Time series-based driving behavior anomaly detection method and related device thereof

Abstract

The embodiment of the application belongs to the technical field of financial technology, is applied to vehicle driving detection, can assist in market research and analysis of vehicle accident identification and vehicle insurance pricing, and relates to a driving behavior anomaly detection method based on time series and related equipment thereof, which comprises analyzing and predicting whether driving behavior is abnormal by combining an ARIMA model and vehicle driving trajectory data, in the ARIMA model training stage, performing stationarity and non-white noise tests on the training set, so that the time series data sequence corresponding to each element in the finally retained driving trajectory data set meets the test requirements, and the anomaly detection accuracy of the ARIMA model is ensured. By combining an artificial intelligence model and real-time detection, whether the driving behavior of a target vehicle is abnormal is comprehensively analyzed, compared with the existing stopping and checking method, the method is more intelligent and scientific, and whether the driving behavior of a vehicle being driven is abnormal can be analyzed in combination with a high-speed real-time snapshot device, so that the occurrence of accidents is minimized.

Description

Technical Field

[0001] This application relates to the field of financial technology and can assist in market research and analysis for vehicle accident identification and auto insurance pricing. In particular, it relates to a time-series-based method for detecting abnormal driving behavior and related equipment. Background Technology

[0002] With the development of science and technology, acquiring data and using data modeling to solve corresponding problems has become a very common technical means. Due to data protection laws, private cars cannot be equipped with drunk driving detection devices like commercial vehicles. Therefore, the current monitoring of driving behavior in private cars mainly relies on manual detection by traffic police and the necessary instruments. If these detection devices are removed, the monitoring still relies on binary classification algorithms based on basic differences such as the time of the report and the time of the accident. Furthermore, manual investigation and negotiation by phone are still required to ultimately determine whether the driver has exhibited abnormal driving behavior.

[0003] Currently, the rise of artificial intelligence (AI) technology provides technical support for intelligent data processing. However, in the identification of abnormal driving behavior, there is still a lack of an analysis and processing method that can combine AI technology to intelligently and automatically replace the traditional manual detection methods used by traffic police. Therefore, the identification of abnormal driving behavior is not yet intelligent and automated enough. Summary of the Invention

[0004] The purpose of this application is to propose a time-series-based method for detecting abnormal driving behavior and related equipment, in order to solve the problem that the existing technology is not intelligent and automated enough in the identification of abnormal driving behavior.

[0005] To address the aforementioned technical problems, this application provides a time-series-based method for detecting abnormal driving behavior, employing the following technical solution:

[0006] A time-series-based method for detecting abnormal driving behavior includes the following steps:

[0007] Real-time acquisition of driving trajectory data of the vehicle to be detected, wherein the driving trajectory data includes each collection time point and the corresponding vehicle location information at each collection time point;

[0008] The driving trajectory data is arranged in a time sequence according to a preset time interval to obtain a time sequence data sequence;

[0009] The time-series data sequence is input into a pre-trained ARIMA model to detect abnormal driving behavior.

[0010] Repeat the above detection steps and calculate the ratio of vehicles exhibiting abnormal driving behavior using a probability algorithm;

[0011] If the ratio is greater than a preset ratio threshold, the vehicle's registration information is obtained, and a reminder for key attention is sent to relevant departments.

[0012] Furthermore, the step of arranging the driving trajectory data in time sequence according to a preset time interval specifically includes:

[0013] Based on the preset time interval and the various collection time points, the driving trajectory data is preprocessed, wherein the preprocessing includes data augmentation or data deletion.

[0014] Select the vehicle location information corresponding to any collection time point as the initial time series data;

[0015] According to the preset time interval, the vehicle location information corresponding to the next time series point is obtained sequentially as the target time series data;

[0016] The target time series data corresponding to each time series point is arranged according to the preset time series data arrangement format.

[0017] Furthermore, the step of preprocessing the driving trajectory data based on the preset time interval and the various collection time points specifically includes:

[0018] Obtain the time interval between two adjacent acquisition time points in each of the acquisition time points;

[0019] Determine the relationship between the preset time interval and the time interval between two adjacent acquisition time points;

[0020] If the preset time interval is greater than the time interval between two adjacent collection time points, the driving trajectory data is scaled and the vehicle location information is deleted.

[0021] If the preset time interval is less than the time interval between two adjacent collection time points, the driving trajectory data is stretched and the vehicle location information is amplified.

[0022] Furthermore, after performing the step of obtaining the time-series data sequence, the method further includes:

[0023] According to the preset filtering rules, abnormal data in the time series data sequence is filtered out and cleared. The preset filtering rules are specifically as follows: obtain all vehicle location information in the time series data sequence, calculate the relationship between the current vehicle location information and the average value of all vehicle location information using the variance value calculation method, and determine whether the variance value is greater than a preset variance threshold. If it is greater, the current vehicle location information is marked as abnormal data.

[0024] Furthermore, before performing the step of inputting the time-series data sequence into the pre-trained ARIMA model for driving behavior anomaly detection, the method further includes:

[0025] Obtain a pre-labeled driving trajectory dataset, wherein a portion of the driving trajectory dataset consists of normal driving trajectory data and another portion consists of abnormal driving trajectory data;

[0026] According to the preset time interval, each driving trajectory data in the driving trajectory dataset is arranged in time sequence to obtain a time sequence data sequence with the same number of elements as the driving trajectory dataset.

