Air conditioner load demand response potential analysis method and system

Abstract

The invention discloses an air conditioner load demand response potential analysis method and system. The method comprises the following steps: acquiring historical load and influence factor data of a building air conditioner system, and performing correlation analysis and screening to obtain key influence factor characteristics; constructing a first air conditioner load prediction model based on the screening result and a historical load curve, and pre-training the first air conditioner load prediction model; the pre-training model is adjusted through the historical load and feature data of the target building and the real-time collected data, and a second air conditioner load prediction model is obtained; and based on the second air conditioner load prediction model, the final demand response potential value is obtained by changing the input air conditioner temperature set value and carrying out difference calculation on the model prediction result output. According to the method, different environments and requirements can be adapted, cooperative work is realized by utilizing high-performance computing resources of the cloud and real-time data of the terminal side based on a transfer learning framework, and the prediction precision and the operation efficiency of the model are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of air conditioning system control and optimization, and in particular to an air conditioning load demand response potential analysis method and system. Background Art

[0002] The load demand response of traditional building air-conditioning systems faces the challenges of complex environmental changes and energy consumption. In order to improve the intelligent response capability of air-conditioning systems, an advanced air-conditioning load demand response potential analysis method is proposed based on the transfer learning framework and combined with deep learning technology. Through comprehensive analysis of historical load and influencing factor data, as well as prediction and fine-tuning with the help of deep learning models, this method can accurately estimate the demand response potential of air-conditioning systems in real-time scenarios, thereby achieving effective use of energy. Summary of the invention

[0003] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides an air conditioning load demand response potential analysis method and system to solve the problem that the load demand response of the traditional building air conditioning system cannot adapt to complex environmental changes and energy efficient utilization.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, an embodiment of the present invention provides an air conditioning load demand response potential analysis method, comprising: obtaining historical load and influencing factor data of a building air conditioning system, performing correlation analysis on the influencing factors based on the historical load data set, and screening key influencing factor features;

[0008] Constructing a first air conditioning load prediction model based on the screening results and the historical load curve, and pre-training the first air conditioning load prediction model using the historical load and influencing factor data of the building air conditioning system;

[0009] The pre-trained model is adjusted using the historical load and characteristic data of the target building and the real-time collected data to obtain a second air conditioning load prediction model;

[0010] Based on the second air-conditioning load prediction model, the final demand response potential value is obtained by changing the input air-conditioning temperature setting value and performing a difference calculation on the model prediction result output.

[0011] As a preferred solution of the air conditioning load demand response potential analysis method of the present invention, wherein: performing correlation analysis on the influencing factors based on the historical load data set, screening the key influencing factor features includes: using the Spearman rank correlation coefficient to analyze the correlation between the influencing factors and the load, quantifying the degree of influence of each feature on the load demand, and screening to obtain the minimum dimensional feature set, the calculation formula of which is expressed as:

[0012]

[0013] Among them, ρ is the correlation coefficient, d is the position difference corresponding to the two variables, and n is the number of load samples.

[0014] As a preferred solution of the air conditioning load demand response potential analysis method of the present invention, wherein: constructing a first air conditioning load prediction model based on the screening results and the historical load curve includes: the first air conditioning load prediction model structure includes a convolutional neural network, a bidirectional long short-term memory network, a flattening layer and a fully connected layer;

[0015] The convolutional neural network is composed of a one-dimensional convolutional layer, a one-dimensional maximum pooling layer and a random dropout layer. The convolutional neural network extracts spatial features and local patterns in the original load data through the convolutional layer, and is used for load classification prediction through the combination of multi-scale features;

[0016] The bidirectional long short-term memory network is composed of several pairs of independent LSTM layers and random loss layers, and the bidirectional long short-term memory network processes input prediction data through several pairs of independent LSTM layers.

[0017] As a preferred solution of the air-conditioning load demand response potential analysis method described in the present invention, wherein: pre-training the first air-conditioning load prediction model using the historical load and influencing factor data of the building air-conditioning system includes: the first air-conditioning load prediction model is pre-trained using an adaptive learning rate mechanism and batch normalization technology.

