Method, device, storage medium and computer equipment for monitoring trajectory of operating vehicles

Abstract

The present invention discloses a trajectory monitoring method, device, storage medium and computer equipment for operating vehicles. The method includes: obtaining the trajectory data and business type of the vehicle to be monitored; determining the trajectory prediction model corresponding to the vehicle to be monitored according to the business type of the vehicle to be monitored, wherein the trajectory prediction model is obtained based on the training of the isolation forest algorithm and the random forest algorithm; inputting the trajectory data of the vehicle to be monitored into the trajectory prediction model to obtain the trajectory abnormality probability value of the vehicle to be monitored. The above method obtains the trajectory prediction model based on the training of the isolation forest algorithm and the random forest algorithm, and predicts the trajectory abnormality probability value of the vehicle to be monitored through the trajectory prediction model, which can timely and accurately monitor the probability of trajectory abnormality of the operating vehicle, thereby effectively improving the timeliness and accuracy of the driving trajectory monitoring of the operating vehicle.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence and big data processing technology, and in particular to a trajectory monitoring method, device, storage medium and computer equipment for operating vehicles. Background Art

[0002] With the rapid development of my country's automobile industry and the continuous improvement of people's living standards, cars have gradually become a very important means of transportation in people's lives. In this context, car rental services are becoming more and more popular in the market because they cater to the temporary car needs of some users.

[0003] Generally speaking, operating vehicles (rented vehicles) can be divided into many types according to different business types, such as long-term rental, short-term rental, online rental and online self-operation, etc. Various types of operating vehicles may encounter various problems during operation, such as loss, modification and damage, etc. In order to avoid such problems, the lessor usually installs GPS devices on the operating vehicles, and then uses the GPS devices to monitor the driving trajectory and vehicle location of the operating vehicles in real time.

[0004] In the prior art, vehicle lessors usually determine whether there are operational risks in operating vehicles by monitoring the trajectory of operating vehicles, such as whether they often enter dangerous sections of road, whether they often stop near second-hand car markets or repair shops, and whether they travel too far. However, this simple trajectory monitoring method cannot help vehicle lessors timely and accurately understand the actual operational risks of operating vehicles. In addition, since there are many types of operating vehicles and the driving trajectories of each type are different, it is more difficult to monitor operating vehicles and reduces the overall accuracy of operating vehicle trajectory monitoring, which ultimately causes property losses to vehicle lessors. Summary of the invention

[0005] In view of this, the present application provides a trajectory monitoring method, device, storage medium and computer equipment for operating vehicles, the main purpose of which is to solve the technical problem that the driving trajectory monitoring of operating vehicles is poor in timeliness and accuracy.

[0006] According to a first aspect of the present invention, a trajectory monitoring method for an operating vehicle is provided, the method comprising:

[0007] Obtain the trajectory data and business type of the vehicle to be monitored;

[0008] According to the business type of the vehicle to be monitored, a trajectory prediction model corresponding to the vehicle to be monitored is determined, wherein the trajectory prediction model is obtained by training based on an isolation forest algorithm and a random forest algorithm;

[0009] The trajectory data of the vehicle to be monitored is input into the trajectory prediction model to obtain the abnormal probability value of the trajectory of the vehicle to be monitored.

[0010] According to a second aspect of the present invention, there is provided a trajectory monitoring device for an operating vehicle, the device comprising:

[0011] A data acquisition module is used to obtain the trajectory data and business type of the vehicle to be monitored;

[0012] A model determination module is used to determine a trajectory prediction model corresponding to the vehicle to be monitored according to the business type of the vehicle to be monitored, wherein the trajectory prediction model is obtained by training based on an isolation forest algorithm and a random forest algorithm;

[0013] The data processing module is used to input the trajectory data of the vehicle to be monitored into the trajectory prediction model to obtain the trajectory abnormality probability value of the vehicle to be monitored.

[0014] According to a third aspect of the present invention, there is provided a storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the above-mentioned trajectory monitoring method for operating vehicles.

[0015] According to a fourth aspect of the present invention, there is provided a computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned method for monitoring the trajectory of operating vehicles when executing the program.

