Abnormal driving track identification method and device, electronic equipment and storage medium

By combining GPS telemetry data flow, image classification model and statistical indicators, abnormal driving trajectories are identified, and the problem of low recognition accuracy in the prior art is solved, and the accuracy of recognition is improved.

CN118968156APending Publication Date: 2024-11-15北京宏瓴科技发展有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411015762.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-26
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

In the prior art, the accuracy of abnormal driving trajectory recognition is low, and it is impossible to effectively distinguish between normal driving strokes caused by traffic jams and false driving trajectories caused by stationary vehicles.

Method used

By obtaining the vehicle's GPS telemetry data stream, multiple driving strokes are determined, and the driving trajectory image is input into the image classification model, and the probability value, coefficient of variation and Hearst index are combined to identify whether the driving trajectory is an abnormal driving trajectory.

Benefits of technology

提高了异常行驶轨迹的识别准确率,能够更有效地区分正常和异常行驶轨迹,减少虚假轨迹的干扰。

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118968156A_ABST
    Figure CN118968156A_ABST
Patent Text Reader

Abstract

The invention provides an abnormal driving track identification method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining a GPS telemetry data flow of a vehicle, and determining a plurality of driving journeys based on the GPS telemetry data flow; for any target driving stroke in the multiple driving strokes, obtaining a driving track image of the target driving stroke through the GPS telemetry data stream corresponding to the target driving stroke; inputting the driving track image into an image classification model to obtain a probability value that the driving track of the target driving stroke output by the image classification model is a normal driving track; and determining whether the driving track of the target driving stroke is an abnormal driving track or not based on the probability value that the driving track of the target driving stroke is a normal driving track and the variable coefficient and the Hurst index corresponding to the GPS telemetry data flow corresponding to the target driving stroke. By adopting the technical scheme, the problem of low recognition accuracy of the abnormal driving track in the prior art is solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of information processing technology, and in particular to a method, device, electronic device and storage medium for identifying abnormal driving trajectories. Background Art

[0002] The On-Board Diagnostics (OBD) system used to collect the vehicle's movement trajectory often uploads Global Positioning System (GPS) telemetry data with slight drift in longitude and latitude when the vehicle is stationary due to objective factors such as hardware signals. The GPS telemetry data formed by the drift has the same characteristics as the GPS telemetry data during driving. Therefore, the data access system in the Internet of Vehicles platform will accept the drifting GPS telemetry data and form the vehicle's driving trajectory. These driving trajectories are false driving trajectories. In the actual process, the false driving trajectories formed by the drift need to be identified to avoid interfering with the vehicle's real driving trajectory.

[0003] In the prior art, the false driving trajectory is identified by combining the drift GPS data uploaded by the OBD device with a selected threshold, and matching the average speed, distance, driving time, etc. of the driving trip as characteristic factors with each selected rule to identify abnormal driving trajectories.

[0004] However, the existing false driving trajectory recognition method has a low accuracy rate and is unable to determine whether a trip with an average driving speed that is too low is a normal driving trip caused by objective factors such as traffic jams or a false driving trajectory caused by a stationary vehicle due to drift points reported by the OBD device. Summary of the invention

[0005] The present invention provides a method for identifying an abnormal driving trajectory, which is used to solve the problem of low accuracy in identifying abnormal driving trajectories in the prior art and improves the accuracy in identifying abnormal driving trajectories.

[0006] In a first aspect, the present invention provides a method for identifying an abnormal driving trajectory, comprising: obtaining a GPS telemetry data stream of a vehicle, and determining multiple driving trips based on the GPS telemetry data stream; for any target driving trip among the multiple driving trips, obtaining a driving trajectory image of the target driving trip through the GPS telemetry data stream corresponding to the target driving trip; inputting the driving trajectory image into an image classification model to obtain a probability value output by the image classification model that the driving trajectory of the target driving trip is a normal driving trajectory; based on the probability value that the driving trajectory of the target driving trip is a normal driving trajectory, and the coefficient of variation and the Hurst index corresponding to the GPS telemetry data stream corresponding to the target driving trip, determining whether the driving trajectory of the target driving trip is an abnormal driving trajectory; wherein the image classification model is trained based on sample driving trajectory images with probability value labels.

