Industrial load data enhancement method based on cyclic convolution generative adversarial network

Abstract

The embodiment of the invention provides an industrial load data enhancement method based on a cyclic convolution generative adversarial network. The method comprises the following steps: acquiring historical industrial load data, and dividing the historical industrial load data into a training set and a test set; constructing a generator and a discriminator based on the multilayer perceptron network, forming a GAN model by the generator and the discriminator, and constructing a loss function of the generator and a loss function of the GAN model; on the basis of the training set, optimization is carried out through parameters of the generator, and an optimized generator is obtained; based on the optimized output data of the generator, parameters of a discriminator are optimized through a loss function of a GAN model, and an optimized discriminator is obtained; the optimized generator and the optimized discriminator form an optimized GAN model; and inputting the test set into the optimized GAN model, and outputting a data enhancement result. In this way, the complex data processing capacity of the model can be enhanced, the load space-time relation can be better captured, then a reliable basis is provided, and it is guaranteed that power grid planning is reasonable, economical and efficient.

Description

Technical Field

[0001] The present invention generally relates to the field of power systems, and more specifically, to an industrial load data enhancement method based on a cyclic convolutional generative adversarial network. Background Art

[0002] At present, recurrent convolutional generative adversarial network technology has been widely used in various data expansion fields. Among them, the RCGAN model can be used to generate power dispatch plans. By learning the power grid operation data (such as unit power generation data, AC interconnection line data, DC interconnection line data, flexible load data, energy storage dispatch data, etc.), the generator can simulate the power dispatch plan under different circumstances, and the discriminator can distinguish the generated dispatch plan from the real historical dispatch plan.

[0003] The existing recurrent convolutional generative adversarial network technology has high requirements for data volume. When data is scarce, the accuracy of traditional methods decreases or even fails. Insufficient data diversity also affects accuracy. For example, the exponential smoothing optimization method requires a large amount of data, and some methods have difficulty in processing complex data features, resulting in a lack of adaptability. Summary of the invention

[0004] According to an embodiment of the present invention, an industrial load data enhancement scheme based on a cyclic convolutional generative adversarial network is provided. This scheme can enhance the model's ability to process complex data and better capture the temporal and spatial relationship of loads, thereby providing a reliable basis for power system planning and ensuring that power grid planning is reasonable, economical, and efficient.

[0005] In a first aspect of the present invention, a method for industrial load data enhancement based on a cyclic convolutional generative adversarial network is provided. The method comprises:

[0006] Obtain historical industrial load data, and divide the historical industrial load data into a training set and a test set;

[0007] Constructing a generator and a discriminator based on a multi-layer perceptron network, wherein the generator and the discriminator constitute a GAN model, and constructing a loss function of the generator and a loss function of the GAN model;

[0008] Based on the training set, the parameters of the generator are optimized through the generator's loss function to obtain the optimized generator;

[0009] Based on the output data of the optimized generator, the parameters of the discriminator are optimized through the loss function of the GAN model to obtain an optimized discriminator; the optimized generator and the optimized discriminator constitute an optimized GAN model;

[0010] The test set is input into the optimized GAN model to output the enhanced results of the historical industrial load data.

[0011] Further, the generator includes a first input layer, a first hidden layer and a first output layer in sequence;

[0012] The first hidden layer is used to perform a convolution operation, including:

[0013]

[0014] Among them, G(x i ) is the output of the generator convolution layer; x i is the current processing data; f is the improved Tanh-ReLU activation function; K is the convolution kernel set; θ G is the generator parameter; P data (x i ) is the real data; b is the bias term.

[0015] Further, the discriminator includes a second input layer, a second hidden layer and a second output layer in sequence;

[0016] The second hidden layer is used to perform a convolution operation, including:

[0017]

[0018] Among them, G'(x i ) is the output of the discriminator convolutional layer; x i is the current processing data; f is the improved Tanh-ReLU activation function; K is the convolution kernel set; θ D is the discriminator parameter; P data is real data; b is the bias term.

[0019] Furthermore, the improved Tanh-ReLU activation function includes:

[0020]

[0021] Among them, x is the data of the input activation function; a is the Tanh-ReLU activation function parameter.