[0027] Using the preset filtering rules, abnormal data is removed from each time series data sequence to obtain the input sequence;

[0028] The input sequences are sequentially fed into a pre-built ARIMA model for incremental classification training to obtain normal or abnormal classification results for each input sequence.

[0029] The pre-trained ARIMA model is obtained when the classification result matches the labeling result.

[0030] Furthermore, after performing the step of arranging each driving trajectory data in the driving trajectory dataset according to the preset time interval to obtain a time-series data sequence with the same number of elements as the driving trajectory dataset, the method further includes:

[0031] Beforehand, the time series data sequence corresponding to each element in the driving trajectory dataset is subjected to stationarity test and non-white noise test. Specifically, the ADF test method is used to test the stationarity of the time series data sequence corresponding to each element in the driving trajectory dataset, and the point value distribution probability statistics method is used to test the non-white noise of the time series data sequence corresponding to each element in the driving trajectory dataset.

[0032] Based on the stationarity test results and the non-white noise test results, time series data sequences that do not meet the anomaly detection requirements are removed, while time series data sequences that do meet the anomaly detection requirements are retained.

[0033] Furthermore, the step of inputting the time-series data sequence into the pre-trained ARIMA model for driving behavior anomaly detection specifically includes:

[0034] The time series data sequence is input into the pre-trained ARIMA model to obtain the model output results;

[0035] If the model output is normal, then the driving behavior corresponding to the time series data sequence is normal.

[0036] If the model output is abnormal, then the driving behavior corresponding to the time series data sequence is abnormal.

[0037] The step of repeatedly performing the above detection steps and calculating the ratio of abnormal driving behavior in the vehicle using a probability algorithm specifically includes:

[0038] The total number of repeated detections and the number of times the model output was an anomaly were counted.

[0039] The ratio of abnormal driving behavior of the vehicle is calculated based on the probability algorithm, the total number of repeated detections, and the number of times the model output result is abnormal.

[0040] To address the aforementioned technical problems, this application also provides a time-series-based driving behavior anomaly detection device, employing the following technical solution:

[0041] A time-series-based abnormal driving behavior detection device, comprising:

[0042] The data acquisition module is used to acquire the driving trajectory data of the vehicle to be detected in real time. The driving trajectory data includes each collection time point and the corresponding vehicle location information at each collection time point.

[0043] The time-series data conversion module is used to arrange the driving trajectory data into time-series data according to a preset time interval to obtain a time-series data sequence;

[0044] The ARIMA model detection module is used to input the time series data sequence into the pre-trained ARIMA model to detect abnormal driving behavior.

[0045] The probability algorithm calculation module is used to repeatedly execute the above detection steps and calculate the ratio of abnormal driving behavior of the vehicle through a probability algorithm.

[0046] The abnormal driving alert module is used to obtain the vehicle's registration information and send a reminder to relevant departments to pay close attention if the ratio is greater than a preset ratio threshold.

[0047] To address the aforementioned technical problems, this application also provides a computer device that employs the following technical solution:

[0048] A computer device includes a memory and a processor, wherein the memory stores computer-readable instructions, and the processor executes the computer-readable instructions to implement the steps of the time-series-based abnormal driving behavior detection method described above.

[0049] To address the aforementioned technical problems, this application also provides a computer-readable storage medium, employing the technical solution described below:

[0050] A computer-readable storage medium storing computer-readable instructions, which, when executed by a processor, implement the steps of the time-series-based abnormal driving behavior detection method described above.

[0051] Compared with the prior art, the embodiments of this application have the following main advantages:

[0052] The time-series-based driving behavior anomaly detection method described in this application acquires driving trajectory data of the vehicle to be detected in real time; arranges the driving trajectory data into time-series data according to a preset time interval to obtain a time-series data sequence; inputs the time-series data sequence into a pre-trained ARIMI model for driving behavior anomaly detection; repeats the above detection steps, calculates the ratio of the vehicle exhibiting driving behavior anomalies using a probability algorithm; if the ratio is greater than a preset threshold, obtains the vehicle's registration information and sends a key attention reminder to relevant departments. This application considers combining the ARIMI model and the vehicle's driving trajectory data to analyze and predict whether driving behavior is abnormal. Simultaneously, during the ARIMI model training phase, stationarity and non-white noise tests are performed on the training set to ensure that the time-series data sequence corresponding to each element in the final retained driving trajectory dataset meets the test requirements, thus guaranteeing the anomaly detection accuracy of the ARIMI model. By combining artificial intelligence models with real-time detection, the system comprehensively analyzes whether the driving behavior of target vehicles is abnormal. Compared with existing methods of stopping and screening vehicles, this approach is more intelligent and scientific. It can combine high-speed real-time capture equipment to analyze whether there are any abnormal driving behaviors in moving vehicles, thereby minimizing the occurrence of major traffic accidents. Attached Figure Description

[0053] To more clearly illustrate the solutions in this application, the accompanying drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0054] Figure 1 This is an exemplary system architecture diagram to which this application can be applied;

[0055] Figure 2 A flowchart of an embodiment of the time-series-based abnormal driving behavior detection method according to this application;

[0056] Figure 3 yes Figure 2 A flowchart of a specific embodiment of step 202 shown;