[0007] Optionally, based on the probability value that the driving trajectory of the target driving itinerary is a normal driving trajectory, and the coefficient of variation and Hurst index corresponding to the GPS telemetry data stream corresponding to the target driving itinerary, determining whether the driving trajectory of the target driving itinerary is an abnormal driving trajectory includes: determining that the driving trajectory of the target driving itinerary is a normal driving trajectory when any one of the first condition, the second condition and the third condition is met; wherein the first condition is that the probability value is greater than the first probability threshold; the second condition is that the probability value is greater than the second probability threshold and the coefficient of variation is less than the first coefficient of variation threshold; the third condition is that the probability value is greater than the third probability threshold and the coefficient of variation is less than the second coefficient of variation threshold and the Hurst index is greater than the Hurst index threshold; the first probability threshold is greater than the second probability threshold and the second probability threshold is greater than the third probability threshold; the first coefficient of variation threshold is greater than the second coefficient of variation threshold.

[0008] Optionally, the method further includes: when the first condition, the second condition and the third condition are not met, determining that the driving trajectory of the target driving trip is an abnormal driving trajectory.

[0009] Optionally, determining multiple driving journeys based on a GPS telemetry data stream includes: acquiring ignition time information and flameout time information of the vehicle; and segmenting the GPS telemetry data stream based on the ignition time information and the flameout time information to obtain multiple driving journeys.

[0010] Optionally, the GPS telemetry data stream is segmented based on the ignition time information and the flameout time information, including: for any ignition moment, determining the time interval between the ignition moment and the most recent flameout moment, wherein the most recent flameout moment is earlier than the ignition moment; when the time interval between the most recent flameout moment and the ignition moment is greater than a time threshold, segmenting the GPS telemetry data stream at the most recent flameout moment.

[0011] Optionally, the image classification model is a VIT model.

[0012] In a second aspect, the present invention provides an abnormal driving trajectory identification device, which includes: a first processing module, used to obtain the GPS telemetry data stream of the vehicle, and determine multiple driving trips based on the GPS telemetry data stream; an acquisition module, used to obtain a driving trajectory image of a target driving trip for any target driving trip among the multiple driving trips through the GPS telemetry data stream corresponding to the target driving trip; a second processing module, used to input the driving trajectory image into an image classification model, and obtain a probability value of the driving trajectory of the target driving trip output by the image classification model being a normal driving trajectory; a first determination module, used to determine whether the driving trajectory of the target driving trip is an abnormal driving trajectory based on the probability value of the driving trajectory of the target driving trip being a normal driving trajectory, and the coefficient of variation and Hurst index corresponding to the GPS telemetry data stream corresponding to the target driving trip; wherein the image classification model is trained based on sample driving trajectory images with probability value labels.

[0013] In a third aspect, the present invention further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for identifying abnormal driving trajectories as described above is implemented.

[0014] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-mentioned methods for identifying abnormal driving trajectories.

[0015] In a fifth aspect, the present invention further provides a computer program product, comprising a computer program, which, when executed by a processor, implements any of the above-mentioned methods for identifying abnormal driving trajectories.

[0016] The abnormal driving trajectory identification method, device, electronic device and storage medium provided by the present invention determine multiple driving trips through the GPS telemetry data stream of the vehicle, and for any target driving trip in the multiple driving trips, obtain the driving trajectory image of the target driving trip through the GPS telemetry data stream corresponding to the target driving trip, and then input the driving trajectory image into the image classification model to obtain the probability value of the driving trajectory of the target driving trip output by the image classification model as a normal driving trajectory; based on the probability value of the driving trajectory of the target driving trip as a normal driving trajectory, and the coefficient of variation and Hurst index corresponding to the GPS telemetry data stream corresponding to the target driving trip, determine whether the driving trajectory of the target driving trip is an abnormal driving trajectory. By combining the probability value, the coefficient of variation and the Hurst index to jointly identify whether the driving trajectory is an abnormal driving trajectory, the problem of low recognition accuracy of abnormal driving trajectories in the prior art is solved, and the recognition accuracy of abnormal driving trajectories is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 It is a flow chart of the abnormal driving trajectory identification method provided by the present invention.

