Vehicle-mounted air conditioner abnormal shutdown prediction method and system, electronic equipment and storage medium
By acquiring and aggregating the target signals of the vehicle and using the vehicle-mounted air conditioner abnormal shutdown prediction model for prediction, the problem of the in-house air conditioner abnormal shutdown in the existing technology is solved, and accurate prediction and early warning is achieved, and user experience is improved.
Patent Information
- Application Number
- CN202411334575.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-24
- Publication Date
- 2024-12-20
AI Technical Summary
The prior art cannot directly predict whether the vehicle air conditioner will be shut down abnormally, resulting in the inability to effectively monitor and early warning.
By obtaining various target signals reported by the vehicle within the preset time period, aggregating the data to obtain descriptive statistics, and making predictions based on the trained vehicle air conditioner abnormal shutdown prediction model, including binary and multi-classification models to determine downtime risks and exception categories.
Accurate prediction of abnormal shutdown of vehicle air conditioners is achieved, prediction efficiency is improved, data is reduced, and user experience is improved and maintenance is timely.
Smart Images
Figure CN119167178A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of artificial intelligence technology, and in particular to a method, system, electronic device and storage medium for predicting abnormal shutdown of a vehicle air conditioner. Background Art
[0002] As a frequently used device in the car, the importance of car air conditioning in modern cars is self-evident. Therefore, the user experience of car air conditioning will greatly affect the user's evaluation of the quality of the car. For automobile manufacturers or car air conditioning manufacturers, it is necessary to actively monitor the operating status of the car air conditioning and actively issue risk warnings before the car air conditioning stops abnormally to improve user satisfaction.
[0003] At present, the real-time data of the relevant indicators of the compressor of the vehicle air conditioner is usually collected, and then the real-time data is classified using a rule system based on expert experience to determine the fault type of the vehicle air conditioner. However, this method relies on expert experience, resulting in low classification accuracy. In addition, this method classifies the fault type of the vehicle air conditioner, and does not directly predict whether the vehicle air conditioner will shut down abnormally, resulting in the inability to monitor whether the vehicle air conditioner will shut down abnormally. Summary of the invention
[0004] The present application provides a method, system, electronic device and storage medium for predicting abnormal shutdown of a vehicle air conditioner, so as to solve the problem in the prior art that whether the vehicle air conditioner will shut down abnormally is not directly predicted, resulting in the inability to monitor whether the vehicle air conditioner will shut down abnormally.
[0005] The present application provides a method for predicting abnormal shutdown of a vehicle air conditioner, the method comprising: obtaining target information of multiple vehicles, the target information comprising multiple types of target signals reported by the vehicles within a preset time period, the target signal being a signal pre-selected from the signals reported by the vehicles for predicting abnormal shutdown of the vehicle air conditioner;
[0006] Performing aggregation processing on each type of the target signal respectively to obtain a descriptive statistic, wherein the descriptive statistic is used to describe the characteristics or distribution of any type of the target signal within a preset time period, and the descriptive statistic corresponds to the type of the target signal;
[0007] Based on the descriptive statistics and the trained vehicle air conditioner abnormal shutdown prediction model, the vehicle air conditioner abnormal shutdown prediction is performed.
[0008] In one embodiment of the present application, the step of performing aggregation processing on each type of the target signal to obtain descriptive statistics includes:
[0009] Obtain the lower quartile, median, upper quartile, interquartile range, standard deviation, variance, range and mean of the target signal of any type within a preset time period, wherein the interquartile range is the difference between the upper quartile and the lower quartile, and the range is the difference between the maximum value and the minimum value of the target signal of the current type within the preset time period, and the target signal is a numerical signal;
[0010] The lower quartile, the median, the upper quartile, the interquartile range, the standard deviation, the variance, the range and the mean are determined as descriptive statistics of the target signal of the current category; and the target signals of the remaining categories are traversed to obtain the descriptive statistics of the target signals of the remaining categories.
[0011] In one embodiment of the present application, based on the descriptive statistics and the trained vehicle air conditioner abnormal shutdown prediction model, the step of predicting the vehicle air conditioner abnormal shutdown includes:
[0012] Input the descriptive statistics into the binary classification model of the trained vehicle air conditioner abnormal shutdown prediction model to predict the vehicle air conditioner abnormal shutdown and obtain a prediction result, wherein the prediction result is whether the vehicle air conditioner of the current vehicle has an abnormal shutdown risk or has no abnormal shutdown risk;
[0013] When the prediction result is that the vehicle air conditioner of the current vehicle is at risk of shutdown, the descriptive statistics are input into the multi-classification model in the vehicle air conditioner abnormal shutdown prediction model to perform vehicle air conditioner abnormality classification to obtain abnormal classification results; the prediction result and the abnormal classification results are both used for abnormal shutdown warning of the vehicle air conditioner.