[0016] The present invention provides a trajectory monitoring method, device, storage medium and computer equipment for operating vehicles. The method first obtains the trajectory data and business type of the vehicle to be monitored, then determines the trajectory prediction model corresponding to the vehicle to be monitored according to the business type, and finally obtains the trajectory abnormality probability value of the vehicle to be monitored through the trajectory prediction model. In the above method, the trajectory prediction model is obtained based on the training of the isolation forest algorithm and the random forest algorithm, wherein the isolation forest algorithm can separate the normal trajectory and the abnormal trajectory in the sample, thereby providing sufficient predicted sample data for the random forest algorithm, and the random forest algorithm can accurately predict the abnormal probability of the trajectory data based on the predicted sample data, thereby timely and accurately monitoring the probability of abnormal trajectory of the operating vehicle, thereby effectively improving the timeliness and accuracy of the driving trajectory monitoring of the operating vehicle.

[0017] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0019] Figure 1 A schematic diagram of a process flow of a trajectory monitoring method for an operating vehicle provided by an embodiment of the present invention is shown.

[0020] Figure 2 A schematic diagram of a flow chart of a method for training a trajectory prediction model provided by an embodiment of the present invention is shown.

[0021] Figure 3 A schematic structural diagram of a trajectory monitoring device for an operating vehicle provided in an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0022] The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other without conflict.

[0023] In one embodiment, Figure 1 As shown, a trajectory monitoring method for operating vehicles is provided, which is described by taking the method applied to a computer device such as a server as an example, and includes the following steps:

[0024] 101. Obtain the trajectory data and business type of the vehicle to be monitored.

[0025] Among them, the vehicle to be monitored refers to an operating vehicle that needs to be tracked, and the operating vehicle refers to a vehicle provided to the vehicle lessee or to an individual driver by the vehicle operator or the vehicle lessor in the form of leasing, operating lease, self-operation, etc. Generally speaking, operating vehicles can be divided into a variety of different business types according to their different leasing, operating lease, and self-operation methods, such as long-term rental, short-term rental, online rental, and online self-operation, etc. In this embodiment, the computer equipment can collect the trajectory data of the vehicle to be monitored in real time through a GPS device installed on the vehicle to be monitored, and can obtain the current business type of the vehicle to be monitored in the database according to the vehicle identification (such as the license plate number) of the vehicle to be monitored. It can be understood that the number of vehicles to be monitored can be one or more, and the business type of the vehicle to be monitored can change according to the actual change of the operating mode. In order to facilitate the explanation of the specific meaning of the business type of the operating vehicle, the following definitions of several business types are provided for reference.

[0026] In this embodiment, long-term lease refers to a vehicle lease contract of more than 6 months, and the lessees are mainly enterprises and institutions. During the contract period, the vehicle is used as a business vehicle for enterprises and institutions, and the driver is arranged by the lessee himself, which has nothing to do with the lessor. Short-term lease refers to a vehicle lease contract of no more than 60 days, and the lessees are mainly individual customers. During the contract period, the vehicle is used by the lessee's customer. Online booking and leasing refers to the vehicle belonging to the lessor, and individual drivers rent vehicles on the online car-hailing platform, conduct online car-hailing business, and get remuneration and pay rent. Online booking and self-operation refers to the vehicle belonging to the online car-hailing platform, the platform recruits and trains drivers on its own, the drivers conduct online car-hailing business, and the platform pays the drivers. It should be noted that in addition to the above-mentioned business types, operating vehicles can also have other business types, which are not listed one by one in this embodiment.

[0027] 102. Determine a trajectory prediction model corresponding to the vehicle to be monitored according to the business type of the vehicle to be monitored, wherein the trajectory prediction model is trained based on an isolation forest algorithm and a random forest algorithm.

[0028] Among them, the trajectory prediction model refers to a model for predicting the trajectory of an operating vehicle. The input of the model is the trajectory data of the operating vehicle, and the output is the trajectory anomaly probability value of the operating vehicle. The probability of anomaly in the trajectory of the monitored vehicle can be known through the trajectory anomaly probability value. That is, the higher the trajectory anomaly probability value, the higher the operating risk of the monitored vehicle, and vice versa.

