Vehicle battery feed reason detection method and device, vehicle and storage medium

Abstract

The invention relates to the technical field of vehicles, in particular to a vehicle battery feed reason detection method and device, a vehicle and a storage medium, and the method comprises the steps: carrying out the data preprocessing of collected operation data, obtaining the to-be-labeled data of the vehicle, carrying out the data analysis of the to-be-labeled data, selecting the target labeled data meeting a preset condition, and carrying out the labeling of the target labeled data, the method comprises the following steps: obtaining label data and label-free data of a vehicle, training a preset machine model based on the label data to obtain a mapping relationship between the label data and labels, receiving new vehicle operation data and performing data preprocessing, inputting the new vehicle operation data and the label-free data into the trained preset machine model, and performing data preprocessing on the new vehicle operation data. And obtaining a feed reason prediction result of the vehicle based on the mapping relationship between the label data and the labels. Therefore, the problems of low detection efficiency, insufficient accuracy and the like of the battery feed reason are solved, and the accurate judgment of the battery feed reason is realized by collecting the multi-source data of the vehicle and combining the deep learning algorithm.

Description

Technical Field

[0001] The present application relates to the technical field of vehicles, and particularly relates to a method, a device, a vehicle and a storage medium for detecting the reasons for vehicle battery power failure. Background Art

[0002] Vehicle power failure refers to the situation where the vehicle cannot start due to insufficient battery power or other electrical system failures. This phenomenon is relatively common in daily driving, which not only affects the user's travel plan but also may lead to serious safety risks. Therefore, in order to avoid problems such as low comfort and low safety caused by vehicle power failure, it is very necessary to accurately detect the reasons for vehicle power failure.

[0003] In the related art, the detection methods for vehicle power failure mainly rely on the experience of relevant technical personnel and simple test tools for detection. For example, a multimeter is used to measure the battery voltage or the power failure situation of the vehicle is detected through a remote monitoring system.

[0004] However, the above detection methods have many limitations. For example, when relying on the experience of relevant technical personnel for detection, since manual detection takes a long time, the detection efficiency is low. In the face of complex faults, it is necessary to try many times to locate the problem, resulting in insufficient detection accuracy; when detecting through a remote monitoring system, usually only a basic health status report is provided, lacking in-depth fault diagnosis capabilities. Therefore, the detection accuracy is also insufficient and needs to be solved urgently. Summary of the Invention

[0005] The present application provides a method, a device, a vehicle and a storage medium for detecting the reasons for vehicle battery power failure to solve problems such as low detection efficiency and insufficient accuracy of the reasons for battery power failure.

[0006] The first aspect of the present application provides a method for detecting the reasons for vehicle battery power failure, including the following steps:

[0007] Collect the operation data of the vehicle, where the operation data includes the operation data of the vehicle in a normal driving state and the operation data of the vehicle in a power failure state;

[0008] Perform data preprocessing on the operation data to obtain the data to be labeled of the vehicle, and perform data analysis on the data to be labeled, select the target labeled data that meets the preset conditions for labeling to obtain the labeled data and the unlabeled data of the vehicle, and train a preset machine model based on the labeled data. After the preset machine model is trained, obtain the mapping relationship between the labeled data and the labels;

[0009] Determine whether new vehicle operation data is received. If the new vehicle operation data is received, preprocess the new vehicle operation data, and use the preprocessed new vehicle operation data and the unlabeled data of the vehicle as input samples to input into the trained preset machine model, so as to obtain the prediction result of the power feeding reason of the vehicle based on the mapping relationship between the label data and the label.

[0010] According to an embodiment of the present application, the collecting the operation data of the vehicle includes:

[0011] Collect the battery voltage data, battery current data, battery temperature data, electrical system state data of the vehicle, environmental data of the vehicle, and feedback data of the user.

[0012] According to an embodiment of the present application, the preprocessing the operation data to obtain the to-be-labeled data of the vehicle includes:

[0013] Clean the operation data, remove the abnormal data of the operation data and fill the missing data of the operation data;

[0014] Normalize the operation data after data cleaning, and extract features from the normalized operation data to obtain the to-be-labeled data of the vehicle.

[0015] According to an embodiment of the present application, the performing data analysis on the to-be-labeled data and selecting the target labeled data that meets the preset conditions for labeling includes:

[0016] Identify the confidence level of the to-be-labeled data;

[0017] Determine whether there is target labeled data in the confidence levels of the to-be-labeled data whose confidence level is less than the preset confidence level threshold;

[0018] If there is target labeled data in the confidence levels of the to-be-labeled data whose confidence level is less than the preset confidence level threshold, determine that the target labeled data meets the preset conditions, and label the target labeled data.

