CN114911501A - Traffic accident recognition prediction method, device and computer readable storage medium - Google Patents
Traffic accident recognition prediction method, device and computer readable storage medium Download PDFInfo
- Publication number
- CN114911501A CN114911501A CN202210559958.9A CN202210559958A CN114911501A CN 114911501 A CN114911501 A CN 114911501A CN 202210559958 A CN202210559958 A CN 202210559958A CN 114911501 A CN114911501 A CN 114911501A
- Authority
- CN
- China
- Prior art keywords
- traffic accident
- target data
- vehicle
- preset condition
- driving behavior
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/60—Software deployment
- G06F8/65—Updates
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
- G08G1/0133—Traffic data processing for classifying traffic situation
Abstract
The invention discloses a traffic accident recognition and prediction method, equipment and a computer readable storage medium, and belongs to the technical field of intelligent driving. The driving behavior of the virtual driver and/or the driving behavior of the human driver are judged whether to reach the preset condition or not; if yes, transmitting preset target data back to a background server associated with the vehicle; acquiring a software upgrading package transmitted by the background server, wherein the software upgrading package is integrated by an algorithm obtained by training the background server according to the target data; and upgrading according to the software upgrading package so as to identify and predict the traffic accident. The invention solves the problem of lower accuracy when the existing intelligent driving technology identifies and predicts the traffic accident, and realizes the technical effect of improving the traffic accident identification and prediction performance of the intelligent driving technology.
Description
Technical Field
The invention relates to the technical field of intelligent driving, in particular to a traffic accident recognition and prediction method, traffic accident recognition and prediction equipment and a computer readable storage medium.
Background
Under the support of the internet of vehicles and the artificial intelligence technology, the existing intelligent driving technology can assist a driver to drive safely, and even realize automatic driving under the condition of no driver. The technology can help avoid potential safety hazards such as drunk driving and fatigue driving, and has great significance for reducing driver errors and improving safety.
In intelligent driving technology, the recognition and prediction of traffic accidents are essential functions thereof. In order to identify and predict traffic accidents, different solutions are proposed in the prior art, such as vehicle trajectory prediction, traffic condition prediction, collision prediction technologies and the like based on a high-definition map, a high-definition camera or a millimeter wave radar, and although the technologies improve the identification and prediction of the traffic accidents, the technologies mostly simply use the distance between a vehicle and a target as a prediction judgment standard, and the situations capable of identification and prediction are few, and the actual road traffic condition has extremely high complexity and uncertainty, so that the existing intelligent driving technology still has the problem of low accuracy when identifying and predicting the traffic accidents.
Disclosure of Invention
The invention mainly aims to provide a traffic accident recognition and prediction method, equipment and a computer readable storage medium, and aims to solve the problem that the existing intelligent driving technology is low in accuracy when recognizing and predicting traffic accidents.
In order to achieve the above object, the present invention provides a traffic accident recognition and prediction method, which is applied to a vehicle, and comprises the following steps:
judging whether the driving behavior of the virtual driver and/or the driving behavior of the human driver reach a preset condition or not;
if yes, transmitting preset target data back to a background server associated with the vehicle;
acquiring a software upgrading package transmitted by the background server, wherein the software upgrading package is integrated by an algorithm obtained by training the background server according to the target data;
and upgrading according to the software upgrading package to identify and predict the traffic accident.
Optionally, the step of determining whether the driving behavior of the virtual driver meets a preset condition includes:
acquiring first behavior information generated by the driving behavior of the virtual driver, and judging whether the first behavior information reaches a first preset condition;
and if the first behavior information reaches a first preset condition, judging that the driving behavior of the virtual driver reaches the preset condition.
Optionally, the step of determining whether the driving behavior of the human driver meets a preset condition includes:
acquiring second behavior information generated by the driving behavior of the human driver, and judging whether the second behavior information reaches a second preset condition or not;
and if the second behavior information reaches a second preset condition, judging that the driving behavior of the human driver reaches the preset condition.