[0029] Specifically, the trajectory prediction model can be different according to the type of operating vehicle, that is, each business type corresponds to a corresponding trajectory prediction model. In this way, the accuracy of trajectory monitoring of operating vehicles of different business types can be improved. In this embodiment, before model training, the sample data can be first classified, and then the corresponding trajectory prediction model can be trained according to the data samples corresponding to each classified business type, so as to obtain the trajectory prediction model corresponding to each business type. In addition, the trajectory prediction model corresponding to each business type can be obtained after training through the isolation forest algorithm and the random forest algorithm. Among them, the isolation forest algorithm is an unsupervised anomaly detection method suitable for continuous data, that is, the abnormal data in the sample can be separated without labeled samples; the random forest algorithm is an algorithm that integrates multiple decision trees into a classification model through ensemble learning. Specifically, the random splitting sample strategy of the isolation forest algorithm can be used to accurately find the isolated abnormal sample data in a large amount of data, that is, to find the abnormal trajectory data in a large amount of historical trajectory data, and then use the trained isolation forest model to perform binary classification on a large number of historical trajectory samples, that is, to divide the historical trajectory data into normal trajectories and abnormal trajectories. Finally, the classified historical trajectory data is used as samples of the random forest to train multiple decision trees in the random forest, and finally a trajectory prediction model that can predict the probability value of trajectory anomaly is obtained.

[0030] 103. Input the trajectory data of the vehicle to be monitored into the trajectory prediction model to obtain an abnormal probability value of the trajectory of the vehicle to be monitored.

[0031] Specifically, after determining the trajectory prediction model corresponding to the vehicle to be monitored, the computer device can input the trajectory data of the vehicle to be monitored into the trajectory prediction model to obtain the trajectory abnormality probability value of the vehicle to be monitored. Then, the computer device can display or output the trajectory abnormality probability value so that the user can promptly know the abnormal condition of the trajectory of the vehicle to be monitored. In addition, an alarm threshold can also be set, and an alarm prompt can be automatically issued when the trajectory abnormality probability value of the vehicle to be monitored exceeds the alarm threshold, or an alarm prompt can be automatically issued when the trajectory abnormality probability value of the vehicle to be monitored continues to increase, etc. It can be understood that there can be many application scenarios for the trajectory abnormality probability value, and this embodiment will not cite examples one by one here. Conventional data output methods in the prior art are all applicable to this embodiment.

[0032] The trajectory monitoring method for operating vehicles provided in this embodiment first obtains the trajectory data and business type of the vehicle to be monitored, then determines the trajectory prediction model corresponding to the vehicle to be monitored according to the business type, and finally obtains the trajectory abnormality probability value of the vehicle to be monitored through the trajectory prediction model. In the above method, the trajectory prediction model is obtained based on the training of the isolation forest algorithm and the random forest algorithm, wherein the isolation forest algorithm can separate the normal trajectory and the abnormal trajectory in the sample, thereby providing sufficient predicted sample data for the random forest algorithm, and the random forest algorithm can accurately predict the abnormal probability of the trajectory data based on the predicted sample data, thereby timely and accurately monitoring the probability of trajectory abnormality of the operating vehicle, thereby effectively improving the timeliness and accuracy of the driving trajectory monitoring of the operating vehicle.

[0033] In one embodiment, step 101 can be implemented in the following manner: first, the location data of the vehicle to be monitored is collected in real time by a GPS device installed on the vehicle to be monitored, then the driving trajectory of the vehicle to be monitored is generated based on the real-time collected location data of the vehicle to be monitored, and finally, multiple positioning points are selected from the driving trajectory, and the trajectory data of the vehicle to be monitored is generated based on the longitude and latitude coordinate values ​​of the multiple positioning points. Among them, the GPS device refers to the equipment installed by the car rental party on the vehicle for position monitoring, which is generally divided into active and passive GPS. The active GPS needs to be connected to the vehicle battery, while the passive one does not; the driving trajectory refers to the trajectory route formed during the vehicle's driving process, that is, the line formed by the position of the car at each moment; the trajectory data refers to the data composed of the coordinate values ​​of multiple positioning points extracted from the driving trajectory. In this embodiment, the trajectory data of the vehicle to be monitored can be generated by extracting the longitude and latitude coordinate values ​​of the positioning points, thereby reducing the calculation amount of the trajectory data and improving the timeliness of the trajectory monitoring.

