A method and storage medium for identifying abnormal vehicle wakeup caused by after-installed OBD equipment

Abstract

The present invention discloses a method and storage medium for identifying abnormal awakening of a vehicle caused by an after-installed OBD device. The method comprises acquiring vehicle data with a timestamp; performing abnormal value processing on the acquired vehicle data to obtain pre-processed data; screening suspicious vehicles that are abnormally awakened due to OBD; recording the start time and end time of the abnormal awakening of the suspicious vehicle; recording the time period in which the abnormal awakening occurs and the number of abnormal awakenings of the vehicle during the time period; calculating the average abnormal awakening duration of the time period in which the maximum number of abnormal awakenings of the suspicious vehicle occurs, and counting the mode of the abnormal awakening duration in the time period in which the maximum number of abnormal awakenings occurs; if the difference between the average abnormal awakening duration and the mode of the abnormal awakening duration is less than 1s, it is determined that the vehicle is abnormally awakened due to OBD. The present invention can effectively prevent the situation in which the vehicle is out of power due to after-installed equipment installed by the user, and reduce the probability of the vehicle being out of power.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle quality monitoring, and in particular to a method and a storage medium for identifying abnormal vehicle awakening caused by an aftermarket OBD device. Background Art

[0002] If the car is not in an extreme environment, such as extremely cold temperatures or high latitude and high altitude areas, the possibility of abnormal discharge of the battery after the car is turned off is small. However, once the car battery is low in power, its overall life will be greatly reduced, the user's cost of use will increase, and the user experience will be greatly reduced.

[0003] Generally speaking, the main reasons for a car's battery drain can be divided into the following categories: The first category is the user's car usage habits, such as using the car's computer to listen to music or watch movies after parking, or forgetting to turn off the lights after turning off the engine. The second category is leakage in the vehicle's own electrical equipment. The third category is equipment installed by the user or others. Among these devices, one category is power supply equipment, such as a driving recorder, and the other category is OBD interface equipment, such as a navigation host, GPS positioning, etc.

[0004] For the first type of problem, it is common because user habits are uncontrollable. However, most current car designs have power management modules, which will detect the current battery power status of the vehicle in real time before the user leaves the vehicle. If the power drops to a certain threshold, it will remind the user or directly cut it off. Therefore, the first type of problem has been effectively solved. For the second type of problem, before the vehicle is launched, the vehicle manufacturer will do a lot of testing and road tests under various working conditions, so the probability of occurrence is low. The third type of problem is also the user's own behavior, which is also uncontrollable and common. Since the after-installed equipment is not controlled by the vehicle system, there is currently no effective way to solve it. Generally, after the vehicle is out of power, professional after-sales maintenance personnel will check the power supply lines of the vehicle one by one to determine the problem. This is not only inefficient, but also cannot avoid the occurrence of power outages.

[0005] Therefore, if the situation in which the vehicle is abnormally awakened by aftermarket equipment can be effectively identified before the vehicle runs out of power, it can serve as an early warning, reminding the user to judge whether the vehicle has abnormal leakage behavior based on the abnormal awakening of the vehicle. This can effectively solve the third type of problem and greatly reduce the possibility of the vehicle running out of power. Summary of the invention

[0006] In view of the above-mentioned deficiencies in the prior art, the purpose of the present invention is to solve the problem in the prior art that the abnormal power consumption of the user's after-installed OBD device is not monitored, resulting in an increased possibility of vehicle power outage, and to provide a method and storage medium for identifying abnormal wake-up of the vehicle caused by after-installed OBD devices, which can monitor the abnormal power consumption of the user's after-installed OBD device, effectively prevent the vehicle from running out of power due to the user's after-installed device, and reduce the probability of vehicle power outage.

[0007] This can provide early warning of abnormal power consumption, reduce the probability of vehicle power outage, reduce user costs, and improve user experience.

[0008] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0009] A method for identifying abnormal vehicle wake-up caused by an after-installed OBD device comprises the following steps:

[0010] Acquire vehicle data with a timestamp; wherein the vehicle data includes a vehicle network sleep flag and a vehicle network wake-up flag;

[0011] Perform outlier processing on the acquired vehicle data, delete samples with invalid values, and fill the missing values ​​with the average value to obtain preprocessed data;

[0012] Based on the pre-processed data, the suspicious vehicles with abnormal wake-up caused by OBD are screened out according to the OBD high-frequency wake-up differentiation model;

[0013] Record the start and end time of the abnormal awakening of the suspicious vehicle, and calculate the duration of the abnormal awakening; at the same time, record the time period of the abnormal awakening and the number of abnormal awakenings of the vehicle during the time period;

[0014] Calculate the average abnormal awakening duration in the time period with the maximum number of abnormal awakenings of the suspicious vehicle, and count the mode of the abnormal awakening duration in the time period with the maximum number of abnormal awakenings;

[0015] If the difference between the average abnormal awakening duration and the mode of abnormal awakening duration is less than 1 s, the vehicle is determined to be a vehicle with abnormal awakening caused by OBD.

