Transformer temperature prediction model training and transformer temperature prediction method

Abstract

The embodiment of the invention discloses a transformer temperature prediction model training method and a transformer temperature prediction method. The method comprises the following steps: acquiring a data set for model training, and preprocessing the data set; wherein the data set comprises transformer operation data and external environment data; constructing a preset transformer temperature prediction model based on the long short-term memory network and the variable parameter neurodynamics optimization model; and carrying out model training on the preset transformer temperature prediction model by adopting the data set, and carrying out dynamic optimization on the long and short-term memory network by adopting a variable parameter neurodynamics optimization model until a converged target transformer temperature prediction model is obtained. By adopting the scheme, the variable parameter neurodynamics optimization model dynamically optimizes the long short-term memory network to obtain the convergent target transformer temperature prediction model, and when different operation conditions and environment changes are coped, stable prediction performance is still kept, and the robustness and adaptability of the model are remarkably improved.

Description

Technical Field

[0001] The embodiments of the present invention relate to the technical field of power equipment monitoring and prediction, and in particular to a transformer temperature prediction model training and a transformer temperature prediction method. Background Art

[0002] With the continuous growth of electricity demand and the aging of power equipment, the reliable operation of transformers has become increasingly important. As a key equipment in the power system, the heating condition of the transformer directly affects its operating life and safety. Traditional transformer heating monitoring and prediction methods mainly rely on regular maintenance and empirical formulas, but these methods have certain limitations and it is difficult to accurately predict the heating condition of the transformer in real time.

[0003] In recent years, the development of machine learning and neural network technology has provided new solutions for transformer temperature prediction. Among them, long short-term memory networks are widely used in power system state prediction due to their superior performance in processing time series data. However, standard long short-term memory networks may have problems with insufficient prediction accuracy and robustness when dealing with complex nonlinear problems.

[0004] Therefore, how to improve the accuracy of transformer temperature prediction is a technical problem that needs to be solved urgently by those skilled in the art. Summary of the invention

[0005] The embodiment of the present invention provides a transformer temperature prediction model training and a transformer temperature prediction method, so as to achieve stable transformer temperature prediction performance when dealing with different operating conditions and environmental changes, and significantly improve the robustness and adaptability of the model.

[0006] In a first aspect, an embodiment of the present invention provides a transformer temperature prediction model training method, comprising:

[0007] Acquire a data set for model training and preprocess the data set; wherein the data set includes transformer operation data and external environment data;

[0008] Construct a preset transformer temperature prediction model based on long short-term memory network and variable parameter neural dynamics optimization model;

[0009] The preset transformer temperature prediction model is trained using a data set, and the long short-term memory network is dynamically optimized using a variable parameter neural dynamics optimization model until a converged target transformer temperature prediction model is obtained.

[0010] In a second aspect, an embodiment of the present invention provides a transformer temperature prediction method, comprising:

[0011] Determine the transformer target data to be predicted; the target data includes transformer operation data and external environment data;

[0012] The target data is input into the target transformer temperature prediction model, the temperature of the transformer to be predicted is predicted, the heating state of the transformer to be predicted is output and an abnormal warning is issued.

[0013] In a third aspect, an embodiment of the present invention provides a transformer temperature prediction model training device, comprising:

[0014] A data set acquisition module, used to acquire a data set for model training and preprocess the data set; wherein the data set includes transformer operation data and external environment data;

[0015] An initial model building module for building a preset transformer temperature prediction model based on a long short-term memory network and a variable-parameter neural dynamics optimization model;

[0016] The target model training module is used to use the data set to perform model training on the preset transformer temperature prediction model, and use the variable parameter neural dynamics optimization model to dynamically optimize the long short-term memory network until a converged target transformer temperature prediction model is obtained.

[0017] In a fourth aspect, an embodiment of the present invention further provides a transformer temperature prediction device, comprising:

[0018] A target data determination module is used to determine the target data of the transformer to be predicted; the target data includes transformer operation data and external environment data;

[0019] The temperature prediction module is used to input the target data into the target transformer temperature prediction model, perform temperature prediction on the transformer to be predicted, output the heating state of the transformer to be predicted and issue an abnormal warning.