[0019] According to an embodiment of the present application, the training the preset machine model based on the label data includes:

[0020] Divide the label data to obtain a training set, a validation set, and a test set of the label data;

[0021] Set the initial parameters of the preset machine model, and use the training set of the label data to train the preset machine model to obtain the training result of the preset machine model;

[0022] Based on the training results, adjust the initial parameters of the preset machine model using the validation set to optimize the learning performance of the preset machine model, and evaluate the learning performance of the preset machine model according to the test set.

[0023] According to the method for detecting the cause of vehicle battery power feeding in the embodiments of the present application, after preprocessing the collected operation data, the data to be labeled of the vehicle is obtained. Then, data analysis is performed on the data to be labeled, and target labeled data that meets preset conditions is selected for labeling to obtain the labeled data and unlabeled data of the vehicle. Based on the labeled data, a preset machine model is trained to obtain the mapping relationship between the labeled data and the labels. After receiving new vehicle operation data and performing data preprocessing, the new vehicle operation data and the unlabeled data are used as inputs to the trained preset machine model to obtain the prediction result of the cause of battery power feeding of the vehicle based on the mapping relationship between the labeled data and the labels. Thus, problems such as low detection efficiency and insufficient accuracy of the cause of battery power feeding are solved. By collecting multi-source data of the vehicle and combining deep learning algorithms, accurate judgment of the cause of battery power feeding is realized.

[0024] The second aspect of the embodiments of the present application provides a device for detecting the cause of vehicle battery power feeding, including:

[0025] An acquisition module, configured to acquire operation data of a vehicle, where the operation data includes operation data of the vehicle in a normal driving state and operation data of the vehicle in a power feeding state;

[0026] A labeling module, configured to preprocess the operation data to obtain the data to be labeled of the vehicle, perform data analysis on the data to be labeled, select target labeled data that meets preset conditions for labeling to obtain the labeled data and unlabeled data of the vehicle, train a preset machine model based on the labeled data, and obtain the mapping relationship between the labeled data and the labels after the preset machine model is trained;

[0027] An acquisition module, configured to determine whether new vehicle operation data is received. If the new vehicle operation data is received, preprocess the new vehicle operation data, and use the preprocessed new vehicle operation data and the unlabeled data of the vehicle as input samples to input to the trained preset machine model to obtain the prediction result of the cause of battery power feeding of the vehicle based on the mapping relationship between the labeled data and the labels by using the trained preset machine model.

[0028] According to an embodiment of the present application, the acquisition module is specifically configured to:

[0029] Collect the battery voltage data, battery current data, battery temperature data, electrical system status data of the vehicle, environmental data of the vehicle, and feedback data of the user.

[0030] According to an embodiment of the present application, the annotation module is specifically configured to:

[0031] Perform data cleaning on the operation data, remove abnormal data from the operation data, and fill in missing data of the operation data;

[0032] Perform data normalization on the operation data after data cleaning, and perform feature extraction on the operation data after normalization processing to obtain the data to be annotated of the vehicle.

[0033] According to an embodiment of the present application, the annotation module is specifically configured to:

[0034] Identify the confidence level of the data to be annotated;

[0035] Determine whether there is target annotation data in the confidence levels of the data to be annotated whose confidence level is less than a preset confidence level threshold;

[0036] If there is target annotation data in the confidence levels of the data to be annotated whose confidence level is less than the preset confidence level threshold, determine that the target annotation data meets the preset conditions, and annotate the target annotation data.

[0037] According to an embodiment of the present application, the annotation module is specifically configured to:

[0038] Divide the tag data to obtain a training set, a validation set, and a test set of the tag data;

[0039] Set initial parameters of the preset machine model, and use the training set of the tag data to train the preset machine model to obtain a training result of the preset machine model;

[0040] Based on the training result, use the validation set to adjust the initial parameters of the preset machine model to optimize the learning performance of the preset machine model, and evaluate the learning performance of the preset machine model according to the test set.

[0041] The detection device for the vehicle battery power supply reason according to the embodiment of the present application preprocesses the collected operation data to obtain the data to be labeled of the vehicle, analyzes the data to be labeled, selects the target labeled data that meets the preset conditions for labeling, obtains the labeled data and unlabeled data of the vehicle, and trains the preset machine model based on the labeled data to obtain the mapping relationship between the labeled data and the labels. After receiving the new vehicle operation data and preprocessing the data, the new vehicle operation data and the unlabeled data are used as inputs to the trained preset machine model to obtain the prediction result of the battery power supply reason of the vehicle based on the mapping relationship between the labeled data and the labels. Thus, the problems of low detection efficiency and insufficient accuracy of the battery power supply reason are solved. By collecting multi-source data of the vehicle and combining deep learning algorithms, accurate judgment of the battery power supply reason is realized.