[0019] Figure 2 This is the driving track image provided by the present invention.

[0020] Figure 3 It is a schematic diagram of the structure of the abnormal driving trajectory identification device provided by the present invention.

[0021] Figure 4 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] It should be noted that, in the embodiments of the present invention, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0024] It should be noted that, in this article, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device including the element. In addition, it should be pointed out that the scope of the method and device in the embodiment of the present invention is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in reverse order according to the functions involved, for example, the described method may be performed in an order different from that described, and various steps may also be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0025] In order to clearly describe the technical solutions of the embodiments of the present invention, in the embodiments of the present invention, the words "first", "second", etc. are used to distinguish between the same items or similar items with basically the same functions and effects. Those skilled in the art can understand that the words "first", "second", etc. are not limiting the quantity and execution order.

[0026] Figure 1 FIG. 1 is a flow chart of the abnormal driving trajectory identification method provided by the present invention. Figure 1 As shown, the method includes the following steps S102-S108: Step S102: Obtain the GPS telemetry data stream of the vehicle, and determine multiple driving trips based on the GPS telemetry data stream.

[0027] As an optional embodiment, the GPS telemetry data stream may be uploaded by the OBD device, and the GPS telemetry data stream includes data information such as time, longitude, latitude, altitude, angle, speed, and number of satellites.

[0028] As an optional embodiment, a GPS telemetry data stream can be expressed as {"lat":29.564978, "lon":106.57399, "alt":156.0, "t":"2024-06-20 00:07:50", "s":32.9, "sc":9.0, "d":70.0}, wherein lat represents longitude, lon represents latitude, alt represents altitude, t represents the time when the current GPS telemetry data stream is obtained, s represents angle, sc represents speed, and d represents the number of satellites.

[0029] In an exemplary embodiment, determining multiple driving trips based on the GPS telemetry data stream can be achieved by following the steps S11-S12: Step S11: Acquire the ignition time information and the ignition time information of the vehicle.

[0030] As an optional embodiment, the ignition time information and the ignition time information may be acquired through a sensor.

[0031] Step S12: Segment the GPS telemetry data stream based on the ignition time information and the ignition time information to obtain multiple driving trips.

[0032] As an optional embodiment, the distance corresponding to the time from one ignition on to one ignition off may be regarded as a driving journey to segment the GPS telemetry data stream.

[0033] In an exemplary embodiment, segmenting the GPS telemetry data stream based on the ignition time information and the ignition time information can be achieved through the following steps S21-S22: Step S21: for any ignition time, determine the time interval between the ignition time and the most recent flameout time, wherein the most recent flameout time is earlier than the ignition time.

[0034] As an optional embodiment, assuming that the first ignition time is 7:10, the first flameout time is 7:40, the second ignition time is 8:30, and the second flameout time is 8:35, the second ignition time is taken as the basis, and the time interval with the most recent flameout time is determined to be 50 minutes. When judging the time interval between the ignition time and the most recent flameout time, it is the time interval between the current ignition time and the last flameout time. By limiting the most recent flameout time to be earlier than the ignition time, it can avoid taking the situation of flameout immediately after ignition into account, which will introduce errors in advance for the subsequent identification of abnormal driving trajectories.

[0035] Step S22: when the time interval between the most recent ignition-off time and the ignition-on time is greater than the time threshold, the GPS telemetry data stream is segmented at the most recent ignition-off time.

[0036] As an optional embodiment, the time threshold can be set by comprehensively considering driving behavior characteristics and traffic regulations. For example, if you need to rest for 20 minutes after driving for four hours continuously, you can use 20 minutes as the time threshold for segmentation.

[0037] As an optional embodiment, assuming that the first ignition time is 7:10, the first shutdown time is 7:40, the second ignition time is 8:30, and the second shutdown time is 8:35. At this time, the time interval between the second ignition time and the first shutdown time is greater than 20 minutes. Therefore, the GPS telemetry data stream is split at 7:40 corresponding to the first shutdown time. Since the vehicle is in the process of driving, each moment may bring changes in information such as longitude, latitude and altitude. Therefore, the split telemetry data stream contains multiple longitude, latitude, altitude, time, angle, speed and number of satellites.

