Industrial capacitor capacitance prediction method based on LSTM model and related equipment

Abstract

The invention provides an industrial capacitor capacitance prediction method based on an LSTM model and related equipment. The method comprises the following steps: firstly, collecting test data of a target industrial capacitor based on a capacitor capacitance test system; then training the improved long short-term memory network model to obtain a trained long short-term memory network model; finally, test data are input into the trained long-short-term memory network model for capacitance prediction, a capacitance prediction result is obtained, and the capacitance prediction result is used for evaluating the aging trend and the remaining life of the target industrial capacitor. Compared with the prior art, the simple test system is used for controlling and collecting the test data of the target industrial capacitor, the capacitance value of the target industrial capacitor is predicted through the improved long and short-term memory network and the test data, the spatial characteristics of the historical data of the industrial capacitor and the time correlation of the time sequence data can be extracted, and the accuracy of the test data is improved. And the prediction precision is improved.

Description

Technical Field

[0001] The present invention relates to the field of industrial testing technology, and in particular to an industrial capacitor capacitance prediction method based on an LSTM model and related equipment. Background Art

[0002] Industrial capacitors are common electronic components in industrial equipment. Their performance and lifespan are crucial to the stability and reliability of industrial equipment. The aging of industrial capacitors may lead to reduced equipment performance, increased failure rates, and increased maintenance costs. Therefore, accurately predicting the aging trend and service life of industrial capacitors is crucial for maintenance and operation management.

[0003] In recent years, data-driven life prediction methods have become mainstream. A historical database is established for the monitoring data of the entire operation process of the capacitor through testing equipment, and a corresponding prediction model is established through the historical database. Then, combined with the monitoring data of the actual operation of the capacitor, its life prediction is completed.

[0004] At present, the laboratory test equipment of domestic capacitor manufacturers is basically customized by instrument manufacturers. By assembling multiple testers and customizing functional requirements, the purpose of automation, long-term operation, and a large number of tests can be achieved. The test equipment is large in size and the cost of obtaining test data is high. In addition, the current traditional data-driven life prediction method cannot simultaneously consider the spatial characteristics of industrial capacitor historical data and the time correlation of time series data, and the prediction accuracy is not high. Summary of the invention

[0005] The present invention provides an industrial capacitor capacitance prediction method and related equipment based on an LSTM model, the purpose of which is to improve the prediction accuracy.

[0006] In order to achieve the above object, the present invention provides an industrial capacitor capacitance prediction method based on LSTM model, comprising:

[0007] Step 1, collecting test data of a target industrial capacitor based on a capacitance value test system;

[0008] Step 2, training the improved long short-term memory network model to obtain a trained long short-term memory network model;

[0009] Step 3: Input the test data into the trained long short-term memory network model to predict the capacitance and obtain the capacitance prediction result. The capacitance prediction result is used to evaluate the aging trend and remaining life of the target industrial capacitor.

[0010] Furthermore, the capacitance value test system includes a test subsystem, a monitoring subsystem, and a data management cloud platform;

[0011] A test subsystem, used to test the target industrial capacitor and obtain test data;

[0012] The monitoring subsystem is used to monitor the test process of the test subsystem and obtain monitoring data;

[0013] Data management cloud platform, used to process test data and monitoring data.

[0014] Further, the test subsystem includes an industrial computer, a tester, and an environmental temperature chamber;

[0015] Industrial computer, used to run the test software and establish communication with the tester for data transmission;

[0016] A tester for collecting the capacitance value and leakage current of the target capacitor;

[0017] Ambient temperature chamber, used to provide a test environment for the target capacitor.

[0018] Further, the tester includes a capacitance test energy box, a test board, and an Ethernet switch;

[0019] Capacitor test energy box, used to provide working power and electrical stress for the test board and target capacitor;

[0020] A test board is used to collect electrical stress data, capacitance value and leakage current as test data;

[0021] Ethernet switch, used to establish a data transmission channel between the test board and the data management cloud platform.

[0022] Furthermore, the improved long short-term memory network model includes an input layer, a convolution layer, a pooling layer, a dropout layer, a first long short-term memory layer, a second long short-term memory layer, a third long short-term memory layer, an attention layer, a fully connected layer, and an output layer, which are connected in sequence.

[0023] Furthermore, before inputting the test data into the trained LSTM network model for capacity prediction, the following steps are also included:

[0024] Clean the test data to obtain cleaned test data;

[0025] The cleaned test data is normalized to obtain normalized test data.