[0022] Furthermore, the generator parameters are optimized based on the training set by using the generator's loss function to obtain an optimized generator, including:

[0023] The training set is input into the generator, and the parameters of the generator are optimized through the generator's loss function to obtain the output of the generator under different parameters;

[0024] The parameters of the discriminator remain unchanged, and the discriminator is used to judge the accuracy of the output of the generator under different parameters. When the generator exit condition is met, the generator corresponding to the output with the highest accuracy is used as the optimized generator.

[0025] Furthermore, the loss function of the generator includes:

[0026]

[0027] in, is the loss function of the generator; N is the total number of data; x i is the current processing data; i is the index of the current processing data; G(x i ) is the output of the generator convolutional layer.

[0028] Furthermore, the output data of the optimized generator is used to optimize the parameters of the discriminator through the loss function of the GAN model to obtain the optimized discriminator, including:

[0029] Keep the parameters of the optimized generator unchanged, input the output data of the optimized generator into the discriminator, optimize the parameters of the discriminator through the loss function of the GAN model, and obtain the output of the discriminator under different parameters;

[0030] When the discriminator exit condition is reached, the discriminator with the highest accuracy is used as the optimized discriminator.

[0031] Furthermore, the loss function of the GAN model includes:

[0032]

[0033] in, is the loss function of the GAN model; is the intermediate expectation of the real data; D(x i ) is the optimized discriminator; is the network intermediate expectation; D(G'(x i )) is the convolutional layer decision function; G'(x i ) is the output of the discriminator convolutional layer.

[0034] In a second aspect of the present invention, an industrial load data enhancement device based on a cyclic convolutional generative adversarial network is provided. The device comprises:

[0035] An acquisition module is used to acquire historical industrial load data and divide the historical industrial load data into a training set and a test set;

[0036] A construction module is used to construct a generator and a discriminator based on a multi-layer perceptron network, wherein the generator and the discriminator constitute a GAN model, and a loss function of the generator and a loss function of the GAN model are constructed;

[0037] The generator optimization module optimizes the parameters of the generator based on the training set through the generator's loss function to obtain the optimized generator;

[0038] A discriminator optimization module, used to optimize the parameters of the discriminator based on the output data of the optimized generator through the loss function of the GAN model to obtain an optimized discriminator; the optimized generator and the optimized discriminator constitute an optimized GAN model;

[0039] The enhancement module is used to input the test set into the optimized GAN model and output the enhancement result of the historical industrial load data.

[0040] In a third aspect of the present invention, an electronic device is provided. The electronic device has at least one processor; and a memory connected to the at least one processor in communication; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can perform the method of the first aspect of the present invention.

[0041] It should be understood that the contents described in the summary of the invention are not intended to limit the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The above and other features, advantages and aspects of the embodiments of the present invention will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, wherein:

[0043] Figure 1 A flowchart of an industrial load data enhancement method based on a cyclic convolutional generative adversarial network according to an embodiment of the present invention is shown;

[0044] Figure 2 It shows an iterative optimization flow chart of a generator according to an embodiment of the present invention;

[0045] Figure 3 A flowchart of iterative optimization of a discriminator according to an embodiment of the present invention is shown;

[0046] Figure 4 A block diagram of an industrial load data enhancement device based on a cyclic convolutional generative adversarial network according to an embodiment of the present invention is shown;

[0047] Figure 5 A block diagram of an exemplary electronic device capable of implementing embodiments of the present invention is shown;

[0048] Among them, 500 is an electronic device, 501 is a computing unit, 502 is a ROM, 503 is a RAM, 504 is a bus, 505 is an I / O interface, 506 is an input unit, 507 is an output unit, 508 is a storage unit, and 509 is a communication unit. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] In addition, the term "and / or" in this article is only a description of the association relationship between the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0051] In the present invention, the GAN model is composed of a generator and a discriminator. The optimized GAN model is obtained by optimizing the parameters of the generator and the discriminator, and then the data is input into the optimized GAN model to obtain enhanced output historical industrial load data. The above method can quickly and accurately extract data features, thereby judging the authenticity of the input data and improving the accuracy of data filling.