[0016] Furthermore, the abnormal awakening determination method is:

[0017] When the vehicle network sleep flag changes from 0 to 1, if the vehicle network wake-up flag changes from 0 to 1, it is determined that a wake-up occurs;

[0018] If there is no active awakening within 2 minutes before or after the awakening time, the current awakening is determined to be an abnormal awakening.

[0019] Furthermore, the acquired vehicle data with timestamps also includes unlocking signal, locking signal, driver's door lock status and anti-theft status. When the unlocking signal, locking signal and driver's door lock status change, and the anti-theft status changes from armed state to disarmed state, it is determined to be active wake-up.

[0020] Furthermore, the method for determining the end of abnormal wakeup is:

[0021] After a wake-up occurs, if the vehicle network sleep flag changes from 0 to 1, it is determined that the abnormal wake-up is over;

[0022] Record the end time of the current abnormal wakeup, and subtract the start time of the abnormal wakeup from the end time of the abnormal wakeup to get the duration of the abnormal wakeup; at the same time, record the time period of the abnormal wakeup, which is divided into one hour per time period, and one day is divided into 24 time periods.

[0023] In the specific implementation, vehicle data of 1,000 vehicles for one month is extracted from the Internet of Vehicles big data platform, and the maximum number of abnormal wake-ups per hour and the time period with the maximum number of abnormal wake-ups are used as feature variables. The OBD high-frequency wake-up distinction model is obtained by training through a machine learning clustering algorithm.

[0024] Further, obtaining the OBD high-frequency wake-up differentiation model includes the following steps:

[0025] The maximum number of abnormal wake-ups for each vehicle is used as the first dimension of the features used for model training, and the time period of the maximum number of abnormal wake-ups is used as the second dimension. The time period feature is mapped to 1-24, with one hour as a period.

[0026] Select the KMEANS algorithm and adjust the algorithm hyperparameter k, k is selected from 2 to 9, and the model with the optimal silhouette coefficient is obtained;

[0027] The category with the largest number of maximum abnormal awakening times is selected as the class of OBD high-frequency awakening, thereby obtaining the OBD high-frequency awakening distinction model.

[0028] The present invention also provides a storage medium, wherein the storage medium stores one or more programs, and when the one or more programs are executed by a processor, the steps of the method for identifying abnormal vehicle awakening caused by a post-installed OBD device are executed.

[0029] Compared with the prior art, the present invention has the following beneficial effects:

[0030] The method provided by the present invention is to identify the access of the after-installed OBD device based on the relevant data of the vehicle's internal network wake-up and sleep. The data mainly used include the vehicle startup and shutdown data, the user's vehicle entry and exit operation data, the vehicle network wake-up and sleep sign data, the abnormal power consumption identification rules formulated by the vehicle network sleep and rest logic, and the regular wake-up identification algorithm, and analyze whether the vehicle is equipped with an OBD device and the OBD device causes the vehicle to wake up abnormally. It can effectively prevent the situation where the vehicle is out of power due to the user's after-installed equipment, reduce the probability of vehicle power outage, and reduce the user's use cost; at the same time, it can provide a strong basis for the after-sales processing of the after-sales team. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 The present invention is a logic flow chart of a method for identifying abnormal vehicle awakening caused by an after-installed OBD device.

[0032] Figure 2 This is a flow chart of abnormal wakeup determination and recording in the present invention.

[0033] Figure 3 This is a flow chart for confirming abnormal vehicle awakening caused by OBD in the present invention. DETAILED DESCRIPTION

[0034] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0035] Embodiment 1

[0036] This embodiment discloses a method for identifying abnormal vehicle awakening caused by an aftermarket OBD device.

[0037] A method for identifying abnormal vehicle wakeup caused by after-market OBD devices, see Figure 1 , including the following steps,

[0038] Acquire vehicle data with a timestamp; wherein the vehicle data includes a vehicle network sleep flag and a vehicle network wake-up flag.