[0014] In one embodiment of the present application, the training steps of the vehicle air conditioner abnormal shutdown prediction model include:
[0015] Obtain a training set, the training set comprising: evenly distributed positive sample data and negative sample data, the positive sample data comprising: target information samples reported by vehicles without abnormal shutdown of the vehicle air conditioner, and corresponding positive sample labels, the negative sample data comprising: target information samples reported by vehicles with abnormal shutdown of the vehicle air conditioner, corresponding negative sample labels, and abnormal category labels;
[0016] Performing aggregation processing on the positive sample data and the negative sample data respectively to obtain a first descriptive statistic sample of each type of signal in the positive sample data and a second descriptive statistic sample of each type of signal in the negative sample data;
[0017] Inputting the first descriptive statistics sample and the second descriptive statistics sample into the binary classification model of the vehicle air conditioner abnormal shutdown prediction model to perform vehicle air conditioner abnormal shutdown prediction, and obtaining an output result of the binary classification model;
[0018] The binary classification model is trained based on the output result and the corresponding sample label, where the sample label is the positive sample label or the negative sample label.
[0019] In one embodiment of the present application, the training step of the vehicle air conditioner abnormal shutdown prediction model further includes:
[0020] Acquire a negative sample data set, wherein the negative sample data set includes a plurality of negative sample data, and abnormal category labels of the negative sample data in the negative sample data set are evenly distributed;
[0021] Inputting the negative sample data into the multi-classification model of the vehicle air conditioner abnormal shutdown prediction model to perform vehicle air conditioner abnormality classification to obtain an abnormality category prediction result;
[0022] The multi-classification model is trained based on the abnormal category prediction results and the corresponding abnormal category labels.
[0023] In one embodiment of the present application, the binary classification model and the multi-classification model are both decision tree models;
[0024] The abnormality category prediction results include: startup failure, high temperature abnormality, load voltage overvoltage, load voltage undervoltage, excessive load and abnormal speed.
[0025] In one embodiment of the present application, the step of obtaining target information of multiple vehicles includes:
[0026] The target information of multiple vehicles is obtained from a data warehouse. The target information is reported to the cloud platform by the vehicles and stored in the data warehouse by the cloud platform.
[0027] The present application also provides a vehicle air conditioner abnormal shutdown prediction system, the system comprising: a target information acquisition module, used to acquire target information of multiple vehicles, the target information comprising multiple types of target signals reported by the vehicles within a preset time period, the target signal being a signal pre-selected from the signals reported by the vehicles for predicting abnormal shutdown of the vehicle air conditioner;
[0028] A data processing module, used for performing aggregation processing on each type of the target signal to obtain descriptive statistics, wherein the descriptive statistics are used to describe the characteristics or distribution of any type of the target signal within a preset time period, and the descriptive statistics correspond to the type of the target signal;
[0029] The prediction module is used to predict the abnormal shutdown of the vehicle air conditioner based on the descriptive statistics and the trained abnormal shutdown prediction model of the vehicle air conditioner.
[0030] The present application also provides an electronic device, comprising a processor, a memory and a communication bus; the communication bus is used to connect the processor and the memory; the processor is used to execute a computer program stored in the memory to implement a method provided in any one of the above embodiments.
[0031] The present application also provides a computer-readable storage medium on which a computer program is stored, wherein the computer program is used to enable a computer to execute the method provided in any of the above embodiments.
[0032] Beneficial effects of the present application: The present application proposes a method, system, electronic device and storage medium for predicting abnormal shutdown of vehicle air conditioner. The method obtains target information of multiple vehicles, and the target information includes multiple types of target signals reported by the vehicle within a preset time period. The target signal is a signal pre-selected from the signals reported by the vehicle for predicting abnormal shutdown of the vehicle air conditioner; each type of target signal is aggregated to obtain a descriptive statistic, which is used to describe the characteristics or distribution of any type of target signal within a preset time period, and the descriptive statistic corresponds to the type of the target signal; based on the descriptive statistic and the trained vehicle air conditioner abnormal shutdown prediction model, the vehicle air conditioner abnormal shutdown prediction is performed. The method can better realize the abnormal shutdown prediction of the vehicle air conditioner, facilitate the relevant personnel to intervene in the vehicle air conditioner maintenance in time, and avoid the problem of poor user experience caused by abnormal shutdown of the vehicle air conditioner. In addition, the method can also solve the problem of too much data in the prediction process, that is, by aggregating each type of target signal, a descriptive statistic is obtained, and based on the descriptive statistic and the vehicle air conditioner abnormal shutdown prediction model, the vehicle air conditioner abnormal shutdown prediction is performed. Compared with the large number of target signals in the preset time period, descriptive statistics can reduce the amount of data to a large extent. For example: assuming that there are n types of target signals reported by the vehicle in the preset time period, and the reporting frequency of each target signal in the preset time period is 1 second / time, then the number of target signals of any type is 1×the number of seconds in the target time period, and the total number of target signals is 1×the number of seconds in the target time period×n. This method can reduce the amount of data to a large extent and improve the prediction efficiency by obtaining the descriptive statistics of each type of target signal and using the descriptive statistics as the basis for the subsequent prediction of abnormal shutdown of the vehicle air conditioner. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] Figure 1 A flow chart of a method for predicting abnormal shutdown of a vehicle air conditioner provided in one embodiment of the present application;
[0034] Figure 2 A schematic diagram of a process for reporting data in a method for predicting abnormal shutdown of an on-board air conditioner provided in an embodiment of the present application;
[0035] Figure 3 A schematic diagram of a two-stage prediction process in a vehicle air conditioner abnormal shutdown prediction method provided in an embodiment of the present application;
[0036] Figure 4 A schematic diagram of the structure of a vehicle air conditioner abnormal shutdown prediction system provided in one embodiment of the present application;
[0037] Figure 5 A schematic diagram of the overall prediction process of a vehicle air conditioner abnormal shutdown prediction system provided in one embodiment of the present application;
[0038] Figure 6 A schematic diagram of the structure of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION
[0039] The following describes the embodiments of the present application through specific examples, and those skilled in the art can easily understand other advantages and effects of the present application from the contents disclosed in this specification. The present application can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.