[0052] Figure 1 A flowchart of an industrial load data enhancement method based on a cyclic convolutional generative adversarial network according to an embodiment of the present invention is shown.

[0053] The method includes:

[0054] S101. Obtain historical industrial load data, and divide the historical industrial load data into a training set and a test set.

[0055] Specifically, historical industrial load data refers to the data of asymmetric current amplitude, voltage amplitude and phase in the three-phase power system, which is caused by the access of single-phase equipment, nonlinear load operation, etc. This asymmetric data will produce negative sequence components and affect the performance of the motor. Monitoring this data can take compensation measures and is also conducive to optimizing the grid topology and capacity configuration.

[0056] S102. Constructing a generator and a discriminator based on a multi-layer perceptron network, wherein the generator and the discriminator constitute a GAN (Generative Adversarial Network) model, and constructing a loss function of the generator and a loss function of the GAN model.

[0057] Specifically, a multilayer perceptron (MLP) network is used to construct the generator and discriminator, that is, to map multiple input data sets to a single output data set.

[0058] In this embodiment, the generator includes a first input layer, a first hidden layer and a first output layer in sequence;

[0059] The first hidden layer is used to perform a convolution operation, including:

[0060]

[0061] Among them, G(x i ) is the output of the generator convolutional layer, that is, the output of the first hidden layer; x i is the current processing data; f is the improved Tanh-ReLU activation function; K is the convolution kernel set; θ G is the generator parameter; P data (x i ) is the real data; b is the bias term.

[0062] Specifically, the workflow of the generator is: the training set data is used as the input data of the first input layer of the generator; then the data of the first input layer is passed to the first hidden layer, and the first hidden layer performs a nonlinear transformation on the data through the Tanh-ReLU activation function; finally, the output data of the first hidden layer is passed to the first output layer, and a data set is output, where the data set includes the training set data and the filled data.

[0063] For small sample data sets, the generator generates new data samples by learning the distribution of real data, which not only increases the richness of the data but also prevents the model from overfitting.

[0064] In this embodiment, the discriminator includes a second input layer, a second hidden layer, and a second output layer in sequence;

[0065] The second hidden layer is used to perform a convolution operation, including:

[0066]

[0067] Among them, G'(x i ) is the output of the discriminator convolutional layer, that is, the output of the second hidden layer; x iis the current processing data; f is the improved Tanh-ReLU activation function; K is the convolution kernel set; θ D is the discriminator parameter; P data is the real data; b is the bias term.

[0068] Specifically, the workflow of the discriminator is as follows: the data set output by the generator is used as the input data of the first input layer of the discriminator; the data of the first input layer is then passed into the first hidden layer, and the first hidden layer performs a nonlinear transformation on the data through the Tanh-ReLU activation function; finally, the output data of the first hidden layer is passed into the first output layer, and the accuracy is output, where the accuracy is the average percentage error between the padded data and the training set data in the data set output by the generator. The smaller the error, the higher the accuracy.

[0069] The discriminator provides feedback signals to the generator, distinguishing true from false data while learning the distribution characteristics of real data. The discriminator can discover the inherent patterns of the data by constantly comparing the generated data with the real data. The discriminator's discrimination results can be used as a criterion to measure the quality of the data generated by the generator, which helps the entire GAN model converge to a stable state.

[0070] In this embodiment, the improved Tanh-ReLU activation function includes:

[0071]

[0072] Among them, x is the data of the input activation function; a is the Tanh-ReLU activation function parameter.

[0073] Tanh-ReLU (Hyperbolic Tangent-Rectified Linear Unit) is an improved function based on the traditional Relu (Rectified Linear Unit), which makes the function defined when it is less than or equal to 0, improving the shortcoming of the traditional Relu function that all values ​​are set to 0, thus solving the problem of neural saturation and improving the efficiency of calculation.

[0074] S103. Based on the training set, the parameters of the generator are optimized through the loss function of the generator to obtain an optimized generator, such as Figure 2 As shown, including:

[0075] The training set is input into the generator, and the parameters of the generator are optimized through the generator's loss function to obtain the output of the generator under different parameters;

[0076] The parameters of the discriminator remain unchanged, and the discriminator is used to judge the accuracy of the output of the generator under different parameters. When the generator exit condition is met, the generator corresponding to the output with the highest accuracy is used as the optimized generator.