[0038] Step S104: for any target driving trip among the multiple driving trips, a driving track image of the target driving trip is obtained through a GPS telemetry data stream corresponding to the target driving trip.

[0039] As an optional embodiment, the acquired GPS telemetry data stream is shown in Table 1 below.

[0040] Table 1 - GPS telemetry data flow

[0041] After the data stream in Table 1 is segmented according to the time threshold, the data set corresponding to a driving trip is obtained as follows: [29.564, 106.573, 156., 07:00, 32.9, 9.0, 70.0

[0042] 29.565, 106.574, 156.0, 07:10, 40.2, 9.0, 55.0

[0043] 30.156,108.032,156.0,07:25,44.0,9.0,48.0

[0044] 32.667, 117.655, 157.1, 07:55, 38.0, 9.0, 46.0]

[0045] The latitude and longitude features extracted from the above data set are converted into a two-dimensional array: [29.564, 106.573], [29.565, 106.574], [30.156, 108.032], and [32.667, 117.655].

[0046] Then, the map is directly obtained through the website, and the coordinate points corresponding to the above two-dimensional data are marked on the map. Based on all the marked points, the driving track image corresponding to the above two-dimensional data is obtained, such as Figure 2 As shown, it is the driving track image provided by the present invention.

[0047] Step S106: Input the driving trajectory image into the image classification model to obtain the probability value of the driving trajectory of the target driving trip output by the image classification model being a normal driving trajectory; wherein the image classification model is trained based on sample driving trajectory images with probability value labels.

[0048] In an exemplary embodiment, the image classification model is a VIT model.

[0049] As an optional embodiment, the driving trajectory image can be loaded using opencv, converted into a digital vector and input into the VIT model to obtain the probability value output by the model.

[0050] Step S108: Determine whether the driving trajectory of the target driving itinerary is an abnormal driving trajectory based on the probability value that the driving trajectory of the target driving itinerary is a normal driving trajectory and the coefficient of variation and Hurst index corresponding to the GPS telemetry data stream corresponding to the target driving itinerary.

[0051] In an exemplary embodiment, based on the probability value that the driving trajectory of the target driving trip is a normal driving trajectory, and the coefficient of variation and the Hurst exponent corresponding to the GPS telemetry data stream corresponding to the target driving trip, determining whether the driving trajectory of the target driving trip is an abnormal driving trajectory includes: When any one of the first condition, the second condition and the third condition is met, the driving trajectory of the target driving journey is determined to be a normal driving trajectory.

[0052] The first condition is that the probability value is greater than a first probability threshold.

[0053] As an optional embodiment, the first condition is that the probability value is greater than 0.92 (ie, the first probability threshold).

[0054] The second condition is that the probability value is greater than the second probability threshold and the coefficient of variation is less than the first coefficient of variation threshold.

[0055] As an optional embodiment, the second condition is that the probability value is greater than 0.8 (ie, the second probability threshold value) and the coefficient of variation is less than 0.6 (ie, the first coefficient of variation threshold value).

[0056] The third condition is that the probability value is greater than the third probability threshold, the coefficient of variation is less than the second coefficient of variation threshold, and the Hurst exponent is greater than the Hurst exponent threshold.

[0057] As an optional embodiment, the third condition is that the probability value is greater than 0.6 (ie, the third probability threshold), the coefficient of variation is less than 0.4 (ie, the second coefficient of variation threshold), and the Hurst index is greater than 0.6 (ie, the Hurst index threshold).

[0058] The first probability threshold is greater than the second probability threshold and the second probability threshold is greater than the third probability threshold. The first coefficient of variation threshold is greater than the second coefficient of variation threshold.