[0034] In one embodiment, the method of selecting a positioning point to generate trajectory data in step 101 can be implemented in the following manner: first, the driving trajectory is intercepted into n segments, and a positioning point is selected from each segment of the driving trajectory to obtain n positioning points of the driving trajectory, and then the n positioning points are arranged in ascending order according to the chronological order of the acquisition time of the n positioning points to obtain the arrangement sequence number of the n positioning points, and finally the arrangement sequence number of the n positioning points is used as the weight to perform weighted processing on the absolute values ​​of the latitude and longitude coordinates of the n positioning points to obtain the trajectory data of the vehicle to be monitored. In this embodiment, the driving trajectory can be intercepted on average or randomly, and the positioning point can take the median or a random position number, which is not limited in this embodiment. It is understandable that from the perspective of improving the accuracy of the positioning point value, the driving trajectory can be averaged, such as 10 segments on average, and the positioning point can take the median of the longitude and latitude coordinate values ​​of each segment of the trajectory, and then the longitude and latitude coordinate values ​​of the n positioning points are weighted according to the arrangement sequence, such as the longitude*1, latitude*1 of the first positioning point, the longitude*2, latitude*2 of the second positioning point, and so on, and finally the trajectory data of the vehicle to be monitored is obtained. It should be noted that the probability of anomalies of a trajectory at the starting position is small, while the probability of anomalies near the end position will increase. This embodiment can improve the accuracy of abnormal trajectory detection by weighting the coordinate values ​​of the positioning points.

[0035] In one embodiment, the method for obtaining the absolute value of the longitude and latitude coordinates in step 101 can be implemented in the following manner: obtain the Euclidean distance center point of the city where the vehicle to be detected is located, and calculate the distance between the Euclidean distance center point and the preset city center point. When the distance between the Euclidean distance center point and the city center point is less than the preset deviation value, the absolute value of the longitude and latitude coordinates of the n positioning points is obtained according to the difference between the longitude and latitude coordinates of the n positioning points and the Euclidean distance center point; when the distance between the Euclidean distance center point and the city center point is not less than the preset deviation value, the absolute value of the longitude and latitude coordinates of the n positioning points is obtained according to the difference between the longitude and latitude coordinates of the n positioning points and the city center point. Among them, the Euclidean distance center point refers to the position point that minimizes the total Euclidean distance between the elements in the data set. This point can be used to identify the geographic center or density center of a group of elements. The city center point can be preset. For example, the longitude and latitude of the location of the municipal government can be taken as the city center point. In this embodiment, the actual situation of each city is different, so the center point of the city will be different. For example, for some cities, the vehicle activity area is relatively concentrated and mainly around the city. At this time, the city center can be used as the basis for calculating the absolute value of the longitude and latitude coordinates; while for other cities, the vehicle activity area is relatively scattered. In this case, the Euclidean distance center point can be used as the basis for calculating the absolute value of the longitude and latitude coordinates, thereby avoiding the problem of data skew. This embodiment sets the basis for the absolute value of the longitude and latitude coordinates by the distance between the Euclidean distance center point and the city center point, which can improve the accuracy of the positioning point coordinate value, thereby improving the accuracy of vehicle trajectory monitoring in different urban situations.

[0036] In one embodiment, Figure 2 As shown, the training method of the trajectory prediction model in step 102 can be implemented in the following way:

[0037] 201. Acquire multiple pieces of historical trajectory data of multiple operating vehicles, and classify the multiple pieces of historical trajectory data according to the business type of each operating vehicle to obtain multiple pieces of historical trajectory data of each business type.

[0038] 202. According to multiple historical trajectory data of each business type, a trajectory judgment model corresponding to each business type is trained through an isolation forest algorithm.

[0039] 203. Input each piece of historical trajectory data of each service type into a trajectory judgment model of the corresponding service type to obtain a trajectory judgment result of each piece of historical trajectory data, wherein the trajectory judgment result includes normal and abnormal.

[0040] 204. According to the multiple historical trajectory data of each business type and the trajectory judgment result of each historical trajectory data, a trajectory prediction model corresponding to each business type is trained through a random forest algorithm.

