LSTM-based method and device for predicting corrosion degree of switch sliding bed plate

Abstract

The application discloses a switch slide bed plate corrosion degree prediction method and device based on LSTM, acquires historical corrosion characteristic parameters of a switch slide bed plate, the historical corrosion characteristic parameters including corrosion area and average corrosion speed of the switch slide bed plate; inputs the historical corrosion characteristic parameters into a pre-trained switch slide bed plate corrosion degree prediction model based on LSTM, and outputs the corrosion area and average corrosion speed of the switch slide bed plate corresponding to different time points in a future time period; according to the corrosion area and average corrosion speed of the switch slide bed plate corresponding to different time points in the future time period, the average values of the corrosion area and average corrosion speed of the switch slide bed plate in the future time period are respectively calculated; according to the average values of the corrosion area and average corrosion speed of the switch slide bed plate in the future time period, and in combination with a preset corrosion evaluation standard, a switch slide bed plate corrosion degree evaluation value is obtained. The application is characterized by solving the problems of low prediction accuracy, sensitivity to the number of data samples and large error.

Description

Technical Field

[0001] This invention belongs to the field of fault diagnosis technology, specifically relating to a method and device for predicting the corrosion degree of turnout slide plates based on LSTM. Background Technology

[0002] As a fundamental sector of the economy, railways have always received high attention and strong support. The turnout slide plate is a key component ensuring safe railway operation. As a crucial structure enabling trains to switch tracks or cross over, it is considered one of the weakest links in the track system. The slide plate is an important component of the turnout, providing support for the switch rail and frog rail, and directly affecting the safe operation of high-speed trains.

[0003] Turnout slides are exposed to the elements year-round, making them highly susceptible to corrosion from wind, sand, rain, snow, and debris from trains. This corrosion not only affects the slides' lifespan but also increases the resistance to turning the turnout. Turnout slides are a critical component ensuring safe railway operation, so the consequences of corrosion are significant. Failure to promptly detect potential safety hazards in turnout slides will severely impact train safety. Therefore, determining the degree of corrosion of turnout slides is crucial for ensuring the safe and reliable operation of high-speed railways.

[0004] Currently, the common method for predicting the corrosion degree of turnout slide plates is to use a grey model. This involves studying the grey prediction model to develop an improved version, thereby achieving accurate predictions of turnout slide plate corrosion. However, this method suffers from low accuracy, is highly sensitive to the number of data samples, and exhibits significant errors. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention provides a method and device for predicting the corrosion degree of turnout slide plates based on LSTM, aiming to solve the problems of low prediction accuracy, sensitivity to the number of data samples, and large errors.

[0006] To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution:

[0007] A method for predicting the corrosion degree of turnout slide plates based on LSTM includes:

[0008] Obtain historical corrosion characteristic parameters of the turnout slide plate, including the corrosion area and average corrosion rate of the turnout slide plate;

[0009] The historical corrosion characteristic parameters are input into a pre-trained LSTM-based model for predicting the corrosion degree of turnout slide plates, and the corrosion area and average corrosion rate of the turnout slide plates at different times in the future time period are output.

[0010] Based on the corrosion area and average corrosion rate of the turnout slide plate at different times within the future time period, the average values ​​of the corrosion area and average corrosion rate of the turnout slide plate within the future time period are calculated respectively.

[0011] The corrosion degree assessment value of the turnout slide plate is obtained based on the average corrosion area and average corrosion rate of the turnout slide plate within the future time period, combined with the preset corrosion evaluation criteria.

[0012] Furthermore, the method for obtaining the corrosion area of ​​the turnout slide plate includes:

[0013] Images of the turnout slide plate during the corrosion process were acquired;

[0014] The image during the corrosion process is binarized.

[0015] The corrosion area of ​​the turnout slide plate is obtained by calculating the proportion of white pixels to all pixels in the binarized image.

[0016] Furthermore, the method for obtaining the average corrosion rate of the turnout slide plate includes:

[0017] The mass of the turnout slide plate was collected during the corrosion process;

[0018] The average corrosion rate of the turnout slide plate was calculated using the weight gain method based on the mass during the corrosion process.