[0026] Furthermore, the test data is cleaned to obtain cleaned test data, including:

[0027] Remove outliers from test data;

[0028] The 3σ criterion and moving sliding window technology are used to detect the test data after removing outliers to obtain outliers, and the outliers are replaced by the mean of the sliding window to obtain the cleaned test data.

[0029] The present invention also provides an industrial capacitor capacitance prediction device based on the LSTM model, comprising:

[0030] An acquisition module, used for acquiring test data of a target industrial capacitor based on a capacitance value test system;

[0031] A training module, used for training the improved long short-term memory network model to obtain a trained long short-term memory network model;

[0032] The prediction module is used to input the test data into the trained long short-term memory network model to predict the capacitance and obtain the capacitance prediction result. The capacitance prediction result is used to evaluate the aging trend and remaining life of the target industrial capacitor.

[0033] The present invention also provides a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, a method for predicting the capacitance of an industrial capacitor based on an LSTM model is implemented.

[0034] The present invention also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, a method for predicting the capacitance of an industrial capacitor based on an LSTM model is implemented.

[0035] The above scheme of the present invention has the following beneficial effects:

[0036] The present invention first collects test data of target industrial capacitors based on a capacitance test system; then trains an improved long short-term memory network model to obtain a trained long short-term memory network model; finally, the test data is input into the trained long short-term memory network model for capacitance prediction to obtain a capacitance prediction result, and the capacitance prediction result is used to evaluate the aging trend and remaining life of the target industrial capacitor; compared with the prior art, the present invention uses a simple test system to control the collection of test data of the target industrial capacitor, predicts the capacitance of the target industrial capacitor through the improved long short-term memory network and test data, and can extract the spatial characteristics of the historical data of the industrial capacitor and the time correlation of the time series data, thereby improving the prediction accuracy.

[0037] Other beneficial effects of the present invention will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a schematic diagram of a flow chart of an embodiment of the present invention;

[0039] Figure 2 This is a principle block diagram of a capacitor value testing system according to an embodiment of the present invention;

[0040] Figure 3 Schematic diagram of the structure of the improved long short-term memory network model in an embodiment of the present invention;

[0041] Figure 4 The flowchart of the training of the improved long short-term memory network model in the embodiment of the present invention;

[0042] Figure 5 Schematic diagram of the structure of an industrial capacitor capacitance prediction device in an embodiment of the present invention;

[0043] Figure 6 Schematic diagram of the structure of a terminal device in an embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, the following will be described in detail with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are part of the embodiments of the present invention, rather than 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.

[0045] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0046] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installation", "connection" and "connection" should be understood in a broad sense, for example, it can be a locking connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0047] In addition, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0048] In view of the existing problems, the present invention provides an industrial capacitor capacitance prediction method based on an LSTM model and related equipment.

[0049] like Figure 1 As shown, an embodiment of the present invention provides an industrial capacitor capacitance prediction method based on an LSTM model, comprising:

[0050] Step 1, collecting test data of a target industrial capacitor based on a capacitance value test system;

[0051] Step 2, training the improved long short-term memory network model to obtain a trained long short-term memory network model;

[0052] Step 3: Input the test data into the trained long short-term memory network model to predict the capacitance and obtain the capacitance prediction result. The capacitance prediction result is used to evaluate the aging trend and remaining life of the target industrial capacitor.

[0053] Most preferably, Figure 2 As shown, the capacitance value test system includes a test subsystem, a monitoring subsystem, and a data management cloud platform;

[0054] A test subsystem, used to test the target industrial capacitor and obtain test data;

[0055] The monitoring subsystem is used to monitor the test process of the test subsystem in the long-term unattended state and obtain monitoring data to ensure the safe operation of the capacitance test system;

[0056] The data management cloud platform is used to process test data and monitoring data to realize functions such as data storage, analysis, comparison, statistics, and visualization.

[0057] Most preferably, the test subsystem includes an industrial computer, a tester, and an environmental temperature chamber;

[0058] Industrial computer, used to run the test software and establish communication with the tester for data transmission, so as to realize functions such as command and data interaction, data storage, tester parameter configuration, and calibration;

[0059] A tester, for collecting a capacitance value and a leakage current of a target capacitor according to a received command;

[0060] The environmental temperature box is used to provide a test environment for the target capacitor. The test environment includes four environments: normal temperature, low temperature, high temperature and accelerated temperature.