[0020] In a fifth aspect, an embodiment of the present invention further provides an electronic device, the electronic device comprising:

[0021] one or more processors;

[0022] A storage device for storing one or more programs;

[0023] When the one or more programs are executed by the one or more processors, the one or more processors implement the transformer temperature prediction model training method or transformer temperature prediction method described in any embodiment of the present invention.

[0024] In a sixth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the transformer temperature prediction model training method or transformer temperature prediction method described in any embodiment of the present invention.

[0025] In a seventh aspect, an embodiment of the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements the transformer temperature prediction model training method or the transformer temperature prediction method as described in any embodiment of the present invention.

[0026] The embodiment of the present invention provides a transformer temperature prediction model training and transformer temperature prediction method, which obtains transformer operation data and external environment data for model training and performs preprocessing; constructs a preset transformer temperature prediction model based on a long short-term memory network and a variable parameter neural dynamics optimization model; uses a data set to perform model training on the preset transformer temperature prediction model, and uses a variable parameter neural dynamics optimization model to dynamically optimize the long short-term memory network until a converged target transformer temperature prediction model is obtained. The technical solution of the embodiment of the present invention is adopted, and the transformer temperature prediction model training method based on the variable parameter neural dynamics optimization long short-term memory network is obtained to obtain a converged target transformer temperature prediction model, which can better capture the nonlinear characteristics and complex time dependencies in the transformer operating state, and effectively improve the accuracy of transformer temperature prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Other features, objects and advantages of the present invention will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings. The drawings are only for the purpose of illustrating preferred embodiments and are not to be considered as limiting the present invention. Also, the same reference symbols are used throughout the drawings to represent the same parts. In the drawings:

[0028] Figure 1 is a flow chart of a transformer temperature prediction model training method provided by Embodiment 1 of the present invention;

[0029] Figure 2 is a flow chart of a transformer temperature prediction model training method provided by Embodiment 2 of the present invention;

[0030] Figure 3 This is a flow chart of dynamically optimizing a long short-term memory network using a variable parameter neural dynamics optimization model provided by the second embodiment of the present invention;

[0031] Figure 4 is a flow chart of a transformer temperature prediction method provided by Embodiment 3 of the present invention;

[0032] Figure 5is a flow chart of another transformer temperature prediction method provided by Embodiment 3 of the present invention;

[0033] Figure 6 It is a structural schematic diagram of a transformer temperature prediction model training device provided by a fourth embodiment of the present invention;

[0034] Figure 7 It is a structural schematic diagram of a transformer temperature prediction device provided by Embodiment 5 of the present invention;

[0035] Figure 8 It is a structural schematic diagram of an electronic device provided in Embodiment 6 of the present invention. DETAILED DESCRIPTION

[0036] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.

[0037] Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow charts describe the operations (or steps) as sequential processes, many of the operations (or steps) therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but can also have additional steps not included in the accompanying drawings. The process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0038] Among them, the acquisition, storage, use and processing of data in the technical solution of this application are in compliance with the relevant provisions of national laws and regulations. It should be noted that in the embodiments of this application, some existing solutions in the industry such as certain software, components or models may be mentioned, which should be considered as exemplary, and their purpose is only to illustrate the feasibility of the implementation of the technical solution of this application, but it does not mean that the applicant has or will necessarily use the solution.

[0039] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein.

[0040] Embodiment 1

[0041] Figure 1This is a flowchart of a transformer temperature prediction model training method provided in an embodiment of the present invention. This embodiment is applicable to the case of building a transformer temperature prediction model. The method of this embodiment can be executed by a transformer temperature prediction model training device, and the device can be implemented in hardware and / or software. The device can be configured in a server for transformer temperature prediction model training. The method specifically includes the following steps:

[0042] S110: Obtain a data set for model training, and preprocess the data set.