[0019] Furthermore, the training method for the LSTM-based turnout slide plate corrosion prediction model includes:

[0020] Obtain a sample set, which includes several sample data, including the corrosion area and average corrosion rate of the turnout slide plate at different times;

[0021] The initial LSTM network model is trained using the sample set to obtain the LSTM-based prediction model for the corrosion degree of the turnout slide plate.

[0022] Further, the step of obtaining the corrosion degree assessment value of the turnout slide plate based on the average value of the corrosion area and average corrosion rate of the turnout slide plate within the future time period, combined with a preset corrosion evaluation standard, includes:

[0023] The average corrosion area of ​​the turnout slide plate within the future time period is compared with the preset corrosion area rating rule for the turnout slide plate to obtain the corrosion area rating of the turnout slide plate within the future time period.

[0024] The average corrosion rate of the turnout slide plate within the future time period is compared with the preset corrosion resistance rating rule for the turnout slide plate to obtain the corrosion resistance level of the turnout slide plate within the future time period.

[0025] The corrosion degree assessment value of the turnout slide plate is calculated based on the corrosion area level and corrosion resistance level of the turnout slide plate in the future time period.

[0026] Furthermore, the corrosion area rating rule for the turnout slide plate includes:

[0027] Different corrosion area ranges correspond to different corrosion area grades;

[0028] The corrosion resistance rating rules for the turnout slide plates include:

[0029] Different average corrosion rate ranges correspond to different corrosion resistance levels.

[0030] Further, the step of calculating the corrosion degree assessment value of the turnout slide plate based on the corrosion area level and corrosion resistance level of the turnout slide plate within the future time period includes:

[0031] Set an upper limit for the evaluation value;

[0032] Different corrosion area levels correspond to different scoring values;

[0033] Different ranges of corrosion resistance grades correspond to different rating values;

[0034] The corrosion degree assessment value of the turnout slide plate is obtained by subtracting the score value corresponding to the corrosion area level from the upper limit of the assessment value, and then subtracting the score value corresponding to the corrosion resistance level.

[0035] A LSTM-based device for predicting the corrosion level of turnout slide plates includes:

[0036] The acquisition module is used to acquire historical corrosion characteristic parameters of the turnout slide plate, including the corrosion area and average corrosion rate of the turnout slide plate.

[0037] The prediction module is used to input the historical corrosion feature parameters into a pre-trained LSTM-based corrosion prediction model for turnout slide plates, and output the corrosion area and average corrosion rate of the turnout slide plates at different times in the future time period.

[0038] The calculation module is used to calculate the average value of the corrosion area and average corrosion rate of the turnout slide plate within the future time period, based on the corrosion area and average corrosion rate of the turnout slide plate at different times within the future time period.

[0039] The evaluation module is used to obtain an evaluation value of the corrosion degree of the turnout slide plate based on the average value of the corrosion area and average corrosion rate of the turnout slide plate within the future time period, and in combination with a preset corrosion evaluation standard.

[0040] An apparatus includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the LSTM-based method for predicting the corrosion degree of turnout slide plates.

[0041] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the LSTM-based method for predicting the corrosion degree of turnout slide plates.

[0042] Compared with the prior art, the present invention has at least the following beneficial effects:

[0043] This invention provides a method for predicting the corrosion degree of turnout slide plates based on LSTM. Utilizing historical corrosion characteristic parameters of the turnout slide plates, the corrosion area and average corrosion rate are input into a pre-trained LSTM-based model for predicting the corrosion degree of turnout slide plates. This yields the corrosion area and average corrosion rate of the turnout slide plates at different times within a future time period. Then, based on the corrosion area and average corrosion rate at different times within the future time period, the average values ​​of the corrosion area and average corrosion rate for the turnout slide plates within that future time period are calculated. Finally, based on the average values ​​of the corrosion area and average corrosion rate within the future time period, and combined with a preset corrosion evaluation standard, an assessment value for the corrosion degree of the turnout slide plates is obtained. The predicted corrosion area and average corrosion rate parameters for the turnout slide plates have high accuracy and small error, and can adapt to different types of sequential data.

[0044] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the specific embodiments of the present invention, the drawings used in the description of the specific embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0046] Figure 1 This is a flowchart of a method for predicting the corrosion degree of turnout slide plates based on LSTM, according to an embodiment of the present invention.