[0057] Figure 4 yes Figure 3 A flowchart of a specific embodiment of step 301 shown;

[0058] Figure 5 A flowchart of an embodiment of obtaining an ARIMI classification model through training according to this application;

[0059] Figure 6 A schematic diagram of the structure of an embodiment of the time-series-based abnormal driving behavior detection device according to this application;

[0060] Figure 7 yes Figure 6 A schematic diagram of a specific embodiment of 602 is shown;

[0061] Figure 8 yes Figure 7 A schematic diagram of a specific embodiment of 701 is shown;

[0062] Figure 9 A schematic diagram of the structure of a specific embodiment of the ARIMA model pre-training module according to this application;

[0063] Figure 10 A schematic diagram of the structure of an embodiment of the computer device according to this application. Detailed Implementation

[0064] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings of this application, are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings of this application are used to distinguish different objects, not to describe a particular order.

[0065] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0066] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0067] like Figure 1 As shown, system architecture 100 may include terminal devices 101, 102, and 103, a network 104, and a server 105. Network 104 serves as the medium for providing communication links between terminal devices 101, 102, and 103 and server 105. Network 104 may include various connection types, such as wired or wireless communication links, or fiber optic cables, etc.

[0068] Users can use terminal devices 101, 102, and 103 to interact with server 105 via network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101, 102, and 103, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social media platform software, etc.

[0069] Terminal devices 101, 102, and 103 can be various electronic devices with displays and support web browsing, including but not limited to smartphones, tablets, e-book readers, MP3 players (Moving Picture Experts Group Audio Layer III), MP4 players (Moving Picture Experts Group Audio Layer IV), laptops, and desktop computers, etc.

[0070] Server 105 can be a server that provides various services, such as a backend server that supports the pages displayed on terminal devices 101, 102, and 103.

[0071] It should be noted that the time-series-based abnormal driving behavior detection method provided in this application is generally executed by a server / terminal device, and correspondingly, the time-series-based abnormal driving behavior detection device is generally set in the server / terminal device.

[0072] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0073] Continue to refer to Figure 2 The diagram illustrates a flowchart of an embodiment of the time-series-based driving behavior anomaly detection method according to this application. The time-series-based driving behavior anomaly detection method includes the following steps:

[0074] Step 201: Acquire the driving trajectory data of the vehicle to be detected in real time, wherein the driving trajectory data includes each collection time point and the corresponding vehicle location information at each collection time point.

[0075] In this embodiment, the step of acquiring the driving trajectory data of the vehicle to be detected in real time can be implemented through the interface of an external third-party data acquisition platform. For example, the driving trajectory data of the vehicle to be detected can be acquired in real time through the data provision interface of an external GPS vehicle positioning and monitoring system. The driving trajectory data includes each collection time point and the corresponding vehicle location information at each collection time point. The vehicle location information can be latitude and longitude data collected by the GPS vehicle positioning and monitoring system.

[0076] Step 202: Arrange the driving trajectory data into time sequence according to a preset time interval to obtain a time sequence data sequence.

[0077] Continue to refer to Figure 3 , Figure 3 yes Figure 2 A flowchart of a specific embodiment of step 202 shown includes:

[0078] Step 301: Based on the preset time interval and the various collection time points, preprocess the driving trajectory data, wherein the preprocessing includes data amplification or data deletion.

[0079] Continue to refer to Figure 4 , Figure 4 yes Figure 3 A flowchart of a specific embodiment of step 301 shown includes:

[0080] Step 401: Obtain the time interval between two adjacent collection time points in each collection time point;

[0081] Step 402: Determine the relationship between the preset time interval and the time interval between two adjacent acquisition time points;

[0082] Step 403: If the preset time interval is greater than the time interval between two adjacent collection time points, then the driving trajectory data is scaled and the vehicle location information is deleted.

[0083] Step 404: If the preset time interval is less than the time interval between two adjacent collection time points, then the driving trajectory data is stretched and the vehicle location information is amplified.

[0084] In this embodiment, the data augmentation and data reduction processes are two opposing methods with the same goal: to ensure that the collected driving trajectory data meets the requirements of time-series processing. One approach is data augmentation. Specifically, if the interval between different collection points in the collected driving trajectory data is greater than the preset time interval, the driving trajectory data will be sparse. In this case, the driving trajectory data needs to be stretched, and the vehicle position information needs to be augmented to obtain driving trajectory data that meets the time-series processing requirements. The other approach is data reduction. Specifically, if the interval between different collection points in the collected driving trajectory data is less than the preset time interval, the driving trajectory data will be dense. In this case, the driving trajectory data needs to be scaled, and the vehicle position information needs to be reduced to obtain driving trajectory data that meets the time-series processing requirements.

[0085] By stretching or scaling the driving trajectory data to meet the corresponding time data sequence requirements, the high availability of the driving trajectory data is ensured.

[0086] Step 302: Select the vehicle location information corresponding to any collection time point as the initial time series data;

[0087] Step 303: According to the preset time interval, sequentially obtain the vehicle location information corresponding to the next time series point as the target time series data;

[0088] Step 304: Arrange the target time series data corresponding to each time series point according to the preset time series data arrangement format.

[0089] In this embodiment, the data after permutation is obtained as a time-series data sequence.

