Transmission high temperature warning method, training method, device and storage medium

Abstract

This application relates to the field of automotive safety testing and provides a method, training method, device, and storage medium for high-temperature warning of transmissions. The warning method includes: acquiring vehicle network data within a set time period; based on the vehicle network data, performing temperature prediction using a pre-trained bidirectional recurrent neural network model to obtain a temperature prediction result for the transmission after a target time; and issuing a high-temperature warning for the transmission when the temperature prediction result indicates that the probability of high temperature in the transmission exceeds a threshold. This solution avoids irreversible high-temperature damage to the transmission.

Description

Technical Field

[0001] This application belongs to the field of automotive safety testing, and in particular relates to a method, training method, device and storage medium for high temperature warning of transmission. Background Technology

[0002] The automotive transmission is a crucial component of the vehicle's drivetrain. Its primary function is to alter the engine torque and drive wheel speed range to adapt to various complex driving conditions, such as starting, acceleration, and uphill driving, while also ensuring the engine operates under optimal conditions. Therefore, the transmission's performance significantly impacts the overall performance of the vehicle's drivetrain. Research and practical experience have revealed that low transmission efficiency or component damage is not solely related to the structure of individual components but also to the internal heat generation during operation. Due to the compact structure, high torque transmission capacity, and complex operating conditions of automotive transmissions, excessively high internal temperatures can severely affect the performance of the lubricating oil and the condition of components, ultimately impacting the transmission's lifespan and transmission efficiency. Therefore, the transmission's thermal equilibrium state has a significant impact on its performance.

[0003] Currently, transmission fluid temperature is mainly controlled by monitoring real-time fluid temperature using temperature sensors. When the fluid temperature exceeds a warning threshold, an alarm is triggered to alert the user. However, managing transmission fluid temperature through a monitoring-feedback-processing workflow is inherently delayed. By the time the sensor detects a high fluid temperature, the effects on the transmission have already occurred and are irreversible. In other words, existing temperature sensor monitoring methods cannot predict in advance whether the transmission will experience overheating during driving, leading to irreversible damage and reduced transmission lifespan.

[0004] Therefore, a more effective method for high-temperature warning of transmissions is needed. Summary of the Invention

[0005] This application provides a method, training method, device, and storage medium for high-temperature warning of transmissions, in order to solve the problem that temperature sensor monitoring methods cannot prevent transmission damage before high-temperature hazards occur, leading to transmission damage and reduced transmission life.

[0006] The first aspect of this application provides a training method for a high-temperature warning system for a transmission, comprising:

[0007] A first data segment extraction operation is performed on historical vehicle network time-series data to obtain a first dataset and a third dataset; wherein the time period corresponding to the first dataset is earlier than the time period corresponding to the third dataset.

[0008] The first dataset is labeled based on the third dataset to obtain a first labeling result. The first labeling result is used to indicate whether the labeled data in the first dataset corresponds to the gearbox high temperature warning signal.

[0009] A second data segmentation operation is performed on the historical vehicle network time-series data to obtain a second dataset and a fourth dataset; wherein the time period corresponding to the second dataset is later than the time period corresponding to the fourth dataset.

[0010] The second dataset is labeled based on the fourth dataset to obtain a second labeling result. The second labeling result is used to indicate whether the labeled data in the second dataset corresponds to the gearbox high temperature warning signal.

[0011] The first dataset with the first annotation result and the second dataset with the second annotation result are used as historical vehicle network data samples to train the bidirectional recurrent neural network model until the bidirectional recurrent neural network model converges.

[0012] A second aspect of this application provides a method for early warning of high temperature in a transmission, including:

[0013] Acquire vehicle network data within a specified time period;

[0014] Based on the vehicle network data, a pre-trained bidirectional recurrent neural network model is used to predict the temperature of the transmission after a target time. The target time is the time after the set time period. The bidirectional recurrent neural network model is trained based on historical vehicle network data samples, which include a first dataset and a second dataset. The first annotation result of the first dataset is based on the annotation of a third dataset. The first annotation result indicates whether the annotated data in the first dataset corresponds to a transmission high-temperature warning signal. The time period corresponding to the first dataset is earlier than the time period corresponding to the third dataset. The second annotation result of the second dataset is based on the annotation of a fourth dataset. The second annotation result indicates whether the annotated data in the second dataset corresponds to a transmission high-temperature warning signal. The time period corresponding to the second dataset is later than the time period corresponding to the fourth dataset.

[0015] If the temperature prediction result indicates that the probability of the transmission being overheated exceeds a threshold, a high temperature warning signal for the target transmission is output.

[0016] A third aspect of this application provides a training device for high-temperature warning of a transmission, comprising:

[0017] The first acquisition module is used to perform a first data segment extraction operation on historical vehicle network time-series data to obtain a first dataset and a third dataset; wherein the time period corresponding to the first dataset is earlier than the time period corresponding to the third dataset.

[0018] The first annotation module is used to annotate the first dataset based on the third dataset to obtain a first annotation result. The first annotation result is used to indicate whether the annotated data in the first dataset corresponds to the gearbox high temperature warning signal in the third dataset.

[0019] The second acquisition module is used to perform a second data segmentation operation on the historical vehicle network time-series data to obtain a second dataset and a fourth dataset; wherein the time period corresponding to the second dataset is later than the time period corresponding to the fourth dataset.

[0020] The second annotation module is used to annotate the second dataset based on the fourth dataset to obtain a second annotation result. The second annotation result is used to indicate whether the annotated data in the second dataset corresponds to the gearbox high temperature warning signal in the fourth dataset.

[0021] The model training module is used to train the bidirectional recurrent neural network model using the first dataset corresponding to the first annotation result and the second dataset corresponding to the second annotation result as historical vehicle network data samples until the bidirectional recurrent neural network model converges.