[0018] As a preferred solution of the air conditioning load demand response potential analysis method of the present invention, wherein: using the historical load and characteristic data of the target building and the real-time collected data to adjust the pre-trained model to obtain the second air conditioning load prediction model includes: fixing the weight and bias parameters of the convolution layer, based on the historical load and characteristic data set of the target building, retraining the second air conditioning load prediction model for other layers;

[0019] The second air-conditioning load prediction model is deployed on the edge intelligent terminal and calculated using a lightweight model compression algorithm.

[0020] As a preferred solution of the air conditioning load demand response potential analysis method of the present invention, wherein: based on the second air conditioning load prediction model, by changing the input air conditioning temperature setting value and performing a difference calculation on the model prediction result output to obtain the final demand response potential value includes:

[0021] Determine the air conditioning temperature setting range according to the user comfort constraint, simulate the user's demand response event by inputting different air conditioning temperature setting values ​​into the second air conditioning load prediction model, and set the air conditioning load prediction value of 27 degrees Celsius as the baseline;

[0022] By changing the temperature setting value of the model input, the demand response potential value under different temperature setting values ​​is obtained by subtracting each set of model output from the load forecast value of the baseline.

[0023] In a second aspect, the present invention provides an air conditioning load demand response potential analysis system, comprising: a data acquisition and analysis module, for acquiring historical load and influencing factor data of a building air conditioning system, performing correlation analysis on the influencing factors based on the historical load data set, and screening key influencing factor characteristics;

[0024] A model building pre-training module is used to build a first air conditioning load prediction model based on the screening results and the historical load curve, and to pre-train the first air conditioning load prediction model using the historical load and influencing factor data of the building air conditioning system;

[0025] A model adjustment module, used to adjust the pre-trained model using the historical load and characteristic data of the target building and the real-time collected data to obtain a second air conditioning load prediction model;

[0026] The demand response analysis module is used to obtain a final demand response potential value based on the second air conditioning load prediction model by changing the input air conditioning temperature setting value and performing a difference calculation on the model prediction result output.

[0027] As a preferred solution of the air conditioning load demand response potential analysis system of the present invention, the demand response analysis module includes:

[0028] Determine the air conditioning temperature setting range according to the user comfort constraint, simulate the user's demand response event by inputting different air conditioning temperature setting values ​​into the second air conditioning load prediction model, and set the air conditioning load prediction value of 27 degrees Celsius as the baseline;

[0029] By changing the temperature setting value of the model input, the demand response potential value under different temperature setting values ​​is obtained by subtracting each set of model output from the load forecast value of the baseline.

[0030] In a third aspect, the present invention provides a computing device, comprising:

[0031] Memory and processor;

[0032] The memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions. When the computer executable instructions are executed by the processor, the steps of the air conditioning load demand response potential analysis method are implemented.

[0033] In a fourth aspect, the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the air-conditioning load demand response potential analysis method.

[0034] Compared with the prior art, the present invention has the following beneficial effects: the present invention obtains the historical load and influencing factor data of the building air-conditioning system, performs correlation analysis on the influencing factors based on the historical load data set, and screens the key influencing factor characteristics; constructs the first air-conditioning load prediction model based on the screening results and the historical load curve, and pre-trains the first air-conditioning load prediction model using the historical load and influencing factor data of the building air-conditioning system; adjusts the pre-trained model using the historical load and characteristic data of the target building and the real-time collected data to obtain the second air-conditioning load prediction model; based on the second air-conditioning load prediction model, the final demand response potential value is obtained by changing the input air-conditioning temperature setting value and performing a difference calculation on the output of the model prediction result. It can adapt to different environments and needs, and realizes collaborative work based on the migration learning framework by using the high-performance computing resources of the cloud and the real-time data on the terminal side, which effectively improves the prediction accuracy and operation efficiency of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:

[0036] Figure 1 A method flow chart of an air conditioning load demand response potential analysis method and system according to an embodiment of the present invention;