[0039] The acquired vehicle data is processed for outliers, samples with invalid values ​​are deleted, and missing values ​​are filled with the average value to obtain preprocessed data.

[0040] Based on the preprocessed data, the suspicious vehicles with abnormal wake-up caused by OBD are screened out according to the OBD high-frequency wake-up differentiation model.

[0041] The start and end time of the abnormal awakening of the suspicious vehicle are recorded, and the duration of the abnormal awakening is calculated; at the same time, the time period of the abnormal awakening and the number of abnormal awakenings of the vehicle during the time period are recorded.

[0042] Calculate the average abnormal awakening duration in the time period with the maximum number of abnormal awakenings of suspicious vehicles, and count the mode of abnormal awakening duration in the time period with the maximum number of abnormal awakenings.

[0043] If the difference between the average abnormal awakening duration and the mode of abnormal awakening duration is less than 1 s, the vehicle is determined to be a vehicle with abnormal awakening caused by OBD.

[0044] In specific implementation, the abnormal awakening determination method is:

[0045] After the vehicle network sleep flag changes from 0 to 1, if the vehicle network wake-up flag changes from 0 to 1, it is determined that a wake-up occurs.

[0046] If there is no active awakening within 2 minutes before or after the awakening time, the current awakening is determined to be an abnormal awakening.

[0047] The acquired vehicle data with timestamp also includes unlocking signal, locking signal, driver's door lock status and anti-theft status. When the unlocking signal, locking signal and driver's door lock status change, and the anti-theft status changes from armed state to disarmed state, it is determined as active wake-up.

[0048] Furthermore, the method for determining the end of abnormal wakeup is:

[0049] After a wake-up occurs, if the vehicle network sleep flag changes from 0 to 1, it is determined that the abnormal wake-up is over;

[0050] Record the end time of the current abnormal wakeup, and subtract the start time of the abnormal wakeup from the end time of the abnormal wakeup to get the duration of the abnormal wakeup; at the same time, record the time period of the abnormal wakeup, which is divided into one hour per time period, and one day is divided into 24 time periods.

[0051] The number of abnormal wake-ups in each time period of each vehicle is summed up, and the time period with the most abnormal wake-ups is selected to obtain the maximum number of abnormal wake-ups of the suspicious vehicle. The sum of the duration of each abnormal wake-up in the time period with the most abnormal wake-ups is divided by the total number of abnormal wake-ups in the time period to obtain the average abnormal wake-up duration of the time period with the maximum number of abnormal wake-ups of the suspicious vehicle.

[0052] During implementation, vehicle data of 1,000 vehicles for one month is extracted from the Internet of Vehicles big data platform. The maximum number of abnormal wake-ups per hour and the time period when the maximum number of abnormal wake-ups occurs are used as feature variables. The OBD high-frequency wake-up differentiation model is obtained through training through a machine learning clustering algorithm, which specifically includes:

[0053] The maximum number of abnormal wake-up times for each vehicle is used as the first dimension of the features used for model training, and the time period in which the maximum number of abnormal wake-up times occurs is used as the second dimension. The time period features are mapped to 1-24, with one period for each hour.

[0054] Select the KMEANS algorithm and adjust the algorithm hyperparameter k, with k ranging from 2 to 9, to obtain the model with the optimal silhouette coefficient.

[0055] The category with the largest number of maximum abnormal awakening times is selected as the class of OBD high-frequency awakening, thereby obtaining the OBD high-frequency awakening distinction model.

[0056] The method provided by the present invention is to identify the access of the after-installed OBD device based on the relevant data of the vehicle's internal network wake-up and sleep. The data mainly used mainly include the vehicle startup and shutdown data, the user's vehicle entry and exit operation data, the vehicle network wake-up and sleep sign data, the abnormal power consumption identification rules formulated by the vehicle network sleep and rest logic, and the regular wake-up identification algorithm, and analyze whether the vehicle is equipped with an OBD device and the OBD device causes the vehicle to wake up abnormally. It can effectively prevent the situation where the vehicle is out of power due to the user's after-installed equipment, and at the same time provide a strong basis for the after-sales processing of the after-sales team.

[0057] Embodiment 2

[0058] In order to further illustrate the actual effect of the present invention on identifying abnormal vehicle awakening caused by an after-installed OBD device, this embodiment discloses a method and application scenario for identifying abnormal vehicle awakening caused by an after-installed OBD device based on the first embodiment.