Optionally, the step of determining whether the driving behavior of the virtual driver and the driving behavior of the human driver meet the preset condition includes:
acquiring behavior difference information of the driving behavior of the virtual driver and the driving behavior of the human driver, and judging whether the behavior difference information reaches a third preset condition;
and if the behavior difference information reaches a third preset condition, judging that the driving behavior of the virtual driver and the driving behavior of the human driver reach the preset condition.
Optionally, the step of returning the preset target data to the background server associated with the vehicle includes:
determining a point in time at which the driving behavior of the virtual driver and/or the driving behavior of the human driver reaches the preset condition;
and acquiring target data acquired within a preset time before and after the time point, and transmitting the target data back to a background server associated with the vehicle.
Optionally, the step of transmitting the target data back to a background server associated with the vehicle includes:
classifying the target data according to the condition type of the preset condition;
and returning the classified target data to a preset data storage center so that a background server associated with the vehicle can download the target data from the data storage center.
The invention also provides another traffic accident identification and prediction method, which is applied to a background server associated with a vehicle and comprises the following steps:
acquiring target data returned by the vehicle, and cleaning the target data to obtain an effective data set;
training a traffic accident recognition prediction algorithm according to the effective data set to obtain a trained algorithm;
and integrating according to the trained algorithm to obtain a software upgrading package, and transmitting the software upgrading package to the vehicle.
Optionally, the step of cleaning the target data to obtain an effective data set includes:
acquiring the type of the target data, and respectively putting the target data into a first data set and a second data set according to the type of the target data;
and respectively cleaning the data in the first data set and the second data set to obtain the effective data set.
In addition, the present invention also provides a traffic accident recognition and prediction apparatus, characterized in that the apparatus comprises: the traffic accident recognition and prediction system comprises a memory, a processor and a traffic accident recognition and prediction program stored on the memory and operable on the processor, the traffic accident recognition and prediction program being configured to implement the steps of the traffic accident recognition and prediction method as described above.
In addition, the invention also provides a computer readable storage medium, which is characterized in that the computer readable storage medium is stored with a traffic accident identification and prediction program, and the traffic accident identification and prediction program is executed by a processor to realize the steps of the traffic accident identification and prediction method.
The driving behavior of the virtual driver and/or the driving behavior of the human driver are judged whether to reach the preset condition or not; if yes, transmitting preset target data back to a background server associated with the vehicle; acquiring a software upgrading package transmitted by the background server, wherein the software upgrading package is integrated by a traffic accident algorithm obtained by training the background server according to the buried point transmission data; and upgrading according to the software upgrading package to identify and predict the traffic accident.
Compared with the existing mode of predicting the traffic accident based on the condition rule, the invention carries out data return buried point design through respective driving behaviors of the vehicle virtual driver and/or the human driver, thereby returning target data to the background for training the traffic accident prediction algorithm, upgrading the algorithm for predicting the traffic accident aiming at the vehicle, effectively improving the training mode of the traffic accident recognition prediction algorithm of the vehicle, solving the problem of low accuracy of the traffic accident recognition prediction in the existing intelligent driving technology, and realizing the technical effect of improving the traffic accident recognition prediction performance in the intelligent driving technology.
Drawings
FIG. 1 is a schematic structural diagram of a traffic accident identification and prediction device in a hardware operating environment according to an embodiment of the present invention;
FIG. 2 is a flow chart illustrating an embodiment of a traffic accident recognition and prediction method according to the present invention;
fig. 3 is a flowchart illustrating another traffic accident identification and prediction method according to an embodiment of the present invention.
The implementation, functional features and advantages of the present invention will be further described with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
It should be understood that although the terms first, second, third, etc. may be used herein to describe various information, such information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope herein. The word "if," as used herein, may be interpreted as "at … …" or "when … …" or "in response to a determination," depending on the context. The terms "or," "and/or," "including at least one of the following," and the like, as used herein, are to be construed as inclusive or mean any one or any combination.