[0042] The third aspect embodiment of the present application provides a vehicle, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor executes the program to implement the detection method for the vehicle battery power supply reason as described in the above embodiment.

[0043] The fourth aspect embodiment of the present application provides a computer-readable storage medium, and the computer-readable storage medium stores computer instructions, and the computer instructions are used to cause the computer to execute the detection method for the vehicle battery power supply reason as described in the above embodiment.

[0044] The fifth aspect embodiment of the present application provides a computer program product, including a computer program, and the computer program is executed to be used to implement the detection method for the vehicle battery power supply reason as described in the above embodiment.

[0045] The additional aspects and advantages of the present application will be partially given in the following description, partially become obvious from the following description, or be understood through the practice of the present application. Description of the Drawings

[0046] The above and / or additional aspects and advantages of the present application will become obvious and easy to understand from the following description of the embodiments in conjunction with the drawings, where:

[0047] Figure 1 It is a flowchart of a detection method for a vehicle battery power supply reason according to an embodiment of the present application;

[0048] Figure 2 It is a flowchart of the overall method according to an embodiment of the present application;

[0049] Figure 3 It is a schematic diagram of the data preprocessing process according to an embodiment of the present application;

[0050] Figure 4An exemplary diagram of a detection device for reasons of vehicle battery power feeding according to an embodiment of the present application;

[0051] Figure 5 A schematic structural diagram of a vehicle according to an embodiment of the present application. Detailed implementation manners

[0052] Embodiments of the present application will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present application, and should not be construed as a limitation to the present application.

[0053] A detection method, device, vehicle and storage medium for reasons of vehicle battery power feeding according to an embodiment of the present application will be described below with reference to the accompanying drawings. In view of the problems of low detection efficiency and insufficient accuracy of battery power feeding reasons mentioned in the above background art, the present application provides a detection method for reasons of vehicle battery power feeding. In this method, after preprocessing the collected operation data, the to-be-labeled data of the vehicle is obtained. Then, data analysis is performed on the to-be-labeled data, and the target labeled data that meets the preset conditions is selected for labeling to obtain the labeled data and unlabeled data of the vehicle. And a preset machine model is trained based on the labeled data to obtain the mapping relationship between the labeled data and the labels. After receiving new vehicle operation data and performing data preprocessing, the new vehicle operation data and the unlabeled data are used as inputs to the trained preset machine model to obtain the prediction result of the power feeding reason of the vehicle based on the mapping relationship between the labeled data and the labels. Thus, problems such as low detection efficiency and insufficient accuracy of battery power feeding reasons are solved, and by collecting multi-source data of the vehicle and combining deep learning algorithms, accurate judgment of the reasons for battery power feeding is realized.

[0054] Specifically, Figure 1 A flowchart of a detection method for reasons of vehicle battery power feeding provided by an embodiment of the present application.

[0055] As Figure 1 shown, the detection method for reasons of vehicle battery power feeding includes the following steps:

[0056] In step S101, the operation data of the vehicle is collected, where the operation data includes the operation data of the vehicle in a normal driving state and the operation data of the vehicle in a power feeding state.

[0057] According to an embodiment of the present application, collecting the operation data of the vehicle includes: collecting the battery voltage data, battery current data, battery temperature data, electrical system state data of the vehicle, environmental data of the vehicle, and feedback data of the user.

[0058] Specifically, to address the deficiencies of vehicle power feed diagnosis methods (such as using a multimeter) and remote monitoring systems in high-end vehicles in related technologies, which lead to low efficiency, insufficient accuracy, or lack of in-depth fault diagnosis capabilities in vehicle power feed diagnosis, embodiments of this application collect a large amount of data on the vehicle in normal operation and power feed states, and use machine learning algorithms to train an efficient diagnosis model, thereby achieving a fast and accurate determination of the cause of vehicle power feed. The specific steps include data collection, data preprocessing, model training, and model application.