[0022] A fourth aspect of this application provides a high-temperature warning device for a transmission, comprising:

[0023] The acquisition module is used to acquire vehicle network data within a set time period;

[0024] The prediction module is used to predict the temperature of the transmission after a target time by using a pre-trained bidirectional recurrent neural network model based on the vehicle network data. The target time is the time after the set time period. The bidirectional recurrent neural network model is trained based on historical vehicle network data samples, which include a first dataset and a second dataset. The first annotation result of the first dataset is obtained based on the annotation of a third dataset. The first annotation result indicates whether the annotated data in the first dataset corresponds to a transmission high-temperature warning signal. The time period corresponding to the first dataset is earlier than the time period corresponding to the third dataset. The second annotation result of the second dataset is obtained based on the annotation of a fourth dataset. The second annotation result indicates whether the annotated data in the second dataset corresponds to a transmission high-temperature warning signal. The time period corresponding to the second dataset is later than the time period corresponding to the fourth dataset.

[0025] The early warning module is used to output a target transmission high temperature early warning signal when the temperature prediction result indicates that the probability of the transmission being overheated exceeds a threshold.

[0026] A fifth aspect of this application provides a terminal including 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 method as described in the first or second aspect.

[0027] A sixth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method as described in the first or second aspect.

[0028] The fifth aspect of this application provides a computer program product that, when run on a terminal, causes the terminal to perform the steps of the method described in the first aspect.

[0029] As can be seen from the above, this application obtains vehicle network data within a set time period, uses a pre-trained bidirectional recurrent neural network model to predict the temperature of the transmission after a target time, and issues an effective high-temperature warning when the temperature is too high. The pre-trained model is a bidirectional recurrent neural network model trained on a first dataset labeled with data from an earlier time period and a second dataset labeled with data from a later time period. Based on the pre-trained bidirectional recurrent neural network model, this process uses existing vehicle network data to predict the transmission temperature data at future times, ensuring the effectiveness and accuracy of data prediction, thereby achieving effective and accurate prevention in advance when the transmission has not yet suffered high-temperature damage, and avoiding irreversible damage to the transmission due to high temperature. Attached Figure Description

[0030] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 This is a flowchart of a training method for a high-temperature warning system for a transmission provided in an embodiment of this application;

[0032] Figure 2 This is a schematic diagram illustrating data processing using an exemplary model structure based on a bidirectional recurrent neural network model provided in this application embodiment;

[0033] Figure 3 This is a data partitioning diagram of a training method for high-temperature warning of a transmission provided in an embodiment of this application;

[0034] Figure 4 This is a flowchart of a high-temperature warning method for a transmission provided in an embodiment of this application;

[0035] Figure 5 This is a structural diagram of a training device for high-temperature warning of a gearbox provided in an embodiment of this application;

[0036] Figure 6 This is a structural diagram of a high-temperature warning device for a gearbox provided in an embodiment of this application;

[0037] Figure 7 This is a structural diagram of a terminal provided in an embodiment of this application. Detailed Implementation

[0038] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0039] It should be understood that, when used in this specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0040] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of the application. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0041] It should also be further understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0042] As used in this specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrases "if determined" or "if [described condition or event] is detected" may be interpreted, depending on the context, as "once determined," "in response to determination," "once [described condition or event] is detected," or "in response to detection of [described condition or event]."

[0043] In specific implementations, the terminals described in the embodiments of this application include, but are not limited to, other portable devices such as mobile phones, laptop computers, or tablet computers with touch-sensitive surfaces (e.g., touchscreen displays and / or touchpads). It should also be understood that in some embodiments, the device is not a portable communication device, but a desktop computer with touch-sensitive surfaces (e.g., touchscreen displays and / or touchpads).

[0044] The following discussion describes terminals that include displays and touch-sensitive surfaces. However, it should be understood that terminals may include one or more other physical user interface devices such as physical keyboards, mice, and / or joysticks.

[0045] The terminal supports a variety of applications, such as one or more of the following: drawing applications, presentation applications, word processing applications, website creation applications, disc burning applications, spreadsheet applications, game applications, telephone applications, video conferencing applications, email applications, instant messaging applications, exercise support applications, photo management applications, digital camera applications, digital camcorder applications, web browsing applications, digital music player applications, and / or digital video player applications.

[0046] Various applications that can run on a terminal can use at least one common physical user interface device, such as a touch-sensitive surface. One or more functions of the touch-sensitive surface and the corresponding information displayed on the terminal can be adjusted and / or changed between and / or within applications. In this way, the terminal's common physical architecture (e.g., the touch-sensitive surface) can support various applications with user interfaces that are intuitive and transparent to the user.

[0047] It should be understood that the sequence number of each step in this embodiment does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of this application embodiment.

[0048] This application provides a technical solution for a high-temperature warning method for transmissions to address the problem that existing methods of detecting transmission temperature using temperature sensors cannot prevent damage to the transmission before it suffers high temperatures, thus reducing its lifespan.

[0049] To illustrate the technical solution described in this application, specific embodiments are provided below.

[0050] See Figure 1 , Figure 1 This is a flowchart illustrating a training method for a high-temperature warning system for a transmission, as provided in an embodiment of this application. Figure 1 As shown, a training method for a high-temperature warning system for a transmission includes the following steps:

[0051] Step 101: Perform a first data segment extraction operation on the historical vehicle network time series data to obtain the first dataset and the third dataset.

[0052] In this application embodiment, "high temperature" refers to a temperature whose value is higher than a set threshold.

[0053] The time period corresponding to the first dataset is earlier than the time period corresponding to the third dataset.

[0054] Historical vehicle-to-everything (V2X) time-series data consists of historical V2X data, which are related in chronological order.

[0055] The first data segment extraction operation specifically involves extracting two datasets corresponding to each data segment from the historical vehicle network time-series data in chronological order. The time period corresponding to the first dataset extracted in chronological order is before the time period corresponding to the second dataset.

[0056] Specifically, when extracting historical vehicle-to-everything (V2X) time-series data, data extraction can be performed based on time periods to obtain datasets for different time periods. Each dataset contains V2X time-series data for the corresponding time period.

[0057] The time period can be divided into two segments: one segment is defined by dividing a set duration forward and backward from a set historical time, and the other segment is defined by dividing a set duration forward from a set historical time.