[0037] Figure 2 A deep learning model architecture diagram of an air conditioning load demand response potential analysis method and system according to an embodiment of the present invention;

[0038] Figure 3 A schematic diagram of a transfer learning strategy of an air conditioning load demand response potential analysis method and system according to an embodiment of the present invention;

[0039] Figure 4A comparison diagram of load prediction before terminal response of an air conditioning load demand response potential analysis method and system according to an embodiment of the present invention;

[0040] Figure 5 A load prediction comparison diagram after terminal-side response of an air-conditioning load demand response potential analysis method and system according to an embodiment of the present invention;

[0041] Figure 6 A comparison diagram of load prediction before and after response of an air conditioning load demand response potential analysis method and system according to an embodiment of the present invention;

[0042] Figure 7 A comparison diagram of load prediction before and after response of an air conditioning load demand response potential analysis method and system according to an embodiment of the present invention;

[0043] Figure 8 A comparison diagram of the actual load before and after the response of an air conditioning load demand response potential analysis method and system according to an embodiment of the present invention;

[0044] Fig. 9 An adjustable potential curve diagram of an air conditioning load demand response potential analysis method and system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0046] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0047] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0048] The present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0049] At the same time, in the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper, lower, inner and outer" are based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as limiting the present invention. In addition, the terms "first, second or third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0050] In the present invention, unless otherwise clearly specified and limited, the terms "install, connect, connect" should be understood in a broad sense, for example: it can be a fixed connection, a detachable connection or an integral connection; it can also be a mechanical connection, an electrical connection or a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0051] Example 1

[0052] Reference Figure 1 to Figure 3 , is an embodiment of the present invention, which provides an air conditioning load demand response potential analysis method, comprising:

[0053] S100: Obtain historical load and influencing factor data of the building air conditioning system, perform correlation analysis on the influencing factors based on the historical load data set, and screen key influencing factor characteristics;

[0054] S200: constructing a first air conditioning load prediction model based on the screening results and the historical load curve, and pre-training the first air conditioning load prediction model using the historical load and influencing factor data of the building air conditioning system;

[0055] S300: Using the historical load and characteristic data of the target building and the real-time collected data to adjust the pre-trained model to obtain a second air conditioning load prediction model;

[0056] S400: Obtain a final demand response potential value based on a second air conditioning load prediction model by changing an input air conditioning temperature setting value and performing a difference calculation on the model prediction result output.

[0057] It should be noted that the historical load and influencing factor data of the building air-conditioning system are obtained by deploying sensor networks and Internet of Things technology to build a comprehensive actual building air-conditioning system data collection platform. Based on this platform, the historical load and various potential influencing factor data of the target building are obtained in real time, including indoor and outdoor temperature, humidity, wind speed, pedestrian flow and other multi-dimensional information, to achieve high-frequency data collection of the operating status of the air-conditioning system, ensuring the comprehensiveness and timeliness of the data.

[0058] It should also be noted that before conducting correlation analysis on influencing factors based on historical load data sets, data preprocessing is required to remove data points containing missing values ​​or outliers to ensure the stability of model training; secondly, the data is normalized to map the data of each dimension to a unified scale to prevent certain features from having a dominant influence on the model. In addition, smoothing technology is used to denoise the load curve, thereby improving the interpretability of the data and the robustness of the model.

[0059] In the embodiment of the present application, the correlation analysis of the influencing factors is performed based on the historical load data set, and the screening of key influencing factor features includes: using the Spearman rank correlation coefficient to analyze the correlation between the influencing factors and the load, quantifying the degree of influence of each feature on the load demand, and screening to obtain the minimum dimensional feature set, the calculation formula of which is expressed as:

[0060]

[0061] Among them, ρ is the correlation coefficient, d is the position difference corresponding to the two variables, and n is the number of load samples.