[0090] In this embodiment, after performing the step of obtaining the time-series data sequence, the method further includes: filtering out abnormal data in the time-series data sequence according to a preset filtering rule and performing a clearing process. Specifically, the preset filtering rule is as follows: obtaining all vehicle location information in the time-series data sequence, calculating the relationship between the current vehicle location information and the average value of all vehicle location information using a variance value calculation method, determining whether the variance value is greater than a preset variance threshold, and if it is greater, marking the current vehicle location information as abnormal data.

[0091] The filtering rules are used to remove abnormal data from the time-series data sequence, ensuring the availability of the input time-series data sequence and reducing the impact of abnormal data on the detection results.

[0092] Step 203: Input the time-series data sequence into the pre-trained ARIMA model to perform abnormal driving behavior detection.

[0093] In this embodiment, before performing the step of inputting the time-series data sequence into the pre-trained ARIMI model to detect abnormal driving behavior, the method further includes: obtaining an ARIMI classification model through training.

[0094] The ARIMA (Autoregressive Integrated Moving Average) model is a time series forecasting analysis method used to predict changes in an object over time based on time series data. Therefore, this application considers combining the ARIMA model with vehicle driving trajectory data to analyze and predict whether abnormal driving behavior exists. The aim is to identify vehicles with abnormal driving behavior early in the prediction phase by analyzing the vehicle's driving trajectory data, thereby reducing traffic accidents.

[0095] Continue to refer to Figure 5 The diagram illustrates a flowchart of an embodiment of obtaining an ARIMI classification model through training according to this application. The method of obtaining an ARIMI classification model through training includes the following training steps:

[0096] Step 501: Obtain a pre-labeled driving trajectory dataset, wherein a portion of the driving trajectory dataset consists of normal driving trajectory data and another portion consists of abnormal driving trajectory data;

[0097] Step 502: According to the preset time interval, arrange each driving trajectory data in the driving trajectory dataset into time sequence data to obtain a time sequence data sequence with the same number of elements as the driving trajectory dataset;

[0098] In this embodiment, after performing the step of arranging each driving trajectory data in the driving trajectory dataset according to the preset time interval to obtain a time-series data sequence with the same number of elements as the driving trajectory dataset, the method further includes: performing a stationarity test and a non-white noise test on the time-series data sequence corresponding to each element in the driving trajectory dataset in advance. Specifically, the ADF test method is used to perform a stationarity test on the time-series data sequence corresponding to each element in the driving trajectory dataset, and the point value distribution probability statistics method is used to perform a non-white noise test on the time-series data sequence corresponding to each element in the driving trajectory dataset; based on the stationarity test results and the non-white noise test results, time-series data sequences that do not meet the anomaly detection requirements are removed, and time-series data sequences that meet the anomaly detection requirements are retained.

[0099] In this embodiment, the step of using the ADF (Augmented Dickey-Fuller test, unit root) method to test the stationarity of the time-series data sequence corresponding to each element in the driving trajectory dataset specifically includes: determining whether the time-series data sequence has a unit root; if no unit root exists, the time-series data sequence is stationary; if a unit root exists, the time-series data sequence is non-stationary. Accordingly, a stationary time-series data sequence meets the anomaly detection requirements, while a non-stationary time-series data sequence does not meet the anomaly detection requirements.

[0100] In this embodiment, the step of performing non-white noise testing on the time-series data sequence corresponding to each element in the driving trajectory dataset using the point value distribution probability statistics method specifically includes: judging the distribution of vehicle position information in the time-series data sequence; if the vehicle position information in the time-series data sequence changes according to a preset pattern, then the anomaly detection requirement is met; if the vehicle position information in the time-series data sequence is too scattered, then the anomaly detection requirement is not met. Since anomaly detection is performed on vehicle driving trajectory data, it is obvious that data reflecting vehicle position information changing according to a certain pattern meets the anomaly detection requirement. If the vehicle position information in the time-series data sequence is too scattered and has no pattern to follow, it is obviously fake data and does not meet the anomaly detection requirement.

[0101] By performing stationarity and non-white noise tests on the time-series data sequences corresponding to each element in the driving trajectory dataset, it is ensured that the time-series data sequences corresponding to each element in the final retained driving trajectory dataset meet the test requirements. At the same time, it also avoids the impact of useless data on model training and ensures the anomaly detection accuracy of the ARIMI model.

[0102] Step 503: Using the preset filtering rules, abnormal data is cleared from each time series data sequence to obtain the input sequence;

[0103] Step 504: Input the input sequences sequentially into the pre-built ARIMA model for incremental classification training to obtain the normal or abnormal classification results corresponding to each input sequence.

[0104] Step 505, continue until the classification result matches the labeling result, and obtain the pre-trained ARIMA model.

[0105] In this embodiment, the step of inputting the time-series data sequence into the pre-trained ARIMA model to detect abnormal driving behavior specifically includes: inputting the time-series data sequence into the pre-trained ARIMA model and obtaining the model output result; if the model output result is normal, then the driving behavior corresponding to the time-series data sequence is normal; if the model output result is abnormal, then the driving behavior corresponding to the time-series data sequence is abnormal.

[0106] Step 204: Repeat the above detection steps and calculate the ratio of abnormal driving behavior of the vehicle using a probability algorithm.