[0047] Figure 2 In the figure, (a) is a comparison chart of corrosion rate fitting, and (b) is a comparison chart of corrosion area prediction fitting.

[0048] Figure 3 In the figure, (a) represents the difference between the predicted corrosion area and the actual corrosion area, and (b) represents the difference between the predicted corrosion rate and the actual corrosion rate.

[0049] Figure 4 This is an evaluation diagram of the corrosion stage of the slide plate in an embodiment of the present invention. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] This invention provides a method for predicting the corrosion degree of turnout slide plates based on LSTM, which specifically includes the following steps:

[0052] Step 1: Obtain historical corrosion characteristic parameters of the turnout slide plate, including the corrosion area and average corrosion rate of the turnout slide plate.

[0053] As a preferred embodiment, the method for obtaining the corrosion area of ​​the turnout slide plate is as follows:

[0054] a. Acquire images of the turnout slide plate during the corrosion process; for example, use a camera and an electron microscope to acquire image information of the turnout slide plate during the corrosion process.

[0055] b. Perform binarization processing on the image during the corrosion process;

[0056] c. Calculate the proportion of white pixels to all pixels in the binarized image to obtain the corrosion area of ​​the turnout slide plate.

[0057] As a preferred embodiment, the method for obtaining the average corrosion rate of the turnout slide plate is as follows:

[0058] ① Collect the mass of the turnout slide plate during the corrosion process;

[0059] ②Based on the mass during the corrosion process, the average corrosion rate of the turnout slide plate is calculated using the weight gain method.

[0060] Step 2: Input the historical corrosion characteristic parameters into the pre-trained LSTM-based turnout slide plate corrosion degree prediction model, and output the corrosion area and average corrosion rate of the turnout slide plate at different times in the future time period.

[0061] For example, output the corrosion area and average corrosion rate of the turnout slide plate for the following times in the next 200 hours: 2h, 8h, 20h, 32h, 44h, 56h, 68h, 80h, 92h, 104h, 116h, 128h, 140h, 152h, 164h, 176h, 188h, and 200h.

[0062] Preferably, before inputting the historical corrosion characteristic parameters into the pre-trained LSTM-based turnout slide plate corrosion prediction model, the method further includes:

[0063] The historical corrosion characteristic parameters are normalized.

[0064] In other words, the normalized historical corrosion characteristic parameters are then input into the pre-trained LSTM-based turnout slide plate corrosion prediction model.

[0065] In this embodiment, the training method for the LSTM-based turnout slide plate corrosion prediction model is as follows:

[0066] A. Obtain a sample set, which includes several sample data, including the corrosion area and average corrosion rate of the turnout slide plate at different times.

[0067] For example, the sample set includes the corrosion area and average corrosion rate of turnout slide plates after 2 hours, 8 hours, 20 hours, 32 hours, 44 hours, 56 hours, 68 hours, 80 hours, and 92 hours. The corrosion area and average corrosion rate of turnout slide plates at 104h, 116h, 128h, 140h, 152h, 164h, 176h, 188h, and 200h were recorded. The sample set was divided into training and testing datasets, where:

[0068] The training dataset includes the corrosion area and average corrosion rate of 60% of the turnout slide plates after 2 hours, 8 hours, 20 hours, 32 hours, 44 hours, 56 hours, 68 hours, 80 hours, and 92 hours. Corrosion area and average corrosion rate of turnout slide plates at 104h, 116h, 128h, 140h, 152h, 164h, 176h, 188h, and 200h.

[0069] The test dataset includes the corrosion area and average corrosion rate of 40% of the turnout slide plates after 2 hours, 8 hours, 20 hours, 32 hours, 44 hours, 56 hours, 68 hours, 80 hours, and 92 hours. Corrosion area and average corrosion rate of turnout slide plates at 104h, 116h, 128h, 140h, 152h, 164h, 176h, 188h, and 200h.

[0070] B. Use the sample set to train an initial LSTM network model to obtain the LSTM-based prediction model for the corrosion degree of the turnout slide plate;

[0071] For example, an initial LSTM network model is trained using the training dataset. After training, the test dataset is fed into the trained LSTM-based model for predicting the corrosion level of turnout slide plates to obtain the corrosion information prediction results at each time point within 200 hours, and then output them.