[0077] Specifically, the iterative optimization process of the generator is:

[0078] Step 1: Fix the parameters of the discriminator, adjust the parameters of the generator through the loss function of the generator, and use the output of the adjusted generator as the input of the discriminator.

[0079] Step 2: Record the accuracy of the discriminator output and the current number of iterations.

[0080] Step 3: Determine whether the exit condition of the generator iterative optimization is met. After exiting the iterative optimization, the generator corresponding to the output with the highest accuracy is used as the optimized generator. The specific exit condition of the generator iterative optimization is as follows:

[0081] (1) If the current number of iterations reaches the maximum number of iterations, the generator with the highest accuracy previously recorded is used as the optimized generator and the iteration is exited; the maximum number of iterations is generally 50.

[0082] (2) If the current number of iterations has not reached the maximum number of iterations, continue to repeat steps 1-3 until (1) is satisfied.

[0083] After the generator is optimized, it can generate data that is closer to the real data distribution, improving the generalization ability of the entire generative adversarial network (GAN) model.

[0084] In this embodiment, the loss function of the generator includes:

[0085]

[0086] in, is the loss function of the generator; N is the total number of data; i is the index of the currently processed data; x i is the current processing data; G(x i ) is the output of the generator convolutional layer.

[0087] The optimization of the loss function helps the generator converge more stably during the training process, controls the learning rate of the generator, avoids too fast or too slow parameter updates, makes the generator gradually approach the optimal state, and prevents oscillation or divergence in the training process. When the loss function adjusts the parameters of the generator to improve the quality of the data it generates, the discriminator will also adjust accordingly, so that the entire model reaches a dynamic balance and improves the stability and robustness of the model.

[0088] S104, based on the output data of the optimized generator, the parameters of the discriminator are optimized through the loss function of the GAN model to obtain an optimized discriminator, such as Figure 3 As shown; the optimized generator and the optimized discriminator constitute an optimized GAN model, including:

[0089] Keep the parameters of the optimized generator unchanged, input the output data of the optimized generator into the discriminator, optimize the parameters of the discriminator through the loss function of the GAN model, and obtain the output of the discriminator under different parameters;

[0090] When the discriminator exit condition is reached, the discriminator with the highest accuracy is used as the optimized discriminator.

[0091] Specifically, the iterative optimization process of the discriminator is:

[0092] Step 1: Fix the parameters of the optimized generator, adjust the parameters of the discriminator through the loss function of the GAN model, and use the output of the optimized generator as the input of the adjusted discriminator.

[0093] Step 2: Record the accuracy of the adjusted discriminator output and the current number of iterations.

[0094] Step 3: Determine whether the exit condition of the iterative optimization of the discriminator is met. After exiting the iterative optimization, the discriminator with the highest accuracy is used as the optimized discriminator. The specific exit condition of the iterative optimization of the discriminator is as follows:

[0095] (1) If the current accuracy reaches a certain threshold, the current discriminator is used as the optimized discriminator and the iteration is skipped; the threshold is generally 1.85%.

[0096] (2) If the current number of iterations reaches the maximum number of iterations, the discriminator with the highest accuracy previously recorded is used as the optimized discriminator and the iteration is skipped; the maximum number of iterations is generally 50.

[0097] (3) If both conditions (1) and (2) are not met, continue to repeat the above steps 1-3 until either (1) or (2) is met.

[0098] The optimization of the discriminator provides more valuable feedback to the generator. When the discriminator becomes more sensitive, the generator can learn how to generate data closer to real data from more accurate judgments. In addition, the optimized discriminator makes the model more robust and can maintain good discrimination performance in the presence of noise or data changes to a certain extent.

[0099] In this embodiment, the loss function of the GAN model includes:

[0100]

[0101] in, is the loss function of the GAN model; is the intermediate expectation of the real data; D(x i ) is the optimized discriminator; is the network intermediate expectation; D(G'(x i )) is the convolutional layer decision function; G'(x i ) is the output of the discriminator convolutional layer.