[0058] Specifically, in this step, the first dataset corresponds to a time period, and the third dataset corresponds to another time period. The first time period can be obtained by dividing the time period forward and backward based on a set historical time point, and extracting the data from that time period. Then, the third dataset can be obtained by dividing the time period forward and backward based on a set historical time point after the first time period, and extracting the data from that time period.

[0059] Alternatively, based on a set historical moment, divide the time period forward by a set duration as one of the aforementioned time periods, extract the data from this time period to obtain the first dataset, and then based on a historical moment after the set historical moment, divide the time period forward by a set duration as another of the aforementioned time periods, extract the data from this time period to obtain the third dataset.

[0060] Alternatively, based on a set historical moment, divide the time period forward by a set duration as one of the aforementioned time periods, extract the data from this time period to obtain the first dataset, and then based on a historical moment after this time period, divide the time period forward by a set duration as another of the aforementioned time periods, extract the data from this time period to obtain the third dataset.

[0061] In one example, the interval between two time periods is set to a certain duration.

[0062] In this embodiment, the bidirectional recurrent neural network model is trained based on historical vehicle network data samples.

[0063] Among them, combined Figure 2 As shown in one example, during the training phase, the bidirectional recurrent neural network model may include a forward BiLSTM (Bi-directional Long Short-Term Memory) module, a reverse BiLSTM module, a linear fusion module, a fully connected layer, and a softmax (normalized exponential function) layer.

[0064] In this embodiment, the bidirectional recurrent neural network model needs to be trained first. During the model training phase, training samples can be constructed first. Among them, the historical vehicle network data samples used as training samples include a first dataset and a second dataset.

[0065] The first dataset is derived from the annotation of the third dataset, and the first and third datasets are in chronological order. This chronological order means that the time period corresponding to the data in the first dataset is earlier than the time period corresponding to the data in the third dataset.

[0066] When extracting historical vehicle network time-series data, two datasets can be extracted in chronological order and two datasets can be extracted in reverse chronological order.

[0067] The process involves extracting two datasets in chronological order. The time period corresponding to one dataset is after the time period corresponding to the other dataset. The dataset with the later time period needs to be used to label the dataset with the earlier time period.

[0068] Correspondingly, it is also necessary to extract two datasets in reverse chronological order, with the time period corresponding to one dataset preceding the time period corresponding to the other dataset. The data in the dataset with the earlier time needs to be used to label the data in the dataset with the later time.

[0069] In this step, historical vehicle-to-everything (V2X) time-series data is extracted into datasets, and data journeys are divided. Optionally, each extracted dataset may contain V2X data of a fixed time period.

[0070] Step 102: Label the first dataset based on the third dataset to obtain the first labeling result.

[0071] The first annotation result is used to indicate whether the annotated data in the first dataset corresponds to the high temperature warning signal of the transmission.

[0072] The first annotation result is, for example, a data label.

[0073] Taking the first dataset as an example, the data annotation involves using the data in the third dataset as a reference to label the data in the first dataset with true or false values ​​to indicate whether it corresponds to a high temperature warning signal for the transmission in the future. The data labels are exemplified as true or false, 0 or 1, etc.

[0074] In one example, the time periods corresponding to the first and third datasets are the same length.

[0075] Typically, vehicle-to-everything (V2X) data in a connected vehicle system is generated by various devices within the vehicle uploading data at set time intervals. Therefore, when time periods are of equal length, the number of time points within each time period is the same, and each time point in the dataset corresponds to a specific V2X data sample. Thus, a correspondence between the data in two datasets within two time periods of equal length can be established based on the temporal sequence of their time points.

[0076] In this case, correspondingly, the first dataset is labeled based on the third dataset, including:

[0077] The data in the first dataset and the third dataset are compared sequentially according to time.

[0078] If the third dataset contains a transmission high temperature warning signal, then the data in the first dataset whose time sequence corresponds to the transmission high temperature warning signal is marked as true;

[0079] If the third dataset does not contain a high-temperature warning signal for the transmission, then the data in the corresponding time series in the first dataset will be marked as false.

[0080] The algorithm iterates through the data in both datasets chronologically. If a transmission oil temperature warning signal appears in the reference dataset, the corresponding data in the dataset to be labeled as true. If no transmission oil temperature warning signal appears in the reference dataset, the corresponding data in the dataset to be labeled as false. Furthermore, if a transmission oil temperature warning signal appears in the dataset to be labeled, that data entry is deleted.

[0081] Step 103: Perform a second data segment extraction operation on the historical vehicle network time series data to obtain the second dataset and the fourth dataset.

[0082] The time period corresponding to the second dataset is later than the time period corresponding to the fourth dataset.

[0083] The second data segment extraction operation specifically involves extracting two data segments from the historical vehicle network time-series data in both forward and reverse order. The time period corresponding to the first data segment extracted in reverse order is after the time period corresponding to the second data segment.

[0084] Specifically, when extracting historical vehicle-to-everything (V2X) time-series data, data extraction can be performed based on time periods to obtain datasets for different time periods. Each dataset contains V2X time-series data for the corresponding time period.

[0085] The time period can be divided into two segments: one segment is defined by dividing a set duration forward and backward from a set historical time, and the other segment is defined by dividing a set duration forward from a set historical time.

[0086] Specifically, in this step, the second dataset corresponds to one time period, and the fourth dataset corresponds to another time period. The first time period can be divided forward and backward by a set historical time as a reference, and the data of the first time period can be extracted to obtain the second dataset. The second time period can be divided forward and backward by a set historical time before the first time period as a reference, and the data of the second time period can be extracted to obtain the fourth dataset.

[0087] Alternatively, based on a set historical moment, divide the time period forward by a set duration as one of the aforementioned time periods, extract the data from this time period to obtain the second dataset, and then based on a historical moment before this time period, divide the time period forward by a set duration as another of the aforementioned time periods, extract the data from this time period to obtain the fourth dataset.

[0088] Alternatively, based on a set historical moment, divide the time period forward as a set duration, extract the data from this time period to obtain the second dataset, and then based on a historical moment before this time period, divide the time period forward as another set duration, extract the data from this time period to obtain the fourth dataset.

