A Quantile Transformer-based industrial park energy consumption interval prediction method and system

Abstract

The present application proposes a method and system for predicting energy consumption intervals in industrial parks based on a quantile Transformer, the method comprising: obtaining industrial energy consumption data, preprocessing the industrial energy consumption data, wherein the preprocessing process includes data cleaning and normalization; calculating the average deviation of the preprocessed industrial energy consumption data relative to the historical seasonal average energy consumption every day, and seasonally adjusting the preprocessed industrial energy consumption data according to the average deviation; inputting the seasonally adjusted industrial energy consumption data into a trained Transformer model to obtain a normalized prediction value, wherein the Transformer model adopts a self-attention mechanism and a quantile loss function; and denormalizing the normalized prediction value to obtain a denormalized predicted energy consumption value. The present application adopts a Transformer model combined with a quantile loss function, which can better capture and reflect the complexity and changing trends of industrial energy consumption data, and the introduction of a seasonal adjustment method can reduce the impact of seasonal changes on energy consumption forecasts.

Description

Technical Field

[0001] The present application relates to industrial energy consumption prediction technology, and in particular to a method and system for predicting energy consumption intervals in industrial parks based on a quantile Transformer. Background Art

[0002] With the acceleration of global industrialization and the increasing shortage of energy resources, industrial energy consumption management and optimization have become important issues of concern to governments, enterprises and research institutions. Effective energy consumption forecasting is a key step in achieving energy consumption management and optimization. It plays a vital role in formulating reasonable energy procurement strategies, optimizing production plans, reducing energy costs and meeting environmental protection requirements.

[0003] However, industrial energy consumption forecasting faces a series of challenges:

[0004] (1) Data complexity: Industrial energy consumption data are usually highly nonlinear, seasonal, and cyclical, and are affected by multiple factors (such as production activities, climate conditions, market demand, etc.), making it difficult for traditional forecasting methods to capture the inherent laws of the data.

[0005] (2) Abnormal and extreme situations: Abnormal and extreme situations often occur in industrial production processes, such as equipment failure, energy supply interruption, emergencies, etc. These factors will have unpredictable impacts on energy consumption, further increasing the difficulty of prediction.

[0006] (3) Seasonal changes: Industrial energy consumption is often affected by seasonal changes, such as heating in winter and cooling in summer. These seasonal factors can mask or distort the true trend of energy consumption, leading to inaccurate forecast results.

[0007] (4) Model complexity: Traditional energy consumption prediction models are often based on simple statistical methods or linear regression models. These models are often unable to handle high-dimensional, nonlinear, and dynamically changing industrial energy consumption data, resulting in limited prediction accuracy and robustness. Summary of the invention

[0008] The present application aims to solve one of the technical problems in the related art at least to some extent.

[0009] To this end, the first purpose of this application is to propose a quantile Transformer-based industrial park energy consumption interval prediction method. By combining the Transformer architecture, quantile loss function and seasonal adjustment technology, it can effectively handle the complexity, anomalies and extreme situations, and seasonal changes of industrial energy consumption data, thereby improving the accuracy, robustness and adaptability of energy consumption prediction.

[0010] The second purpose of this application is to propose an industrial park energy consumption interval prediction system based on a quantile Transformer.

[0011] The third objective of the present application is to provide an electronic device.

[0012] A fourth objective of the present application is to provide a computer-readable storage medium.

[0013] To achieve the above purpose, the first embodiment of the present application proposes a method for predicting energy consumption intervals of industrial parks using a quantile Transformer, including:

[0014] Acquire industrial energy consumption data, and preprocess the industrial energy consumption data, wherein the preprocessing process includes data cleaning and normalization;

[0015] Calculating the average deviation of the pre-processed industrial energy consumption data from the historical seasonal average energy consumption on a daily basis, and seasonally adjusting the pre-processed industrial energy consumption data according to the average deviation;

[0016] The seasonally adjusted industrial energy consumption data is input into the trained Transformer model to obtain the normalized prediction value. The Transformer model uses the self-attention mechanism and the quantile loss function.

[0017] The normalized predicted value is subjected to denormalization processing to obtain a denormalized predicted energy consumption value.