[0040] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present application, and thus the drawings only show components related to the present application rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed at will, and the component layout may also be more complicated.
[0041] In the following description, a large number of details are discussed to provide a more thorough explanation of the embodiments of the present application. However, it is obvious to those skilled in the art that the embodiments of the present application can be implemented without these specific details. In other embodiments, well-known structures and devices are shown in the form of block diagrams rather than in detail to avoid making the embodiments of the present application difficult to understand.
[0042] Combine the following Figures 1 to 6 , the vehicle air conditioner abnormal shutdown prediction method, system, electronic device and storage medium provided in this application are explained.
[0043] See also Figure 1 , Figure 1A flow chart of a method for predicting abnormal shutdown of a vehicle air conditioner provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, the method includes:
[0044] S110: Acquire target information of multiple vehicles, wherein the target information includes multiple types of target signals reported by the vehicles within a preset time period, wherein the target signals are pre-selected signals from the signals reported by the vehicles and are used to predict abnormal shutdown of the vehicle air conditioner.
[0045] It should be noted that vehicles, such as new energy vehicles, are equipped with a large number of sensors on their bodies. Each sensor reports the collected signal to the cloud platform through the data reporting system on the vehicle side (the data reporting system includes a data bus and a data sending device, etc.) via the Internet of Things center, and then the cloud platform stores the signal in the data warehouse. Figure 2 As shown, sensors, data reporting system, IoT center, cloud platform and data warehouse are connected in sequence.
[0046] Since there are many sensors in the vehicle, there are thousands of types of signals reported by the sensors. In order to improve the accuracy of the prediction of abnormal shutdown of the vehicle air conditioner, this embodiment makes full use of the signals collected by the vehicle sensors, that is, multiple types of signals are selected from the large number of signals reported by the vehicle as target signals, and the target signals are used as the basis for subsequent prediction of abnormal shutdown of the vehicle air conditioner. This method can firstly improve the utilization rate of the signals reported by the vehicle, and secondly, it can make the dimensions of the basis used when judging whether the vehicle air conditioner will shut down abnormally more comprehensive, which helps to improve the accuracy of the prediction of abnormal shutdown of the vehicle air conditioner. Compared with the method of using signals with fewer dimensions as the basis for predicting abnormal shutdown of the vehicle air conditioner, the method of obtaining target information in this embodiment is more helpful to improve the accuracy of the prediction of abnormal shutdown of the vehicle air conditioner.
[0047] Taking new energy vehicles as an example, there are 149 target signals (thermal related signals) for vehicles, which can be used to predict abnormal shutdown of vehicle air conditioners. The target signal list is shown in Table 1 below:
[0048] Table 1 List of target signals
[0049]
[0050] As shown in Table 1, the target signals include: signals involved in information exchange of thermal management or temperature control systems related to vehicle control units (such as cooling fan status signals, engine operation status signals, and heater control signals, etc., 29 signals), signals related to the status information of the central control unit (such as engine speed signals and vehicle speed signals, etc., 23 signals), signals of battery-related sensors (such as battery pack temperature signals, etc., 16 signals), temperature signals detected by temperature controllers (such as air conditioning outlet temperature signals, transmission system oil temperature signals, and engine exhaust temperature signals, etc., 7 signals), signals related to the air conditioning system (such as air conditioning compressor status signals and fan control signals, etc., 54 signals), signals related to the body control module status (such as steering angle signals, brake status signals, and door status signals, etc., 8 signals), and setting signals of the in-vehicle infotainment system (such as volume control signals, sound source selection signals, and frequency adjustment signals, etc., 12 signals). There are a total of 149 target signals, and the data format of each signal is numerical.
[0051] By acquiring the above 149 types of target signals, the dimension of the target signal can be well expanded, which helps to improve the accuracy of subsequent abnormal shutdown prediction of the vehicle air conditioner. It can be understood that from the perspective of abnormal shutdown prediction of the vehicle air conditioner, one type of target signal is equivalent to one dimension. By acquiring the above multiple types of target signals, it can help to predict the abnormal shutdown of the vehicle air conditioner.
[0052] It should be mentioned that the preset time period can be set according to actual conditions, such as the previous day, the previous 2 days, etc. Assuming that the preset time period is the previous day, and the vehicle's sensor uploads the collected signal once every 1 second, the size of the data finally obtained from the data warehouse is 149×1×24×60×60=12873600.
[0053] S120: performing aggregation processing on each type of the target signal respectively to obtain a descriptive statistic, wherein the descriptive statistic is used to describe the characteristics or distribution of any type of the target signal within a preset time period, and the descriptive statistic corresponds to the type of the target signal.