[0041] In the above embodiment, multiple historical trajectory data of multiple regions, multiple business types, and multiple operating vehicles can be collected first, and then the multiple historical trajectory data can be classified according to the business type to obtain the historical trajectory data under each business type, and then the trajectory judgment model and trajectory prediction model corresponding to each business type can be trained respectively according to the classified historical trajectory data. Among them, the trajectory judgment model is obtained by training the isolation forest algorithm, and the trajectory prediction model is obtained by training the random forest algorithm. Specifically, the trajectory judgment model can perform abnormal trajectory prejudgment on the collected historical trajectory data, so as to obtain the trajectory judgment result (normal or abnormal) of each historical trajectory data, and train multiple decision trees of the random forest according to the prejudged labeled historical trajectory data (the label is the trajectory judgment result of the historical trajectory data), so as to obtain the trajectory prediction model corresponding to each business type. The above embodiment can effectively improve the accuracy of vehicle trajectory monitoring under each business type by classifying the historical trajectory data. In addition, the present embodiment obtains the trajectory prediction model by first training the isolation forest and then training the random forest, which can also effectively reduce the workload of labeling sample data, thereby improving the speed of model training.

[0042] In one embodiment, the method for obtaining a trajectory judgment model through isolation forest algorithm training can be implemented in the following manner: first, a plurality of historical trajectory sample groups are respectively extracted from a plurality of historical trajectory data of each business type, wherein each historical trajectory sample group contains a predetermined number of historical trajectory data, and then, through a recursive algorithm, a historical trajectory data is randomly selected between the maximum value and the minimum value of each historical trajectory sample group to split the historical trajectory sample group until each historical trajectory sample group cannot be split any further, thereby obtaining a plurality of binary trees corresponding to each business type, and finally, the plurality of binary trees corresponding to each business type are integrated to obtain a trajectory judgment model corresponding to each business type.

[0043] In one embodiment, the method for obtaining a trajectory prediction model through random forest algorithm training can be implemented in the following manner: first, a sample set of the trajectory prediction model is constructed according to multiple historical trajectory data of each business type and the trajectory judgment result of each historical trajectory data, and then multiple historical trajectory sample groups are randomly extracted from the training set with replacement, and a decision tree is trained according to each historical trajectory sample group to obtain multiple decision trees corresponding to each business type, and finally, the multiple decision trees corresponding to each business type are integrated to obtain a trajectory prediction model corresponding to each business type.

[0044] Further, as Figure 1 , Figure 2 The specific implementation of the method shown in the embodiment provides a trajectory monitoring device for operating vehicles, such as Figure 3 As shown, the device includes: a data acquisition module 31, a model determination module 32 and a data processing module 33.

[0045] The data acquisition module 31 can be used to obtain the trajectory data and business type of the vehicle to be monitored;

[0046] The model determination module 32 may be used to determine a trajectory prediction model corresponding to the vehicle to be monitored according to the service type of the vehicle to be monitored, wherein the trajectory prediction model is obtained by training based on an isolation forest algorithm and a random forest algorithm;

[0047] The data processing module 33 may be used to input the trajectory data of the vehicle to be monitored into the trajectory prediction model to obtain an abnormal probability value of the trajectory of the vehicle to be monitored.

[0048] In a specific application scenario, the data acquisition module 31 can be used to collect the position data of the vehicle to be monitored in real time through a GPS device installed on the vehicle to be monitored, generate the driving trajectory of the vehicle to be monitored based on the real-time collected position data of the vehicle to be monitored, select multiple positioning points from the driving trajectory, and generate the trajectory data of the vehicle to be monitored based on the latitude and longitude coordinate values ​​of the multiple positioning points.

[0049] In a specific application scenario, the data acquisition module 31 can be used to intercept the driving trajectory into n segments, and select a positioning point from each segment of the driving trajectory to obtain n positioning points of the driving trajectory, and arrange the n positioning points in ascending order according to the chronological order of the collection time of the n positioning points to obtain the arrangement number of the n positioning points, and use the arrangement number of the n positioning points as the weight to perform weighted processing on the absolute values ​​of the latitude and longitude coordinates of the n positioning points to obtain the trajectory data of the vehicle to be monitored.

[0050] In a specific application scenario, the data acquisition module 31 can be specifically used to obtain the Euclidean distance center point of the city where the vehicle to be detected is located, and calculate the distance between the Euclidean distance center point and the preset city center point. When the distance between the Euclidean distance center point and the city center point is less than the preset deviation value, the absolute values ​​of the longitude and latitude coordinates of the n positioning points are obtained based on the difference between the longitude and latitude coordinate values ​​of the n positioning points and the Euclidean distance center point. When the distance between the Euclidean distance center point and the city center point is not less than the preset deviation value, the absolute values ​​of the longitude and latitude coordinates of the n positioning points are obtained based on the difference between the longitude and latitude coordinate values ​​of the n positioning points and the city center point.