[0018] Optionally, the obtaining of industrial energy consumption data and preprocessing of the industrial energy consumption data includes:

[0019] Collect the industrial energy consumption data from the industrial park, and identify and correct missing values ​​and abnormal values ​​in the industrial energy consumption data;

[0020] The industrial energy consumption data after data cleaning is standardized to the range of [0,1], and the expression is:

[0021]

[0022] in, It is the industrial energy consumption data after data cleaning. and They are the minimum and maximum values ​​of the industrial energy consumption data after data cleaning. It is the normalized industrial energy consumption data.

[0023] Optionally, the calculation of the average deviation of the preprocessed industrial energy consumption data relative to the historical seasonal average energy consumption per day is expressed as:

[0024]

[0025] in, It is The average deviation of the day from the historical seasonality, For the Normalized industrial energy consumption data for the day, It is the seasonal average energy consumption during the same historical period.

[0026] Optionally, the pre-processed industrial energy consumption data is seasonally adjusted according to the average deviation, and the expression is:

[0027]

[0028] in, is the seasonally adjusted industrial energy consumption data. for A collection of .

[0029] Optionally, the self-attention mechanism includes:

[0030]

[0031] in, is a matrix of queries, keys, and values, is the dimension of the key.

[0032] Optionally, the self-attention mechanism includes:

[0033]

[0034] in, is a matrix of queries, keys, and values, is the dimension of the key.

[0035] Optionally, performing denormalization processing on the normalized predicted value to obtain a denormalized predicted energy consumption value includes:

[0036]

[0037] in, is the denormalized predicted energy consumption value.

[0038] To achieve the above purpose, the second embodiment of the present application proposes an industrial park energy consumption interval prediction system based on a quantile transformer, including:

[0039] A preprocessing module, used to obtain industrial energy consumption data and preprocess the industrial energy consumption data, wherein the preprocessing process includes data cleaning and normalization;

[0040] A seasonal adjustment module, used to calculate the average deviation of the pre-processed industrial energy consumption data relative to the historical seasonal average energy consumption every day, and seasonally adjust the pre-processed industrial energy consumption data according to the average deviation;

[0041] The prediction module is used to input seasonally adjusted industrial energy consumption data into the trained Transformer model to obtain normalized prediction values. The Transformer model uses a self-attention mechanism and a quantile loss function.

[0042] The denormalization module is used to perform denormalization processing on the normalized predicted value to obtain a denormalized predicted energy consumption value.

[0043] To achieve the above-mentioned purpose, the third aspect of the present application provides an electronic device, including: a processor, and a memory communicatively connected to the processor;

[0044] The memory stores computer-executable instructions;

[0045] The processor executes the computer-executable instructions stored in the memory to implement the method as described in any one of the first aspects above.

[0046] To achieve the above-mentioned purpose, the fourth aspect embodiment of the present application proposes a computer-readable storage medium, in which computer execution instructions are stored. When the computer execution instructions are executed by a processor, they are used to implement the method as described in any one of the above-mentioned first aspects.

[0047] The technical solution provided by the embodiments of the present application brings at least the following beneficial effects:

[0048] (1) The present invention adopts the Transformer model combined with the quantile loss function, which can better capture and reflect the complexity and changing trend of industrial energy consumption data, thereby achieving more accurate energy consumption prediction;

[0049] (2) By using different quantile levels to predict energy consumption, this application can provide industrial enterprises with an energy consumption prediction range, which helps decision makers to more comprehensively evaluate prediction errors and uncertainties, thereby formulating more reasonable energy management strategies;

[0050] (3) The seasonal adjustment method introduced in the present invention can effectively process and reduce the energy consumption forecast deviation caused by seasonal changes, and improve the accuracy and stability of the forecast;

[0051] (4) By combining the Transformer model and the quantile loss function, this application can effectively handle anomalies and extreme situations in industrial energy consumption data and improve the robustness and adaptability of the model.

[0052] (5) This application can help industrial enterprises to carry out energy planning and management more effectively, thereby maximizing energy utilization and minimizing energy consumption by providing accurate energy consumption forecasts and forecast ranges.

[0053] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0055] Figure 1 It is a flowchart of a method for predicting energy consumption intervals in an industrial park using a quantile Transformer according to an embodiment of the present application;

[0056] Figure 2 It is a block diagram of an industrial park energy consumption interval prediction system based on a quantile transformer according to an embodiment of the present application;

[0057] Figure 3 It is a block diagram of an electronic device. DETAILED DESCRIPTION

[0058] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0059] The following describes a method and device for predicting energy consumption intervals of an industrial park using a quantile Transformer according to an embodiment of the present application with reference to the accompanying drawings.