[0061] Most preferably, the tester includes a capacitance test energy box, a test board, and an Ethernet switch;

[0062] The capacitor test energy box is used to provide working power and electrical stress for the test board and the target capacitor. It is set outside the ambient temperature box as an independent entity and is electrically connected to the test board and the target capacitor through cables;

[0063] A test board is used to collect electrical stress data, capacitance value and leakage current as test data;

[0064] Ethernet switch, used to establish a data transmission channel between the test board and the data management cloud platform.

[0065] According to the distribution characteristics of industrial capacitor capacitance data, the embodiment of the present invention improves the long short-term memory network, adopts a single-step prediction method, uses the actual value corresponding to the previous moment and the historical moment data as data input for prediction, and uses the predicted value as output. For example, the test data of Mubao capacitors in the past three days are arranged in chronological order to form a sequence X n =[X (1) ,X (2) ,X (3) ] as input, and get the next moment X from the output layer (4) Compared with using the predicted value as the input for subsequent predictions, this improved long short-term memory network model can reduce the impact of inaccurate previous prediction results on subsequent prediction results, thus having better robustness and effectiveness.

[0066] Specifically, if Figure 3 As shown, the improved long short-term memory network model includes an input layer, a convolution layer, a pooling layer, a dropout layer, a first long short-term memory layer, a second long short-term memory layer, a third long short-term memory layer, an attention layer, a fully connected layer, and an output layer, which are connected in sequence.

[0067] The input layer is used to receive input test data. In the embodiment of the present invention, the test data of the past three days is taken as an example, that is, the dimension of the input layer input is 3;

[0068] The convolution layer is used to extract the features of the test data using a one-dimensional convolution kernel. The size of the convolution kernel is set to 1, which means that feature extraction is performed on the test data at each time point to obtain feature extraction results.

[0069] The pooling layer is used to use a one-dimensional maximum pooling layer to reduce the dimension of the feature extraction result, improve the abstraction ability of the model, and take the maximum value of the pooling window;

[0070] The Dropout layer is used to prevent data overfitting;

[0071] The first long short-term memory layer, the second long short-term memory layer, and the third long short-term memory layer are all used to extract spatial features and temporal features from the feature extraction results after pooling;

[0072] The attention layer is used to calculate the attention weights and multiply the extracted spatial features and temporal features with the attention weights to implement the attention mechanism, so that the model can focus on the more important parts of the time series;

[0073] The fully connected layer is used to integrate the features output by the attention layer;

[0074] The output layer is used to output the capacitance prediction result with a dimension of 1, that is, the capacitance value for the next day.

[0075] In the embodiment of the present invention, the number of layers and the number of neurons of the long short-term memory are determined by parameter tuning.

[0076] Specifically, if Figure 4 As shown, step 2 includes:

[0077] The Adam algorithm is selected as the optimization algorithm of the improved long short-term memory network model. The root mean square error (RMSE) is used as the loss function to conduct experimental analysis on the number of hidden layer nodes (hidden size), batch size (batch size), learning rate (learning-rate), and number of iterations (epoch) to determine the optimal parameters of the model;

[0078] The determination coefficient R2 is selected as the evaluation index. The determination coefficient (R2) is also called the goodness of fit. Its value range is between 0 and 1. The higher the value of R2, the stronger the model's ability to explain the data and the better the fitting effect. The calculation formula is:

[0079]

[0080] The formula represents the ratio of 1 minus the sum of the squares of the differences between the actual value and the predicted value to the sum of the squares of the differences between the actual value and the average of the actual values;

[0081] Initialize the parameters of the network and input the data to be used for training;

[0082] The network performs forward propagation, the weight coefficient remains unchanged, and the output of the previous layer is used as the input of the next layer, and the input information is transmitted from the input layer to the hidden layer, and then from the hidden layer to the output layer;

[0083] The network performs back propagation, how to calculate the error evaluation index, feed the error value back to the input layer, and continuously adjust the network weights;

[0084] The Adam optimizer optimizes the network weights and checks whether the network model converges. This process needs to be repeated until the loss function is reduced to a certain extent. At this time, the network weights with the minimum training loss are obtained and the model training is completed.

[0085] Specifically, before inputting the test data into the trained LSTM network model for capacity prediction, the following steps are also included:

[0086] Clean the test data to obtain cleaned test data;

[0087] The cleaned test data is normalized to obtain normalized test data.