[0059] As an optional embodiment, extracting features lat, lon, alt, s, sc, d from the GPS telemetry data stream to construct a feature array can be expressed as: [29.565767, 106.575944, 158.0, 50.9, 7.0, 76.0]

[0060] [29.566202, 106.578076, 161.0, 48.40, 8.0, 89.0]

[0061] [29.566178, 106.578336, 161.0, 46.1, 8.0, 102.0]

[0062] [29.566048, 106.578743, 162.0, 37.6, 8.0, 119.0]

[0063] [29.560493, 106.583854, 166.0, 31.5, 4.0, 193.0]

[0064] The above feature data is reduced in dimension using PCA principal component analysis, and the resulting 1-dimensional feature array can be expressed as: [2.04273038e-02, 2.04450692e-02, 2.05628068e-02, 1.99340395e-02,

[0065] 1.87769270e-02, 1.79449424e-02, 1.75829090e-02, 1.74759654e-02,

[0066] 1.77306300e-02, 1.80679758e-02, 1.84595179e-02, 1.87478148e-02,

[0067] 1.90132237e-02, 1.86792749e-02, 1.72535546e-02, 1.56657202e-02]

[0068] As an optional embodiment, the coefficient of variation (CV) is a measure of relative standard deviation, used to indicate the degree of dispersion of a data set. It is the ratio of the standard deviation to the mean value. It can be calculated using the following formula:

[0069] in, represents the standard deviation of a 1D feature array, Represents the mean of a 1-D feature array.

[0070] As an optional embodiment, the Hurst index is a statistical indicator for evaluating the smoothness and autocorrelation of a time series. The calculation of the Hurst index is mainly based on the rescaled range (R / S) analysis method. It reflects the long-term correlation and self-similarity in time series data.

[0071] The singular coefficient of the above 1-dimensional feature array is calculated, and the singular coefficient is 0.3390. The Hurst index is calculated based on the above 1-dimensional feature array, and the Hurst index is 0.4379.

[0072] As an optional embodiment, assuming that the probability value is 0.7, the singular coefficient is 0.3390, and the Hurst index is 0.6932, the probability value is greater than 0.6, the singular coefficient is less than 0.4, and the Hurst index is greater than 0.6, and the above third condition is met, then the trajectory is a normal driving trajectory.

[0073] In an exemplary embodiment, the method further includes: when the first condition, the second condition, and the third condition are not satisfied, determining that the driving trajectory of the target driving trip is an abnormal driving trajectory.

[0074] As an optional embodiment, assuming that the probability value is 0.7, the singularity coefficient is 0.3390, and the Hurst index is 0.4379, the probability value is greater than 0.6, the singularity coefficient is less than 0.4, but the Hurst index is less than 0.6 and does not meet any of the above conditions, then the trajectory is an abnormal driving trajectory.

[0075] In the above steps S102-S108, multiple driving trips are determined through the vehicle's GPS telemetry data stream. For any target driving trip in the multiple driving trips, a driving trajectory image of the target driving trip is obtained through the GPS telemetry data stream corresponding to the target driving trip, and then the driving trajectory image is input into the image classification model to obtain the probability value of the driving trajectory of the target driving trip output by the image classification model being a normal driving trajectory; based on the probability value of the driving trajectory of the target driving trip being a normal driving trajectory, and the coefficient of variation and Hurst index corresponding to the GPS telemetry data stream corresponding to the target driving trip, it is determined whether the driving trajectory of the target driving trip is an abnormal driving trajectory. By combining the probability value, the coefficient of variation and the Hurst index to jointly identify whether the driving trajectory is an abnormal driving trajectory, the problem of low recognition accuracy of abnormal driving trajectories in the prior art is solved, and the recognition accuracy of abnormal driving trajectories is improved.

[0076] The executor of each step in the above method may be an abnormal driving trajectory identification device, which may be implemented by software and / or hardware. The device may be integrated into an electronic device, which may be a terminal device (such as a smart phone, a personal computer, etc.), or a server (such as a local server or a cloud server, or a server cluster, etc.), or a processor, or a chip, etc.

[0077] Obviously, the above-described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. In order to better understand the above method, the above process is described below in conjunction with the embodiments, but the technical solutions of the embodiments of the present invention are not limited. Specifically: The present invention provides a method for jointly identifying abnormal driving trajectories based on probability value, coefficient of variation and Hurst exponent, which specifically comprises the following steps: Step 1: Use a time-window-based streaming computing method to split the GPS telemetry data stream uploaded by the OBD device into driving trips. The splitting method is based on the information such as car ignition and shutdown reported by the sensor combined with the time threshold. The time threshold is selected based on the comprehensive driving behavior characteristics and traffic rules such as fatigue driving regulations.