[0060] Figure 1 is a flowchart of a method for predicting energy consumption intervals in an industrial park using a quantile transformer according to an embodiment of the present application. Figure 1 As shown, the method comprises the following steps:

[0061] Step 101, obtain industrial energy consumption data, and preprocess the industrial energy consumption data, wherein the preprocessing process includes data cleaning and normalization.

[0062] In some embodiments, energy consumption data of various parts or the entire production process of an industrial plant are collected through sensor networks, smart meters or other monitoring equipment. These data generally include information such as energy consumption in a time series and equipment operating status.

[0063] It is understandable that in order to ensure the accuracy, completeness, consistency and reliability of the data and to improve the quality and accuracy of data analysis, the original industrial energy consumption data must first be cleaned, that is, to identify and correct missing values ​​and outliers in the industrial energy consumption data.

[0064] As a possible implementation method, the missing value processing process is to identify and process missing values, and use interpolation (mean, previous value and subsequent value, linear interpolation, model prediction filling, etc.) or direct elimination according to the situation; the outlier identification process is to identify and process outliers through statistical analysis methods, which may be erroneous records or special events that need to be marked separately.

[0065] In a possible embodiment, the statistical analysis method is any one of a box plot, a Z-score, an IQR score method, or other statistical analysis methods.

[0066] Then, the industrial energy consumption data after data cleaning is normalized. The purpose of this is to scale the data to the same scale, eliminate the dimension, and facilitate model learning. It is usually used for comparison between features.

[0067] As a possible implementation method, the industrial energy consumption data after data cleaning is standardized to the range of [0,1], and the expression is:

[0068]

[0069] in, It is the industrial energy consumption data after data cleaning. and They are the minimum and maximum values ​​of the industrial energy consumption data after data cleaning. It is the normalized industrial energy consumption data.

[0070] In a possible embodiment, consider an industrial energy consumption data set, which contains daily average energy consumption data for one year, with a total of 365 data points. To simplify the example, the following data is used for example:

[0071]

[0072] The industrial energy consumption data after preprocessing is:

[0073]

[0074] Step 102, calculating the average deviation of the preprocessed industrial energy consumption data relative to the historical seasonal average energy consumption every day, and performing seasonal adjustment on the preprocessed industrial energy consumption data according to the average deviation.

[0075] In the embodiment of the present application, in order to eliminate the seasonal impact and focus on the energy consumption pattern of non-seasonal changes, the average deviation of the pre-processed industrial energy consumption data relative to the historical seasonal average energy consumption is first calculated, and the expression is:

[0076]

[0077] in, It is The average deviation of the day from the historical seasonality, For the Normalized industrial energy consumption data for the day, It is the seasonal average energy consumption during the same historical period.

[0078] After obtaining the average deviation of the preprocessed industrial energy consumption data relative to the historical seasonal average energy consumption every day, the preprocessed industrial energy consumption data is adjusted according to the seasonal deviation to remove the seasonal effect and make the model focus on non-seasonal changes. The expression is:

[0079]

[0080] in, is the seasonally adjusted industrial energy consumption data. for A collection of .

[0081] In a possible embodiment, assuming that the historical seasonal average energy consumption in the same period is 120, the seasonal deviation is:

[0082]

[0083] Seasonally adjusted industrial energy consumption data are:

[0084]

[0085] Step 103, input the seasonally adjusted industrial energy consumption data into the trained Transformer model to obtain a normalized prediction value, wherein the Transformer model adopts a self-attention mechanism and a quantile loss function.

[0086] In an embodiment of the present application, the adjusted data is sent to a pre-trained Transformer model. This model uses a self-attention mechanism to focus on important parts of the input sequence, improves the model's sensitivity to changes in energy consumption, and adopts a quantile loss function as the optimization target. It is sensitive to the predicted distribution and is particularly suitable for estimating specific quantiles, such as the median, rather than just the mean.

[0087] It should be noted that the self-attention mechanism is one of the core components of the Transformer model, which allows the model to consider its internal dependencies when processing different parts of the input sequence. Its representation is:

[0088]

[0089] in, is a matrix of queries, keys, and values ​​that come from a linear transformation of the input sequence. Each row in the matrix corresponds to a position in the input sequence, and the columns represent the feature dimensions; is the dimension of the key, and the denominator under the square root represents a scaling factor, which is used for scale normalization to prevent the dot product result from being too large or too small due to the dimension size; the softmax function is used to convert the result of the dot product into a probability distribution to ensure that the sum of the attention weights is 1, thereby selecting the most important information for each position.