[0072] In other words:

[0073] ① Obtain the corresponding corrosion characteristic value parameters and normalize them;

[0074] ② Divide the sample set into a 6:4 ratio, establish a prediction model based on the data characteristics, and initialize the model parameters;

[0075] ③ Train the LSTM network model using the training dataset and continuously update the model parameters using gradient descent until the prediction accuracy requirements are met;

[0076] ④ Input the corrosion feature parameters into the trained prediction model and output the predicted value of the feature parameters at each time point within 200 hours.

[0077] In this embodiment, the input layer of the initial LSTM network model provides a prediction value through an LSTM with 32 units. The model has a batch size of 1, an iteration count of 100, uses MSE as the loss function, Adam as the optimization algorithm, a learning rate of 0.005, and a dataset in CSV format.

[0078] Step 3: Based on the corrosion area and average corrosion rate of the turnout slide plate at different times within the future time period, calculate the average value of the corrosion area and average corrosion rate of the turnout slide plate within the future time period.

[0079] Step 4: Based on the average corrosion area and average corrosion rate of the turnout slide plate within the future time period, and in conjunction with the preset corrosion evaluation criteria, obtain the corrosion degree assessment value of the turnout slide plate.

[0080] Specifically, based on the corrosion degree assessment value of the turnout slide plate, and referring to the corrosion stage evaluation chart of the slide plate (such as...), Figure 4 As shown), the degree of corrosion of the turnout slide plate can be predicted. The corrosion stage evaluation diagram includes:

[0081] During the first stage of corrosion, the turnout slide plate can function normally.

[0082] In the second stage of corrosion, the corrosion of the turnout slide plate deepens, requiring increased inspection efforts.

[0083] In the third stage of corrosion, the turnout slide plates show severe corrosion, requiring timely repair or replacement.

[0084] In one embodiment, the corrosion degree assessment value of the turnout slide plate is obtained based on the average value of the corrosion area and average corrosion rate of the turnout slide plate within the future time period, combined with a preset corrosion evaluation standard, as follows:

[0085] First, the average corrosion area of ​​the turnout slide plate within the future time period is compared with the preset corrosion area rating rule for the turnout slide plate to obtain the corrosion area rating of the turnout slide plate within the future time period.

[0086] Secondly, the average corrosion rate of the turnout slide plate within the future time period is compared with the preset corrosion resistance rating rule for the turnout slide plate to obtain the corrosion resistance level of the turnout slide plate within the future time period.

[0087] Finally, based on the corrosion area level and corrosion resistance level of the turnout slide plate in the future time period, the corrosion degree assessment value of the turnout slide plate is calculated.

[0088] Preferably, the corrosion area rating rule for the turnout slide plate is as follows:

[0089] Different corrosion area ranges correspond to different corrosion area grades. More specifically, please refer to Table 1:

[0090] Table 1 Corrosion Area Rating of Turnout Slide Plates (Rule Level)

[0091]

[0092] Preferably, as shown in Table 2, the corrosion resistance rating rules for the turnout slide plate include:

[0093] Different average corrosion rate ranges correspond to different corrosion resistance grades. For more specific details, please refer to Table 2:

[0094] Table 2 Corrosion Resistance Rating Rules for Turnout Slide Plates

[0095]

[0096] In a preferred embodiment, the corrosion degree assessment value of the turnout slide plate is calculated based on the corrosion area level and corrosion resistance level of the turnout slide plate within the future time period, as follows:

[0097] First, make the following settings:

[0098] Set an upper limit for the evaluation value; for example, the upper limit for the evaluation value is set to 100.

[0099] Different corrosion area levels correspond to different scoring values. For example, the corrosion area levels of turnout slide plates are divided into five levels: 0, 1-3, 4-6, 5-9, and 10. The score is 0 for level 10, 10 for level 5-9, 20 for level 4-6, -30 for level 1-3, and 50 for level 0.

[0100] Different corrosion resistance levels correspond to different scoring values. For example, the corrosion resistance levels are divided into six levels: 1, 2-3, 4-5, 6-7, 8-9, and 10. The score for reaching level 1 is -20 points, the score for reaching level 2-3 is 10 points, the score for reaching level 4-5 is 10 points, the score for reaching level 6-9 is 0 points, and the score for reaching level 10 is 30 points.