[0051] In a specific application scenario, the device also includes a model training module 34, which can be specifically used to obtain multiple historical trajectory data of multiple operating vehicles, and classify the multiple historical trajectory data according to the business type of each operating vehicle to obtain multiple historical trajectory data of each business type. Based on the multiple historical trajectory data of each business type, a trajectory judgment model corresponding to each business type is trained through an isolation forest algorithm, and each historical trajectory data of each business type is input into the trajectory judgment model of the corresponding business type to obtain a trajectory judgment result of each historical trajectory data, wherein the trajectory judgment result includes normal and abnormal. Based on the multiple historical trajectory data of each business type and the trajectory judgment result of each historical trajectory data, a trajectory prediction model corresponding to each business type is trained through a random forest algorithm.

[0052] In a specific application scenario, the model training module 34 can be specifically used to extract multiple historical trajectory sample groups from multiple historical trajectory data of each business type, wherein each historical trajectory sample group contains a predetermined number of historical trajectory data, and through a recursive algorithm, randomly select a historical trajectory data between the maximum value and the minimum value of each historical trajectory sample group to split the historical trajectory sample group until each historical trajectory sample group can no longer be split, thereby obtaining multiple binary trees corresponding to each business type, and integrating the multiple binary trees corresponding to each business type to obtain a trajectory judgment model corresponding to each business type.

[0053] In a specific application scenario, the model training module 34 can be used to construct a sample set of the trajectory prediction model based on multiple historical trajectory data of each business type and the trajectory judgment results of each historical trajectory data, randomly extract multiple historical trajectory sample groups from the training set with replacement, and train a decision tree based on each historical trajectory sample group to obtain multiple decision trees corresponding to each business type, and integrate the multiple decision trees corresponding to each business type to obtain a trajectory prediction model corresponding to each business type.

[0054] It should be noted that for other corresponding descriptions of the functional units involved in the trajectory monitoring device for an operating vehicle provided in this embodiment, reference can be made to Figure 1 , Figure 2 The corresponding description in will not be repeated here.

[0055] Based on the above Figure 1 , Figure 2 The method shown in the embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the above-mentioned Figure 1 , Figure 2 The trajectory monitoring method of the operating vehicle is shown.

[0056] Based on this understanding, the technical solution of the present application can be embodied in the form of a software product. The software product to be identified can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present application.

[0057] Based on the above Figure 1 , Figure 2 The method shown, and Figure 3 The embodiment of the trajectory monitoring device for operating vehicles shown in the figure, in order to achieve the above purpose, this embodiment also provides a physical device for operating vehicle trajectory monitoring, which can be a personal computer, a server, a smart phone, a tablet computer, a smart watch, or other network devices, etc. The physical device includes a storage medium and a processor, the storage medium is used to store a computer program, and the processor is used to execute the computer program to achieve the above Figure 1 , Figure 2 The method shown.

[0058] Optionally, the physical device may also include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a WI-FI module, etc. The user interface may include a display, an input unit such as a keyboard, etc., and the optional user interface may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (such as a WI-FI interface), etc.

[0059] Those skilled in the art will appreciate that the physical device structure for trajectory monitoring of an operating vehicle provided in this embodiment does not constitute a limitation on the physical device, and may include more or fewer components, or a combination of certain components, or different component arrangements.

[0060] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages the hardware of the physical device and the software resources to be identified, and supports the operation of the information processing program and other software and / or programs to be identified. The network communication module is used to realize the communication between the components inside the storage medium, and the communication with other hardware and software in the information processing physical device.

[0061] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented by means of software plus the necessary general hardware platform, or by hardware. By applying the technical solution of the present application, the trajectory data and business type of the vehicle to be monitored are first obtained, and then the trajectory prediction model corresponding to the vehicle to be monitored is determined according to the business type, and finally the trajectory abnormality probability value of the vehicle to be monitored is obtained through the trajectory prediction model. Compared with the prior art, the above method can timely and accurately monitor the probability of trajectory abnormalities of operating vehicles, thereby effectively improving the timeliness and accuracy of driving trajectory monitoring of operating vehicles.