[0078] Step 2: Load the map data, upload the GPS longitude and latitude to draw the real track, and take a screenshot to get the driving track picture.

[0079] Step 3: Label the driving trajectory image, use the deep learning VIT image classification model to predict the driving trajectory image, and obtain the probability value of the driving trajectory in the driving trajectory image output by the VIT model being a normal trajectory.

[0080] Step 4: Extract the longitude, latitude, altitude, angle, speed, number of satellites and other features from the telemetry data stream, use PCA principal component analysis technology to reduce the dimension, and calculate the coefficient of variation and Hurst exponent.

[0081] Step 5: Combine the probability value, coefficient of variation and Hurst index to comprehensively consider the rule chain to determine whether the travel trajectory is a normal driving trajectory.

[0082] The rule chain is: Rule 1: A probability value > 0.92 indicates a normal driving trajectory.

[0083] Rule 2: Probability value > 0.8 and coefficient of variation < 0.6 indicate a normal driving trajectory.

[0084] Rule 3: Probability value > 0.6, coefficient of variation < 0.4, and Hurst exponent > 0.6 indicate a normal driving trajectory.

[0085] Rule 4: Failure to meet the above rules indicates an abnormal driving trajectory.

[0086] The recognition scheme in the present invention greatly improves the recognition accuracy, and at the same time uses the VIT pre-training model to further reduce the development and maintenance costs.

[0087] The abnormal driving trajectory recognition device provided by the present invention is described below. The abnormal driving trajectory recognition device described below and the abnormal driving trajectory recognition method described above can correspond to each other.

[0088] Figure 3 Schematic diagram of the structure of the abnormal driving trajectory identification device provided by the present invention. Figure 3 As shown, the device comprises: The first processing module 302 is used to obtain a GPS telemetry data stream of the vehicle and determine a plurality of driving trips based on the GPS telemetry data stream.

[0089] The acquisition module 304 is used to obtain a driving track image of any target driving trip among the multiple driving trips through a GPS telemetry data stream corresponding to the target driving trip.

[0090] The second processing module 306 is used to input the driving trajectory image into the image classification model to obtain a probability value of the driving trajectory of the target driving trip output by the image classification model being a normal driving trajectory.

[0091] A first determination module 308 is used to determine whether the driving trajectory of the target driving trip is an abnormal driving trajectory based on the probability value that the driving trajectory of the target driving trip is a normal driving trajectory, and the coefficient of variation and the Hurst index corresponding to the GPS telemetry data stream corresponding to the target driving trip; Among them, the image classification model is trained based on sample driving trajectory images with probability value labels.

[0092] The above-mentioned abnormal driving trajectory identification device determines multiple driving trips through the vehicle's GPS telemetry data stream, and for any target driving trip in the multiple driving trips, obtains the driving trajectory image of the target driving trip through the GPS telemetry data stream corresponding to the target driving trip, and then inputs the driving trajectory image into the image classification model to obtain the probability value of the driving trajectory of the target driving trip output by the image classification model as a normal driving trajectory; based on the probability value of the driving trajectory of the target driving trip as a normal driving trajectory, and the coefficient of variation and Hurst index corresponding to the GPS telemetry data stream corresponding to the target driving trip, determine whether the driving trajectory of the target driving trip is an abnormal driving trajectory. By combining the probability value, the coefficient of variation and the Hurst index to jointly identify whether the driving trajectory is an abnormal driving trajectory, the problem of low recognition accuracy of abnormal driving trajectories in the prior art is solved, and the recognition accuracy of abnormal driving trajectories is improved.

[0093] Optionally, the first determination module 308 is also used to determine that the driving trajectory of the target driving itinerary is a normal driving trajectory when any one of the first condition, the second condition and the third condition is met; wherein the first condition is that the probability value is greater than the first probability threshold; the second condition is that the probability value is greater than the second probability threshold and the coefficient of variation is less than the first coefficient of variation threshold; the third condition is that the probability value is greater than the third probability threshold and the coefficient of variation is less than the second coefficient of variation threshold and the Hurst index is greater than the Hurst index threshold; the first probability threshold is greater than the second probability threshold and the second probability threshold is greater than the third probability threshold; the first coefficient of variation threshold is greater than the second coefficient of variation threshold.