[0101] Finally, by subtracting the score corresponding to the corrosion area level from the upper limit of the evaluation value, and then subtracting the score corresponding to the corrosion resistance level, the corrosion degree evaluation value of the turnout slide plate is obtained.

[0102] The following detailed explanation of specific embodiments will provide a more detailed description of the present invention.

[0103] Example

[0104] To demonstrate the effectiveness of this method, the corrosion of a turnout slide plate was simulated using Q235 steel as the substrate in a neutral salt spray environment. The substrate material dimensions were 100mm × 100mm × 10mm. The cladding layer material consisted of Cu powder and Ni powder mixed in a 9:1 mass ratio. The Cu powder used was spherical with a smooth surface, a particle size of 38μm, and a purity of 99%. The Ni powder used was also spherical with a particle size of 45μm and a purity of 99%. After completing the chemical plating of graphite powder, the nickel-plated graphite is made into copper-based graphite self-lubricating material through powder metallurgy. Then, sodium carboxymethyl cellulose binder is used to uniformly mix the cladding powder (copper-based graphite composite self-lubricating material) into a paste and applied to the surface of the Q235 steel plate substrate. Laser cladding is performed with parameters of 1000W laser power, 200mm / s scanning speed, 2.5mm spot diameter and 50% overlap. Then, a 5% sodium chloride solution is continuously sprayed in the form of a spray tower, so that the salt spray settles onto six prepared test specimens (100mm×100mm×10mm).

[0105] After the corrosion test, images of the Q235 substrate during the corrosion process were acquired using a camera and electron microscope. After binarization of the images, the proportion of white pixels to the total number of pixels in the binary image was calculated to determine the corrosion area. The average corrosion rate was calculated using the weighting method, yielding a corrosion area and average corrosion rate of 0.25 mm for the Q235 substrate. 2 and 0.2g / m 2 ·h.

[0106] After normalization, the corrosion rate is input into a pre-trained LSTM-based model for predicting the corrosion degree of turnout slide plates. The LSTM-based model outputs the corrosion area and average corrosion rate at various times (2h, 8h, 20h, 32h, 44h, 56h, 68h, 80h, 92h, 104h, 116h, 128h, 140h, 152h, 164h, 176h, 188h, and 200h), respectively. ), 44h(0.26,0.26), 56h(0.26,0.28), 68h(0.29,0.30), 80h(0.29,0.32), 92h(0.30,0.36), 104h(0.32,0.36), 116h(0.33,0.36), 1 28h(0.33,0.38), 140h(0.34,0.40), 152h(0.34,0.40), 164h(0.35,0.40), 176h(0.36,0.40), 188h(0.36,0.41), 200h(0.36,0.42)

[0107] The average corrosion area and average corrosion value of the Q235 steel substrate at 18 time points were calculated to be 0.306 mm. 2 and 0.326g / m 2 ·h.

[0108] Based on the corrosion area rating rules and corrosion resistance rating rules for turnout slide plates, as well as the different scoring values ​​corresponding to different corrosion areas and corrosion resistance levels, the corrosion level of the base material Q235 steel is obtained. The corrosion area corresponds to level 4, the average corrosion rate corresponds to level 5, and the calculated evaluation score is 70 points.

[0109] Comparison Figure 4 The corrosion stage assessment diagram of the slide plate shows that the selected turnout slide plate is in the first stage of corrosion, indicating that the slide plate can work normally.

[0110] Comparative Example

[0111] The following comparative experiments on different corrosion prediction methods are used to verify the method of the present invention.

[0112] Step 1: First, a gray-scale prediction model is used to predict the corrosion area and average corrosion rate of the Q235 steel substrate. Based on the sample dataset of collected feature parameters, the original data of the Q235 steel substrate is input into the model for first-order accumulation processing. Then, gray parameters are calculated to establish differential equations, and the equations are solved to calculate the prediction model of the first-order accumulation sequence. Finally, cumulative subtraction and restoration processing is performed to calculate the prediction model of the original data sequence. The corrosion status of the Q235 steel substrate is predicted and analyzed using the gray-scale prediction model, as shown in Table 2. It can be seen that the gray-scale prediction model results fit the actual values ​​relatively well, but some values ​​have large errors, and the prediction results are unstable. This situation may be because the gray-scale prediction model itself does not have self-learning ability, and the gray-scale model has a better prediction effect on increasing trends, while the corrosion process of the turnout slide plate is affected by multiple factors, and the trend is not stable.