[0089] In one example, the interval between two time periods is set to a certain duration.

[0090] The second dataset was obtained by labeling the fourth dataset, and the second and fourth datasets are in reverse chronological order. Inverse chronological order means that the time period corresponding to the data in the second dataset is later than the time period corresponding to the data in the fourth dataset.

[0091] In this way, the data in the first dataset is labeled with reference to historical data that occurred after it, and the data in the second dataset is labeled with reference to historical data that occurred before it.

[0092] In the process of taking the above data screenshots, specifically in one example, such as Figure 3 As shown, the current time is defined as T, and the data within the time interval p from time T onwards is the observation period data. The data from time T+p+d to time T+p+d+q is the performance period data, where d is the interval time. Thus, data from time T to time T+p can be extracted sequentially as the first dataset (i.e., the observation period data), and data from time T+p+d to time T+p+d+q can be extracted as the third dataset (i.e., the performance period data). The observation period data is then labeled based on the performance period data.

[0093] Similarly, data from time Tp to time T can be extracted in reverse chronological order to form the second dataset, and data from time Tpdq to time Tpd can be extracted to form the fourth dataset. The data in the second dataset can then be labeled based on the data in the fourth dataset.

[0094] In one example, a first data segment extraction operation is performed on historical vehicle-to-everything (V2X) time-series data to obtain a first dataset and a third dataset, including:

[0095] The first dataset is obtained by extracting a data segment from the first time period from the historical vehicle network time series data, and the third dataset is obtained by extracting a data segment from the second time period.

[0096] Correspondingly, a second data segment extraction operation is performed on the historical vehicle-to-everything (V2X) time-series data to obtain the second dataset and the fourth dataset, including:

[0097] The second dataset is obtained by extracting the data segment of the second time period from the historical vehicle network time series data, and the fourth dataset is obtained by extracting the data segment of the first time period.

[0098] The second time period is after the first time period and has a set time interval between them.

[0099] This process involves extracting data from the first and second time periods of historical vehicle-to-everything (V2X) time-series data. When different time periods are used for extraction, the data in the first time period forms either sample data or reference data for labeling the sample data.

[0100] It facilitates the extraction and use of data from historical vehicle network time-series data, and is also able to learn the time-series mapping relationship features between dataset samples with corresponding time periods when training the model bidirectionally, thereby improving the model training effect.

[0101] Step 104: Label the second dataset based on the fourth dataset to obtain the second labeling result.

[0102] The second annotation result is used to indicate whether the annotated data in the second dataset corresponds to the gearbox high temperature warning signal.

[0103] The second annotation result is, for example, a data label.

[0104] Taking the second dataset as an example, the data annotation involves using the data in the fourth dataset as a reference to label the data in the second dataset with true or false values ​​to indicate whether it corresponded to the high temperature warning signal of the gearbox in the past. The data labels are exemplified as true or false, 0 or 1, etc.

[0105] In one example, the second and fourth datasets each correspond to the same time period.

[0106] Typically, vehicle-to-everything (V2X) data in a connected vehicle system is generated by various devices within the vehicle uploading data at set time intervals. Therefore, when time periods are of equal length, the number of time points within each time period is the same, and each time point in the dataset corresponds to a specific V2X data sample. Thus, a correspondence between the data in two datasets within two time periods of equal length can be established based on the temporal sequence of their time points.

[0107] In this case, correspondingly, the second dataset is labeled based on the fourth dataset, including:

[0108] The data in the second dataset and the fourth dataset are compared sequentially according to time.

[0109] If the fourth dataset contains a transmission high temperature warning signal, then the data in the second dataset whose time sequence corresponds to the transmission high temperature warning signal is marked as true;

[0110] If the fourth dataset does not contain a high-temperature warning signal for the transmission, then the data in the corresponding time series in the second dataset will be marked as false.

[0111] During this process, the data annotation information in the first and second datasets is used to assist in supervising model training, thereby improving the model convergence speed and prediction accuracy.

[0112] Finally, we obtain historical vehicle network data samples containing the labeled first and second datasets, which are the model training samples.

[0113] Once the model training samples are constructed, the model needs to be trained based on these samples.

[0114] The algorithm iterates through the data in both datasets chronologically. If a transmission oil temperature warning signal appears in the reference dataset, the corresponding data in the dataset to be labeled as true. If no transmission oil temperature warning signal appears in the reference dataset, the corresponding data in the dataset to be labeled as false. Furthermore, if a transmission oil temperature warning signal appears in the dataset to be labeled, that data entry is deleted.

[0115] In this process, the annotation results corresponding to the first and second datasets are used to assist in the implementation of supervised model training, thereby improving the model convergence speed and prediction accuracy.

[0116] Step 105: Use the first dataset with the first annotation result and the second dataset with the second annotation result as historical vehicle network data samples to train the bidirectional recurrent neural network model until the bidirectional recurrent neural network model converges.

[0117] The first dataset with the first annotation result and the second dataset with the second annotation result, obtained by the above annotation method, are used as data training samples to implement model training. This enables the trained samples to learn the mapping relationship between the vehicle network data in the previous period and whether the transmission high temperature warning occurs in the later period through bidirectional loop learning, thus ensuring the effectiveness and accuracy of data prediction.

[0118] Finally, we obtain historical vehicle network data samples, which include a first dataset with the first annotation results and a second dataset with the second annotation results, i.e., model training samples.

[0119] Once the model training samples are constructed, the model needs to be trained based on these samples.

[0120] Specifically, taking a bidirectional recurrent neural network, which includes a forward BiLSTM module, a backward BiLSTM module, a linear fusion layer, a fully connected layer, and a softmax layer, as an example.

[0121] The forward and reverse BiLSTM modules share the same network structure and operating principle. The difference lies in the input data: the forward BiLSTM module receives a sequential sequence of vehicle network data, while the reverse BiLSTM module receives a reverse sequence. The output of the reverse BiLSTM module is then reversed again to produce an intermediate feature vector. This intermediate feature vector is then linearly fused with the feature vector from the forward BiLSTM module to obtain the true / false classification result.