[0059] A method for identifying abnormal vehicle wake-up caused by an after-installed OBD device comprises the following steps:

[0060] First, extract the corresponding number and latitude of data sets from the Internet of Vehicles big data platform, and perform data preprocessing, abnormal wake-up judgment and recording, see Figure 2 , including the following steps:

[0061] A. Data extraction: Extract 1,000 vehicles from the Internet of Vehicles big data platform, user vehicle usage data for one month and vehicle network heartbeat data, including engine running status EngineStatus, power status feedback PowerStatusFeedback, lock signal LockSignal, driver's side door lock status DriverDoorLockStatus, anti-theft status KeyAlarmStatus, unlock signal UnlockSignal, vehicle network sleep flag Sleepflag, and vehicle network wake-up flag Awakeflag. Each data is timestamped.

[0062] B. Data preprocessing: The data extracted in the data extraction phase are processed for outliers. The first principle of outlier processing is to delete samples with invalid values. The second principle is to fill missing values ​​with the average value.

[0063] C. Determination of an abnormal wakeup: The abnormal power consumption feature consists of the number of abnormal wakeups and the duration of the abnormal wakeup. When the vehicle network sleep flag Sleepflag changes from 0 to 1, and the vehicle network wakeup flag Awakeflag changes from 0 to 1, it is determined that a wakeup occurs at this moment. Within 2 minutes before and after the wakeup moment, it is determined whether there is user active wakeup data. The active wakeup data includes changes in the unlocking signal UnlockSignal, the locking signal LockSignal, the driver's side door lock status DriverDoorLockStatus, and the anti-theft status KeyAlarmStatus from the armed state to the disarmed state. If no active wakeup data is found, the current wakeup is determined to be an abnormal wakeup, and the start time of the current abnormal wakeup is recorded. After that, if the vehicle network sleep flag Sleepflag changes from 0 to 1, it is determined that the abnormal wakeup ends, and the end time of the current abnormal wakeup is recorded. The start time of the abnormal wakeup is subtracted from the end time of the abnormal wakeup to obtain the duration of the abnormal wakeup. At the same time, the time period of the abnormal wakeup is recorded. The time period is divided into one hour, and one day is divided into 24 sections.

[0064] D. Calculate the maximum number of abnormal wake-ups in each hour: add up the number of abnormal wake-ups in each time period for each vehicle, and obtain the time period with the largest number of abnormal wake-ups and the maximum number of abnormal wake-ups in that time period.

[0065] Secondly, based on the maximum number of abnormal wake-up times per hour and the time period when the maximum number of abnormal wake-up times occurred as feature variables, a machine learning clustering algorithm is used for training to obtain a classification model for vehicles with high-frequency abnormal wake-up caused by aftermarket OBD devices. The average wake-up time and the mode of abnormal wake-up time of vehicles with high frequency abnormal wake-up are calculated to determine whether the vehicle is an aftermarket OBD vehicle. See Figure 3 , including the following steps:

[0066] Training data preparation for high-frequency abnormal wake-up vehicle classification algorithm: The maximum number of abnormal wake-up times for each vehicle is used as the first dimension of the features used for model training, the time period with the maximum number of abnormal wake-up times is used as the second dimension, and the time period features are mapped to 1~24, with one section per hour.

[0067] B. Algorithm selection and training: Select the KMEANS algorithm and adjust the algorithm hyperparameter k. Select k from 2 to 9 to obtain the model with the optimal silhouette coefficient. Select the category with the largest number of abnormal wake-up times as the class of OBD high-frequency wake-up, thereby obtaining the OBD high-frequency wake-up distinction model.

[0068] C. Finally determine whether the vehicle is abnormally awakened due to OBD: Use the OBD high-frequency wake-up distinction model to screen out highly suspicious vehicles that are abnormally awakened due to OBD, and calculate the average abnormal wake-up duration in the time period with the maximum number of abnormal wake-ups for these vehicles. The average number of abnormal wake-ups is equal to the sum of the duration of each abnormal wake-up in the time period with the maximum number of abnormal wake-ups, divided by the number of abnormal wake-ups in the time period with the maximum number of abnormal wake-ups. At the same time, count the mode of abnormal wake-up duration in the time period with the maximum number of abnormal wake-ups, that is, the time with the longest abnormal wake-up duration. If the difference between the average number of abnormal wake-ups and the mode of abnormal wake-up duration is less than 1s, the vehicle is determined to be an abnormally awakened vehicle due to OBD.

[0069] Embodiment 3

[0070] This embodiment discloses a storage medium based on the first embodiment.