[0043] The data set includes transformer operation data and external environment data; the transformer operation data includes load current and transformer temperature; the external environment data includes ambient temperature and wind speed. The preprocessing may refer to standardizing the data set to ensure the quality and consistency of the data.

[0044] S120. Construct a preset transformer temperature prediction model based on a long short-term memory network and a variable parameter neural dynamics optimization model.

[0045] The Long Short-Term Memory (LSTM) network is a time-based recurrent neural network, which is mainly composed of a forget gate, a memory gate, and an output gate. The forget gate is responsible for determining how much of the cell state at the previous moment is retained to the cell state at the current moment; the input gate is responsible for determining how much of the current moment's input is retained to the cell state at the current moment; and the output gate is responsible for determining how much output the cell state at the current moment has. The input layer of the LSTM network accepts feature vectors, and multiple LSTM layers are used to capture long-term dependencies in time series. Dropout layers are used between LSTM layers to reduce overfitting.

[0046] The variable parameter neural dynamics optimization model is an optimization method based on neural networks. By introducing non-negative slack variables, time-varying linear equations and inequality equations are transformed into mixed nonlinear equations, and then a nonlinear dynamic model is constructed for problem solving. This model no longer relies on special activation functions to achieve finite time convergence, but achieves finite convergence from the model construction itself, which means that higher solution accuracy can be achieved in a shorter time.

[0047] As an emerging optimization method, the variable parameter neural dynamics optimization model provided by the embodiment of the present invention has the characteristics of fast convergence speed, high precision and strong robustness. Combining the variable parameter neural dynamics optimization model with the long short-term memory network can effectively improve the prediction performance of the LSTM network, especially in dealing with transformer temperature prediction problems with nonlinear and time-varying characteristics, showing significant advantages.

[0048] S130, using a data set to perform model training on the preset transformer temperature prediction model, and using a variable parameter neural dynamics optimization model to dynamically optimize the long short-term memory network until a converged target transformer temperature prediction model is obtained.

[0049] The collected transformer data set is divided and model training is performed, and the variable parameter neural dynamics optimization model is used to dynamically optimize the long short-term memory network to obtain the optimal values ​​of the transformer temperature prediction model parameters; when the optimal values ​​of the transformer temperature prediction model parameters are obtained, a converged target transformer temperature prediction model is obtained. The optimal values ​​of the parameters are obtained by training and solving the variable parameter neural dynamics optimization model.

[0050] The embodiment of the present invention provides a transformer temperature prediction model training method, which obtains a data set for model training and preprocesses the data set; wherein the data set includes transformer operation data and external environment data; constructs a preset transformer temperature prediction model based on a long short-term memory network and a variable parameter neural dynamics optimization model; uses the data set to perform model training on the preset transformer temperature prediction model, and uses the variable parameter neural dynamics optimization model to dynamically optimize the long short-term memory network until a converged target transformer temperature prediction model is obtained. The technical solution of the embodiment of the present invention is adopted, and the transformer temperature prediction model training method based on the variable parameter neural dynamics optimization long short-term memory network is obtained to obtain a converged target transformer temperature prediction model, which can better capture the nonlinear characteristics and complex time dependencies in the transformer operation state, and effectively improve the accuracy of transformer temperature prediction.

[0051] Embodiment 2

[0052] Figure 2 The flowchart of a transformer temperature prediction model training method provided in an embodiment of the present invention. The embodiment of the present invention further optimizes the above embodiment on the basis of the above embodiment, and the embodiment of the present invention can be combined with various optional solutions in one or more of the above embodiments. Figure 2 As shown, the transformer temperature prediction model training method provided in the embodiment of the present invention may include the following steps:

[0053] S210: Obtain a data set for model training, and preprocess the data set.

[0054] Among them, the transformer operation data and external environment data used for model training are obtained and preprocessed.

[0055] As an optional but non-limiting implementation, the step of obtaining a data set for model training and preprocessing the data set includes but is not limited to steps A1-A2:

[0056] Step A1: Acquire initial transformer operation data and external environment data for transformer temperature prediction model training; wherein the transformer operation data includes load current and transformer temperature, and the external environment data includes ambient temperature and wind speed.