[0102] In practical applications, data may be affected by noise or change to a certain extent. The discriminator parameters optimized by the loss function can enable the discriminator to better cope with these situations. The optimized discriminator parameters can make the model have better generalization ability. When faced with a new data set, the discriminator can use the common features learned during the optimization process to judge the authenticity of the data.

[0103] S105: Input the test set into the optimized GAN model, and output the enhanced results of the historical industrial load data.

[0104] Specifically, for power system monitoring data, such as voltage and current waveforms, expanding samples through the GAN model can help train more robust fault diagnosis models and improve the ability to identify complex electrical faults.

[0105] According to the embodiments of the present invention, the following advantages are achieved:

[0106] (1) The present invention improves the activation function of the original GAN ​​model so that it is defined when the data is less than or equal to 0. The new activation function can ensure that the gradient is transmitted more effectively during the back propagation process, so that the model reaches a stable state faster. The optimized GAN model automatically normalizes the output of neurons, thereby reducing the drastic fluctuations in parameter updates of the generator and discriminator during training and enhancing the stability of the model.

[0107] (2) The optimized GAN model has a flexible structure and can be adjusted according to different data characteristics and requirements. The model can quickly adapt and accurately optimize when faced with various complex power load changes (such as random load fluctuations, seasonal changes, etc.). At the same time, the model is also easy to combine with other technologies or methods to further improve performance and has good scalability.

[0108] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described order of actions, because according to the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0109] The above is an introduction to a method embodiment. The following is a further explanation of the scheme of the present invention through a device embodiment having the same inventive concept as the method in the aforementioned embodiment.

[0110] like Figure 4 As shown, the device 400 includes:

[0111] The acquisition module 410 is used to acquire historical industrial load data and divide the historical industrial load data into a training set and a test set.

[0112] Construction module 420, constructs a generator and a discriminator based on a multi-layer perceptron network, wherein the generator and the discriminator constitute a GAN model, and constructs a loss function of the generator and a loss function of the GAN model.

[0113] The generator optimization module 430 optimizes the parameters of the generator through the loss function of the generator based on the training set to obtain an optimized generator.

[0114] The discriminator optimization module 440 is used to optimize the parameters of the discriminator based on the output data of the optimized generator through the loss function of the GAN model to obtain an optimized discriminator; the optimized generator and the optimized discriminator constitute an optimized GAN model.

[0115] The enhancement module 450 is used to input the test set into the optimized GAN model and output the enhancement result of the historical industrial load data.

[0116] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the described module can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0117] In the technical solution of the present invention, the acquisition, storage and application of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0118] According to an embodiment of the present invention, the present invention further provides an electronic device.

[0119] Figure 5A schematic block diagram of an electronic device 500 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, 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 invention described and / or required herein.

[0120] The electronic device 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the electronic device 500 can also be stored. The computing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0121] Multiple components in the electronic device 500 are connected to the I / O interface 505, including: an input unit 506, such as a keyboard, a mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a disk, an optical disk, etc.; and a communication unit 509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 509 allows the electronic device 500 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0122] The computing unit 501 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 501 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 501 performs the various methods and processes described above, such as methods S101 to S105. For example, in some embodiments, methods S101 to S105 may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 508. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 500 via the ROM 502 and / or the communication unit 509. When the computer program is loaded into the RAM 503 and executed by the computing unit 501, one or more steps of the methods S101 to S105 described above may be performed. Alternatively, in other embodiments, the computing unit 501 may be configured to execute methods S101 to S105 in any other appropriate manner (eg, by means of firmware).

[0123] 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.

[0124] The program code for implementing the method of the present invention 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, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0125] In the context of the present invention, a machine-readable medium may be a tangible medium that may contain or store a program for use by an instruction execution system, device or equipment or used in combination 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, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, 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), 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.

[0126] 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).

[0127] The systems and techniques described herein may be implemented in a computing system that includes back-end 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 front-end 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 back-end components, middleware components, or front-end 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), and the Internet.