[0094] Optionally, the device further includes a second determination module, configured to determine that the driving trajectory of the target driving trip is an abnormal driving trajectory when the first condition, the second condition and the third condition are not met.

[0095] Optionally, the first processing module 302 is further used to obtain ignition time information and ignition off time information of the vehicle; and to segment the GPS telemetry data stream based on the ignition time information and the ignition off time information to obtain multiple driving trips.

[0096] Optionally, the first processing module 302 is also used to determine the time interval between the ignition moment and the most recent flameout moment for any ignition moment, wherein the most recent flameout moment is earlier than the ignition moment; when the time interval between the most recent flameout moment and the ignition moment is greater than a time threshold, the GPS telemetry data stream is split at the most recent flameout moment.

[0097] Optionally, the image classification model is a VIT model.

[0098] Figure 4 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 4 As shown, the electronic device may include: a processor 410, a communications interface 420, a memory 430 and a communication bus 440, wherein the processor 410, the communications interface 420 and the memory 430 communicate with each other through the communication bus 440. The processor 410 may call the logic instructions in the memory 430 to execute the abnormal driving trajectory recognition method, which includes: A GPS telemetry data stream of the vehicle is obtained, and a plurality of driving trips are determined based on the GPS telemetry data stream.

[0099] For any target driving trip among the multiple driving trips, a driving track image of the target driving trip is obtained through a GPS telemetry data stream corresponding to the target driving trip.

[0100] The driving trajectory image is input into the image classification model to obtain the probability value of the driving trajectory of the target driving trip output by the image classification model being a normal driving trajectory.

[0101] Based on the probability value that the driving trajectory of the target driving trip is a normal driving trajectory, and the coefficient of variation and the Hurst exponent corresponding to the GPS telemetry data stream corresponding to the target driving trip, it is determined whether the driving trajectory of the target driving trip is an abnormal driving trajectory.

[0102] Among them, the image classification model is trained based on sample driving trajectory images with probability value labels.

[0103] In addition, the logic instructions in the above-mentioned memory 430 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0104] On the other hand, the present invention further provides a computer program product, the computer program product includes a computer program, the computer program can be stored in a non-transitory computer-readable storage medium, when the computer program is executed by a processor, the computer can execute the abnormal driving trajectory identification method provided by the above methods, the method includes: A GPS telemetry data stream of the vehicle is obtained, and a plurality of driving trips are determined based on the GPS telemetry data stream.

[0105] For any target driving trip among the multiple driving trips, a driving track image of the target driving trip is obtained through a GPS telemetry data stream corresponding to the target driving trip.

[0106] The driving trajectory image is input into the image classification model to obtain the probability value of the driving trajectory of the target driving trip output by the image classification model being a normal driving trajectory.

[0107] Based on the probability value that the driving trajectory of the target driving trip is a normal driving trajectory, and the coefficient of variation and the Hurst exponent corresponding to the GPS telemetry data stream corresponding to the target driving trip, it is determined whether the driving trajectory of the target driving trip is an abnormal driving trajectory.

[0108] Among them, the image classification model is trained based on sample driving trajectory images with probability value labels.

[0109] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method for identifying abnormal driving trajectories provided by the above methods is implemented, and the method includes: A GPS telemetry data stream of the vehicle is obtained, and a plurality of driving trips are determined based on the GPS telemetry data stream.

[0110] For any target driving trip among the multiple driving trips, a driving track image of the target driving trip is obtained through a GPS telemetry data stream corresponding to the target driving trip.

[0111] The driving trajectory image is input into the image classification model to obtain the probability value of the driving trajectory of the target driving trip output by the image classification model being a normal driving trajectory.

[0112] Based on the probability value that the driving trajectory of the target driving trip is a normal driving trajectory, and the coefficient of variation and the Hurst exponent corresponding to the GPS telemetry data stream corresponding to the target driving trip, it is determined whether the driving trajectory of the target driving trip is an abnormal driving trajectory.

[0113] Among them, the image classification model is trained based on sample driving trajectory images with probability value labels.