[0107] In this embodiment, the step of repeatedly performing the above detection steps and calculating the ratio of abnormal driving behavior of the vehicle using a probability algorithm specifically includes: counting the total number of repeated detections and the number of times the model output result is abnormal; and calculating the ratio of abnormal driving behavior of the vehicle based on the probability algorithm, the total number of repeated detections, and the number of times the model output result is abnormal.

[0108] In this embodiment, the driving trajectory data of the vehicle under test is acquired in real time by repeatedly executing the above detection steps. This ensures that a vehicle can be analyzed multiple times for different driving trajectories throughout the entire driving process. Through multiple data analyses and anomaly detections, the driving behavior of the target vehicle can be comprehensively identified to determine whether there are any abnormalities. This is more scientific and reasonable, and no longer relies solely on a single anomaly detection to determine whether there are any abnormalities in the vehicle's driving. It also better reflects the humane approach of traffic management departments in law enforcement and avoids administrative errors caused by accidental phenomena.

[0109] Step 205: If the ratio is greater than a preset ratio threshold, obtain the vehicle's registration information and send a reminder to relevant departments to pay close attention.

[0110] If the ratio is greater than the preset threshold value when the above detection steps are repeated, it indicates that the target vehicle frequently exhibits abnormal driving behavior. Therefore, sending a key attention reminder to relevant departments can not only alert them to the source of danger in advance, but also contact the person in charge as soon as possible to correct the driving behavior, thereby reducing the occurrence of dangerous accidents to a certain extent.

[0111] Furthermore, by combining artificial intelligence models with real-time detection, the system comprehensively analyzes whether there are any abnormalities in the driving behavior of target vehicles. Compared with existing methods of stopping and screening vehicles, this approach is more intelligent and scientific. It can combine high-speed real-time capture equipment to analyze whether there are any abnormal driving behaviors in vehicles in motion, thereby minimizing the occurrence of major traffic accidents.

[0112] Furthermore, by combining artificial intelligence models with real-time detection, a comprehensive analysis of the target vehicle's driving behavior can be conducted to identify any abnormalities. This can serve as prior data to provide some support for car insurance pricing and enrich the dimensions of factors influencing car insurance pricing.

[0113] This application acquires the driving trajectory data of the vehicle to be detected in real time; arranges the driving trajectory data into a time-series data sequence according to a preset time interval; inputs the time-series data sequence into a pre-trained ARIMA model for abnormal driving behavior detection; repeats the above detection steps, and calculates the ratio of vehicles exhibiting abnormal driving behavior using a probability algorithm; if the ratio is greater than a preset threshold, the registration information of the vehicle is obtained, and a key attention reminder is sent to relevant departments. This application considers combining the ARIMA model and the vehicle's driving trajectory data to analyze and predict whether abnormal driving behavior exists. Simultaneously, during the ARIMA model training phase, stationarity and non-white noise tests are performed on the training set to ensure that the time-series data sequence corresponding to each element in the final retained driving trajectory dataset meets the test requirements, thus guaranteeing the anomaly detection accuracy of the ARIMA model. By combining artificial intelligence models and real-time detection, a comprehensive analysis of whether the target vehicle's driving behavior is abnormal is conducted. Compared to existing vehicle interception and screening methods, this approach is more intelligent and scientific, and can combine high-speed real-time capture equipment to analyze whether a moving vehicle exhibits abnormal driving behavior, minimizing the occurrence of major traffic accidents.

[0114] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0115] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0116] In this embodiment, the ARIMA model and vehicle driving trajectory data are combined to analyze and predict whether there are any abnormalities in driving behavior. Simultaneously, during the ARIMA model training phase, stationarity and non-white noise tests are performed on the training set to ensure that the time-series data sequence corresponding to each element in the final retained driving trajectory dataset meets the test requirements, thus guaranteeing the anomaly detection accuracy of the ARIMA model. By combining artificial intelligence models with real-time detection, the method comprehensively analyzes whether the target vehicle's driving behavior is abnormal. Compared to existing vehicle interception and screening methods, this approach is more intelligent and scientific, and can combine high-speed real-time capture equipment to analyze whether there are any abnormalities in the driving behavior of vehicles in motion, thereby minimizing the occurrence of major traffic accidents.

[0117] Further reference Figure 6 As a response to the above Figure 2 The implementation of the method shown in this application provides an embodiment of a time-series-based abnormal driving behavior detection device, which is similar to... Figure 2 Corresponding to the method embodiments shown, this device can be specifically applied to various electronic devices.

[0118] like Figure 6 As shown, the time-series-based abnormal driving behavior detection device 600 described in this embodiment includes: a data acquisition module 601, a time-series data conversion module 602, an ARIMA model detection module 603, a probability algorithm calculation module 604, and an abnormal driving alert module 605. Wherein:

[0119] The data acquisition module 601 is used to acquire the driving trajectory data of the vehicle to be detected in real time. The driving trajectory data includes each collection time point and the corresponding vehicle location information at each collection time point.

[0120] The time-series data conversion module 602 is used to arrange the driving trajectory data into time-series data according to a preset time interval to obtain a time-series data sequence;

[0121] The ARIMA model detection module 603 is used to input the time series data sequence into the pre-trained ARIMA model to detect abnormal driving behavior.