[0054] It should be noted that, in order to reduce the amount of data, this embodiment obtains the corresponding descriptive statistics by performing aggregation processing on each type of target signal. Also assuming that the preset time period is the previous day, the vehicle's sensor uploads the collected signal once every 1 second. Based on this, the data format of the target signal before aggregation processing is shown in Table 2 below:
[0055] Table 2 Data format of target signal before aggregation processing
[0056]
[0057] As shown in Table 2, before the aggregation process, the data format of the target signal includes vehicle identification, data generation time and date, etc., and its data size is 149×1×24×60×60.
[0058] After aggregation processing, the data format is as shown in Table 3 below:
[0059] Table 3 Data format of target signal after aggregation processing
[0060]
[0061] Among them, m represents the size of the descriptive statistics corresponding to each type of target signal. Assuming that there are 8 descriptive statistics corresponding to each type of target signal, the size of the data after aggregation processing is 149×8=1192, etc. Obviously, this method greatly reduces the amount of data.
[0062] S130: Based on the descriptive statistics and the trained vehicle air conditioner abnormal shutdown prediction model, perform vehicle air conditioner abnormal shutdown prediction.
[0063] It should be noted that the vehicle air conditioner abnormal shutdown prediction model is a model based on machine learning. By using machine learning to predict the abnormal shutdown of the vehicle air conditioner, the prediction accuracy can be effectively improved. Compared with the rule model, the vehicle air conditioner abnormal shutdown prediction model uses machine learning methods to more easily mine non-intuitive relationships between features, automatically perform feature learning, have strong adaptability and generalization capabilities, and low cost. Through the above steps, the prediction of abnormal shutdown of the vehicle air conditioner can be better realized, so that relevant personnel can go to the door in time to repair the vehicle air conditioner, avoid abnormal shutdown of the vehicle air conditioner, and effectively improve the user experience.
[0064] In some embodiments, the step of performing aggregation processing on each type of the target signal to obtain descriptive statistics includes:
[0065] 1. Obtain the lower quartile, median, upper quartile, interquartile range, standard deviation, variance, range and mean of the target signal of any type within a preset time period, wherein the interquartile range is the difference between the upper quartile and the lower quartile, and the range is the difference between the maximum and minimum values of the target signal of the current type within the preset time period, and the target signal is a numerical signal.
[0066] It should be noted that the lower quantile is determined by arranging any type of target signal from small to large according to its value, and determining the value at the 25% position as the lower quantile of the target signal of the current type. The median refers to the value in the middle position after arranging any type of target signal from small to large according to its value. The upper quartile is determined by arranging any type of target signal from small to large according to its value, and determining the value at the 75% position as the upper quartile of the target signal of the current type.
[0067] 2. Determine the lower quartile, the median, the upper quartile, the interquartile range, the standard deviation, the variance, the range and the mean as the descriptive statistics of the target signal of the current type; traverse the target signals of the remaining types to obtain the descriptive statistics of the target signals of the remaining types.
[0068] It should be noted that by obtaining the lower quartile, median, upper quartile, interquartile range, standard deviation, variance, range and mean of each type of target signal through statistics, and determining the lower quartile, median, upper quartile, interquartile range, standard deviation, variance, range and mean as descriptive statistics, it can help in the subsequent prediction of abnormal shutdown of vehicle air conditioners and effectively reduce the amount of data.
[0069] In some embodiments, based on the descriptive statistics and the trained vehicle air conditioner abnormal shutdown prediction model, the step of predicting the vehicle air conditioner abnormal shutdown includes:
[0070] 1. Input the descriptive statistics into the binary classification model of the trained vehicle air conditioner abnormal shutdown prediction model to predict the vehicle air conditioner abnormal shutdown and obtain a prediction result, wherein the prediction result is whether the vehicle air conditioner of the current vehicle has an abnormal shutdown risk or no abnormal shutdown risk.
[0071] 2. When the prediction result indicates that the vehicle air conditioner of the current vehicle is at risk of shutdown, the descriptive statistics are input into the multi-classification model in the vehicle air conditioner abnormal shutdown prediction model to perform vehicle air conditioner abnormality classification to obtain abnormal classification results; the prediction result and the abnormal classification results are both used for abnormal shutdown warning of the vehicle air conditioner.
[0072] It should be noted that the above embodiment divides the prediction process of the vehicle air conditioner abnormal shutdown prediction model into two stages, one is binary classification and the other is multi-classification. By first performing binary classification and then multi-classification, after determining that the vehicle's vehicle air conditioner has an abnormal shutdown risk, the abnormal category of the vehicle can be classified with high accuracy.
[0073] Figure 3The following is a flow chart of two-stage prediction in the method for predicting abnormal shutdown of a vehicle air conditioner provided in one embodiment of the present application. Figure 3 As shown in the figure, after obtaining the descriptive statistics, the descriptive statistics are input into the binary classification model to predict the abnormal shutdown of the vehicle air conditioner and obtain the prediction result. After that, it is determined whether there will be an abnormal shutdown, that is, whether there is an abnormal shutdown risk. If there is, the descriptive statistics are input into the multi-classification model to perform abnormal classification of the vehicle air conditioner and obtain the abnormal classification result. If not, no processing is performed to save computing resources.
[0074] In some embodiments, the method further includes: when the prediction result indicates that the vehicle air conditioner of the current vehicle has a risk of abnormal shutdown, issuing a warning message, wherein the warning message is used to instruct relevant personnel to perform on-site maintenance to improve user experience.