[0057] Step A2: Standardize the initial transformer operation data and external environment data to obtain target data required for executing transformer temperature prediction model training.

[0058] The transformer temperature data set is established based on a large amount of data recorded during simulation experiments and power grid operation. The load current, transformer temperature, ambient temperature and wind speed for transformer temperature prediction model training are obtained and a data set is constructed; the data set is standardized to ensure data quality and consistency.

[0059] S220. Construct a transformer temperature prediction model based on long short-term memory network.

[0060] The long short-term memory network unit consists of a memory unit C t and three gating mechanisms (forget gate f t , input gate i t , output gate O t ). The calculation process for each time step is as follows:

[0061] The forget gate determines the memory unit C of the previous time step t-1 How much information is retained to the current time step can be expressed as:

[0062] f t =σ(W f ×[h t-1 , x t ]+b f )

[0063] Among them, f t is the output of the forget gate, x t is the input data of the current time step, h t-1 is the hidden state of the previous time step, w f represents the weight matrix, b f represents the bias term, and σ represents the Sigmoid activation function.

[0064] The input gate determines how much new information is stored in the memory cell. It consists of two parts: one is the activation value i of the input gate. t , and the other is the content of the candidate memory cell It can be expressed as:

[0065]

[0066] Among them, w i and w C represents the weight matrix, b i and b C Represents the bias term.

[0067] Combine the forget gate and the input gate to update the memory unit C of the current time step t , which can be expressed as:

[0068]

[0069] The output gate determines the hidden state h of the current time step t , which can be expressed as:

[0070] O t =σ(w o ×[h t-1 , x t ]+b O )

[0071] h t =O t ×tanh(C t )

[0072] Among them, C t is the state of the memory cell at the current moment, C t-1 is the state of the memory cell at the last moment, O t is the output result of the activation function Sigmoid of the output gate, h t is the output at the current moment.

[0073] The LSTM network effectively controls the flow of information between time steps by introducing forget gates, input gates, and output gates. The above formula describes in detail the calculation process of the LSTM unit at each time step. Through these steps, LSTM can remember and utilize key information over a long period of time.

[0074] The LSTM network prevents gradient vanishing and explosion by selectively forgetting or memorizing relevant time series information through these three basic gating units, and can better capture the long-term dependencies in time series data, which is suitable for nonlinear data prediction. Compared with other neural networks, the LSTM network is more suitable for transformer temperature prediction.

[0075] S230, introducing a variable parameter neural dynamics optimization model into the transformer temperature prediction model to construct a preset transformer temperature prediction model.

[0076] Among them, the variable parameter neural dynamics optimization model is used to dynamically adjust the network parameters of the long short-term memory network.

[0077] As an optional but non-limiting implementation, the variable parameter neural dynamics optimization model is characterized as follows:

[0078]

[0079] Among them, t∈R represents continuous time; e(t)∈R represents the deviation function between the actual output of the network and the expected output; λ∈R, λ>0, represents the constant that controls the convergence speed; φ(.) represents the mapping function, where if the mapping function satisfies the monotonically increasing odd function condition, the model error continues to decrease.

[0080] Taking the input of the embodiment of the present invention as x, the model output is Take ω∈R as an example. ω∈R represents all learnable parameters in the long short-term memory network model, and the relationship between the three can be expressed as:

[0081]

[0082] The network optimization method of deep learning generally seeks the optimal parameter matrix to minimize the objective function, where the objective function is often the loss function of the model, which can be expressed as:

[0083]

[0084] in, It can be some common loss functions, such as mean square error loss or cross entropy loss. So we have:

[0085]

[0086] Specifically, in order to maintain the general premise, λ is taken as 0.1; in order to reduce the amount of network calculation, the mapping function is taken as φ(x) = x. According to the chain rule, the left side of the above formula is equivalent to the following form:

[0087]

[0088] We can further obtain:

[0089]

[0090] Here For ω The generalized inverse matrix of . Considering that the model optimization is implemented on a digital computer, the above continuous time expression needs to be discretized. Combined with the Euler discrete formula, it can be transformed into:

[0091]

[0092] in Represents discrete step length, k>0, k∈R represents discrete time. In actual implementation, the step length is 0.1.