[0059] Specifically, as Figure 2 shown, since data collection is the basis for the implementation of the entire method, obtaining rich data from various sources can provide the necessary input for subsequent analysis and modeling, thereby ensuring that the collected data covers various situations of the vehicle in normal operation and power feed states, which helps to improve the comprehensiveness and accuracy of the model. Therefore, embodiments of this application first need to collect the operation data of the vehicle in the normal driving state and the operation data of the vehicle in the power feed state. For example, through the vehicle's built-in sensors, such as a battery voltage sensor to collect the vehicle's battery voltage data, a current sensor to collect the vehicle's battery current data, a temperature sensor, etc. to collect the vehicle's battery temperature data, through the OBD (On-Board Diagnostics) to collect the vehicle's electrical system status data and the vehicle's environmental data, and through a mobile application to collect user feedback data, including the vehicle's usage situation, maintenance records, etc.

[0060] Among them, the types of collected data mainly include battery voltage and current changes, environmental parameters (such as temperature, humidity, etc., external factors that may affect battery performance), vehicle status, etc. Data collection combines real-time collection and event-triggered collection to ensure the integrity and timeliness of the data.

[0061] In step S102, the operation data is preprocessed to obtain the vehicle's data to be labeled, and the data to be labeled is analyzed. The target labeled data that meets the preset conditions is selected for labeling to obtain the vehicle's labeled data and unlabeled data. Based on the labeled data, a preset machine model is trained, and after the preset machine model is trained, the mapping relationship between the labeled data and the labels is obtained.

[0062] According to an embodiment of this application, preprocessing the operation data to obtain the vehicle's data to be labeled includes: cleaning the operation data to remove abnormal data from the operation data and filling in the missing data of the operation data; performing data normalization on the operation data after data cleaning, and extracting features from the normalized operation data to obtain the vehicle's data to be labeled.

[0063] Among them, both the preset conditions and the preset machine model can be set by those skilled in the art according to actual test requirements, and no specific limitation is made here.

[0064] Specifically, data preprocessing is one of the key steps before using machine learning methods, which directly affects the training effect and final performance of the preset machine model. Therefore, after collecting the operation data of the vehicle in the normal driving state and the operation data of the vehicle in the power-off state, in order to ensure the accuracy of the data, it is necessary to further perform data preprocessing on the collected operation data.

[0065] Specifically, as Figure 3 shown, first of all, it is necessary to perform data cleaning on the collected operation data, which mainly includes two steps: removing outliers and handling missing values, that is, removing the abnormal data of the operation data and filling the missing data of the operation data. Among them, the reason for removing outliers is that outliers may have a negative impact on the training of the preset machine model, resulting in overfitting or performance degradation of the preset machine model. The reason for handling missing values is that missing values will affect the integrity of the data and the accuracy of the preset machine model, and need to be reasonably processed. Among them, the method for removing outliers generally uses the box plot method, that is, using the quartiles (Q1 and Q3) and the interquartile range (IQR, Interquartile Range) to identify and remove outliers. For example, data points less than Q1 - 1.5 * IQR or greater than Q3 + 1.5 * IQR are regarded as outliers. The specific method is as follows: (1) Calculate the first quartile (Q1, that is, the 25th percentile) and the third quartile (Q3, that is, the 75th percentile) of the data set; (2) The interquartile range is defined as the difference between Q3 and Q1, that is, IQR = Q3 - Q1; (3) Determine the lower and upper bounds of the outliers based on IQR. The lower bound is: Q1 - 1.5 * IQR, and the upper bound is: Q3 + 1.5 * IQR; (4) Identify the outliers in the data according to the upper and lower bounds. Any value below the lower bound, that is, less than Q1 - 1.5 * IQR, or above the upper bound, that is, greater than Q3 + 1.5 * IQR, is considered an outlier; (5) Delete the outliers. The method for handling missing values generally uses the filling method, that is, filling the missing values with the mean, median, mode or interpolation method.

[0066] Secondly, in order to help map the values of different features to the same scale, thereby improving the training efficiency and stability of the preset machine model, it is necessary to further perform data normalization processing on the cleaned operation data. For example, Min - Max, Z - score normalization, etc. can be used for normalization, so as to accelerate the convergence speed of optimization algorithms such as gradient descent based on the normalization processing, and at the same time avoid the deviation caused by the difference in feature scales.

[0067] Finally, feature extraction is performed on the normalized operation data to extract features highly correlated with the target variable, obtaining the data to be labeled for the vehicle, so as to improve the prediction performance of the preset machine model. For example, statistical feature methods can be used, such as calculating statistical features such as the mean, variance, maximum value, minimum value, kurtosis, and skewness of the battery voltage and current; calculating dynamic features such as the change rate and fluctuation range of the voltage and current.

[0068] It should be noted that the above outlier processing can be computationally implemented based on Python (Scikit-learn, Keras, TensorFlow libraries), R language, etc.