[0122] Specifically, the output of the inverse BiLSTM module needs to be reversed again before being output as follows: Only the intermediate feature vectors can be compared with the feature vectors output by the forward BiLSTM module. The true / false classification result is obtained by linear fusion.

[0123] In one example, the bidirectional recurrent neural network model includes forward training and backward training processes. Forward training uses a first dataset, and backward training uses a second dataset.

[0124] During forward training, the data at time t-1 is processed by the network and then flows to time t, and then through the network again to time t+1. Therefore, the information state at time t+1 is related to the information state at time t, and the information state at time t is related to the information state at time t-1.

[0125] Similarly, during the reverse training of the model, the information state at time t-1 is related to the information state at time t. The information state at time t is related to the information state at time t+1. This process realizes the bidirectional fusion of sample features in the bidirectional recurrent neural network model, which can effectively utilize the temporal information of vehicle network data and improve the accuracy of prediction.

[0126] The linear fusion module linearly fuses the forward intermediate feature vector and the inverse intermediate feature vector to obtain a fusion result. There are multiple fusion results. The fully connected layer classifies based on multiple fusion results, and the softmax layer normalizes the fused data after classification and outputs the prediction result as a value in the range of 0-1.

[0127] That is, the model's predicted output is a value in the range of 0-1. When the predicted value is closer to 1, the prediction is true, indicating that the transmission will overheat in the later period; when the predicted value is closer to 0, the prediction is false, indicating that the transmission will not overheat in the later period.

[0128] In one example, a first dataset with first annotation results and a second dataset with second annotation results are used as historical vehicle network data samples to train a bidirectional recurrent neural network model, including:

[0129] Encode the first dataset with the first annotation result and the second dataset with the second annotation result;

[0130] The encoded first dataset and the second dataset are used as historical vehicle network data samples and input into the bidirectional recurrent neural network model for training until the bidirectional recurrent neural network model converges.

[0131] Encoding the dataset involves, for example, calculating the mean, maximum, minimum, and variance of historical vehicle network data within the dataset to obtain corresponding data encoding results. This allows for the identification of data features within multi-dimensional historical vehicle network data, enabling better prediction of gearbox temperature in future periods based on the data features of current vehicle network data.

[0132] The first annotation result is used to compare the encoded first dataset with the model's predicted output after inputting the encoded first dataset into the bidirectional recurrent neural network model, and adjust the model parameters based on the comparison result. The process of inputting the encoded first dataset into the parameter-adjusted model, obtaining the model's predicted output, comparing the model's predicted output with the first annotation result, and adjusting the model parameters based on the comparison result is repeated.

[0133] Similarly, the second annotation result is used to compare the model prediction output with the encoded second dataset after inputting it into the bidirectional recurrent neural network model, and adjust the model parameters based on the comparison result. The process of inputting the encoded second dataset into the parameter-adjusted model, obtaining the model prediction output, comparing the model prediction output with the second annotation result, and adjusting the model parameters based on the comparison result is repeated.

[0134] The model is trained using the two sets of training samples mentioned above until the bidirectional recurrent neural network model converges.

[0135] The above implementation process, based on historical vehicle network time-series data, extracts datasets corresponding to different time periods. A first dataset with data annotation based on the earlier time period and a second dataset with data annotation based on the later time period are constructed to obtain model training samples. The bidirectional recurrent neural network model is then trained. When training the model bidirectionally, it can learn the temporal mapping relationship features between dataset samples with corresponding time periods, improve the model training effect, and achieve early, effective and accurate prevention of transmission damage before high temperature damage occurs.

[0136] See Figure 4 , Figure 4 This is a flowchart of a high-temperature warning method for a transmission provided in an embodiment of this application. Figure 4 As shown, a method for high temperature warning of a transmission includes the following steps:

[0137] Step 401: Obtain vehicle network data within the set time period.

[0138] For example, the set time period is specifically a period of time within a set duration before and after the current time, or a period of time within a set duration before the current time.

[0139] The signals corresponding to various data in the vehicle are uploaded to the vehicle network platform at a fixed frequency. Therefore, the vehicle network platform contains data sequences with temporal relationships of different vehicle data. That is, the vehicle network data is the time-series data of the vehicle during its journey, which is uploaded at a fixed frequency.

[0140] Correspondingly, in one example, obtaining vehicle-to-everything (V2X) data within a set time period includes:

[0141] Vehicle data uploaded at a fixed frequency within a set time period is extracted from the vehicle network system and used as the vehicle network data.

[0142] Specifically, the vehicle network data includes vehicle identification number (VIN), data collection time, vehicle ambient temperature, engine coolant temperature, transmission oil temperature, vehicle gear position, clutch slippage indicator, and other data.

[0143] Step 402: Based on vehicle network data, temperature prediction is performed using a pre-trained bidirectional recurrent neural network model to obtain the temperature prediction result of the transmission after the target time.

[0144] The temperature prediction result, as the output of the bidirectional recurrent neural network model, can correspond to two categories: true or false. That is, the temperature prediction result can be either a predicted temperature value exceeding the high temperature threshold, which belongs to the true category, or a predicted temperature value not exceeding the high temperature threshold, which belongs to the false category.

[0145] The target time is the time after a set time period, in order to achieve temperature prediction of future time data based on existing data.

[0146] In one example, this step, based on vehicle-to-everything (V2X) data, uses a pre-trained bidirectional recurrent neural network model to predict the temperature of the transmission at the target time, including:

[0147] The vehicle network data is encoded; the encoded vehicle network data is then input into the pre-trained bidirectional recurrent neural network model for temperature prediction, resulting in a temperature prediction of the transmission at the target time.

[0148] Encoding vehicle network data specifically involves solving for characteristic values ​​such as mean, maximum, minimum, and variance based on the vehicle network data to obtain corresponding data encoding results. This allows for the identification of data features in multi-dimensional historical vehicle network data, enabling better prediction of transmission temperature in future periods based on the data features of current vehicle network data.