[0088] Specifically, the test data is cleaned to obtain cleaned test data, including:

[0089] Remove outliers from test data;

[0090] The 3σ criterion and moving sliding window technology are used to detect the test data after removing outliers to obtain outliers, and the outliers are replaced by the mean of the sliding window to obtain the cleaned test data.

[0091] In the embodiment of the present invention, since the collected test data contains a large amount of noise and spikes, and these abnormal data will interfere with the algorithm training process, it is necessary to perform outlier processing on the test data to remove the outliers in the test data.

[0092] According to the characteristics of industrial capacitance data distribution conforming to time series changes, the embodiment of the present invention adopts the 3σ criterion and the moving sliding window technology to detect the test data after removing outliers; wherein the 3σ criterion means that a set of test data is assumed to contain only random errors, and the standard deviation is obtained by calculation and processing, and an interval is determined according to a certain probability. It is considered that any error exceeding this interval is not a random error but a gross error, and the data containing the error should be eliminated. The 3σ criterion is:

[0093] The probability that the value is distributed in (μ-σ, μ+σ) is 0.6827

[0094] The probability that the value is distributed in (μ-2σ, μ+2σ) is 0.9545

[0095] The probability that the value is distributed in (μ-3σ, μ+3σ) is 0.9973

[0096] In summary, it can be considered that the values ​​are almost all concentrated in the interval (μ-3σ, μ+3σ), and the possibility of exceeding this range is less than 0.3%, so the interval (μ-3σ, μ+3σ) is identified as the confidence interval. In addition, since the traditional 3σ criterion is limited to the processing of sample data with normal or approximately normal distribution, it is not reliable to use the criterion to eliminate gross errors when the data does not conform to the normal distribution or the sample size is small. Therefore, it is considered to use the 3σ principle combined with a moving sliding window to detect outliers and replace the outliers with the mean of the sliding window. After adjusting different parameters (number of windows and standard deviation multiples), the best effect is achieved when the number of windows is 7 and the confidence interval is identified as (μ-2σ, μ+2σ).

[0097] The embodiment of the present invention scales the test data to the interval [0,1] by using the min_max normalization technique, and the calculation formula is:

[0098]

[0099] Among them, Z norm represents the normalized test data, Z is the input test data, and Z min and Z max are the minimum and maximum values ​​of the input test data respectively.

[0100] The embodiment of the present invention first collects test data of the target industrial capacitor based on the capacitance test system; then trains the improved long short-term memory network model to obtain the trained long short-term memory network model; finally, the test data is input into the trained long short-term memory network model to predict the capacitance, and obtain the capacitance prediction result, which is used to evaluate the aging trend and remaining life of the target industrial capacitor; compared with the prior art, the embodiment of the present invention uses a simple test system to control the collection of test data of the target industrial capacitor, and predicts the capacitance of the target industrial capacitor through the improved long short-term memory network and test data, so as to extract the spatial characteristics of the historical data of the industrial capacitor and the time correlation of the time series data, thereby improving the prediction accuracy.

[0101] like Figure 5 As shown, the embodiment of the present invention further provides an industrial capacitor capacitance prediction device 100 based on the LSTM model, and the industrial capacitor capacitance prediction device 100 includes:

[0102] The acquisition module 101 is used to acquire the test data of the target industrial capacitor based on the capacitance value test system;

[0103] A training module 102 is used to train the improved long short-term memory network model to obtain a trained long short-term memory network model;

[0104] The prediction module 103 is used to input the test data into the trained long short-term memory network model to perform capacitance prediction and obtain a capacitance prediction result, which is used to evaluate the aging trend and remaining life of the target industrial capacitor.

[0105] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0106] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0107] The embodiment of the present invention also provides a terminal device, such as Figure 6 As shown, the terminal device D10 of this embodiment includes: at least one processor D100 ( Figure 6 Only one processor is shown in the figure), a memory D101, and a computer program D102 stored in the memory D101 and executable on the at least one processor D100. When the processor D100 executes the computer program D102, the above-mentioned industrial capacitor capacitance prediction method based on the LSTM model is implemented.

[0108] The terminal device D10 may be a computing device such as a desktop computer, a notebook, a PDA, a server, a server cluster, a cloud server, etc. The terminal device may include, but is not limited to, a processor D100 and a memory D101. Those skilled in the art will appreciate that Figure 6 This is only an example of the terminal device D10 and does not constitute a limitation on the terminal device D10. The terminal device D10 may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, it may also include input and output devices, network access devices, etc.