[0062] Those skilled in the art will appreciate that the accompanying drawings are only schematic diagrams of a preferred implementation scenario, and the modules or processes in the accompanying drawings are not necessarily necessary for implementing the present application. Those skilled in the art will appreciate that the modules in the devices in the implementation scenario can be distributed in the devices of the implementation scenario according to the description of the implementation scenario, or can be changed accordingly and located in one or more devices different from the present implementation scenario. The modules of the above-mentioned implementation scenario can be combined into one module, or can be further split into multiple submodules.

[0063] The above serial numbers of this application are only for description and do not represent the advantages and disadvantages of the implementation scenarios. The above disclosure is only a few specific implementation scenarios of this application, but this application is not limited to them, and any changes that can be thought of by technicians in this field should fall within the scope of protection of this application.

Claims

1. A method for monitoring the trajectory of an operating vehicle, characterized in that: The method comprises: Acquire the trajectory data and business type of the vehicle to be monitored, wherein the vehicle to be monitored is an operating vehicle, and the business type is divided according to the leasing, leased, and self-operated modes of the operating vehicle; Wherein, the acquisition of the trajectory data of the vehicle to be monitored includes: collecting the position data of the vehicle to be monitored in real time through a GPS device installed on the vehicle to be monitored; generating the driving trajectory of the vehicle to be monitored according to the real-time collected position data of the vehicle to be monitored; cutting the driving trajectory into n segments, and selecting a positioning point from each segment of the driving trajectory to obtain n positioning points of the driving trajectory; arranging the n positioning points in ascending order according to the sequence of the collection time of the n positioning points to obtain the arrangement sequence numbers of the n positioning points; taking the arrangement sequence numbers of the n positioning points as weights, weighting the absolute values ​​of the latitude and longitude coordinates of the n positioning points to obtain the trajectory data of the vehicle to be monitored; The method for obtaining the absolute values ​​of longitude and latitude coordinates includes: obtaining the Euclidean distance center point of the city where the vehicle to be detected is located, and calculating the distance between the Euclidean distance center point and a preset city center point; when the distance between the Euclidean distance center point and the city center point is less than a preset deviation value, obtaining the absolute values ​​of the longitude and latitude coordinates of the n positioning points according to the difference between the longitude and latitude coordinate values ​​of the n positioning points and the Euclidean distance center point; when the distance between the Euclidean distance center point and the city center point is not less than the preset deviation value, obtaining the absolute values ​​of the longitude and latitude coordinates of the n positioning points according to the difference between the longitude and latitude coordinate values ​​of the n positioning points and the city center point; According to the business type of the vehicle to be monitored, a trajectory prediction model corresponding to the vehicle to be monitored is determined, wherein the training method of the trajectory prediction model includes: for the historical trajectory data corresponding to the vehicle of each business type, randomly splitting the historical trajectory data by an isolation forest algorithm, finding abnormal trajectory data in the historical trajectory data, and dividing the historical trajectory data into normal trajectories and abnormal trajectories, using the divided historical trajectory data as samples of a random forest model, training multiple decision trees in the random forest model, and obtaining the trajectory prediction model; The method of randomly dividing the historical trajectory data by using an isolation forest algorithm to find abnormal trajectory data in the historical trajectory data includes: extracting multiple historical trajectory sample groups from multiple historical trajectory data of each business type, each of the historical trajectory sample groups contains a predetermined number of historical trajectory data; randomly selecting a historical trajectory data between the maximum value and the minimum value of each historical trajectory sample group by a recursive algorithm to divide the historical trajectory sample group until each historical trajectory sample group cannot be divided any further, thereby obtaining multiple binary trees corresponding to each business type; integrating the multiple binary trees corresponding to each business type to obtain a trajectory judgment model corresponding to each business type; and finding abnormal trajectory data in the historical trajectory data by using the trajectory judgment model; The trajectory data of the vehicle to be monitored is input into the trajectory prediction model to obtain a trajectory abnormality probability value of the vehicle to be monitored, wherein the trajectory abnormality probability value is used to characterize the operational risk of the vehicle to be monitored.