[0093] Consider ξ ω The generalized inverse matrix of In the actual network model, the number of parameters is very large, and ξ is solved once in each iteration. ω The inverse matrix of is extremely time-consuming. Here we can replace ξ ω Abstractly understood as a row vector, because ξ ω is the partial derivative of the loss value for each ω, and its relationship does not change due to the expression form. According to the pseudo-inverse solution method of the row vector, we can get:

[0094]

[0095] Substituting the above formula into the iterative formula and further simplifying the calculation of the pseudo-inverse, we can get:

[0096]

[0097] Among them, α K is the equivalent learning rate.

[0098] The embodiment of the present invention introduces a variable parameter neural dynamics optimization model into the transformer temperature prediction model to construct a preset transformer temperature prediction model, which can better capture the nonlinear characteristics and complex time dependencies in the transformer operating state, and effectively improve the accuracy of transformer temperature prediction.

[0099] S240, using a data set to perform model training on the preset transformer temperature prediction model, and using a variable parameter neural dynamics optimization model to dynamically optimize the long short-term memory network until a converged target transformer temperature prediction model is obtained.

[0100] Among them, see Figure 3 , a variable parameter neural dynamics optimization model is used to dynamically optimize the long short-term memory network to establish the optimal transformer temperature prediction model.

[0101] As an optional but non-limiting implementation, the preset transformer temperature prediction model is trained using a data set, and the long short-term memory network is dynamically optimized using a variable parameter neural dynamics optimization model until a converged target transformer temperature prediction model is obtained, including but not limited to steps B1-B5:

[0102] Step B1: Divide the dataset into training set, validation set and test set.

[0103] Step B2: Initialize the network parameters of the long short-term memory network; wherein the initialized network parameters include the number of network layers, the number of hidden layer units, and the learning rate.

[0104] Step B3: Dynamically optimize and solve the network parameters of the long short-term memory network through the variable parameter neural dynamics optimization model, determine the optimal value of the transformer temperature prediction model, and verify it on the verification set based on the solution parameters.

[0105] Step B4: If the verification result meets the preset verification conditions, the converged target transformer temperature prediction model is output.

[0106] Step B5: If the verification result does not meet the preset verification conditions, the network parameters of the long short-term memory network are continuously dynamically optimized and solved until a converged target transformer temperature prediction model is obtained.

[0107] Among them, the transformer operation data set and the external environmental condition data set are obtained, and the data set division and preprocessing operations are performed. For example, the training set and the test set are divided in a ratio of 9:1; the input data includes but is not limited to transformer temperature, load current, voltage, current and ambient temperature, etc., and the embodiment of the present invention does not impose specific restrictions on this. The network parameters of the LSTM network are initialized, and the network parameters are not limited to the number of network layers, the number of hidden layer units and the learning rate, etc. The network parameters of the LSTM network are dynamically optimized and solved through a variable parameter neural dynamics optimization model, and the obtained parameters are verified on the verification set. If the preset verification conditions are met, the optimal prediction model is output; otherwise, the network parameters are optimized and retrained until the preset verification conditions are met; the preset verification conditions may refer to the accuracy of the verification results reaching a certain threshold.

[0108] The embodiment of the present invention provides a transformer temperature prediction model training method, which dynamically optimizes the long short-term memory network by using a variable parameter neural dynamics optimization model to obtain a converged target transformer temperature prediction model. Through the variable parameter neural dynamics optimization model, the long short-term memory network can adaptively adjust parameters, and still maintain stable prediction performance when dealing with different operating conditions and environmental changes, significantly improving the robustness and adaptability of the model. The transformer temperature prediction model provided by the present invention can accurately predict the heating of the transformer and improve the operating safety of power equipment.