[0149] The above process corresponds to the application process of the model. When making data predictions based on the trained model, it is only necessary to collect data within the time period p for feature encoding. The data after feature encoding is then fed into the trained bidirectional recurrent neural network model for data prediction, which can yield the prediction result of whether the gearbox will overheat after time d.

[0150] In the above process, the pre-trained bidirectional recurrent neural network model is used to predict temperature data, enabling rapid prediction of the transmission temperature at future moments based on existing vehicle network data. This allows for effective and accurate prevention of transmission damage before it becomes a problem, avoiding irreversible damage caused by high temperatures.

[0151] The bidirectional recurrent neural network model was trained based on historical vehicle network data samples.

[0152] The historical vehicle networking data sample includes a first dataset and a second dataset. The first annotation result of the first dataset is obtained based on the annotation of the third dataset. The first annotation result is used to indicate whether the annotated data in the first dataset corresponds to the transmission high temperature warning signal. The time period corresponding to the first dataset is earlier than the time period corresponding to the third dataset. The second annotation result of the second dataset is obtained based on the annotation of the fourth dataset. The second annotation result is used to indicate whether the annotated data in the second dataset corresponds to the transmission high temperature warning signal. The time period corresponding to the second dataset is later than the time period corresponding to the fourth dataset.

[0153] For details on how to train a bidirectional recurrent neural network model, please refer to the aforementioned examples; they will not be repeated here.

[0154] Step 403: If the temperature prediction result indicates that the probability of high temperature in the transmission exceeds the threshold, output a high temperature warning signal for the target transmission.

[0155] The target transmission high temperature warning signal is used to indicate the display of transmission high temperature warning information on the vehicle's center console display, or to indicate the illumination of the transmission high temperature warning light, or to indicate the playback of the transmission high temperature warning sound, etc.

[0156] In this embodiment, vehicle network data is used in conjunction with a bidirectional recurrent neural network to predict the likelihood of high transmission temperature in the near future. The vehicle owner is then alerted by a high transmission temperature warning system, which can prevent damage to the transmission and extend its lifespan by taking preventative measures before the transmission experiences high oil temperature hazards.

[0157] See Figure 5 , Figure 5 This is a structural diagram of the training device for high temperature warning of gearbox provided in the embodiments of this application. For ease of explanation, only the parts related to the embodiments of this application are shown.

[0158] The high-temperature warning training device 500 for the transmission includes:

[0159] The first acquisition module 501 is used to perform a first data segment extraction operation on historical vehicle network time-series data to obtain a first dataset and a third dataset; wherein the time period corresponding to the first dataset is earlier than the time period corresponding to the third dataset.

[0160] The first annotation module 502 is used to annotate the first dataset based on the third dataset to obtain a first annotation result. The first annotation result is used to indicate whether the annotated data in the first dataset corresponds to the gearbox high temperature warning signal in the third dataset.

[0161] The second acquisition module 503 is used to perform a second data segmentation operation on the historical vehicle network time series data to obtain a second dataset and a fourth dataset; wherein the time period corresponding to the second dataset is later than the time period corresponding to the fourth dataset.

[0162] The second annotation module 504 is used to annotate the second dataset based on the fourth dataset to obtain a second annotation result. The second annotation result is used to indicate whether the annotated data in the second dataset corresponds to the gearbox high temperature warning signal in the fourth dataset.

[0163] The model training module 505 is used to train the bidirectional recurrent neural network model using the first dataset corresponding to the first annotation result and the second dataset corresponding to the second annotation result as historical vehicle network data samples until the bidirectional recurrent neural network model converges.

[0164] The first acquisition module 501 is specifically used for:

[0165] The first dataset is obtained by extracting a data segment from the first time period from the historical vehicle network time series data, and the third dataset is obtained by extracting a data segment from the second time period.

[0166] The second acquisition module 503 is specifically used for:

[0167] The second dataset is obtained by extracting the data segment of the second time period from the historical vehicle network time series data, and the fourth dataset is obtained by extracting the data segment of the first time period.

[0168] The second time period is after the first time period and has a set time interval between them.

[0169] Wherein, the first dataset and the third dataset each correspond to the same time period; the first annotation module 502 is specifically used for:

[0170] The data in the first dataset and the third dataset are compared sequentially according to time.

[0171] If the third dataset contains a transmission high temperature warning signal, then the data in the first dataset whose time sequence corresponds to the transmission high temperature warning signal is marked as true;

[0172] If the third dataset does not contain a high-temperature warning signal for the transmission, then the data in the corresponding time series in the first dataset will be marked as false.

[0173] Wherein, the second dataset and the fourth dataset each correspond to the same time period; the second annotation module 504 is specifically used for:

[0174] The data in the second dataset and the fourth dataset are compared sequentially according to time.

[0175] If the fourth dataset contains a transmission high temperature warning signal, then the data in the second dataset whose time sequence corresponds to the transmission high temperature warning signal is marked as true;

[0176] If the fourth dataset does not contain a high-temperature warning signal for the transmission, then the data in the corresponding time series in the second dataset will be marked as false.

[0177] The model training module 505 is specifically used for:

[0178] Encode the first dataset with the first annotation result and the second dataset with the second annotation result;

[0179] The encoded first dataset and the second dataset are used as historical vehicle network data samples and input into the bidirectional recurrent neural network model for training until the bidirectional recurrent neural network model converges.

[0180] The training device for high-temperature warning of the transmission provided in this application can realize all the processes of the above-described training method for high-temperature warning of the transmission and achieve the same technical effect. To avoid repetition, it will not be described again here.

[0181] See Figure 6 , Figure 6 This is a structural diagram of the high-temperature warning device for the gearbox provided in the embodiments of this application. For ease of explanation, only the parts related to the embodiments of this application are shown.