[0069] According to an embodiment of the present application, data analysis is performed on the data to be labeled, and target labeled data that meets preset conditions is selected for labeling, including: identifying the confidence level of the data to be labeled; determining whether there is target labeled data in the confidence levels of the data to be labeled whose confidence level is less than the preset confidence threshold; if there is target labeled data in the confidence levels of the data to be labeled whose confidence level is less than the preset confidence threshold, it is determined that the target labeled data meets the preset conditions, and the target labeled data is labeled.

[0070] Among them, the preset confidence threshold can be set by those skilled in the art according to actual test requirements, or can be obtained through a limited number of computer simulations, and no specific limitation is made here.

[0071] Specifically, after the above-mentioned data preprocessing of the vehicle's operation data, further preset machine model training is performed, mainly including data labeling, model selection, model training, and model optimization, which will be introduced in detail one by one below.

[0072] Specifically, first, perform data analysis on the data to be labeled for the vehicle, and select the target labeled data that meets the preset conditions for labeling. For example, label the data that is most important for improving the performance of the preset machine model. Obtain the labeled data and unlabeled data of the vehicle. Data labeling is to assign labels to the training data, clarify the corresponding power feed reason for each piece of data (such as "battery aging", "charging system failure", etc.), so that the preset machine model can learn the correct mapping relationship. Among them, it is necessary to select the target labeled data that meets the preset conditions for labeling. For example, if there is target labeled data with a confidence level lower than the preset confidence threshold among the confidence levels of the data to be labeled, it is determined that the target labeled data meets the preset conditions and label the target labeled data, which can significantly improve the model performance and generalization ability of the model. The confidence level is based on the similarity score between the data to be labeled and the known power feed reason. The preset confidence threshold can be set to X% according to historical data analysis. Usually, the amount of labeled data should account for 10%-30% of the entire data set. The specific proportion depends on the scale and complexity of the data set. This part of the labeled data should cover as many scenarios and situations as possible in order to train a preset machine model that can generalize to unseen data. Data labeling can be done through professional tools or manually. For example, professional data labeling tools such as Labelbox and Amazon Mechanical Turk can be used, or professional personnel can manually label the data according to the vehicle's maintenance records and fault codes, and record the time of each power feed occurrence and the vehicle's status information. Secondly, train the preset machine model based on the labeled data. The labeled content is the power feed reason: such as "battery aging", "charging system failure", "electrical system short circuit", and record the time of each power feed occurrence, mark the vehicle's status at the time of power feed, such as mileage, start times, etc. After the preset machine model training is completed, obtain the mapping relationship between the labeled data and the label, that is, clarify the corresponding power feed reason for each piece of data.

[0073] It should be noted that since there are many types of preset machine models, it is necessary to select a suitable preset machine model to experimentally compare the performance of different models. At the same time, it is also necessary to consider whether the GPU (Graphics Processing Unit) resources of the current training model can support it, so as to select a suitable model.

[0074] For example, the preset machine models mainly include decision trees, random forests or support vector machines. Among them, decision trees are suitable for simple classification tasks and are easy to understand and interpret; random forests can integrate multiple decision trees to improve the stability and accuracy of the model; support vector machines are suitable for high-dimensional data and have good classification performance.

[0075] According to an embodiment of the present application, training a preset machine model based on label data includes: dividing the label data to obtain a training set, a validation set, and a test set of the label data; setting initial parameters of the preset machine model, and training the preset machine model using the training set of the label data to obtain a training result of the preset machine model; based on the training result, adjusting the initial parameters of the preset machine model using the validation set to optimize the learning performance of the preset machine model, and evaluating the learning performance of the preset machine model according to the test set.

[0076] Specifically, as Figure 2 shown, training a preset machine model based on the above-mentioned labeled target annotation data, so as to optimize the model performance by adjusting hyperparameters. The goal of training is to enable the preset machine model to learn to identify different power supply reasons from input features. Training the preset machine model mainly includes training the model and evaluating the model performance. Among them, training the model can use the labeled data set to train the selected preset machine model so that it can learn the rules of power supply reasons from the data, and evaluating the model performance can evaluate the performance of the model through methods such as cross-validation to ensure the generalization ability of the model.

[0077] It should be noted that during the training process of the preset machine model, high-performance computing power or a cloud computing platform with GPU resources is required to improve the computing accuracy.

[0078] Specifically, first, divide the label data to obtain a training set, a validation set, and a test set of the label data. Among them, the training set is the data used to train the preset machine model, the validation set is the data used to adjust hyperparameters and evaluate the model performance, and the test set is the data used to finally evaluate the performance of the preset machine model; second, initialize the preset machine model, mainly including setting the initial parameters of the preset machine model; third, use the training set to train the model and adjust the model parameters to minimize the loss function; finally, evaluate the model performance and save the model. The initial parameters of the preset machine model can be adjusted using the validation set to optimize the learning performance of the preset machine model. Saving the model means saving the trained preset machine model for subsequent use.