[0090] The quantile loss function is based on the relative size of the predicted value and the true value, as well as the selected quantile. , to calculate the loss, the expression is:

[0091]

[0092] in, is the loss value, is the quantile level, is the normalized true value, is the normalized predicted value.

[0093] In one possible embodiment, assuming that the selected quantile level is 0.5, the loss function is:

[0094]

[0095] It should be noted that the Transformer model is used to train the seasonally adjusted industrial energy consumption data. Since model training is a well-known method in the art, all steps of model training are not expanded here, and only the final prediction results are focused on.

[0096] Step 104, performing denormalization processing on the normalized predicted value to obtain a denormalized predicted energy consumption value.

[0097] In the embodiment of the present application, the normalized predicted value is converted back to the original energy consumption value range, and its expression is:

[0098]

[0099] in, is the denormalized predicted energy consumption value.

[0100] In a possible embodiment, assuming that the normalized prediction value predicted by the Transformer model is 0.9, the predicted energy consumption after denormalization is:

[0101]

[0102] The final prediction result is 144.

[0103] It should be noted that, in addition to using a single quantile level for prediction, the present application can also use different quantile levels to predict the prediction interval of energy consumption.

[0104] In a possible embodiment, when the quantile levels are selected as 0.1 and 0.9, the predicted values ​​under the two quantiles may be calculated to represent the predicted energy consumption range.

[0105]

[0106]

[0107] In this example, assuming that the normalized prediction value predicted by the Transformer model is 0.9 and the predicted energy consumption after denormalization is 144, the predicted energy consumption range is:

[0108]

[0109]

[0110] By using different quantile levels (0.1 and 0.9), a range of energy consumption can be predicted, which provides industrial companies with a range for energy management and helps to more comprehensively evaluate and plan energy use.

[0111] In order to implement the above embodiments, the present application also proposes an industrial park energy consumption interval prediction system based on a quantile Transformer.

[0112] Figure 2 1 is a block diagram of an industrial park energy consumption interval prediction system 10 based on a quantile transformer according to an embodiment of the present application, including:

[0113] The preprocessing module 100 is used to obtain industrial energy consumption data and preprocess the industrial energy consumption data, wherein the preprocessing process includes data cleaning and normalization;

[0114] A seasonal adjustment module 200 is used to calculate the average deviation of the pre-processed industrial energy consumption data relative to the historical seasonal average energy consumption every day, and seasonally adjust the pre-processed industrial energy consumption data according to the average deviation;

[0115] A prediction module 300 is used to input seasonally adjusted industrial energy consumption data into a trained Transformer model to obtain a normalized prediction value, wherein the Transformer model adopts a self-attention mechanism and a quantile loss function;

[0116] The denormalization module 400 is used to perform denormalization processing on the normalized predicted value to obtain a denormalized predicted energy consumption value.

[0117] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0118] Figure 3 A schematic block diagram of an example electronic device 700 that can be used to implement an embodiment of the present application is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or required herein.

[0119] like Figure 3 As shown, the device 700 includes a computing unit 701, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 702 or a computer program loaded from a storage unit 708 into a random access memory (RAM) 703. In the RAM 703, various programs and data required for the operation of the device 700 can also be stored. The computing unit 701, the ROM 702, and the RAM 703 are connected to each other via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0120] A number of components in the device 700 are connected to the I / O interface 705, including: an input unit 706, such as a keyboard, a mouse, etc.; an output unit 707, such as various types of displays, speakers, etc.; a storage unit 708, such as a disk, an optical disk, etc.; and a communication unit 709, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 709 allows the device 700 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0121] The computing unit 701 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 701 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 701 performs the various methods and processes described above. For example, in some embodiments, the scheme proposed in the present application may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 708. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 700 via the ROM 702 and / or the communication unit 709. When the computer program is loaded into the RAM 703 and executed by the computing unit 701, one or more steps of the method described above may be performed. Alternatively, in other embodiments, the computing unit 701 may be configured as the method shown in the present application in any other appropriate manner (e.g., by means of firmware).