[0122] The probability algorithm calculation module 604 is used to repeatedly execute the above detection steps and calculate the ratio of abnormal driving behavior of the vehicle through a probability algorithm.

[0123] The abnormal driving alert module 605 is used to obtain the vehicle's registration information and send a reminder to relevant departments to pay close attention if the ratio is greater than a preset ratio threshold.

[0124] Continue to refer to Figure 7 , Figure 7 yes Figure 6 The schematic diagram of a specific embodiment of the time-series data conversion module 602 shown below illustrates that the time-series data conversion module 602 includes: a preprocessing submodule 701, an initial time-series data determination submodule 702, a time-series data sequential acquisition submodule 703, and a time-series data arrangement processing submodule 704. Wherein:

[0125] The preprocessing submodule 701 is used to preprocess the driving trajectory data based on the preset time interval and the various collection time points, wherein the preprocessing includes data amplification or data deletion.

[0126] The initial time series data determination submodule 702 is used to select the vehicle location information corresponding to any acquisition time point as the initial time series data;

[0127] The time-series data acquisition submodule 703 is used to sequentially acquire the vehicle location information corresponding to the next time series point as the target time-series data according to the preset time interval;

[0128] The time series data arrangement and processing submodule 704 is used to arrange the target time series data corresponding to each time series point according to the preset time series data arrangement format.

[0129] Continue to refer to Figure 8 , Figure 8 yes Figure 7 The schematic diagram shows a specific embodiment of the preprocessing submodule 701. The preprocessing submodule 701 includes: a time interval acquisition unit 801, a time interval size determination unit 802, a first preprocessing unit 803, and a second preprocessing unit 804. Wherein:

[0130] The time interval acquisition unit 801 is used to acquire the time interval between two adjacent acquisition time points in each acquisition time point;

[0131] The time interval size determination unit 802 is used to determine the size relationship between the preset time interval and the time interval between two adjacent acquisition time points;

[0132] The first preprocessing unit 803 is used to scale the driving trajectory data and delete the vehicle location information if the preset time interval is greater than the time interval between two adjacent collection time points.

[0133] The second preprocessing unit 804 is used to stretch the driving trajectory data and amplify the vehicle location information if the preset time interval is less than the time interval between two adjacent acquisition time points.

[0134] In some specific embodiments of this application, the time-series-based abnormal driving behavior detection device 600 further includes: an ARIMA model pre-training module, which can be referenced further. Figure 9 , Figure 9 This is a schematic diagram of a specific embodiment of the ARIMA model pre-training module according to this application. The ARIMA model pre-training module includes: a training data acquisition submodule 901, a time-series data sequence generation submodule 902, an outlier data removal submodule 903, a classification model training submodule 904, and a comparison and recognition submodule 905. Wherein:

[0135] The training data acquisition submodule 901 is used to acquire a pre-labeled driving trajectory dataset, wherein a portion of the driving trajectory dataset consists of normal driving trajectory data and another portion consists of abnormal driving trajectory data.

[0136] The time-series data sequence generation submodule 902 is used to perform time-series data arrangement on each driving trajectory data in the driving trajectory dataset according to the preset time interval, so as to obtain a time-series data sequence with the same number of elements as the driving trajectory dataset.

[0137] The abnormal data removal submodule 903 is used to remove abnormal data from each time series data sequence using the preset filtering rules to obtain the input sequence;

[0138] The classification model training submodule 904 is used to sequentially input the input sequences to the pre-built ARIMA model for incremental classification training to obtain the normal or abnormal classification results corresponding to each input sequence.

[0139] The comparison and recognition submodule 905 is used to obtain the pre-trained ARIMA model until the classification result matches the labeling result.

[0140] This application acquires the driving trajectory data of the vehicle to be detected in real time; arranges the driving trajectory data into a time-series data sequence according to a preset time interval; inputs the time-series data sequence into a pre-trained ARIMA model for abnormal driving behavior detection; repeats the above detection steps, and calculates the ratio of vehicles exhibiting abnormal driving behavior using a probability algorithm; if the ratio is greater than a preset threshold, the registration information of the vehicle is obtained, and a key attention reminder is sent to relevant departments. This application considers combining the ARIMA model and the vehicle's driving trajectory data to analyze and predict whether abnormal driving behavior exists. Simultaneously, during the ARIMA model training phase, stationarity and non-white noise tests are performed on the training set to ensure that the time-series data sequence corresponding to each element in the final retained driving trajectory dataset meets the test requirements, thus guaranteeing the anomaly detection accuracy of the ARIMA model. By combining artificial intelligence models and real-time detection, a comprehensive analysis of whether the target vehicle's driving behavior is abnormal is conducted. Compared to existing vehicle interception and screening methods, this approach is more intelligent and scientific, and can combine high-speed real-time capture equipment to analyze whether a moving vehicle exhibits abnormal driving behavior, minimizing the occurrence of major traffic accidents.

[0141] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing related hardware through computer-readable instructions. These computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the methods described above. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, or read-only memory (ROM), or random access memory (RAM).

[0142] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0143] To address the aforementioned technical problems, embodiments of this application also provide a computer device. Please refer to [link / reference needed]. Figure 10 , Figure 10 This is a basic structural block diagram of the computer device in this embodiment.