[0079] Furthermore, in order to further optimize the performance of the obtained preset machine model, mainly by adjusting the hyperparameters of the preset machine model, further improving the performance and generalization ability of the preset machine model. Thus, through the above four steps of data annotation, model selection, model training, and model optimization, the mapping relationship between label data and labels is obtained, and after systematically training the preset machine model, it is ensured that the preset machine model can accurately predict the reasons for vehicle power supply.

[0080] In step S103, it is determined whether new vehicle operation data is received. If new vehicle operation data is received, the new vehicle operation data is preprocessed, and the preprocessed new vehicle operation data and the unlabeled data of the vehicle are used as input samples and input into a trained preset machine model, so as to obtain a prediction result of the power feed reason of the vehicle based on the mapping relationship between the labeled data and the label by using the trained preset machine model.

[0081] Specifically, if new vehicle operation data is received, the received new vehicle operation data is preprocessed through the above steps of data preprocessing, and the preprocessed new vehicle operation data and the unlabeled data of the vehicle are used as input samples and input into a trained preset machine model. Thus, a prediction result of the power feed reason of the vehicle can be obtained based on the mapping relationship between the labeled data and the label by using the trained preset machine model. The specific data preprocessing and model training processes are the same as those of the above method. To avoid redundancy, no specific description is given here.

[0082] Therefore, through large-scale data training in the embodiments of the present application, the preset machine model can identify various complex power feed reasons, improve the accuracy and reliability of diagnosis, and the preset machine model can immediately analyze when the vehicle power feed occurs, provide an immediate diagnosis result, shorten the troubleshooting time, ensure the real-time nature of the feedback of the vehicle power feed reason, and at the same time, through continuous monitoring and analysis of historical data, the model can also predict potential power feed risks and achieve preventive maintenance, reflecting the predictability of the feedback of the vehicle power feed reason.

[0083] According to the method for detecting the reason for the vehicle battery power feed in the embodiments of the present application, after preprocessing the collected operation data, the to-be-labeled data of the vehicle is obtained. The to-be-labeled data is analyzed, and the target labeled data that meets the preset conditions is selected for labeling to obtain the labeled data and unlabeled data of the vehicle. And a preset machine model is trained based on the labeled data to obtain the mapping relationship between the labeled data and the label. After receiving new vehicle operation data and preprocessing it, the new vehicle operation data and the unlabeled data are used as input to the trained preset machine model to obtain a prediction result of the power feed reason of the vehicle based on the mapping relationship between the labeled data and the label. Thus, problems such as low detection efficiency and insufficient accuracy of the battery power feed reason are solved. By collecting multi-source data of the vehicle and combining deep learning algorithms, accurate judgment of the battery power feed reason is realized.

[0084] Next, a device for detecting the reason for the vehicle battery power feed according to an embodiment of the present application will be described with reference to the accompanying drawings.

[0085] Figure 4 It is a block diagram of the device for detecting the reason for the vehicle battery power feed in the embodiments of the present application.

[0086] AsFigure 4 As shown in Figure 4 , the detection device 10 for the reasons of vehicle battery power feeding includes: an acquisition module 100, a labeling module 200, and an obtaining module 300.

[0087] Among them, the acquisition module 100 is used to acquire the operation data of the vehicle, where the operation data includes the operation data of the vehicle in a normal driving state and the operation data of the vehicle in a power feeding state.

[0088] The labeling module 200 is used to perform data preprocessing on the operation data to obtain the data to be labeled of the vehicle, and perform data analysis on the data to be labeled, select the target labeled data that meets the preset conditions for labeling, obtain the labeled data and the unlabeled data of the vehicle, and train a preset machine model based on the labeled data. After the preset machine model is trained, obtain the mapping relationship between the labeled data and the labels.

[0089] The obtaining module 300 is used to determine whether new vehicle operation data is received. If new vehicle operation data is received, perform data preprocessing on the new vehicle operation data, and use the preprocessed new vehicle operation data and the unlabeled data of the vehicle as input samples to input into the trained preset machine model, so as to obtain the prediction result of the vehicle power feeding reason based on the mapping relationship between the labeled data and the labels by using the trained preset machine model.

[0090] According to an embodiment of the present application, the acquisition module 100 is specifically used for:

[0091] Acquire the battery voltage data, battery current data, battery temperature data, electrical system state data of the vehicle, environmental data of the vehicle, and user feedback data of the vehicle.