[0062] In an embodiment of the present application, constructing a first air conditioning load prediction model based on the screening results and the historical load curve includes: the first air conditioning load prediction model structure includes a convolutional neural network, a bidirectional long short-term memory network, a flattening layer, and a fully connected layer;

[0063] The convolutional neural network consists of a one-dimensional convolutional layer, a one-dimensional maximum pooling layer, and a random dropout layer. The convolutional neural network extracts spatial features and local patterns in the original load data through the convolutional layer, and uses the combination of multi-scale features to classify and predict loads.

[0064] The bidirectional long short-term memory network consists of several pairs of independent LSTM layers and random loss layers. The bidirectional long short-term memory network processes the input prediction data through several pairs of independent LSTM layers.

[0065] In an embodiment of the present application, pre-training the first air-conditioning load prediction model using the historical load of the building air-conditioning system and the influencing factor data includes: pre-training the first air-conditioning load prediction model using an adaptive learning rate mechanism and batch normalization technology.

[0066] It should be noted that the pre-training process can enable the model to better adapt to the operating rules of different buildings and reduce the difficulty of training the model on the target building.

[0067] In an embodiment of the present application, the second air conditioning load prediction model is obtained by adjusting the pre-trained model using the historical load and characteristic data of the target building and the real-time collected data, including: fixing the weight and bias parameters of the convolutional layer, and retraining the second air conditioning load prediction model for other layers based on a small sample data set of the historical load and characteristic data of the target building;

[0068] The second air-conditioning load prediction model is deployed on the edge intelligent terminal and calculated using a lightweight model compression algorithm.

[0069] It should be noted that the optimized model is deployed on the edge intelligent terminal, and a lightweight model compression algorithm is used to reduce the size and computational burden of the model, ensuring efficient operation on the edge device. In addition, a distributed system architecture is used to distribute the model's reasoning tasks to multiple terminals for processing, improving the overall concurrency and reliability of the system.

[0070] In the embodiment of the present application, based on the second air conditioning load prediction model, the final demand response potential value is obtained by changing the input air conditioning temperature setting value and performing a difference calculation on the model prediction result output, including:

[0071] Determine the air conditioning temperature setting range according to the user comfort constraint, simulate the user's demand response event by inputting different air conditioning temperature setting values ​​into the second air conditioning load prediction model, and set the air conditioning load prediction value of 27 degrees Celsius as the baseline;

[0072] By changing the temperature setting value of the model input, the demand response potential value under different temperature setting values ​​is obtained by subtracting each set of model output from the load forecast value of the baseline.

[0073] The above is a schematic scheme of an air conditioning load demand response potential analysis method of this embodiment. It should be noted that the technical scheme of this air conditioning load demand response potential analysis system and the technical scheme of the air conditioning load demand response potential analysis method described above belong to the same concept. For details not described in detail in the technical scheme of the air conditioning load demand response potential analysis system in this embodiment, please refer to the description of the technical scheme of the air conditioning load demand response potential analysis method described above.

[0074] In this embodiment, an air conditioning load demand response potential analysis system includes:

[0075] The data acquisition and analysis module is used to obtain the historical load and influencing factor data of the building air-conditioning system, conduct correlation analysis on the influencing factors based on the historical load data set, and screen the key influencing factor characteristics;

[0076] A model building pre-training module is used to build a first air conditioning load prediction model based on the screening results and the historical load curve, and to pre-train the first air conditioning load prediction model using the historical load and influencing factor data of the building air conditioning system;

[0077] A model adjustment module, used to adjust the pre-trained model using the historical load and characteristic data of the target building and the real-time collected data to obtain a second air conditioning load prediction model;

[0078] The demand response analysis module is used to obtain a final demand response potential value based on the second air conditioning load prediction model by changing the input air conditioning temperature setting value and performing a difference calculation on the model prediction result output.

[0079] In the embodiment of the present application, the demand response analysis module includes:

[0080] Determine the air conditioning temperature setting range according to the user comfort constraint, simulate the user's demand response event by inputting different air conditioning temperature setting values ​​into the second air conditioning load prediction model, and set the air conditioning load prediction value of 27 degrees Celsius as the baseline;

[0081] By changing the temperature setting value of the model input, the demand response potential value under different temperature setting values ​​is obtained by subtracting each set of model output from the load forecast value of the baseline.

