considered to be 20 W / m². 2 21W / m 2 22W / m 2 23W / m 2 24W / m 2 25W / m 2 26W / m 2 By analogy, a preset 10-minute timetable for indoor heat generation is determined.

[0179] Sb2. Based on the preset 10-minute indoor heat generation timetable, and the 10-minute air conditioning operation data and meteorological data of the target room in the second time period, construct the second sample dataset.

[0180] It should be understood that when the corrected indoor heating timetable is a 10-minute value, the corresponding sample set should also be a 10-minute sample set; when the corrected indoor heating timetable is an hourly value, the corresponding sample set should also be an hourly sample set. In this way, the dimensions of the sample set correspond to the indoor heating timetable to be corrected, improving the accuracy of the indoor heating timetable correction.

[0181] S105. Based on the second training sample set and the first thermal network model, the thermal network model determination device corrects the indoor heat generation timetable to obtain the corrected indoor heat generation timetable.

[0182] In some embodiments, such as Figure 8 As shown, step S105 is specifically implemented as follows:

[0183] S1051a. Based on the day type, the second training sample set is divided to obtain the training sample subsets corresponding to each day type.

[0184] For example, day types can be divided into weekdays and holidays.

[0185] S1052a. For each day type, the indoor heat generation timetable is corrected based on the training sample subset corresponding to the day type and the first thermal network model, so as to obtain the corrected indoor heat generation timetable corresponding to the day type.

[0186] It should be understood that indoor heat generation in rooms generally differs significantly between weekdays and weekends. For example, buildings such as offices and schools tend to have higher indoor heat generation on weekdays and lower indoor heat generation on weekends; similarly, hotels and shopping malls tend to have lower indoor heat generation on weekdays and higher indoor heat generation on weekends. Therefore, dividing the training samples according to day type yields a revised indoor heat generation timetable for different day types, thereby improving the accuracy of subsequent modeling.

[0187] In some embodiments, such as Figure 9 As shown, step S105 is specifically implemented as follows:

[0188] S1051b. Based on the parameters to be predicted in the thermal network model, the thermal network model determination device determines the objective function of the correction process.

[0189] In some examples, when the parameter to be predicted in the thermal network model is the indoor temperature of the target room, the objective function of the correction process is:

[0190]

[0191] In formula (7), s is the time-by-time value of the indoor heat generation to be corrected, n is the total number of training samples, and t data，i Let t be the actual indoor temperature at time i. pre，i Let be the predicted indoor temperature value at time i. It is the average of the actual indoor temperature.

[0192] In other examples, when the parameter to be predicted in the thermal network model is the air conditioning load of the target room, the objective function of the correction process is:

[0193]

[0194] In formula (8), s is the time-by-time value of the indoor heat generation to be corrected, n is the total number of training samples, and q data，i Let q be the actual air conditioning load of the room at time i. pre，i Let be the predicted air conditioning load of the room at time i. It is the average value of the actual room air conditioning load.

[0195] S1052b: Based on the second training sample set, the thermal network model determination device uses a genetic algorithm to solve the objective function of the correction process and obtain the corrected indoor heat generation timetable.

[0196] In some embodiments, still with Figure 3 For example, in the room shown, Figure 10 As shown, step S1052b above is specifically implemented as follows:

[0197] Sc1, Determine the genetic strategy.

[0198] In some examples, the above-mentioned determination of genetic strategies includes determining the selection operator, crossover operator, and mutation operator in the genetic process.

[0199] Sc2, set the evolutionary generation counter t2 = 0, set the maximum evolutionary generation N2, and generate M2 individuals with the parameter to be identified as the initial population P2(0); where the parameter to be identified is the indoor heat generation.

[0200] Optionally, the above-mentioned M2 individuals that generate the parameters to be identified are used as the initial population P2(0), which can be specifically implemented as follows: M2 individuals that generate the parameters to be identified are randomly used as the initial population P2(0).

[0201] It should be understood that randomly generating the parameters to be identified has a certain probability of covering the optimal solution range of the genetic algorithm.

[0202] Optionally, the M2 individuals that generate the parameters to be identified are used as the initial population P2(0), which can be specifically implemented as follows: the preset indoor heat generation is used as the initial value to assign values ​​to the parameters to be identified.

[0203] It should be understood that the preset indoor heat generation is determined according to the industry standards of relevant departments, and therefore is closer to the optimal solution range of the genetic algorithm than random assignment.