[0144] The computer device 10 includes a memory 10a, a processor 10b, and a network interface 10c that are interconnected via a system bus. It should be noted that only the computer device 10 with components 10a-10c is shown in the figure; however, it should be understood that it is not required to implement all the shown components, and more or fewer components can be implemented alternatively. Those skilled in the art will understand that the computer device described here is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0145] The computer device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer device can interact with the user via a keyboard, mouse, remote control, touchpad, or voice control.

[0146] The memory 10a includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 10a may be an internal storage unit of the computer device 10, such as the hard disk or memory of the computer device 10. In other embodiments, the memory 10a may also be an external storage device of the computer device 10, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 10. Of course, the memory 10a may also include both the internal storage unit and the external storage device of the computer device 10. In this embodiment, the memory 10a is typically used to store the operating system and various application software installed on the computer device 10, such as computer-readable instructions for a time-series-based method for detecting abnormal driving behavior. Furthermore, the memory 10a can also be used to temporarily store various types of data that have already been output or will be output.

[0147] In some embodiments, the processor 10b may be a central processing unit (CPU), controller, microcontroller, microprocessor, or other data processing chip. The processor 10b is typically used to control the overall operation of the computer device 10. In this embodiment, the processor 10b is used to execute computer-readable instructions stored in the memory 10a or to process data, for example, to execute computer-readable instructions for the time-series-based driving behavior anomaly detection method.

[0148] The network interface 10c may include a wireless network interface or a wired network interface, which is typically used to establish communication connections between the computer device 10 and other electronic devices.

[0149] The computer equipment proposed in this embodiment belongs to the field of financial technology and is applied to vehicle driving detection. It can assist in market research and analysis for vehicle accident identification and car insurance pricing. This application acquires the driving trajectory data of the vehicle to be detected in real time; arranges the driving trajectory data into time series data according to a preset time interval to obtain a time series data sequence; inputs the time series data sequence into a pre-trained ARIMA model to detect abnormal driving behavior; repeats the above detection steps and calculates the ratio of abnormal driving behavior of the vehicle through a probability algorithm; if the ratio is greater than a preset ratio threshold, the registration information of the vehicle is obtained and a key attention reminder is sent to relevant departments. This application considers combining the ARIMA model and the vehicle's driving trajectory data to analyze and predict whether there are abnormal driving behaviors. At the same time, during the ARIMA model training phase, stationarity tests and non-white noise tests are performed on the training set to ensure that the time series data sequence corresponding to each element in the final retained driving trajectory dataset meets the test requirements, thus ensuring the anomaly detection accuracy of the ARIMA model. By combining artificial intelligence models with real-time detection, the system comprehensively analyzes whether the driving behavior of target vehicles is abnormal. Compared with existing methods of stopping and screening vehicles, this approach is more intelligent and scientific. It can combine high-speed real-time capture equipment to analyze whether there are any abnormal driving behaviors in moving vehicles, thereby minimizing the occurrence of major traffic accidents.

[0150] This application also provides another embodiment, namely, providing a computer-readable storage medium storing computer-readable instructions that can be executed by a processor to cause the processor to perform the steps of the time-series-based abnormal driving behavior detection method described above.

[0151] The computer-readable storage medium proposed in this embodiment belongs to the field of financial technology and is applied to vehicle driving detection. It can assist in market research and analysis for vehicle accident identification and car insurance pricing. This application acquires the driving trajectory data of the vehicle to be detected in real time; arranges the driving trajectory data into time series according to a preset time interval to obtain a time series data sequence; inputs the time series data sequence into a pre-trained ARIMI model to detect abnormal driving behavior; repeats the above detection steps and calculates the ratio of abnormal driving behavior of the vehicle using a probability algorithm; if the ratio is greater than a preset ratio threshold, the registration information of the vehicle is obtained and a key attention reminder is sent to relevant departments. This application considers combining the ARIMI model and the vehicle's driving trajectory data to analyze and predict whether there are abnormal driving behaviors. At the same time, during the ARIMI model training phase, stationarity and non-white noise tests are performed on the training set to ensure that the time series data sequence corresponding to each element in the final retained driving trajectory dataset meets the test requirements, thus ensuring the anomaly detection accuracy of the ARIMI model. By combining artificial intelligence models with real-time detection, the system comprehensively analyzes whether the driving behavior of target vehicles is abnormal. Compared with existing methods of stopping and screening vehicles, this approach is more intelligent and scientific. It can combine high-speed real-time capture equipment to analyze whether there are any abnormal driving behaviors in moving vehicles, thereby minimizing the occurrence of major traffic accidents.

[0152] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0153] Obviously, the embodiments described above are only some embodiments of this application, not all embodiments. The accompanying drawings show preferred embodiments of this application, but do not limit the patent scope of this application. This application can be implemented in many different forms; rather, the purpose of providing these embodiments is to provide a more thorough and comprehensive understanding of the disclosure of this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of this application's specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the scope of patent protection of this application.