[0082] This embodiment further provides a computing device, which is applicable to the case of the air conditioning load demand response potential analysis method, including:

[0083] Memory and processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the air conditioning load demand response potential analysis method proposed in the above embodiment.

[0084] This embodiment further provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for analyzing the potential for air conditioning load demand response proposed in the above embodiment is implemented.

[0085] The storage medium proposed in this embodiment and the method for analyzing the potential for air conditioning load demand response proposed in the above embodiment belong to the same inventive concept. For technical details not described in detail in this embodiment, please refer to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0086] Through the above description of the implementation methods, the technicians in the relevant field can clearly understand that the present invention can be implemented by means of software and necessary general hardware, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ReadOnly, Memory, ROM), random access memory (RandomAccess Memory, RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform the methods of various embodiments of the present invention.

[0087] Example 2

[0088] Reference Figures 4 to 9 , is an embodiment of the present invention. This embodiment is different from the first embodiment in that this embodiment verifies the beneficial effects of the present invention through testing experiments.

[0089] The cloud data set of the test case in this embodiment is based on the data of a traditional Chinese medicine hospital from January to February and March to April 2015. The preprocessing and partial completion of missing label data are performed. The number of data set samples is 25334 and 25649 respectively.

[0090] In this embodiment, the temperature adjustment interval is set to [24°C, 31°C]. The duration of the demand response of the other part of the terminal model data sample set is set to 30 minutes, and the other settings remain unchanged. The test results of the trained source model after fine-tuning the two parts of the terminal data set are as follows: Figure 4 , Figure 5 shown.

[0091] By changing the temperature setting value of the model input, a series of model outputs were obtained. The demand response potential value under different temperature setting values ​​was obtained by subtracting each set of model outputs from the load forecast value of the baseline. Since the ultimate goal of model verification is to obtain the adjustable forecast potential value of the target terminal user, the key to measuring the accuracy of the model lies in whether the load curve prediction before and after the terminal user's demand response is accurate. The difference between the two is the terminal user's adjustable potential curve. Therefore, only the comparison results of the terminal-side model load prediction and the adjustable potential curve in the test case of the present invention are shown here, as shown in FIG. Figures 6 to 9 As shown;

[0092] in Figure 6 The demand response duration is set to 60 minutes, and the gray area is the predicted adjustable potential value. Figure 7The demand response duration is set to 30 minutes, and the gray area is the predicted adjustable potential value. Figure 8 The demand response time is set to 60 minutes. The gray area is the actual adjustable potential value. Fig. 9 As shown, the final adjustable potential value can be obtained by calculating the average value of the adjustable potential curve, which is about 52.4KW in this test case.

[0093] The present invention is based on a transfer learning framework and combines deep learning technology. Through comprehensive analysis of historical load and influencing factor data, as well as prediction and fine-tuning with the help of deep learning models, it can accurately estimate the demand response potential of the air-conditioning system in real-time scenarios and achieve effective use of energy.

[0094] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for analyzing the potential of air conditioning load demand response, characterized in that: include: Obtain historical load and influencing factor data of the building air conditioning system, perform correlation analysis on the influencing factors based on the historical load data set, and screen key influencing factor characteristics; Constructing a first air conditioning load prediction model based on the screening results and the historical load curve, and pre-training the first air conditioning load prediction model using the historical load and influencing factor data of the building air conditioning system; The pre-trained model is adjusted using the historical load and characteristic data of the target building and the real-time collected data to obtain a second air conditioning load prediction model; Based on the second air-conditioning load prediction model, the final demand response potential value is obtained by changing the input air-conditioning temperature setting value and performing a difference calculation on the model prediction result output.