[0204] Sc3: Calculate the fitness of each individual in the population.

[0205] In some examples, the fitness of each individual in the above-mentioned population is specifically implemented as follows: the indoor heat load to be identified is obtained according to the above formulas (1)-(5), and the fitness of each individual is calculated based on the obtained indoor heat load and objective function. For example, when the parameter to be predicted is the indoor temperature of the room and the objective function is formula (8), the smaller the absolute value of the objective function, the higher the fitness of the corresponding individual. When the parameter to be predicted is the air conditioning load of the room and the objective function is formula (9), the smaller the absolute value of the objective function, the higher the fitness of the corresponding individual.

[0206] Sc4: Apply the selection operator to the population.

[0207] It should be understood that the purpose of selection is to directly pass on optimized individuals to the next generation or to generate new individuals through crossover and then pass them on to the next generation. Selection is based on an assessment of the fitness of individuals within the population. For example, individuals with higher fitness are more likely to be selected for inheritance.

[0208] Sc5: Apply the crossover operator to the population.

[0209] It should be understood that crossover generates new individuals that are then passed on to the next generation, with a probability of producing individuals with better fitness. Crossover is the core step of the genetic algorithm.

[0210] Sc6: Apply the mutation operator to the population.

[0211] It should be understood that applying the mutation operator to the population means changing the gene values ​​at certain loci in the individual strings of the population. Based on this, there is a probability of generating individuals with better fitness. Generally, the mutation probability in genetic algorithms is less than the crossover probability.

[0212] After selection, crossover, and mutation operations, the next generation population P2(t2+1) is obtained from population P2(t2+1). The generation counter t2 is incremented by one after each genetic cycle.

[0213] Sc8. If t2 = N2, then the individual with the highest fitness obtained during the evolutionary process is output as the optimal solution, and the calculation is terminated.

[0214] It should be noted that the genetic algorithm steps provided in this application are merely examples. In actual use, the above genetic algorithm can also incorporate strategies such as cluster extinction, elimination, and population isolation to optimize the genetic algorithm. This application does not impose any limitations on this.

[0215] It should be understood that genetic algorithms are effective at finding globally optimal solutions. Therefore, genetic methods can be used to search for a set of stable and reliable globally optimal solutions for indoor heat generation within the boundary range of the parameters of the thermal network model, thereby correcting the indoor heat generation timetable and making it more accurate. Furthermore, depending on the parameters to be predicted, the objective function of the genetic algorithm during parameter identification will also differ, ensuring that the obtained indoor heat generation timetable is more applicable to the prediction of the corresponding parameters, thus improving the accuracy of the model's predictions.

[0216] S106. The thermal network model determination device substitutes the parameter identification results and the corrected indoor heat generation timetable into the thermal network model to obtain the second thermal network model.

[0217] It should be understood that by substituting the parameter identification results and the corrected indoor heat generation timetable into the heat network model to obtain the second heat network model, the air conditioning load or indoor temperature in the room can be predicted based on the second heat network model. Considering that the indoor heat generation timetable of the second heat network model is a corrected indoor heat generation timetable, which is closer to the actual indoor heat generation situation, the accuracy of the indoor heat generation timetable can be improved during the testing of the second heat network model. Furthermore, the consideration of internal disturbances during the prediction process is more precise, thereby improving the accuracy of the model's predictions.

[0218] In some embodiments, when the thermal network model determination device is applied to room air conditioning load prediction, after the thermal network model of the room to be predicted is trained, if an air conditioning load prediction instruction is received from the terminal device, the device obtains relevant meteorological data of the room to be predicted, such as solar radiation, from the network, and predicts the air conditioning load based on air conditioning operation-related data uploaded by the air conditioner, temperature sensor, or terminal device, such as indoor temperature.

[0219] For example, such as Figure 11 The diagram shown illustrates the air conditioning load prediction results of the heat network model provided in this embodiment. The horizontal axis represents time (in hours), and the vertical axis represents air conditioning load (in watts). The dashed line represents the predicted indoor air conditioning load, and the solid line represents the actual indoor air conditioning load. It can be seen that the air conditioning load predicted by the heat network model provided in this embodiment is quite close to the actual air conditioning load, and the prediction results are relatively accurate.

[0220] Furthermore, the thermal network model determination device can calculate the power consumption of the air conditioning system based on the predicted air conditioning load, and adjust the air conditioning operation mode according to the power consumption of the air conditioning system.