[0109] Embodiment 3

[0110] Figure 4This is a flowchart of a transformer temperature prediction method provided in an embodiment of the present invention. This embodiment is applicable to the case of predicting transformer temperature. The method of this embodiment can be executed by a transformer temperature prediction device, which can be implemented in hardware and / or software. The device can be configured in a transformer temperature prediction server. The method specifically includes the following steps:

[0111] S410: Determine transformer target data to be predicted.

[0112] The target data includes transformer operation data and external environment data.

[0113] S420, inputting the target data into a target transformer temperature prediction model, performing temperature prediction on the transformer to be predicted, outputting the heating state of the transformer to be predicted and issuing an abnormal warning.

[0114] Among them, see Figure 5 , construct a long short-term memory network architecture, and introduce a variable parameter neural dynamics model to adjust the network parameters to enhance the network's learning ability for complex time series data; use a divided data set to train the model, and use a validation set to verify the trained model to obtain a target transformer temperature prediction model. The target transformer temperature prediction model is deployed to the actual transformer monitoring system for temperature prediction; for example, the operating data and external environmental data of the transformer to be predicted are obtained, and the operating data and external environmental data of the transformer to be predicted are input into the target transformer temperature prediction model for prediction, and the heating state of the transformer to be predicted is output, and an abnormal warning is issued when the temperature of the transformer to be predicted is abnormal.

[0115] Optionally, the transformer temperature prediction result of the target transformer temperature prediction model is combined with the actual monitoring data to adjust the transformer operation strategy in real time to prevent overheating and improve the safety and reliability of equipment operation.

[0116] The embodiment of the present invention provides a transformer temperature prediction method, which predicts the transformer operating temperature by constructing a target transformer temperature prediction model based on a long short-term memory network and a variable parameter neural dynamics optimization model, so as to give an early warning when the operating temperature is abnormal. Accurately predicting the heating condition of the transformer prevents the transformer from overheating, provides a scientific basis for the operation and maintenance of the transformer and the prevention of faults, and effectively improves the safety and reliability of the power system.

[0117] Embodiment 4

[0118] Figure 6This is a schematic diagram of the structure of a transformer temperature prediction model training device provided in an embodiment of the present invention. The technical solution of this embodiment can be applied to the case of building a transformer temperature prediction model. The device can be implemented by software and / or hardware and is generally integrated on any electronic device with network communication function, including but not limited to: servers, computers, personal digital assistants and other devices. Figure 6 As shown, the transformer temperature prediction model training device provided in this embodiment may include: a data set acquisition module 610, an initial model construction module 620 and a target model training module 630; wherein,

[0119] The data set acquisition module 610 is used to acquire a data set for model training and preprocess the data set; wherein the data set includes transformer operation data and external environment data;

[0120] An initial model building module 620, for building a preset transformer temperature prediction model based on a long short-term memory network and a variable parameter neural dynamics optimization model;

[0121] The target model training module 630 is used to use the data set to perform model training on the preset transformer temperature prediction model, and use the variable parameter neural dynamics optimization model to dynamically optimize the long short-term memory network until a converged target transformer temperature prediction model is obtained.

[0122] Based on the above embodiment, optionally, the data set acquisition module is specifically used to:

[0123] Acquire initial transformer operation data and external environment data for transformer temperature prediction model training; wherein the transformer operation data includes load current and transformer temperature, and the external environment data includes ambient temperature and wind speed;

[0124] The initial transformer operation data and external environment data are standardized to obtain target data required for executing transformer temperature prediction model training.

[0125] Based on the above embodiment, optionally, the initial model building module is specifically used to:

[0126] Construct a transformer temperature prediction model based on long short-term memory network;

[0127] A variable parameter neural dynamics optimization model is introduced into the transformer temperature prediction model to construct a preset transformer temperature prediction model; wherein the variable parameter neural dynamics optimization model is used to dynamically adjust the network parameters of the long short-term memory network.