[0113] Table 2: Verification Table for Gray Scale Prediction Model

[0114]

[0115] Step 2: Normalize the corrosion area and average corrosion rate of the obtained Q235 steel substrate, divide the dataset into a 6:4 ratio, establish a prediction model based on the data characteristics, and initialize the model parameters; train the slide prediction model using the training set, continuously update the model parameters using gradient descent until the prediction accuracy requirements are met, input the feature parameters of the four types of slides into the trained prediction model, and output the corresponding prediction values; finally, evaluate the model prediction effect, as shown in Table 3.

[0116] Table 3: Evaluation Results of LSTM Model

[0117]

[0118] Step 3: To better compare the results of the grey prediction model and the LSTM prediction network, eight data feature points from the actual corrosion test of the Q235 steel substrate were selected and compared with the predicted data of the two models. To more intuitively show the superiority of the two models in predicting different corrosion characteristic parameters of the slide plate, the predicted values ​​of corrosion area and average corrosion rate of the two models were compared with the actual values ​​of each parameter using comparative graph analysis, such as... Figure 2 As shown in (a) (Note: The vertical axis from top to bottom represents: actual predicted value, LSTM predicted value, and gray predicted value), and (b) (Note: The vertical axis from top to bottom represents: gray predicted value, actual predicted value, and LSTM predicted value), the results show that the gray prediction has a better fitting effect. However, compared with the LSTM prediction results, the LSTM network's predicted value is closer to the actual value, and the trend of change is also closer. This may be because LSTM can update the parameters of the prediction model based on the data, thus obtaining a more ideal prediction result. Figure 3(Note: The left side of the horizontal axis at each time point represents the LSTM prediction value, and the right side represents the gray prediction value.) The results show that all differences in the gray predictions are greater than the maximum difference in the LSTM predictions. Therefore, LSTM is more suitable for predicting corrosion of turnout slide plates.

[0119] Based on the above examples and comparative examples, it can be concluded that the LSTM-based corrosion prediction model for turnout slide plates can accurately predict the corrosion area and average corrosion rate of turnout slide plates within a 200-hour time period. Furthermore, by using the corrosion area rating rules, corrosion resistance rating rules, and different score values ​​corresponding to different corrosion areas and corrosion resistance levels, the corrosion degree of turnout slide plates can be accurately determined.

[0120] A LSTM-based device for predicting the corrosion level of turnout slide plates includes:

[0121] The acquisition module is used to acquire historical corrosion characteristic parameters of the turnout slide plate, including the corrosion area and average corrosion rate of the turnout slide plate.

[0122] The prediction module is used to input the historical corrosion feature parameters into a pre-trained LSTM-based corrosion prediction model for turnout slide plates, and output the corrosion area and average corrosion rate of the turnout slide plates at different times in the future time period.

[0123] The calculation module is used to calculate the average value of the corrosion area and average corrosion rate of the turnout slide plate within the future time period, based on the corrosion area and average corrosion rate of the turnout slide plate at different times within the future time period.

[0124] The evaluation module is used to obtain an evaluation value of the corrosion degree of the turnout slide plate based on the average value of the corrosion area and average corrosion rate of the turnout slide plate within the future time period, and in combination with a preset corrosion evaluation standard.

[0125] In one embodiment of the present invention, a computer device is provided, comprising a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions to achieve a corresponding method flow or corresponding function. The processor described in this embodiment of the present invention can be used to implement an LSTM-based method for predicting the corrosion degree of turnout slide plates.

[0126] In one embodiment of the present invention, a method for predicting the corrosion degree of a turnout slide plate based on LSTM, if implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable storage medium includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data.

[0127] The computer storage medium can be any available medium or data storage device that a computer can access, including but not limited to magnetic storage (e.g., floppy disk, hard disk, magnetic tape, magneto-optical disk (MO)), optical storage (e.g., CD, DVD, BD, HVD), and semiconductor storage (e.g., ROM, EPROM, EEPROM, non-volatile memory (NAND FLASH), solid-state drive (SSD)).