[0075] In some embodiments, the training step of the vehicle air conditioner abnormal shutdown prediction model includes:
[0076] First, a training set is obtained, which includes: evenly distributed positive sample data and negative sample data, the positive sample data includes: target information samples reported by vehicles without abnormal shutdown of the vehicle air conditioner, and corresponding positive sample labels, the negative sample data includes: target information samples reported by vehicles with abnormal shutdown of the vehicle air conditioner, corresponding negative sample labels, and abnormal category labels.
[0077] It should be noted that the data format of the positive sample data and the negative sample data is shown in Table 4 below:
[0078] Table 4 Data format of sample data
[0079]
[0080] In Table 4, it is assumed that the target information sample obtained is the target signal uploaded to the data warehouse by the vehicle the day before. In actual operation, it can also be the target signal uploaded 0.5 days ago, 2 days ago, 3 days ago, etc., which will not be repeated here. When the sample label value is 0, it means that the sample label is a positive sample label, that is, there is no abnormal data. When the sample label value is 1, it means that the sample label is a negative sample label, that is, there is abnormal data. When the abnormal category label is 0, it means no abnormality. When the abnormal category label is 1, it indicates a startup failure. When the abnormal category label is 2, it indicates a high temperature abnormality. When the abnormal category label is 3, it indicates that the load voltage is overvoltage. When the abnormal category label is 4, it indicates that the load voltage is undervoltage. When the abnormal category label is 5, it indicates that the load is too large. When the abnormal category label is 6, it indicates abnormal speed.
[0081] It should be mentioned that the target information sample includes multiple types of target signals reported by the vehicle within a preset time period, such as 149 types of target signals reported by the vehicle the previous day.
[0082] Secondly, the positive sample data and the negative sample data are respectively aggregated to obtain a first descriptive statistic sample of each type of signal in the positive sample data and a second descriptive statistic sample of each type of signal in the negative sample data.
[0083] Specifically, the first descriptive statistics sample includes the lower quantile, median, upper quartile, interquartile range, standard deviation, variance, range and mean of the target signal in the positive sample data. The second descriptive statistics sample includes the lower quantile, median, upper quartile, interquartile range, standard deviation, variance, range and mean of the target signal in the negative sample data. The data format of the first descriptive statistics sample and the second descriptive statistics sample is shown in Table 5 below:
[0084] Table 5 Data format of descriptive statistics samples
[0085]
[0086] It should be noted that, through the above aggregation processing, the first descriptive statistics samples and the second descriptive statistics samples obtained can be divided into training data and test data in equal proportion according to the sample labels. That is, the first descriptive statistics samples and the second descriptive statistics samples in the training data are in equal proportion, and similarly, the first descriptive statistics samples and the second descriptive statistics samples in the test data are in equal proportion. The data ratio of training data to test data can be 8:2, etc.
[0087] Then, the first descriptive statistic sample and the second descriptive statistic sample are input into the binary classification model of the vehicle air conditioner abnormal shutdown prediction model to perform vehicle air conditioner abnormal shutdown prediction and obtain an output result of the binary classification model.
[0088] Specifically, the first descriptive statistic sample and the second descriptive statistic sample in the training data are input into the binary classification model of the vehicle air conditioner abnormal shutdown prediction model to perform vehicle air conditioner abnormal shutdown prediction and obtain the output result of the binary classification model.
[0089] It should be noted that the binary classification model in this embodiment is a decision tree model. The decision tree model uses a search method to find the optimal parameters of the decision tree, and divides the nodes by calculating the information entropy, and finally divides the data into positive classes (no abnormal downtime risk) or negative classes (there is an abnormal downtime risk) at the leaf nodes to obtain the output results. In the specific implementation process, other neural network models can also be used as binary classification models.
[0090] Finally, the binary classification model is trained based on the output results and the corresponding sample labels, where the sample labels are the positive sample labels or the negative sample labels.
[0091] It should be noted that by adopting the above method, the training of the binary classification model can be completed well with high accuracy.
[0092] In some embodiments, the training step of the vehicle air conditioner abnormal shutdown prediction model further includes:
[0093] 1. Obtain a negative sample data set, wherein the negative sample data set includes a plurality of negative sample data, and the abnormal category labels of the negative sample data in the negative sample data set are evenly distributed.
[0094] 2. Inputting the negative sample data into the multi-classification model of the vehicle air conditioner abnormal shutdown prediction model to perform vehicle air conditioner abnormality classification to obtain abnormal category prediction results.
[0095] It should be noted that the multi-classification model in this embodiment also uses a decision tree model to improve the accuracy of classification. In the specific embodiment, other neural network models can also be used as multi-classification models.
[0096] 3. Based on the abnormal category prediction results and the corresponding abnormal category labels, the multi-classification model is trained.
[0097] It should be noted that since the abnormal shutdown of the vehicle air conditioner is an occasional event, the probability of its occurrence is small. Therefore, in the actual process of collecting sample data, most of the collected data are positive sample data, and there are fewer negative sample data, and there is a more serious problem of uneven distribution of positive and negative samples. Therefore, in this embodiment, a two-stage prediction method is used to solve the above-mentioned uneven distribution of positive and negative samples. Specifically, in the first stage, the binary classification model is trained using evenly distributed positive and negative sample data. Since the difficulty of the binary classification is relatively low, the binary classification model can be trained with a relatively small amount of evenly distributed positive and negative sample data. In the second stage, the multi-classification model is trained using negative sample data with evenly distributed abnormal category labels, and the process does not involve positive sample data. Compared with the one-step training method (inputting the training set into the vehicle air conditioner abnormal shutdown prediction model, directly performing vehicle air conditioner abnormal shutdown prediction and vehicle air conditioner abnormal classification, that is, not performing staged prediction), the above-mentioned two-stage prediction method in this embodiment can better avoid problems such as model overfitting caused by uneven distribution of positive and negative samples. It helps to improve the prediction accuracy and has low cost.