[0182] The high-temperature warning device 600 for the transmission includes:

[0183] The acquisition module 601 is used to acquire vehicle network data within a set time period;

[0184] The prediction module 602 is used to predict the temperature of the transmission after a target time by using a pre-trained bidirectional recurrent neural network model based on the vehicle network data. The target time is the time after the set time period. The bidirectional recurrent neural network model is trained based on historical vehicle network data samples, which include a first dataset and a second dataset. The first annotation result of the first dataset is obtained based on the annotation of a third dataset. The first annotation result indicates whether the annotated data in the first dataset corresponds to a transmission high-temperature warning signal. The time period corresponding to the first dataset is earlier than the time period corresponding to the third dataset. The second annotation result of the second dataset is obtained based on the annotation of a fourth dataset. The second annotation result indicates whether the annotated data in the second dataset corresponds to a transmission high-temperature warning signal. The time period corresponding to the second dataset is later than the time period corresponding to the fourth dataset.

[0185] The early warning module 603 is used to output a target transmission high temperature early warning signal when the temperature prediction result indicates that the probability of high temperature of the transmission exceeds a threshold.

[0186] Specifically, the acquisition module 601 is used for:

[0187] Vehicle data uploaded at a fixed frequency within a set time period is extracted from the vehicle network system and used as the vehicle network data.

[0188] The prediction module 602 is specifically used for:

[0189] The vehicle network data is encoded;

[0190] The encoded vehicle network data is input into the pre-trained bidirectional recurrent neural network model for temperature prediction, and the temperature prediction result of the gearbox after the target time is obtained.

[0191] The transmission high temperature warning device provided in this application can realize all the processes of the above-described transmission high temperature warning method embodiments and achieve the same technical effect. To avoid repetition, it will not be described again here.

[0192] Figure 7 This is a structural diagram of a terminal provided in an embodiment of this application. As shown in the figure, the terminal 7 of this embodiment includes: at least one processor 70 ( Figure 7 (Only one is shown in the diagram), memory 71, and computer program 72 stored in said memory 71 and executable on said at least one processor 70, which, when executed, implements the steps in any of the above method embodiments.

[0193] The terminal 7 can be a computing device such as a desktop computer, laptop, handheld computer, or cloud server. The terminal 7 may include, but is not limited to, a processor 70 and a memory 71. Those skilled in the art will understand that... Figure 7 This is merely an example of terminal 7 and does not constitute a limitation on terminal 7. It may include more or fewer components than shown, or combine certain components, or different components. For example, the terminal may also include input / output devices, network access devices, buses, etc.

[0194] The processor 70 can 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. The general-purpose processor can be a microprocessor or any conventional processor.

[0195] The memory 71 can be an internal storage unit of the terminal 7, such as a hard disk or memory of the terminal 7. The memory 71 can also be an external storage device of the terminal 7, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the terminal 7. Furthermore, the memory 71 can include both internal storage units and external storage devices of the terminal 7. The memory 71 is used to store the computer program and other programs and data required by the terminal. The memory 71 can also be used to temporarily store data that has been output or will be output.

[0196] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to 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 embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0197] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0198] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0199] In the embodiments provided in this application, it should be understood that the disclosed devices / terminals and methods can be implemented in other ways. For example, the device / terminal embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling or direct coupling or communication connection may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0200] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0201] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0202] If the integrated module / unit is implemented as 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, all or part of the processes in the methods of the above embodiments 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 medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electrical carrier signals and telecommunication signals.

[0203] The methods described in this application can be implemented in whole or in part by a computer program product. When the computer program product is run on a terminal, the terminal executes the steps in the various method embodiments described above.

[0204] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications 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 this application, and should all be included within the protection scope of this application.

Claims

1. A training method for high-temperature warning of a transmission, characterized in that, include: A first data segment extraction operation is performed on historical vehicle network time-series data to obtain a first dataset and a third dataset; wherein the time period corresponding to the first dataset is earlier than the time period corresponding to the third dataset. The data in the first dataset and the third dataset are compared sequentially according to time. If the third dataset contains a transmission high temperature warning signal, then the data in the first dataset whose time sequence corresponds to the transmission high temperature warning signal is marked as true. If the third dataset does not contain a transmission high temperature warning signal, then the data in the first dataset whose time sequence corresponds to false is marked as false, thus obtaining a first labeling result. The first labeling result is used to indicate whether the labeled data in the first dataset corresponds to the transmission high temperature warning signal. A second data segmentation operation is performed on the historical vehicle network time-series data to obtain a second dataset and a fourth dataset; wherein the time period corresponding to the second dataset is later than the time period corresponding to the fourth dataset. The data in the second dataset and the fourth dataset are compared sequentially according to time. If the fourth dataset contains a transmission high temperature warning signal, then the data in the second dataset whose time sequence corresponds to the transmission high temperature warning signal is marked as true. If the fourth dataset does not contain a transmission high temperature warning signal, then the data in the second dataset whose time sequence corresponds to false is marked as false, thus obtaining a second labeling result. The second labeling result is used to indicate whether the labeled data in the second dataset corresponds to the transmission high temperature warning signal. The first dataset with the first annotation result and the second dataset with the second annotation result are used as historical vehicle network data samples to train the bidirectional recurrent neural network model until the bidirectional recurrent neural network model converges.

2. The method according to claim 1, characterized in that, The first data segment extraction operation is performed on the historical vehicle network time-series data to obtain a first dataset and a third dataset, including: The first dataset is obtained by extracting a data segment from the first time period from the historical vehicle network time series data, and the third dataset is obtained by extracting a data segment from the second time period. The second data segment extraction operation is performed on the historical vehicle network time-series data to obtain a second dataset and a fourth dataset, including: The second dataset is obtained by extracting the data segment of the second time period from the historical vehicle network time series data, and the fourth dataset is obtained by extracting the data segment of the first time period. The second time period is after the first time period and has a set time interval between them.

3. The method according to claim 1, characterized in that, The first dataset and the third dataset each correspond to the same time period.

4. The method according to claim 1, characterized in that, The second dataset and the fourth dataset each correspond to the same time period.

5. The method according to claim 1, characterized in that, The step of using the first dataset with the first annotation result and the second dataset with the second annotation result as historical vehicle network data samples to train the bidirectional recurrent neural network model includes: Encode the first dataset with the first annotation result and the second dataset with the second annotation result; The encoded first dataset and the second dataset are used as historical vehicle network data samples and input into the bidirectional recurrent neural network model for training until the bidirectional recurrent neural network model converges.