2. The air conditioning load demand response potential analysis method according to claim 1, characterized in that: Based on the historical load data set, correlation analysis is performed on the influencing factors, and the key influencing factor features are screened, including: using the Spearman rank correlation coefficient to analyze the correlation between the influencing factors and the load, quantifying the influence of each feature on the load demand, and screening to obtain the minimum dimensional feature set, the calculation formula of which is expressed as follows: Among them, ρ is the correlation coefficient, d is the position difference corresponding to the two variables, and n is the number of load samples.

3. The air conditioning load demand response potential analysis method according to claim 2, characterized in that: Constructing a first air conditioning load prediction model based on the screening results and the historical load curve includes: the first air conditioning load prediction model structure includes a convolutional neural network, a bidirectional long short-term memory network, a flattening layer and a fully connected layer; The convolutional neural network is composed of a one-dimensional convolutional layer, a one-dimensional maximum pooling layer and a random dropout layer. The convolutional neural network extracts spatial features and local patterns in the original load data through the convolutional layer, and is used for load classification prediction through the combination of multi-scale features; The bidirectional long short-term memory network is composed of several pairs of independent LSTM layers and random loss layers, and the bidirectional long short-term memory network processes input prediction data through several pairs of independent LSTM layers.

4. The air conditioning load demand response potential analysis method according to claim 3, characterized in that: Using the historical load of the building air-conditioning system and the influencing factor data to pre-train the first air-conditioning load prediction model includes: the first air-conditioning load prediction model is pre-trained using an adaptive learning rate mechanism and batch normalization technology.

5. The air conditioning load demand response potential analysis method according to claim 4, characterized in that: The second air conditioning load prediction model is obtained by adjusting the pre-trained model using the historical load and characteristic data of the target building and the real-time collected data, including: fixing the weight and bias parameters of the convolution layer, and retraining the second air conditioning load prediction model for other layers based on a small sample data set of the historical load and characteristic data of the target building; The second air-conditioning load prediction model is deployed on the edge intelligent terminal and calculated using a lightweight model compression algorithm.

6. The air conditioning load demand response potential analysis method according to claim 5, characterized in that: Based on the second air conditioning load prediction model, the final demand response potential value is obtained by changing the input air conditioning temperature setting value and performing a difference calculation on the model prediction result output, including: Determine the air conditioning temperature setting range according to the user comfort constraint, simulate the user's demand response event by inputting different air conditioning temperature setting values ​​into the second air conditioning load prediction model, and set the air conditioning load prediction value of 27 degrees Celsius as the baseline; By changing the temperature setting value of the model input, the demand response potential value under different temperature setting values ​​is obtained by subtracting each set of model output from the load forecast value of the baseline.

7. An air conditioning load demand response potential analysis system, characterized in that: include: A data acquisition and analysis module is used to obtain historical load and influencing factor data of the building air conditioning system, perform correlation analysis on the influencing factors based on the historical load data set, and screen key influencing factor characteristics; A model building pre-training module is used to build a first air conditioning load prediction model based on the screening results and the historical load curve, and to pre-train the first air conditioning load prediction model using the historical load and influencing factor data of the building air conditioning system; A model adjustment module, used to adjust the pre-trained model using the historical load and characteristic data of the target building and the real-time collected data to obtain a second air conditioning load prediction model; The demand response analysis module is used to obtain a final demand response potential value based on the second air conditioning load prediction model by changing the input air conditioning temperature setting value and performing a difference calculation on the model prediction result output.

8. The air conditioning load demand response potential analysis system according to claim 7, characterized in that: The demand response analysis module includes: Determine the air conditioning temperature setting range according to the user comfort constraint, simulate the user's demand response event by inputting different air conditioning temperature setting values ​​into the second air conditioning load prediction model, and set the air conditioning load prediction value of 27 degrees Celsius as the baseline; By changing the temperature setting value of the model input, the demand response potential value under different temperature setting values ​​is obtained by subtracting each set of model output from the load forecast value of the baseline.

9. An electronic device, comprising: Memory and processor; The memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions. When the computer executable instructions are executed by the processor, the steps of the air conditioning load demand response potential analysis method described in any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the air-conditioning load demand response potential analysis method according to any one of claims 1 to 6.

Images