[0122] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0123] The program code for implementing the method of the present application can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, implements the functions / operations specified in the flow chart and / or block diagram. The program code can be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0124] In the context of the present application, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0125] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0126] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), the Internet, and a blockchain network.

[0127] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship between the client and the server is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services ("Virtual Private Server", or "VPS" for short). The server may also be a server of a distributed system, or a server combined with a blockchain.

[0128] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this application can be executed in parallel, sequentially or in different orders, as long as the expected results of the technical solution of this application can be achieved, and this document is not limited here.

[0129] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A method for predicting energy consumption intervals in industrial parks based on quantile Transformer, characterized in that: include: Acquire industrial energy consumption data, and preprocess the industrial energy consumption data, wherein the preprocessing process includes data cleaning and normalization; In order to eliminate the seasonal impact and focus on the non-seasonal energy consumption pattern, the average deviation of the pre-processed industrial energy consumption data relative to the historical seasonal average energy consumption is calculated every day, and the pre-processed industrial energy consumption data is seasonally adjusted according to the average deviation to remove the seasonal impact and make the model focus on non-seasonal changes; The seasonally adjusted industrial energy consumption data is input into the trained Transformer model to obtain the normalized prediction value. The Transformer model uses the self-attention mechanism and the quantile loss function. Performing a denormalization process on the normalized predicted value to obtain a denormalized predicted energy consumption value; The average deviation of the preprocessed industrial energy consumption data from the historical seasonal average energy consumption per day is calculated as follows: in, It is The average deviation of the day from the historical seasonality, For the Normalized industrial energy consumption data for the day, is the seasonal average energy consumption during the same period of history; The seasonal adjustment of the pre-processed industrial energy consumption data is performed according to the average deviation, and the expression is: in, is the seasonally adjusted industrial energy consumption data. for A collection of is the normalized industrial energy consumption data; The quantile loss function includes: in, is the loss value, is the quantile level, is the normalized true value, is the normalized predicted value.

2. The method according to claim 1, characterized in that The obtaining of industrial energy consumption data and preprocessing of the industrial energy consumption data include: Collect the industrial energy consumption data from the industrial park, and identify and correct missing values ​​and abnormal values ​​in the industrial energy consumption data; The industrial energy consumption data after data cleaning is standardized to the range of [0,1], and the expression is: in, It is the industrial energy consumption data after data cleaning. and They are the minimum and maximum values ​​of the industrial energy consumption data after data cleaning. It is the normalized industrial energy consumption data.

3. The method according to claim 1, characterized in that The self-attention mechanism includes: in, is a matrix of queries, keys, and values, is the dimension of the key.

4. The method according to claim 1, characterized in that: The denormalizing the normalized predicted value to obtain a denormalized predicted energy consumption value includes: in, is the denormalized predicted energy consumption value.

5. A quantile Transformer-based industrial park energy consumption interval prediction system, characterized in that: include: A preprocessing module, used to obtain industrial energy consumption data and preprocess the industrial energy consumption data, wherein the preprocessing process includes data cleaning and normalization; A seasonal adjustment module is used to calculate the average deviation of the pre-processed industrial energy consumption data from the historical seasonal average energy consumption every day in order to eliminate the seasonal impact and focus on the non-seasonal energy consumption pattern, and seasonally adjust the pre-processed industrial energy consumption data according to the average deviation to remove the seasonal impact and make the model focus on non-seasonal changes; The prediction module is used to input seasonally adjusted industrial energy consumption data into the trained Transformer model to obtain normalized prediction values. The Transformer model uses a self-attention mechanism and a quantile loss function. A denormalization module, used for performing denormalization processing on the normalized predicted value to obtain a denormalized predicted energy consumption value; The average deviation of the preprocessed industrial energy consumption data from the historical seasonal average energy consumption per day is calculated as follows: in, It is The average deviation of the day from the historical seasonality, For the Normalized industrial energy consumption data for the day, is the seasonal average energy consumption during the same period of history; The seasonal adjustment of the pre-processed industrial energy consumption data is performed according to the average deviation, and the expression is: in, is the seasonally adjusted industrial energy consumption data. for A collection of is the normalized industrial energy consumption data; The quantile loss function includes: in, is the loss value, is the quantile level, is the normalized true value, is the normalized predicted value.

6. An electronic device, characterized in that: include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 4 when executed by a processor.

Images