[0092] According to an embodiment of the present application, the labeling module 200 is specifically used for:

[0093] Perform data cleaning on the operation data, remove the abnormal data of the operation data and fill the missing data of the operation data.

[0094] Perform data normalization processing on the operation data after data cleaning, and perform feature extraction on the normalized operation data to obtain the data to be labeled of the vehicle.

[0095] According to an embodiment of the present application, the labeling module 200 is specifically used for:

[0096] Identify the confidence of the data to be labeled.

[0097] Judge whether there is target labeled data with a confidence less than the preset confidence threshold in the confidence of the data to be labeled.

[0098] If there is target annotation data with a confidence level less than the preset confidence level threshold among the confidence levels of the data to be annotated, it is determined that the target annotation data meets the preset conditions, and the target annotation data is annotated.

[0099] According to an embodiment of the present application, the annotation module 200 is specifically configured to:

[0100] Divide the labeled data to obtain a training set, a validation set, and a test set of the labeled data;

[0101] Set the initial parameters of the preset machine model, and use the training set of the labeled data to train the preset machine model to obtain the training result of the preset machine model;

[0102] Based on the training result, use the validation set to adjust the initial parameters of the preset machine model to optimize the learning performance of the preset machine model, and evaluate the learning performance of the preset machine model according to the test set.

[0103] According to the detection device for the vehicle battery power supply reason in the embodiment of the present application, after preprocessing the collected operation data, the data to be annotated of the vehicle is obtained, the data to be annotated is analyzed, and the target annotation data that meets the preset conditions is selected for annotation to obtain the labeled data and unlabeled data of the vehicle, and based on the labeled data, the preset machine model is trained to obtain the mapping relationship between the labeled data and the label. After receiving new vehicle operation data and performing data preprocessing, the new vehicle operation data and the unlabeled data are used as inputs to the trained preset machine model to obtain the prediction result of the vehicle's power supply reason based on the mapping relationship between the labeled data and the label. Thus, problems such as low detection efficiency and insufficient accuracy of the battery power supply reason are solved. By collecting multi-source data of the vehicle and combining deep learning algorithms, accurate judgment of the battery power supply reason is realized.

[0104] Figure 5 The structural schematic diagram of the vehicle provided by the embodiment of the present application. The vehicle may include:

[0105] A memory 501, a processor 502, and a computer program stored on the memory 501 and executable on the processor 502.

[0106] When the processor 502 executes the program, it implements the detection method for the vehicle battery power supply reason provided in the above embodiment.

[0107] Further, the vehicle further includes:

[0108] A communication interface 503 for communication between the memory 501 and the processor 502.

[0109] The memory 501 is used to store a computer program executable on the processor 502.

[0110] The memory 501 may include high-speed RAM memory and may also include non-volatile memory, such as at least one disk memory.

[0111] If the memory 501, the processor 502, and the communication interface 503 are implemented independently, the communication interface 503, the memory 501, and the processor 502 can be interconnected through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 5 only a thick line is shown in the figure, but it does not mean that there is only one bus or one type of bus.

[0112] Optionally, in a specific implementation, if the memory 501, the processor 502, and the communication interface 503 are integrated on a single chip, the memory 501, the processor 502, and the communication interface 503 can communicate with each other through an internal interface.

[0113] The processor 502 may be a Central Processing Unit (CPU), or an Application Specific Integrated Circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0114] This embodiment also provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the above-mentioned method for detecting the reasons for vehicle battery power failure is implemented.

[0115] This embodiment also provides a computer program product, including a computer program, which is executed to implement the method for detecting the reasons for vehicle battery power failure in the above-mentioned embodiment.

[0116] In the description of this specification, the descriptions referring to terms such as "one embodiment", "some embodiments", "examples", "specific examples", or "some examples", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of this application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0117] In addition, the terms "first" and "second" are used only for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of such features. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise specifically defined.

[0118] Any process or method description depicted in a flowchart or otherwise described herein can be understood to represent a module, segment, or portion of code including one or more N executable instructions for implementing a customized logic function or process, and the scope of the preferred embodiments of this application includes additional implementations, where the functions can be executed in a substantially simultaneous manner or in a reverse order according to the functions involved, rather than in the order shown or discussed, which should be understood by those skilled in the art to which the embodiments of this application pertain.

[0119] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in conjunction with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electrical connection part (electronic device) having one or N wirings, a portable computer disk cartridge (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which a program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing as appropriate, and then storing it in a computer memory.