[0071] A storage medium stores one or more programs, and when the one or more programs are executed by a processor, the steps of a method for identifying abnormal vehicle wakeup caused by an after-installed OBD device are executed. The computer-readable storage medium may be a USB flash drive, a hard disk, or other device with a storage function.

[0072] Therefore, the method provided by the present invention, through the identification rules of abnormal power consumption formulated by the whole vehicle network sleep-switch logic and the identification algorithm of regular wake-up, analyzes whether the vehicle is equipped with an OBD device and whether the OBD device causes the vehicle to wake up abnormally, which can effectively prevent the vehicle from running out of power due to user-installed equipment, reduce the probability of vehicle running out of power, and reduce the user's usage cost; at the same time, it can provide a strong basis for the after-sales processing of the after-sales team.

[0073] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit the technical solution. Those skilled in the art should understand that those modifications or equivalent substitutions of the technical solution of the present invention that do not depart from the purpose and scope of the technical solution should be included in the scope of the claims of the present invention.

Claims

1. A method for identifying abnormal vehicle wake-up caused by after-installed OBD equipment, It is characterized in that The following steps are included: Acquire vehicle data with a timestamp; wherein the vehicle data includes a vehicle network sleep flag and a vehicle network wake-up flag; Perform outlier processing on the acquired vehicle data, delete samples with invalid values, and fill the missing values ​​with the average value to obtain preprocessed data; Based on the pre-processed data, the suspicious vehicles with abnormal wake-up caused by OBD are screened out according to the OBD high-frequency wake-up differentiation model; Among them, the vehicle data of 1,000 vehicles for one month is extracted from the Internet of Vehicles big data platform, the maximum number of abnormal wake-ups per hour and the time period when the maximum number of abnormal wake-ups occurs are used as feature variables, and the OBD high-frequency wake-up distinction model is obtained by training through a machine learning clustering algorithm; the following steps are included: The maximum number of abnormal wake-ups for each vehicle is used as the first dimension of the features used for model training, and the time period of the maximum number of abnormal wake-ups is used as the second dimension. The time period feature is mapped to 1-24, with one hour as a period. Select the KMEANS algorithm and adjust the algorithm hyperparameter k, k is selected from 2 to 9, and the model with the optimal silhouette coefficient is obtained; Select the category with the largest number of maximum abnormal wake-up times as the class of OBD high-frequency wake-up, thereby obtaining the OBD high-frequency wake-up distinction model; Record the start and end time of the abnormal awakening of the suspicious vehicle, and calculate the duration of the abnormal awakening; at the same time, record the time period of the abnormal awakening and the number of abnormal awakenings of the vehicle during the time period; Calculate the average abnormal awakening duration in the time period with the maximum number of abnormal awakenings of the suspicious vehicle, and count the mode of the abnormal awakening duration in the time period with the maximum number of abnormal awakenings; If the difference between the average abnormal awakening duration and the mode of abnormal awakening duration is less than 1 s, the vehicle is determined to be a vehicle with abnormal awakening caused by OBD.

2. The method for identifying abnormal vehicle wakeup caused by after-installed OBD equipment according to claim 1, It is characterized in that The abnormal awakening determination method is: When the vehicle network sleep flag changes from 0 to 1, if the vehicle network wake-up flag changes from 0 to 1, it is determined that a wake-up occurs; If there is no active awakening within 2 minutes before or after the awakening time, the current awakening is determined to be an abnormal awakening.

3. The method for identifying abnormal vehicle awakening caused by after-installed OBD equipment according to claim 2, It is characterized in that The acquired vehicle data with timestamp also includes unlocking signal, locking signal, driver's side door lock status and anti-theft status; when the unlocking signal, locking signal and driver's side door lock status change, and the anti-theft status changes from armed state to disarmed state, it is judged as active wake-up.

4. The method for identifying abnormal vehicle awakening caused by after-market OBD equipment according to claim 2, It is characterized in that The method for determining the end of abnormal wakeup is: After a wake-up occurs, if the vehicle network sleep flag changes from 0 to 1, it is determined that the abnormal wake-up is over; Record the end time of the current abnormal wakeup, and subtract the start time of the abnormal wakeup from the end time of the abnormal wakeup to get the duration of the abnormal wakeup; at the same time, record the time period of the abnormal wakeup, which is divided into one hour section, and one day is divided into 24 sections.

5. A storage medium, It is characterized in that The storage medium stores one or more programs, and when the one or more programs are executed by the processor, the steps of the method for identifying abnormal awakening of a vehicle caused by an after-market OBD device as described in any one of claims 1 to 4 are executed.

Images