[0221] In some embodiments, when the thermal network model determination device is applied to room temperature prediction, when the thermal network model determination device receives a temperature prediction instruction from the terminal device, it performs indoor temperature prediction based on the air conditioning supply uploaded by the terminal device or the air conditioning supply provided by the air conditioning system, and the thermal network model trained by the dashed line.

[0222] As can be seen, the above mainly describes the solutions provided by the embodiments of this application from a methodological perspective. To achieve the above functions, the embodiments of this application provide corresponding hardware structures and / or software modules for executing each function. Those skilled in the art should readily recognize that, in conjunction with the modules and algorithm steps of the various examples described in the embodiments disclosed herein, the embodiments of this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention.

[0223] The embodiments of this application can divide the controller into functional modules according to the above method examples. For example, each function can be divided into its own functional module, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware or as a software functional module.

[0224] like Figure 12 As shown, this application embodiment provides a control device for executing the above-described determination control method for a thermal network model. The control device 500 includes:

[0225] The acquisition module 501 is used to construct a first training sample set based on a preset indoor heat generation time table and the air conditioning operation data and meteorological data of the target room within a first historical time period. The indoor heat generation time table is used to record the indoor heat generation of the target room at different times.

[0226] The processing module 502 is used to identify parameters of the thermal network model of the target room based on the first training sample set, and obtain the parameter identification results; and to substitute the parameter identification results and the indoor heat generation time table into the thermal network model to obtain the first thermal network model.

[0227] The acquisition module 501 is also used to construct a second training sample set based on the air conditioning operation data and meteorological data of the target room in the second historical time period.

[0228] The processing module 502 is also used to correct the indoor calorific value timetable based on the second training sample set and the first thermal network model to obtain the corrected indoor calorific value timetable; and to substitute the parameter identification results and the corrected indoor calorific value timetable into the thermal network model to obtain the second thermal network model.

[0229] In some embodiments, the processing module 502 is specifically used to: determine the objective function for parameter identification based on the parameters to be predicted in the thermal network model; and solve the objective function for parameter identification using a genetic algorithm based on the first training sample set to obtain the parameter identification result.

[0230] In some embodiments, when the parameter to be predicted in the thermal network model is the indoor temperature of the target room, the objective function for parameter identification is:

[0231]

[0232] Where θ is the parameter to be identified, n is the total number of training samples, and t data，i Let t be the actual indoor temperature at time i. pre，i Let be the predicted indoor temperature value at time i. It is the average of the actual indoor temperature values;

[0233] Alternatively, when the parameter to be predicted in the thermal network model is the air conditioning load of the target room, the objective function for parameter identification is:

[0234]

[0235] Where θ is the parameter to be identified, n is the total number of training samples, and q data，i Let q be the actual air conditioning load of the room at time i. pre，i Let be the predicted air conditioning load of the room at time i. It is the average value of the actual room air conditioning load.

[0236] In some embodiments, the processing module 502 is specifically used to: determine the objective function of the correction process based on the parameters to be predicted in the thermal network model; and solve the objective function of the correction process using a genetic algorithm based on the second training sample set to obtain the corrected indoor heat generation timetable.

[0237] In some embodiments, when the parameter to be predicted in the thermal network model is the indoor temperature of the target room, the objective function of the correction process is:

[0238]

[0239] Where s is the time-by-time value of the indoor heat generation to be corrected, n is the total number of training samples, and t data，iLet t be the actual indoor temperature at time i. pre，i Let be the predicted indoor temperature value at time i. It is the average of the actual indoor temperature values;

[0240] Alternatively, when the parameter to be predicted in the thermal network model is the air conditioning load of the target room, the objective function of the correction process is:

[0241]

[0242] Where s is the time-series value of the indoor heat generation to be corrected, n is the total number of training samples, and q data，i Let q be the actual air conditioning load of the room at time i. pre，i Let be the predicted air conditioning load of the room at time i. It is the average value of the actual room air conditioning load.

[0243] In some embodiments, the processing module 502 is specifically used to: divide the second training sample set according to the day type to obtain a training sample subset corresponding to each day type; and for each day type, correct the indoor heat generation timetable according to the training sample subset corresponding to the day type and the first thermal network model to obtain the corrected indoor heat generation timetable corresponding to the day type.

[0244] Figure 12 Modules in a module can also be called units; for example, a processing module can be called a processing unit.