[0098] In some embodiments, the abnormality category prediction results include: startup failure, high temperature abnormality, load voltage overvoltage, load voltage undervoltage, excessive load, and abnormal speed. It should be noted that the above abnormality category prediction results basically cover the abnormality categories of the vehicle air conditioner and are relatively comprehensive.
[0099] In some embodiments, the step of acquiring target information of multiple vehicles includes:
[0100] The target information of multiple vehicles is obtained from the data warehouse, and the target information is reported by the vehicle to the cloud platform and stored in the data warehouse by the cloud platform. It should be noted that by obtaining the target information of multiple vehicles from the data warehouse, multi-dimensional target information can be obtained, making subsequent predictions more accurate, and effectively improving the utilization rate of data in the data warehouse, which is more convenient to implement.
[0101] In some embodiments, it also includes: receiving feedback data, that is, data related to whether the prediction results and abnormal classification results are correct. Based on the feedback data, various indicators of the prediction results and abnormal classification results are obtained, such as recall rate (the probability of being predicted as a positive sample among the samples that are actually positive), precision rate (the percentage of the predicted correct results in the total samples), accuracy rate (the probability of being a positive sample among all the samples predicted as positive), and F1 value (the harmonic mean of recall rate and accuracy), etc. Based on the feedback data and these indicators, the abnormal shutdown prediction model of the vehicle air conditioner is optimized.
[0102] The vehicle air conditioner abnormal shutdown prediction system provided by the present application is described below. The vehicle air conditioner abnormal shutdown prediction system described below and the vehicle air conditioner abnormal shutdown prediction method described above can be referenced to each other.
[0103] Please refer to Figure 4 The vehicle air conditioner abnormal shutdown prediction system provided in this embodiment includes:
[0104] The target information acquisition module 410 is used to acquire target information of multiple vehicles, wherein the target information includes multiple types of target signals reported by the vehicles within a preset time period, wherein the target signals are pre-selected signals from the signals reported by the vehicles and are used to predict abnormal shutdown of the vehicle air conditioner;
[0105] A data processing module 420 is used to perform aggregation processing on each type of the target signal to obtain a descriptive statistic, wherein the descriptive statistic is used to describe the characteristics or distribution of any type of the target signal within a preset time period, and the descriptive statistic corresponds to the type of the target signal;
[0106] The prediction module 430 is used to predict the abnormal shutdown of the vehicle air conditioner based on the descriptive statistics and the trained abnormal shutdown prediction model of the vehicle air conditioner. The target information acquisition module 410, the data processing module 420 and the prediction module 430 are connected. The vehicle air conditioner abnormal shutdown prediction system in this embodiment can not only realize the abnormal shutdown prediction of the vehicle air conditioner, but also solve the problem of excessive data volume in the prediction process, with low cost and strong feasibility.
[0107] In some embodiments, the data processing module 420 is specifically used to obtain the lower quartile, median, upper quartile, interquartile range, standard deviation, variance, range and mean of any type of the target signal within a preset time period, the interquartile range is the difference between the upper quartile and the lower quartile, the range is the difference between the maximum value and the minimum value of the target signal of the current type within the preset time period, and the target signal is a numerical signal;
[0108] The lower quartile, the median, the upper quartile, the interquartile range, the standard deviation, the variance, the range and the mean are determined as descriptive statistics of the target signal of the current category; and the target signals of the remaining categories are traversed to obtain the descriptive statistics of the target signals of the remaining categories.
[0109] In some embodiments, the prediction module 430 is specifically used to input the descriptive statistics into the binary classification model of the trained vehicle air conditioner abnormal shutdown prediction model to predict the vehicle air conditioner abnormal shutdown and obtain a prediction result, wherein the prediction result is that the vehicle air conditioner of the current vehicle has an abnormal shutdown risk or no abnormal shutdown risk;
[0110] When the prediction result is that the vehicle air conditioner of the current vehicle is at risk of shutdown, the descriptive statistics are input into the multi-classification model in the vehicle air conditioner abnormal shutdown prediction model to perform vehicle air conditioner abnormality classification to obtain abnormal classification results; the prediction result and the abnormal classification results are both used for abnormal shutdown warning of the vehicle air conditioner.
[0111] like Figure 5 As shown, in some embodiments, it also includes: a model training module 510, which is used to obtain a training set, wherein the training set includes: evenly distributed positive sample data and negative sample data, wherein the positive sample data includes: target information samples reported by vehicles without abnormal shutdown of the vehicle air conditioner, and corresponding positive sample labels, and the negative sample data includes: target information samples reported by vehicles with abnormal shutdown of the vehicle air conditioner, corresponding negative sample labels, and abnormal category labels;
[0112] Performing aggregation processing on the positive sample data and the negative sample data respectively to obtain a first descriptive statistic sample of each type of signal in the positive sample data and a second descriptive statistic sample of each type of signal in the negative sample data;
[0113] Inputting the first descriptive statistics sample and the second descriptive statistics sample into the binary classification model of the vehicle air conditioner abnormal shutdown prediction model to perform vehicle air conditioner abnormal shutdown prediction, and obtaining an output result of the binary classification model;
[0114] The binary classification model is trained based on the output result and the corresponding sample label, where the sample label is the positive sample label or the negative sample label.