[0128] On the basis of the above embodiment, optionally, the variable parameter neural dynamics optimization model is characterized as follows:

[0129]

[0130] Among them, t∈R represents continuous time; e(t)∈R represents the deviation function between the actual output of the network and the expected output; λ∈R, λ>0, represents the constant that controls the convergence speed; φ(.) represents the mapping function, where if the mapping function satisfies the monotonically increasing odd function condition, the model error continues to decrease.

[0131] Based on the above embodiment, optionally, the target model training module is specifically used to:

[0132] Divide the dataset into training set, validation set and test set;

[0133] Initialize the network parameters of the long short-term memory network; wherein the initialized network parameters include the number of network layers, the number of hidden layer units, and the learning rate;

[0134] The network parameters of the long short-term memory network are dynamically optimized and solved through the variable parameter neural dynamics optimization model to determine the optimal value of the transformer temperature prediction model, and then verified on the validation set based on the solved parameters;

[0135] If the verification result meets the preset verification conditions, the converged target transformer temperature prediction model is output;

[0136] If the verification result does not meet the preset verification conditions, the network parameters of the long short-term memory network are continuously dynamically optimized and solved until a converged target transformer temperature prediction model is obtained.

[0137] The transformer temperature prediction model training device provided in the embodiment of the present invention can execute the transformer temperature prediction model training method provided in any embodiment of the present invention mentioned above, and has the corresponding functions and beneficial effects of executing the transformer temperature prediction model training method. For detailed process, please refer to the relevant operations of the transformer temperature prediction model training method in the aforementioned embodiment.

[0138] Embodiment 5

[0139] Figure 7 : is a schematic diagram of the structure of a transformer temperature prediction device provided in an embodiment of the present invention. The technical solution of this embodiment can be applied to the case of transformer temperature prediction. The device can be implemented by software and / or hardware and is generally integrated on any electronic device with network communication function, including but not limited to: servers, computers, personal digital assistants and other devices. Figure 7 As shown, the transformer temperature prediction device provided in this embodiment may include: a target data determination module 710 and a temperature prediction module 720; wherein,

[0140] The target data determination module 710 is used to determine the target data of the transformer to be predicted; the target data includes transformer operation data and external environment data;

[0141] The temperature prediction module 720 is used to input the target data into the target transformer temperature prediction model, perform temperature prediction on the transformer to be predicted, output the heating state of the transformer to be predicted and issue an abnormal warning.

[0142] The transformer temperature prediction device provided in the embodiment of the present invention can execute the transformer temperature prediction method provided in any embodiment of the present invention, and has the corresponding functions and beneficial effects of executing the transformer temperature prediction method. For detailed process, please refer to the relevant operations of the transformer temperature prediction method in the above embodiment.

[0143] Embodiment 6

[0144] Figure 8 1 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. The electronic device 10 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 may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) 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.

[0145] like Figure 8 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0146] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0147] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as transformer temperature prediction model training and transformer temperature prediction method.

[0148] In some embodiments, the transformer temperature prediction model training and the transformer temperature prediction method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the transformer temperature prediction model training and the transformer temperature prediction method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to perform the transformer temperature prediction model training and the transformer temperature prediction method in any other appropriate manner (e.g., by means of firmware).

[0149] 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.

[0150] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may 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.

[0151] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage 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. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. 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.

[0152] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device 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 a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0153] 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), a blockchain network, and the Internet.

[0154] A computing 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 client and server relationship 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.

[0155] Embodiment 7

[0156] An embodiment of the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements the transformer temperature prediction model training and transformer temperature prediction method provided in any embodiment of the present application.

[0157] In the process of implementation, the computer program product can be written in one or more programming languages ​​or a combination thereof to perform the computer program code of the present invention, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0158] 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.

[0159] 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. A transformer temperature prediction model training method, characterized in that: The method comprises: Acquire a data set for model training and preprocess the data set; wherein the data set includes transformer operation data and external environment data; Construct a preset transformer temperature prediction model based on long short-term memory network and variable parameter neural dynamics optimization model; The preset transformer temperature prediction model is trained using a data set, and the long short-term memory network is dynamically optimized using a variable parameter neural dynamics optimization model until a converged target transformer temperature prediction model is obtained.