[0128] 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 of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0129] 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 described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0130] The above specific implementations do not constitute a limitation on the protection scope of the present invention. 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 modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An industrial load data enhancement method based on a cyclic convolutional generative adversarial network, characterized in that: include: Obtain historical industrial load data, and divide the historical industrial load data into a training set and a test set; Constructing a generator and a discriminator based on a multi-layer perceptron network, wherein the generator and the discriminator constitute a GAN model, and constructing a loss function of the generator and a loss function of the GAN model; Based on the training set, the parameters of the generator are optimized through the generator's loss function to obtain the optimized generator; Based on the output data of the optimized generator, the parameters of the discriminator are optimized through the loss function of the GAN model to obtain an optimized discriminator; the optimized generator and the optimized discriminator constitute an optimized GAN model; The test set is input into the optimized GAN model to output the enhanced results of the historical industrial load data.

2. The method according to claim 1, characterized in that The generator includes a first input layer, a first hidden layer and a first output layer in sequence; The first hidden layer is used to perform a convolution operation, including: Among them, G(x i ) is the output of the generator convolution layer; x i is the current processing data; f is the improved Tanh-ReLU activation function; K is the convolution kernel set; θ G is the generator parameter; P data (x i ) is the real data; b is the bias term.

3. The method according to claim 1, characterized in that The discriminator includes a second input layer, a second hidden layer and a second output layer in sequence; The second hidden layer is used to perform a convolution operation, including: Among them, G'(x i ) is the output of the discriminator convolutional layer; x i is the current processing data; f is the improved Tanh-ReLU activation function; K is the convolution kernel set; θ D is the discriminator parameter; P data is the real data; b is the bias term.

4. The method according to claim 2 or 3, characterized in that: The improved Tanh-ReLU activation function includes: Among them, x is the data of the input activation function; a is the Tanh-ReLU activation function parameter.

5. The method according to claim 1, characterized in that The method optimizes the parameters of the generator based on the training set through the loss function of the generator to obtain the optimized generator, including: The training set is input into the generator, and the parameters of the generator are optimized through the generator's loss function to obtain the output of the generator under different parameters; The parameters of the discriminator remain unchanged, and the discriminator is used to judge the accuracy of the output of the generator under different parameters. When the generator exit condition is met, the generator corresponding to the output with the highest accuracy is used as the optimized generator.

6. The method according to claim 1 or 5, characterized in that: The loss function of the generator includes: in, is the loss function of the generator; N is the total number of data; x i is the current processing data; i is the index of the current processing data; G(x i ) is the output of the generator convolutional layer.

7. The method according to claim 1, characterized in that The method optimizes the parameters of the discriminator based on the output data of the optimized generator through the loss function of the GAN model to obtain the optimized discriminator, including: Keep the parameters of the optimized generator unchanged, input the output data of the optimized generator into the discriminator, optimize the parameters of the discriminator through the loss function of the GAN model, and obtain the output of the discriminator under different parameters; When the discriminator exit condition is reached, the discriminator with the highest accuracy is used as the optimized discriminator.

8. The method according to claim 1 or 7, characterized in that: The loss function of the GAN model includes: in, is the loss function of the GAN model; is the intermediate expectation of the real data; D(x i ) is the optimized discriminator; is the network intermediate expectation; D(G'(x i )) is the convolutional layer decision function; G'(x i ) is the output of the discriminator convolutional layer.

9. An industrial load data enhancement device based on a cyclic convolutional generative adversarial network, characterized in that: include: An acquisition module is used to acquire historical industrial load data and divide the historical industrial load data into a training set and a test set; A construction module is used to construct a generator and a discriminator based on a multi-layer perceptron network, wherein the generator and the discriminator constitute a GAN model, and a loss function of the generator and a loss function of the GAN model are constructed; The generator optimization module optimizes the parameters of the generator based on the training set through the generator's loss function to obtain the optimized generator; A discriminator optimization module, used to optimize the parameters of the discriminator based on the output data of the optimized generator through the loss function of the GAN model to obtain an optimized discriminator; the optimized generator and the optimized discriminator constitute an optimized GAN model; The enhancement module is used to input the test set into the optimized GAN model and output the enhancement result of the historical industrial load data.

10. An electronic device comprising at least one processor; and a memory connected to the at least one processor in communication; characterized in that: The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 8.

Images