[0114] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0115] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0116] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for identifying abnormal driving trajectories, characterized in that: include: Obtaining a GPS telemetry data stream of the vehicle, and determining a plurality of driving trips based on the GPS telemetry data stream; For any target driving trip among the multiple driving trips, obtaining a driving track image of the target driving trip through a GPS telemetry data stream corresponding to the target driving trip; Inputting the driving trajectory image into an image classification model to obtain a probability value of the driving trajectory of the target driving trip output by the image classification model being a normal driving trajectory; Determining whether the driving trajectory of the target driving itinerary is an abnormal driving trajectory based on a probability value that the driving trajectory of the target driving itinerary is a normal driving trajectory and a coefficient of variation and a Hurst exponent corresponding to a GPS telemetry data stream corresponding to the target driving itinerary; The image classification model is obtained by training based on sample driving trajectory images with probability value labels.

2. The method for identifying abnormal driving trajectories according to claim 1, characterized in that: The determining whether the driving trajectory of the target driving itinerary is an abnormal driving trajectory based on the probability value that the driving trajectory of the target driving itinerary is a normal driving trajectory and the coefficient of variation and the Hurst index corresponding to the GPS telemetry data stream corresponding to the target driving itinerary includes: When any one of the first condition, the second condition and the third condition is met, determining that the driving trajectory of the target driving trip is a normal driving trajectory; Wherein, the first condition is that the probability value is greater than a first probability threshold; The second condition is that the probability value is greater than a second probability threshold and the coefficient of variation is less than a first coefficient of variation threshold; The third condition is that the probability value is greater than a third probability threshold, the coefficient of variation is less than a second coefficient of variation threshold, and the Hurst index is greater than a Hurst index threshold; The first probability threshold is greater than the second probability threshold and the second probability threshold is greater than the third probability threshold; the first coefficient of variation threshold is greater than the second coefficient of variation threshold.

3. The method for identifying abnormal driving trajectories according to claim 2, characterized in that: The method further comprises: When the first condition, the second condition, and the third condition are not satisfied, the driving trajectory of the target driving trip is determined to be an abnormal driving trajectory.

4. The method for identifying abnormal driving trajectories according to claim 1, characterized in that: The determining of a plurality of driving trips based on the GPS telemetry data stream comprises: Obtaining ignition time information and ignition off time information of the vehicle; The GPS telemetry data stream is segmented based on the ignition time information and the ignition off time information to obtain the multiple driving trips.

5. The method for identifying abnormal driving trajectories according to claim 4, characterized in that: The segmenting of the GPS telemetry data stream based on the ignition time information and the ignition off time information includes: For any ignition moment, determine the time interval between the ignition moment and the most recent flameout moment, wherein the most recent flameout moment is earlier than the ignition moment; When the time interval between the most recent flameout moment and the ignition moment is greater than a time threshold, the GPS telemetry data stream is segmented at the most recent flameout moment.

6. The method for identifying abnormal driving trajectories according to claim 1, characterized in that: The image classification model is a VIT model.

7. A device for identifying abnormal driving trajectories, characterized in that: The device comprises: A first processing module, configured to obtain a GPS telemetry data stream of the vehicle, and determine a plurality of driving trips based on the GPS telemetry data stream; an acquisition module, for obtaining, for any target driving trip among the multiple driving trips, a driving track image of the target driving trip through a GPS telemetry data stream corresponding to the target driving trip; A second processing module is used to input the driving trajectory image into an image classification model to obtain a probability value of the driving trajectory of the target driving trip output by the image classification model being a normal driving trajectory; A first determination module is used to determine whether the driving trajectory of the target driving itinerary is an abnormal driving trajectory based on a probability value that the driving trajectory of the target driving itinerary is a normal driving trajectory, and a coefficient of variation and a Hurst exponent corresponding to a GPS telemetry data stream corresponding to the target driving itinerary; The image classification model is obtained by training based on sample driving trajectory images with probability value labels.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the method for identifying abnormal driving trajectories as described in any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for identifying an abnormal driving trajectory as described in any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for identifying an abnormal driving trajectory as described in any one of claims 1 to 6 is implemented.