[0120] It should be understood that various parts of the present application can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0121] Those of ordinary skill in the art of this technology can understand that all or part of the steps carried by the methods of the above embodiments can be completed by a program instructing relevant hardware, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiments.

[0122] In addition, each functional unit in various embodiments of the present application may be integrated into a processing module, may exist physically alone for each unit, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. When the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0123] The above-mentioned storage medium may be a read-only memory, a magnetic disk, an optical disc, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present application. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. A method for detecting the cause of battery feeding in a vehicle, characterized in that: The following steps are involved: Collecting operation data of the vehicle, wherein the operation data includes operation data of the vehicle in a normal driving state and operation data of the vehicle in a power feeding state; Performing data preprocessing on the operating data to obtain the data to be labeled of the vehicle, performing data analysis on the data to be labeled, selecting target labeled data that meets preset conditions for labeling, obtaining labeled data of the vehicle and unlabeled data of the vehicle, training a preset machine model based on the labeled data, and obtaining a mapping relationship between the labeled data and the label after the preset machine model is trained; Determine whether new vehicle operation data is received. If the new vehicle operation data is received, perform data preprocessing on the new vehicle operation data, and input the new vehicle operation data after data preprocessing and the unlabeled data of the vehicle as input samples into the trained preset machine model, so as to use the trained preset machine model to obtain a prediction result of the power feeding reason of the vehicle based on the mapping relationship between the label data and the label.

2. The method according to claim 1, characterized in that: The collecting of vehicle operation data includes: The battery voltage data, battery current data, battery temperature data, electrical system status data of the vehicle, environmental data of the vehicle and user feedback data of the vehicle are collected.

3. The method according to claim 1, characterized in that: The performing data preprocessing on the operation data to obtain the data to be labeled of the vehicle includes: Performing data cleaning on the operation data to remove abnormal data of the operation data and fill in missing data of the operation data; The cleaned operating data is normalized, and feature extraction is performed on the normalized operating data to obtain the data to be labeled of the vehicle.

4. The method according to claim 1, characterized in that: The step of analyzing the data to be labeled and selecting target labeled data that meets preset conditions for labeling includes: Identifying the confidence level of the data to be labeled; Determine whether there is target annotated data whose confidence is less than a preset confidence threshold in the confidence of the data to be annotated; If there is target annotated data whose confidence is less than the preset confidence threshold value in the confidence of the data to be annotated, it is determined that the target annotated data meets the preset condition, and the target annotated data is annotated.

5. The method according to claim 1, characterized in that: The step of training a preset machine model based on the label data includes: Dividing the labeled data to obtain a training set, a validation set, and a test set of the labeled data; Setting initial parameters of the preset machine model, and training the preset machine model using the training set of the label data to obtain a training result of the preset machine model; Based on the training results, the initial parameters of the preset machine model are adjusted using the validation set to optimize the learning performance of the preset machine model, and the learning performance of the preset machine model is evaluated based on the test set.

6. A device for detecting the cause of battery feeding in a vehicle, characterized in that: include: A collection module, used to collect operating data of the vehicle, wherein the operating data includes operating data of the vehicle in a normal driving state and operating data of the vehicle in a power feeding state; A labeling module, used to perform data preprocessing on the operating data to obtain the data to be labeled of the vehicle, perform data analysis on the data to be labeled, select target labeled data that meets preset conditions for labeling, obtain the labeled data of the vehicle and the unlabeled data of the vehicle, train a preset machine model based on the labeled data, and obtain a mapping relationship between the labeled data and the label after the preset machine model is trained; An acquisition module is used to determine whether new vehicle operation data is received. If the new vehicle operation data is received, data preprocessing is performed on the new vehicle operation data, and the new vehicle operation data after data preprocessing and the unlabeled data of the vehicle are input as input samples to the trained preset machine model, so as to use the trained preset machine model to obtain the power feeding reason prediction result of the vehicle based on the mapping relationship between the label data and the label.

7. The device according to claim 6, characterized in that: The acquisition module is specifically used for: The battery voltage data, battery current data, battery temperature data, electrical system status data of the vehicle, environmental data of the vehicle and user feedback data of the vehicle are collected.

8. The device according to claim 6, characterized in that: The annotation module is specifically used for: Performing data cleaning on the operation data to remove abnormal data of the operation data and fill in missing data of the operation data; The cleaned operating data is normalized, and feature extraction is performed on the normalized operating data to obtain the data to be labeled of the vehicle.

9. A vehicle, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method for detecting the cause of battery power feeding of a vehicle as described in any one of claims 1 to 5.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the method for detecting the reason for battery feeding of a vehicle as described in any one of claims 1 to 5.

Images