6. A method for early warning of high temperature in a transmission, characterized in that, include: Acquire vehicle network data within a specified time period; Based on the vehicle network data, a pre-trained bidirectional recurrent neural network model is used to predict the temperature of the transmission after a target time. The target time is the time after the set time period. The bidirectional recurrent neural network model is trained based on historical vehicle network data samples, which include a first dataset and a second dataset. The first annotation result of the first dataset is based on the annotation of a third dataset. The first annotation result indicates whether the annotated data in the first dataset corresponds to a transmission high-temperature warning signal. The time period corresponding to the first dataset is earlier than the time period corresponding to the third dataset. The second annotation result of the second dataset is based on the annotation of a fourth dataset. The second annotation result indicates whether the annotated data in the second dataset corresponds to a transmission high-temperature warning signal. The time period corresponding to the second dataset is earlier than the time period corresponding to the third dataset. The time interval is later than the time interval corresponding to the fourth dataset; the first annotation result is obtained by comparing the data in the first dataset and the third dataset in chronological order, wherein if the third dataset contains a transmission high temperature warning signal, the data in the first dataset whose time sequence corresponds to the transmission high temperature warning signal is labeled as true, and if the third dataset does not contain a transmission high temperature warning signal, the data in the first dataset whose time sequence corresponds to the transmission high temperature warning signal is labeled as false; the second annotation result is obtained by comparing the data in the second dataset and the fourth dataset in chronological order, wherein if the fourth dataset contains a transmission high temperature warning signal, the data in the second dataset whose time sequence corresponds to the transmission high temperature warning signal is labeled as true, and if the fourth dataset does not contain a transmission high temperature warning signal, the data in the second dataset whose time sequence corresponds to the transmission high temperature warning signal is labeled as false; If the temperature prediction result indicates that the probability of the transmission being overheated exceeds a threshold, a high temperature warning signal for the target transmission is output.

7. The method according to claim 6, characterized in that, The step of predicting the temperature of the transmission after the target time using a pre-trained bidirectional recurrent neural network model based on the vehicle network data includes: The vehicle network data is encoded; The encoded vehicle network data is input into the pre-trained bidirectional recurrent neural network model for temperature prediction, and the temperature prediction result of the gearbox after the target time is obtained.

8. A training device for high-temperature warning of a gearbox, characterized in that, include: The first acquisition module is used to perform a first data segment extraction operation on historical vehicle network time-series data to obtain a first dataset and a third dataset; wherein the time period corresponding to the first dataset is earlier than the time period corresponding to the third dataset. The first annotation module is used to compare the data in the first dataset and the third dataset in chronological order. If the third dataset contains a transmission high temperature warning signal, the data in the first dataset whose time sequence corresponds to the transmission high temperature warning signal is marked as true. If the third dataset does not contain a transmission high temperature warning signal, the data in the first dataset whose time sequence corresponds to the transmission high temperature warning signal is marked as false, thus obtaining a first annotation result. The first annotation result is used to indicate whether the annotated data in the first dataset corresponds to the transmission high temperature warning signal in the third dataset. The second acquisition module is used to perform a second data segmentation operation on the historical vehicle network time-series data to obtain a second dataset and a fourth dataset; wherein the time period corresponding to the second dataset is later than the time period corresponding to the fourth dataset. The second annotation module is used to compare the data in the second dataset and the fourth dataset in chronological order. If the fourth dataset contains a transmission high temperature warning signal, the data in the second dataset whose time sequence corresponds to the transmission high temperature warning signal is marked as true. If the fourth dataset does not contain a transmission high temperature warning signal, the data in the second dataset whose time sequence corresponds to the warning signal is marked as false, thus obtaining a second annotation result. The second annotation result is used to indicate whether the annotated data in the second dataset corresponds to the transmission high temperature warning signal in the fourth dataset. The model training module is used to train the bidirectional recurrent neural network model using the first dataset corresponding to the first annotation result and the second dataset corresponding to the second annotation result as historical vehicle network data samples until the bidirectional recurrent neural network model converges.

9. A high-temperature warning device for a transmission, characterized in that, include: The acquisition module is used to acquire vehicle network data within a set time period; The prediction module is used to predict the temperature of the transmission after a target time by using a pre-trained bidirectional recurrent neural network model based on the vehicle network data. The target time is a time after a set time period. The bidirectional recurrent neural network model is trained based on historical vehicle network data samples, which include a first dataset and a second dataset. The first annotation result of the first dataset is obtained based on the annotation of a third dataset. The first annotation result indicates whether the annotated data in the first dataset corresponds to a transmission high-temperature warning signal. The time period corresponding to the first dataset is earlier than the time period corresponding to the third dataset. The second annotation result of the second dataset is obtained based on the annotation of a fourth dataset. The second annotation result indicates whether the annotated data in the second dataset corresponds to a transmission high-temperature warning signal. The corresponding time period is later than the time period corresponding to the fourth dataset; the first annotation result is obtained by comparing the data in the first dataset and the third dataset in chronological order, wherein if the third dataset contains a transmission high temperature warning signal, the data in the first dataset whose time sequence corresponds to the transmission high temperature warning signal is labeled as true, and if the third dataset does not contain a transmission high temperature warning signal, the data in the first dataset whose time sequence corresponds to the transmission high temperature warning signal is labeled as false; the second annotation result is obtained by comparing the data in the second dataset and the fourth dataset in chronological order, wherein if the fourth dataset contains a transmission high temperature warning signal, the data in the second dataset whose time sequence corresponds to the transmission high temperature warning signal is labeled as true, and if the fourth dataset does not contain a transmission high temperature warning signal, the data in the second dataset whose time sequence corresponds to the transmission high temperature warning signal is labeled as false; The early warning module is used to output a target transmission high temperature early warning signal when the temperature prediction result indicates that the probability of the transmission being overheated exceeds a threshold.

10. A terminal, 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 method as described in any one of claims 1 to 7.

Images