[0245] Figure 12 If the various modules in the process are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, in essence, or the parts that contribute to the prior art, or all or part of the technical solutions, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. Storage media for storing computer software products include: USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, optical disks, and other media capable of storing program code.

[0246] This application also provides a hardware structure diagram of a controller, such as... Figure 13As shown, the controller 2000 includes a processor 2001, and optionally, a memory 2002 and a communication interface 2003 connected to the processor 2001. The processor 2001, memory 2002, and communication interface 2003 are connected via a bus 2004.

[0247] Processor 2001 may be a central processing unit (CPU), a general-purpose processor, a network processor (NP), a digital signal processor (DSP), a microprocessor, a microcontroller, a programmable logic device (PLD), or any combination thereof. Processor 2001 may also be any other device with processing capabilities, such as a circuit, device, or software module. Processor 2001 may also include multiple CPUs, and processor 2001 may be a single-core processor or a multi-core processor. Here, "processor" may refer to one or more devices, circuits, or processing cores used to process data (e.g., computer program instructions).

[0248] The memory 2002 can be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions. It can also be an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), a magnetic disk storage medium, or other magnetic storage device, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer. This application embodiment does not impose any limitations on this. The memory 2002 can exist independently or be integrated with the processor 2001. The memory 2002 may contain computer program code. The processor 2001 executes the computer program code stored in the memory 2002 to implement the control method provided in this application embodiment.

[0249] The communication interface 2003 can be used to communicate with other devices or communication networks (such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.). The communication interface 2003 can be a module, circuit, transceiver, or any device capable of enabling communication.

[0250] Bus 2004 can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. Bus 2004 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 13 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0251] This invention also provides a computer-readable storage medium including computer-executable instructions that, when executed on a computer, cause the computer to perform the method provided in the above embodiments.

[0252] This invention also provides a computer program product that can be directly loaded into a memory and contains software code. After being loaded and executed by a computer, the computer program product can implement the methods provided in the above embodiments.

[0253] Those skilled in the art will recognize that, in one or more of the examples above, the functions described in this invention can be implemented using hardware, software, firmware, or any combination thereof. When implemented in software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or code on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein communication media include any medium that facilitates the transfer of a computer program from one place to another. Storage media can be any available medium accessible to a general-purpose or special-purpose computer.

[0254] Through the above description of the embodiments, those skilled in the art can clearly understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0255] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and other division methods may exist in actual implementation. For example, multiple modules or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the shown or discussed mutual couplings, direct couplings, or communication connections may be through some interfaces; indirect couplings or communication connections between devices or modules may be electrical, mechanical, or other forms. Modules described as separate components may or may not be physically separate; components shown as modules may be one physical module or multiple physical modules, i.e., they may be located in one place or distributed in multiple different places. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0256] Furthermore, the functional units in the various embodiments of this invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the technical solution of the embodiments of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0257] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for determining a thermal network model, characterized in that, The method includes: Based on the preset indoor heat generation timetable and the air conditioning operation data and meteorological data of the target room in the first historical time period, a first training sample set is constructed. The indoor heat generation timetable is used to record the indoor heat generation of the target room at different times. Based on the first training sample set, parameter identification is performed on the thermal network model of the target room to obtain parameter identification results; Substituting the parameter identification results and the indoor heat generation timetable into the heat network model, a first heat network model is obtained; Based on the air conditioning operation data and meteorological data of the target room during the second historical time period, a second training sample set is constructed; Based on the second training sample set and the first thermal network model, the indoor heat generation timetable is corrected to obtain a corrected indoor heat generation timetable; wherein, the correction includes: The second training sample set is divided according to the day type to obtain training sample subsets corresponding to weekdays, rest days, and holidays; For each day type, the indoor heat generation timetable is corrected based on the training sample subset corresponding to the day type and the first thermal network model to obtain the corrected indoor heat generation timetable corresponding to the day type; the objective function of the correction process is determined based on the parameters to be predicted of the thermal network model; the parameters to be predicted are the indoor temperature of the target room or the air conditioning load of the target room. Substituting the parameter identification results and the corrected indoor heat generation timetable into the heat network model, a second heat network model is obtained; Wherein, when the parameter to be predicted in the thermal network model is the indoor temperature of the target room, the objective function for parameter identification is: in, The parameters to be identified The total number of training samples. For the first The actual indoor temperature at any given moment. For the first Predicted indoor temperature at a given time. It is the average of the actual indoor temperature values; Alternatively, when the parameter to be predicted in the thermal network model is the air conditioning load of the target room, the objective function for parameter identification is: in, The parameters to be identified The total number of training samples. For the first The actual air conditioning load of the room at a given time. For the first Predicted air conditioning load for the room at any given time. It is the average value of the actual air conditioning load of the room; When the parameter to be predicted in the thermal network model is the indoor temperature of the target room, the objective function of the correction process is: in, These are the time-by-time values ​​of the indoor heat generation to be corrected. The total number of training samples. For the first The actual indoor temperature at any given moment. For the first Predicted indoor temperature at a given time. It is the average of the actual indoor temperature values; Alternatively, when the parameter to be predicted in the thermal network model is the air conditioning load of the target room, the objective function of the correction process is: in, These are the time-by-time values ​​of the indoor heat generation to be corrected. The total number of training samples. For the first The actual air conditioning load of the room at a given time. For the first Predicted air conditioning load for the room at any given time. It is the average value of the actual air conditioning load of the room.