[0115] In some embodiments, the model training module 510 is further used to obtain a negative sample data set, wherein the negative sample data set includes a plurality of negative sample data, and the abnormal category labels of the negative sample data in the negative sample data set are evenly distributed;
[0116] Inputting the negative sample data into the multi-classification model of the vehicle air conditioner abnormal shutdown prediction model to perform vehicle air conditioner abnormality classification to obtain an abnormality category prediction result;
[0117] The multi-classification model is trained based on the abnormal category prediction results and the corresponding abnormal category labels.
[0118] In some embodiments, the target information acquisition module 410 is specifically used to obtain the target information of multiple vehicles from a data warehouse. The target information is reported by the vehicles to the cloud platform and stored in the data warehouse by the cloud platform.
[0119] Figure 5 This is a schematic diagram of the overall prediction process of the vehicle air conditioner abnormal shutdown prediction system provided by an embodiment of the present application. Figure 5 As shown, first, the model training module 510 trains the vehicle air conditioner abnormal shutdown prediction model. Secondly, the trained vehicle air conditioner abnormal shutdown prediction model is deployed to the prediction module 430. Then, the target information acquisition module 410 acquires the target information of multiple vehicles. After that, the data processing module 420 performs aggregation processing on the target signals of each type respectively to obtain descriptive statistics. Next, the prediction module 430 predicts the abnormal shutdown of the vehicle air conditioner based on the descriptive statistics and the trained vehicle air conditioner abnormal shutdown prediction model. The early warning notification module 520 then makes an early warning notification based on the prediction results, etc. Finally, the effect evaluation module 530 receives the feedback data, and optimizes the vehicle air conditioner abnormal shutdown prediction model based on the feedback data, that is, the model training module 510 retrains the model based on the feedback data. Through the above steps, a closed loop of vehicle air conditioner abnormal shutdown prediction is realized.
[0120] In some embodiments, an electronic device is also provided, which may be a server, and its internal structure is shown in FIG. Figure 6 As shown. The electronic device includes a processor, a memory, a network interface and a database connected via a system bus. The processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile and / or volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the electronic device is used to communicate with an external client via a network connection. When the computer program is executed by the processor, the functions or steps on the server side of the above method are implemented.
[0121] In some embodiments, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program: obtaining target information of multiple vehicles, the target information including multiple types of target signals reported by the vehicles within a preset time period, the target signal being a pre-selected signal from the signals reported by the vehicles for predicting abnormal shutdown of a vehicle air conditioner; performing aggregation processing on each type of target signal to obtain a descriptive statistic, the descriptive statistic being used to describe the characteristics or distribution of any type of target signal within a preset time period, the descriptive statistic corresponding to the type of the target signal; and predicting abnormal shutdown of the vehicle air conditioner based on the descriptive statistics and a trained abnormal shutdown prediction model for the vehicle air conditioner.
[0122] In some embodiments, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented: obtaining target information of multiple vehicles, the target information including multiple types of target signals reported by the vehicles within a preset time period, the target signal being a pre-selected signal from the signals reported by the vehicles for predicting abnormal shutdown of the vehicle air conditioner; performing aggregation processing on each type of target signal to obtain a descriptive statistic, the descriptive statistic is used to describe the characteristics or distribution of any type of target signal within a preset time period, and the descriptive statistic corresponds to the type of the target signal; and predicting abnormal shutdown of the vehicle air conditioner based on the descriptive statistics and a trained abnormal shutdown prediction model for the vehicle air conditioner.
[0123] It should be noted that the above functions or steps that can be implemented by the computer-readable storage medium or electronic device can refer to the relevant descriptions on the server side and the client side in the aforementioned method embodiment. To avoid repetition, they will not be described one by one here.
[0124] The flow chart and block diagram in the accompanying drawings illustrate the possible implementation architecture, function and operation of the method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.
[0125] The above embodiments are merely illustrative of the principles and effects of the present application and are not intended to limit the present application. Anyone familiar with the technology may modify or change the above embodiments without violating the spirit and scope of the present application. Therefore, all equivalent modifications or changes made by a person of ordinary skill in the art without departing from the spirit and technical ideas disclosed in the present application shall still be covered by the claims of the present application.
Claims
1. A method for predicting abnormal shutdown of a vehicle air conditioner, characterized in that: include: Acquiring target information of a plurality of vehicles, the target information comprising a plurality of types of target signals reported by the vehicles within a preset time period, the target signals being pre-selected signals from the signals reported by the vehicles for predicting abnormal shutdown of the vehicle air conditioner; Performing aggregation processing on each type of the target signal respectively to obtain a descriptive statistic, wherein the descriptive statistic is used to describe the characteristics or distribution of any type of the target signal within a preset time period, and the descriptive statistic corresponds to the type of the target signal; Based on the descriptive statistics and the trained vehicle air conditioner abnormal shutdown prediction model, the vehicle air conditioner abnormal shutdown prediction is performed.