2. The method according to claim 1, characterized in that: The step of obtaining a data set for model training and preprocessing the data set includes: Acquire initial transformer operation data and external environment data for transformer temperature prediction model training; wherein the transformer operation data includes load current and transformer temperature, and the external environment data includes ambient temperature and wind speed; The initial transformer operation data and external environment data are standardized to obtain target data required for executing transformer temperature prediction model training.

3. The method according to claim 1, characterized in that The preset transformer temperature prediction model based on the long short-term memory network and the variable parameter neural dynamics optimization model is constructed, including: Construct a transformer temperature prediction model based on long short-term memory network; A variable parameter neural dynamics optimization model is introduced into the transformer temperature prediction model to construct a preset transformer temperature prediction model; wherein the variable parameter neural dynamics optimization model is used to dynamically adjust the network parameters of the long short-term memory network.

4. The method according to claim 3, characterized in that: The variable parameter neural dynamics optimization model is characterized as follows: Among them, t∈R represents continuous time; e(t)∈R represents the deviation function between the actual output of the network and the expected output; λ∈R, λ>0, represents the constant that controls the convergence speed; φ(.) represents the mapping function, where if the mapping function satisfies the monotonically increasing odd function condition, the model error continues to decrease.

5. The method according to claim 1, characterized in that The method of using a data set to train the preset transformer temperature prediction model and using a variable parameter neural dynamics optimization model to dynamically optimize the long short-term memory network until a converged target transformer temperature prediction model is obtained includes: Divide the dataset into training set, validation set and test set; Initialize the network parameters of the long short-term memory network; wherein the initialized network parameters include the number of network layers, the number of hidden layer units, and the learning rate; The network parameters of the long short-term memory network are dynamically optimized and solved through the variable parameter neural dynamics optimization model to determine the optimal value of the transformer temperature prediction model, and then verified on the validation set based on the solved parameters; If the verification result meets the preset verification conditions, the converged target transformer temperature prediction model is output; If the verification result does not meet the preset verification conditions, the network parameters of the long short-term memory network are continuously dynamically optimized and solved until a converged target transformer temperature prediction model is obtained.

6. A transformer temperature prediction method, characterized in that: Based on the target transformer temperature prediction model obtained by the transformer temperature prediction model training method according to any one of claims 1 to 5, the transformer temperature prediction method comprises: Determine the transformer target data to be predicted; the target data includes transformer operation data and external environment data; The target data is input into the target transformer temperature prediction model, the temperature of the transformer to be predicted is predicted, the heating state of the transformer to be predicted is output and an abnormal warning is issued.

7. A transformer temperature prediction model training device, characterized in that: The device comprises: A data set acquisition module, used to acquire a data set for model training and preprocess the data set; wherein the data set includes transformer operation data and external environment data; An initial model building module for building a preset transformer temperature prediction model based on a long short-term memory network and a variable-parameter neural dynamics optimization model; The target model training module is used to use the data set to perform model training on the preset transformer temperature prediction model, and use the variable parameter neural dynamics optimization model to dynamically optimize the long short-term memory network until a converged target transformer temperature prediction model is obtained.

8. A transformer temperature prediction device, characterized in that: The device comprises: A target data determination module is used to determine the target data of the transformer to be predicted; the target data includes transformer operation data and external environment data; The temperature prediction module is used to input the target data into the target transformer temperature prediction model, perform temperature prediction on the transformer to be predicted, output the heating state of the transformer to be predicted and issue an abnormal warning.

9. An electronic device, characterized in that: include: one or more processors; A storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the transformer temperature prediction model training method described in any one of claims 1-5 and the transformer temperature prediction method described in claim 6.

10. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions are used to execute the transformer temperature prediction model training method as described in any one of claims 1 to 5 and the transformer temperature prediction method as described in claim 6 when executed by a computer processor.

Images