Claims

1. A method for detecting abnormal driving behavior based on time series, characterized in that, Includes the following steps: Real-time acquisition of driving trajectory data of the vehicle to be detected, wherein the driving trajectory data includes each collection time point and the corresponding vehicle location information at each collection time point; The driving trajectory data is arranged in a time sequence according to a preset time interval to obtain a time sequence data sequence; The time-series data sequence is input into a pre-trained ARIMA model for driving behavior anomaly detection. Prior to the step of inputting the time-series data sequence into the pre-trained ARIMA model for driving behavior anomaly detection, the method further includes: Obtain a pre-labeled driving trajectory dataset. Specifically, the driving trajectory dataset consists of normal driving trajectory data in one part and abnormal driving trajectory data in the other part. According to the preset time interval, each driving trajectory data in the driving trajectory dataset is arranged into time-series data to obtain a time-series data sequence with the same number of elements as the driving trajectory dataset. Specifically, after performing the step of arranging each driving trajectory data in the driving trajectory dataset into time-series data to obtain a time-series data sequence with the same number of elements as the driving trajectory dataset, the method further includes: performing stationarity and non-white noise tests on the time-series data sequence corresponding to each element in the driving trajectory dataset in advance. Specifically, the ADF test method is used to perform stationarity tests on the time-series data sequence corresponding to each element in the driving trajectory dataset, and the point value distribution probability statistics method is used to perform non-white noise tests on the time-series data sequence corresponding to each element in the driving trajectory dataset. Based on the stationarity test results and non-white noise test results, time-series data sequences that do not meet the anomaly detection requirements are removed, and time-series data sequences that meet the anomaly detection requirements are retained. Using preset filtering rules, abnormal data is removed from time-series data sequences that meet the anomaly detection requirements to obtain the input sequence. The specific filtering method is as follows: obtain all vehicle location information in the time-series data sequence, calculate the relationship between the current vehicle location information and the average value of all vehicle location information using the variance value calculation method, and determine whether the variance value is greater than the preset variance threshold. If it is greater, the current vehicle location information is marked as abnormal data. The input sequences are sequentially fed into a pre-built ARIMA model for incremental classification training to obtain normal or abnormal classification results for each input sequence. The pre-trained ARIMA model is obtained when the classification result matches the labeling result. Repeat the above detection steps and calculate the ratio of vehicles exhibiting abnormal driving behavior using a probability algorithm; If the ratio is greater than a preset ratio threshold, the vehicle's registration information is obtained, and a reminder for key attention is sent to relevant departments.

2. The method for detecting abnormal driving behavior based on time series as described in claim 1, characterized in that, The step of arranging the driving trajectory data in time sequence according to a preset time interval specifically includes: Based on the preset time interval and the various collection time points, the driving trajectory data is preprocessed, wherein the preprocessing includes data augmentation or data deletion. Select the vehicle location information corresponding to any collection time point as the initial time series data; According to the preset time interval, the vehicle location information corresponding to the next time series point is obtained sequentially as the target time series data; The target time series data corresponding to each time series point is arranged according to the preset time series data arrangement format.

3. The time-series-based method for detecting abnormal driving behavior according to claim 2, characterized in that, The step of preprocessing the driving trajectory data based on the preset time interval and the various collection time points specifically includes: Obtain the time interval between two adjacent acquisition time points in each of the acquisition time points; Determine the relationship between the preset time interval and the time interval between two adjacent acquisition time points; If the preset time interval is greater than the time interval between two adjacent collection time points, the driving trajectory data is scaled and the vehicle location information is deleted. If the preset time interval is less than the time interval between two adjacent collection time points, the driving trajectory data is stretched and the vehicle location information is amplified.

4. The time-series-based method for detecting abnormal driving behavior according to any one of claims 1 to 3, characterized in that, The step of inputting the time-series data sequence into the pre-trained ARIMA model to detect abnormal driving behavior specifically includes: The time series data sequence is input into the pre-trained ARIMA model to obtain the model output results; If the model output is normal, then the driving behavior corresponding to the time series data sequence is normal. If the model output is abnormal, then the driving behavior corresponding to the time series data sequence is abnormal. The step of repeatedly performing the above detection steps and calculating the ratio of abnormal driving behavior in the vehicle using a probability algorithm specifically includes: The total number of repeated detections and the number of times the model output was an anomaly were counted. The ratio of abnormal driving behavior of the vehicle is calculated based on the probability algorithm, the total number of repeated detections, and the number of times the model output result is abnormal.

5. A time-series-based abnormal driving behavior detection device, characterized in that, The time-series-based driving behavior anomaly detection device implements the steps of the time-series-based driving behavior anomaly detection method as described in any one of claims 1 to 4, wherein the driving behavior anomaly detection device comprises: The detection data acquisition module is used to acquire the driving trajectory data of the vehicle to be detected in real time. The driving trajectory data includes each collection time point and the corresponding vehicle location information at each collection time point. The time-series data conversion module is used to arrange the driving trajectory data into time-series data according to a preset time interval to obtain a time-series data sequence; The ARIMA model detection module is used to input the time series data sequence into the pre-trained ARIMA model to detect abnormal driving behavior. The probability algorithm calculation module is used to repeatedly execute the above detection steps and calculate the ratio of abnormal driving behavior of the vehicle through a probability algorithm. The abnormal driving alert module is used to obtain the vehicle's registration information and send a reminder to relevant departments to pay close attention if the ratio is greater than a preset ratio threshold.

6. A computer device comprising a memory and a processor, the memory storing computer-readable instructions, wherein the processor, when executing the computer-readable instructions, implements the steps of the time-series-based abnormal driving behavior detection method as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the time-series-based driving behavior anomaly detection method as described in any one of claims 1 to 4.

Images