2. The vehicle air conditioner abnormal shutdown prediction method according to claim 1, characterized in that: The step of performing aggregation processing on the target signal of each type to obtain descriptive statistics includes: Obtain the lower quartile, median, upper quartile, interquartile range, standard deviation, variance, range and mean of the target signal of any type within a preset time period, wherein the interquartile range is the difference between the upper quartile and the lower quartile, and the range is the difference between the maximum value and the minimum value of the target signal of the current type within the preset time period, and the target signal is a numerical signal; The lower quartile, the median, the upper quartile, the interquartile range, the standard deviation, the variance, the range and the mean are determined as descriptive statistics of the target signal of the current category; and the target signals of the remaining categories are traversed to obtain the descriptive statistics of the target signals of the remaining categories.
3. The vehicle air conditioner abnormal shutdown prediction method according to claim 1, characterized in that: Based on the descriptive statistics and the trained vehicle air conditioner abnormal shutdown prediction model, the steps of predicting the vehicle air conditioner abnormal shutdown include: Input the descriptive statistics into the binary classification model of the trained vehicle air conditioner abnormal shutdown prediction model to predict the vehicle air conditioner abnormal shutdown and obtain a prediction result, wherein the prediction result is whether the vehicle air conditioner of the current vehicle has an abnormal shutdown risk or has no abnormal shutdown risk; When the prediction result is that the vehicle air conditioner of the current vehicle is at risk of shutdown, the descriptive statistics are input into the multi-classification model in the vehicle air conditioner abnormal shutdown prediction model to perform vehicle air conditioner abnormality classification to obtain abnormal classification results; the prediction result and the abnormal classification results are both used for abnormal shutdown warning of the vehicle air conditioner.
4. The vehicle air conditioner abnormal shutdown prediction method according to any one of claims 1 to 3, characterized in that: The training steps of the vehicle air conditioner abnormal shutdown prediction model include: Obtain a training set, the training set comprising: evenly distributed positive sample data and negative sample data, the positive sample data comprising: target information samples reported by vehicles without abnormal shutdown of the vehicle air conditioner, and corresponding positive sample labels, the negative sample data comprising: target information samples reported by vehicles with abnormal shutdown of the vehicle air conditioner, corresponding negative sample labels, and abnormal category labels; Performing aggregation processing on the positive sample data and the negative sample data respectively to obtain a first descriptive statistic sample of each type of signal in the positive sample data and a second descriptive statistic sample of each type of signal in the negative sample data; Inputting the first descriptive statistics sample and the second descriptive statistics sample into the binary classification model of the vehicle air conditioner abnormal shutdown prediction model to perform vehicle air conditioner abnormal shutdown prediction, and obtaining an output result of the binary classification model; The binary classification model is trained based on the output result and the corresponding sample label, where the sample label is the positive sample label or the negative sample label.
5. The method for predicting abnormal shutdown of a vehicle air conditioner according to claim 4, characterized in that: The training step of the vehicle air conditioner abnormal shutdown prediction model also includes: Acquire a negative sample data set, wherein the negative sample data set includes a plurality of negative sample data, and abnormal category labels of the negative sample data in the negative sample data set are evenly distributed; Inputting the negative sample data into the multi-classification model of the vehicle air conditioner abnormal shutdown prediction model to perform vehicle air conditioner abnormality classification to obtain an abnormality category prediction result; The multi-classification model is trained based on the abnormal category prediction results and the corresponding abnormal category labels.
6. The method for predicting abnormal shutdown of a vehicle air conditioner according to claim 5, characterized in that: The binary classification model and the multi-classification model are both decision tree models; The abnormality category prediction results include: startup failure, high temperature abnormality, load voltage overvoltage, load voltage undervoltage, excessive load and abnormal speed.
7. The vehicle air conditioner abnormal shutdown prediction method according to claim 1, characterized in that: The steps of obtaining target information of multiple vehicles include: The target information of multiple vehicles is obtained from a data warehouse. The target information is reported to the cloud platform by the vehicles and stored in the data warehouse by the cloud platform.
8. A vehicle air conditioner abnormal shutdown prediction system, characterized in that: include: A target information acquisition module, used to acquire target information of multiple vehicles, wherein the target information includes multiple types of target signals reported by the vehicles within a preset time period, wherein the target signals are pre-selected signals from the signals reported by the vehicles and used to predict abnormal shutdown of the vehicle air conditioner; A data processing module, used for performing aggregation processing on each type of the target signal to obtain descriptive statistics, wherein the descriptive statistics are used to describe the characteristics or distribution of any type of the target signal within a preset time period, and the descriptive statistics correspond to the type of the target signal; The prediction module is used to predict the abnormal shutdown of the vehicle air conditioner based on the descriptive statistics and the trained abnormal shutdown prediction model of the vehicle air conditioner.
9. An electronic device, characterized in that: It comprises a processor, a memory and a communication bus; the communication bus is used to connect the processor and the memory; the processor is used to execute the computer program stored in the memory to implement the vehicle air conditioner abnormal shutdown prediction method as described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and the computer program is used to enable a computer to execute the vehicle air conditioner abnormal shutdown prediction method as described in any one of claims 1 to 7.