[0128] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0129] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0130] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0131] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0132] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for predicting the corrosion degree of turnout slide plates based on LSTM, characterized in that, include: Obtain historical corrosion characteristic parameters of the turnout slide plate, including the corrosion area and average corrosion rate of the turnout slide plate; The method for obtaining the corrosion area of ​​the turnout slide plate includes: acquiring an image of the turnout slide plate during the corrosion process; performing binarization processing on the image during the corrosion process; calculating the proportion of white pixels in the binarized image to the total number of pixels in the image, thereby obtaining the corrosion area of ​​the turnout slide plate. The method for obtaining the average corrosion rate of the turnout slide plate includes: collecting the mass of the turnout slide plate during the corrosion process; and calculating the average corrosion rate of the turnout slide plate using the weight gain method based on the mass during the corrosion process. The historical corrosion characteristic parameters are input into a pre-trained LSTM-based model for predicting the corrosion degree of turnout slide plates, and the corrosion area and average corrosion rate of the turnout slide plates at different times in the future time period are output. Based on the corrosion area and average corrosion rate of the turnout slide plate at different times within the future time period, the average values ​​of the corrosion area and average corrosion rate of the turnout slide plate within the future time period are calculated respectively. The corrosion degree assessment value of the turnout slide plate is obtained based on the average corrosion area and average corrosion rate of the turnout slide plate within the future time period, combined with the preset corrosion evaluation criteria, including: The average corrosion area of ​​the turnout slide plate within the future time period is compared with the preset corrosion area rating rule for the turnout slide plate to obtain the corrosion area rating of the turnout slide plate within the future time period; the corrosion area rating rule for the turnout slide plate includes: different corrosion area ranges correspond to different corrosion area ratings. The average corrosion rate of the turnout slide plate within the future time period is compared with the preset corrosion resistance rating rule for the turnout slide plate to obtain the corrosion resistance level of the turnout slide plate within the future time period; the corrosion resistance rating rule for the turnout slide plate includes: different average corrosion rate ranges correspond to different corrosion resistance levels. Based on the corrosion area level and corrosion resistance level of the turnout slide plate within the future time period, the corrosion degree assessment value of the turnout slide plate is calculated, including: Set an upper limit for the evaluation value; Different corrosion area levels correspond to different scoring values; Different ranges of corrosion resistance grades correspond to different rating values; The corrosion degree assessment value of the turnout slide plate is obtained by subtracting the score value corresponding to the corrosion area level from the upper limit of the assessment value, and then subtracting the score value corresponding to the corrosion resistance level.

2. The method for predicting the corrosion degree of turnout slide plates based on LSTM according to claim 1, characterized in that, The training method for the LSTM-based turnout slide plate corrosion prediction model includes: Obtain a sample set, which includes several sample data, including the corrosion area and average corrosion rate of the turnout slide plate at different times; The initial LSTM network model is trained using the sample set to obtain the LSTM-based prediction model for the corrosion degree of the turnout slide plate.

3. A device for predicting the corrosion degree of turnout slide plates based on LSTM, characterized in that, A method for predicting the corrosion degree of a turnout slide plate based on LSTM as described in claim 1 or 2 includes: The acquisition module is used to acquire historical corrosion characteristic parameters of the turnout slide plate, including the corrosion area and average corrosion rate of the turnout slide plate. The prediction module is used to input the historical corrosion feature parameters into a pre-trained LSTM-based corrosion prediction model for turnout slide plates, and output the corrosion area and average corrosion rate of the turnout slide plates at different times in the future time period. The calculation module is used to calculate the average value of the corrosion area and average corrosion rate of the turnout slide plate within the future time period, based on the corrosion area and average corrosion rate of the turnout slide plate at different times within the future time period. The evaluation module is used to obtain an evaluation value of the corrosion degree of the turnout slide plate based on the average value of the corrosion area and average corrosion rate of the turnout slide plate within the future time period, and in combination with a preset corrosion evaluation standard.

4. An apparatus 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, it implements the steps of the LSTM-based method for predicting the corrosion degree of turnout slide plates as described in claim 1 or 2.

5. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the LSTM-based method for predicting the corrosion degree of turnout slide plates as described in claim 1 or 2.

Images