[0109] The processor D100 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0110] In some embodiments, the memory D101 may be an internal storage unit of the terminal device D10, such as a hard disk or memory of the terminal device D10. In other embodiments, the memory D101 may also be an external storage device of the terminal device D10, such as a plug-in hard disk, a smart memory card (SMC, SmartMedia Card), a secure digital (SD, Secure Digital) card, a flash card (Flash Card), etc. equipped on the terminal device D10. Further, the memory D101 may also include both an internal storage unit of the terminal device D10 and an external storage device. The memory D101 is used to store an operating system, an application program, a boot loader (BootLoader), data and other programs, such as the program code of the computer program, etc. The memory D101 may also be used to temporarily store data that has been output or is to be output.

[0111] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0112] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0113] An embodiment of the present invention further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, a method for predicting the capacitance of an industrial capacitor based on an LSTM model is implemented.

[0114] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the process in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device that can carry the computer program code to the construction device / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a disk or an optical disk.

[0115] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for predicting the capacitance of industrial capacitors based on LSTM model, characterized in that: include: Step 1, collecting test data of a target industrial capacitor based on a capacitance value test system; Step 2, training the improved long short-term memory network model to obtain a trained long short-term memory network model; Step 3: Input the test data into the trained long short-term memory network model to perform capacitance prediction to obtain a capacitance prediction result, which is used to evaluate the aging trend and remaining life of the target industrial capacitor.

2. The method for predicting the capacitance of industrial capacitors based on the LSTM model according to claim 1, characterized in that: The capacitance value testing system includes a testing subsystem, a monitoring subsystem, and a data management cloud platform; The test subsystem is used to test the target industrial capacitor and obtain test data; The monitoring subsystem is used to monitor the testing process of the testing subsystem and obtain monitoring data; The data management cloud platform is used to process the test data and the monitoring data.

3. The industrial capacitor capacitance prediction method based on the LSTM model according to claim 2 is characterized in that: The test subsystem includes an industrial computer, a tester, and an environmental temperature box; The industrial computer is used to run the test software and establish communication with the tester to perform data transmission; The tester is used to collect the capacitance value and leakage current of the target capacitor; The environmental temperature box is used to provide a test environment for the target capacitor.

4. The method for predicting the capacitance of industrial capacitors based on the LSTM model according to claim 3 is characterized in that: The tester includes a capacitance test energy box, a test board, and an Ethernet switch; The capacitance test energy box is used to provide working power and electrical stress for the test board and the target capacitor; The test board is used to collect electrical stress data, capacitance value and leakage current as test data; The Ethernet switch is used to establish a data transmission channel between the test board and the data management cloud platform.

5. The method for predicting the capacitance of industrial capacitors based on the LSTM model according to claim 4 is characterized in that: The improved long short-term memory network model includes an input layer, a convolution layer, a pooling layer, a dropout layer, a first long short-term memory layer, a second long short-term memory layer, a third long short-term memory layer, an attention layer, a fully connected layer, and an output layer which are connected in sequence.

6. The method for predicting the capacitance of industrial capacitors based on the LSTM model according to claim 5, before inputting the test data into the trained long short-term memory network model for capacitance prediction, further comprising: Cleaning the test data to obtain cleaned test data; The cleaned test data is normalized to obtain normalized test data.

7. The method for predicting the capacitance of industrial capacitors based on the LSTM model according to claim 6, characterized in that: The test data is cleaned to obtain cleaned test data, including: removing outliers from the test data; The 3σ criterion and moving sliding window technology are used to detect the test data after removing outliers to obtain outliers, and the outliers are replaced by the mean of the sliding window to obtain the cleaned test data.

8. An industrial capacitor capacitance prediction device based on LSTM model, characterized in that: include: An acquisition module, used for acquiring test data of a target industrial capacitor based on a capacitance value test system; A training module, used for training the improved long short-term memory network model to obtain a trained long short-term memory network model; The prediction module is used to input the test data into the trained long short-term memory network model to perform capacitance prediction to obtain a capacitance prediction result, and the capacitance prediction result is used to evaluate the aging trend and remaining life of the target industrial capacitor.

9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the industrial capacitor capacitance prediction method based on the LSTM model as described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for predicting the capacitance of an industrial capacitor based on an LSTM model as described in any one of claims 1 to 7 is implemented.

Images