2. The method according to claim 1, characterized in that The training method of the trajectory prediction model includes: Acquire multiple historical trajectory data of multiple operating vehicles, and classify the multiple historical trajectory data according to the business type of each of the operating vehicles to obtain multiple historical trajectory data of each business type; Based on multiple historical trajectory data of each business type, the trajectory judgment model corresponding to each business type is trained through the isolation forest algorithm; Input each piece of historical trajectory data of each business type into the trajectory judgment model of the corresponding business type to obtain a trajectory judgment result of each piece of historical trajectory data, wherein the trajectory judgment result includes normal and abnormal; According to the multiple historical trajectory data of each business type and the trajectory judgment result of each historical trajectory data, a trajectory prediction model corresponding to each business type is trained through a random forest algorithm.

3. The method according to claim 2, characterized in that The trajectory prediction model corresponding to each business type is trained by a random forest algorithm based on the multiple historical trajectory data of each business type and the trajectory judgment result of each historical trajectory data, including: Constructing a sample set of a trajectory prediction model according to a plurality of historical trajectory data of each business type and a trajectory judgment result of each historical trajectory data; Multiple historical trajectory sample groups are randomly extracted from the training set with replacement, and a decision tree is trained based on each historical trajectory sample group to obtain multiple decision trees corresponding to each business type; The multiple decision trees corresponding to each business type are integrated to obtain a trajectory prediction model corresponding to each business type.

4. A trajectory monitoring device for an operating vehicle, characterized in that: The device comprises: A data acquisition module is used to acquire the trajectory data and business type of the vehicle to be monitored, wherein the vehicle to be monitored is an operating vehicle, and the business type is divided according to the rental, lease, and self-operation mode of the operating vehicle; The data acquisition module is specifically used to collect the position data of the vehicle to be monitored in real time through the GPS device installed on the vehicle to be monitored; generate the driving trajectory of the vehicle to be monitored according to the real-time collected position data of the vehicle to be monitored; cut the driving trajectory into n sections, and select a positioning point from each section of the driving trajectory to obtain n positioning points of the driving trajectory; arrange the n positioning points in ascending order according to the sequence of collection time of the n positioning points to obtain the arrangement sequence number of the n positioning points; take the arrangement sequence number of the n positioning points as the weight, perform weighted processing on the absolute values ​​of the latitude and longitude coordinates of the n positioning points to obtain the trajectory data of the vehicle to be monitored; The data acquisition module is further used to obtain the Euclidean distance center point of the city where the vehicle to be detected is located, and calculate the distance between the Euclidean distance center point and a preset city center point; when the distance between the Euclidean distance center point and the city center point is less than a preset deviation value, the absolute values ​​of the longitude and latitude coordinates of the n positioning points are obtained according to the difference between the longitude and latitude coordinate values ​​of the n positioning points and the Euclidean distance center point; when the distance between the Euclidean distance center point and the city center point is not less than the preset deviation value, the absolute values ​​of the longitude and latitude coordinates of the n positioning points are obtained according to the difference between the longitude and latitude coordinate values ​​of the n positioning points and the city center point; A model determination module is used to determine the trajectory prediction model corresponding to the vehicle to be monitored according to the business type of the vehicle to be monitored, wherein the training method of the trajectory prediction model includes: for the historical trajectory data corresponding to the vehicle of each business type, randomly splitting the historical trajectory data by an isolation forest algorithm, finding abnormal trajectory data in the historical trajectory data, and dividing the historical trajectory data into normal trajectories and abnormal trajectories, using the divided historical trajectory data as samples of a random forest model, training multiple decision trees in the random forest model, and obtaining the trajectory prediction model; The method of randomly dividing the historical trajectory data by using an isolation forest algorithm to find abnormal trajectory data in the historical trajectory data includes: extracting multiple historical trajectory sample groups from multiple historical trajectory data of each business type, each of the historical trajectory sample groups contains a predetermined number of historical trajectory data; randomly selecting a historical trajectory data between the maximum value and the minimum value of each historical trajectory sample group by a recursive algorithm to divide the historical trajectory sample group until each historical trajectory sample group cannot be divided any further, thereby obtaining multiple binary trees corresponding to each business type; integrating the multiple binary trees corresponding to each business type to obtain a trajectory judgment model corresponding to each business type; and finding abnormal trajectory data in the historical trajectory data by using the trajectory judgment model; The data processing module is used to input the trajectory data of the vehicle to be monitored into the trajectory prediction model to obtain the trajectory abnormality probability value of the vehicle to be monitored, wherein the trajectory abnormality probability value is used to characterize the operational risk of the vehicle to be monitored.

5. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 3 are implemented.

6. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 3 are implemented.

Images