2. The method according to claim 1, characterized in that, The step of identifying parameters of the thermal network model of the target room based on the first training sample set to obtain parameter identification results includes: Based on the parameters to be predicted in the thermal network model, determine the objective function for parameter identification; Based on the first training sample set, a genetic algorithm is used to solve the objective function of the parameter identification to obtain the parameter identification result.

3. A device for determining a thermal network model, characterized in that, The device includes: The acquisition module is used to construct a first training sample set based on a preset indoor heat generation time table and air conditioning operation data and meteorological data of the target room within a first historical time period. The indoor heat generation time table is used to record the indoor heat generation of the target room at different times. The processing module is used to identify parameters of the thermal network model of the target room based on the first training sample set, and obtain parameter identification results; and to substitute the parameter identification results and the indoor heat generation time table into the thermal network model to obtain the first thermal network model. The acquisition module is also used to construct a second training sample set based on the air conditioning operation data and meteorological data of the target room during the second historical time period; The processing module is further configured to correct the indoor heat generation timetable based on the second training sample set and the first thermal network model, to obtain a corrected indoor heat generation timetable; wherein the correction includes: The second training sample set is divided according to the day type to obtain training sample subsets corresponding to weekdays, rest days, and holidays; For each day type, the indoor heat generation timetable is corrected based on the training sample subset corresponding to the day type and the first thermal network model to obtain the corrected indoor heat generation timetable corresponding to the day type; the objective function of the correction process is determined based on the parameters to be predicted of the thermal network model; the parameters to be predicted are the indoor temperature of the target room or the air conditioning load of the target room. Substituting the parameter identification results and the corrected indoor heat generation timetable into the heat network model, a second heat network model is obtained; Wherein, when the parameter to be predicted in the thermal network model is the indoor temperature of the target room, the objective function for parameter identification is: in, The parameters to be identified The total number of training samples. For the first The actual indoor temperature at any given moment. For the first Predicted indoor temperature at a given time. It is the average of the actual indoor temperature values; Alternatively, when the parameter to be predicted in the thermal network model is the air conditioning load of the target room, the objective function for parameter identification is: in, The parameters to be identified The total number of training samples. For the first The actual air conditioning load of the room at a given time. For the first Predicted air conditioning load for the room at any given time. It is the average value of the actual air conditioning load of the room; When the parameter to be predicted in the thermal network model is the indoor temperature of the target room, the objective function of the correction process is: in, These are the time-by-time values ​​of the indoor heat generation to be corrected. The total number of training samples. For the first The actual indoor temperature at any given moment. For the first Predicted indoor temperature at a given time. It is the average of the actual indoor temperature values; Alternatively, when the parameter to be predicted in the thermal network model is the air conditioning load of the target room, the objective function of the correction process is: in, These are the time-by-time values ​​of the indoor heat generation to be corrected. The total number of training samples. For the first The actual air conditioning load of the room at a given time. For the first Predicted air conditioning load for the room at any given time. It is the average value of the actual air conditioning load of the room.

4. The apparatus according to claim 3, characterized in that, The processing module is specifically used for: Based on the parameters to be predicted in the thermal network model, determine the objective function for parameter identification; Based on the first training sample set, a genetic algorithm is used to solve the objective function of the parameter identification to obtain the parameter identification result.

Images