It should be understood that, although the steps in the flowcharts in the embodiments of the present application are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and may be performed in other orders unless otherwise indicated herein. Moreover, at least some of the steps in the figures may include multiple sub-steps or multiple stages that are not necessarily performed at the same time, but may be performed at different times, in different orders, and may be performed alternately or at least partially with respect to other steps or sub-steps of other steps.
In order to more clearly understand the technical features, objects, and effects of the present invention, embodiments of the present invention will now be described with reference to the accompanying drawings.
Referring to fig. 1, fig. 1 is a schematic structural diagram of a hardware-operated traffic accident identification and prediction device according to an embodiment of the present invention.
The traffic accident recognition and prediction device can be a vehicle with a vehicle-mounted intelligent driving system or a background server for algorithm training.
As shown in fig. 1, the traffic accident recognition prediction apparatus may include: a processor 1001, such as a Central Processing Unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Wherein a communication bus 1002 is used to enable connective communication between these components. The user interface 1003 may include a Display screen (Display), an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a WIreless interface (e.g., a WIreless-FIdelity (WI-FI) interface). The Memory 1005 may be a Random Access Memory (RAM) Memory, or may be a Non-Volatile Memory (NVM), such as a disk Memory. The memory 1005 may alternatively be a storage device separate from the processor 1001 described previously.
Those skilled in the art will appreciate that the configuration shown in FIG. 1 does not constitute a limitation of the traffic accident recognition and prediction device, and may include more or fewer components than those shown, or some components in combination, or a different arrangement of components.
As shown in fig. 1, a memory 1005, which is a storage medium, may include therein an operating system, a data storage module, a network communication module, a user interface module, and a traffic accident recognition prediction program.
In the traffic accident recognition prediction apparatus shown in fig. 1, the network interface 1004 is mainly used for data communication with other apparatuses; the user interface 1003 is mainly used for data interaction with a user; the processor 1001 and the memory 1005 of the traffic accident recognition and prediction device of the present invention may be provided in the traffic accident recognition and prediction device, and the traffic accident recognition and prediction device calls the traffic accident recognition and prediction program stored in the memory 1005 through the processor 1001 and performs the following operations:
judging whether the driving behavior of the virtual driver and/or the driving behavior of the human driver reach a preset condition or not;
if yes, transmitting preset target data back to a background server associated with the vehicle;
acquiring a software upgrading package transmitted by the background server, wherein the software upgrading package is integrated by an algorithm obtained by training the background server according to the target data;
and upgrading according to the software upgrading package to identify and predict the traffic accident.
Further, the processor 1001 may be configured to invoke a traffic accident recognition prediction program stored in the memory 1005, and further perform the following operations:
acquiring first behavior information generated by the driving behavior of the virtual driver, and judging whether the first behavior information reaches a first preset condition;
and if the first behavior information reaches a first preset condition, judging that the driving behavior of the virtual driver reaches the preset condition.
Further, the processor 1001 may be configured to invoke a traffic accident recognition prediction program stored in the memory 1005, and further perform the following operations:
acquiring second behavior information generated by the driving behavior of the human driver, and judging whether the second behavior information reaches a second preset condition;
and if the second behavior information reaches a second preset condition, judging that the driving behavior of the human driver reaches the preset condition.
Further, the processor 1001 may be configured to invoke a traffic accident recognition prediction program stored in the memory 1005, and further perform the following operations:
acquiring behavior difference information of the driving behavior of the virtual driver and the driving behavior of the human driver, and judging whether the behavior difference information reaches a third preset condition;
and if the behavior difference information reaches a third preset condition, judging that the driving behavior of the virtual driver and the driving behavior of the human driver reach the preset condition.
Further, the processor 1001 may be configured to invoke a traffic accident recognition prediction program stored in the memory 1005, and further perform the following operations:
determining a point in time at which the driving behavior of the virtual driver and/or the driving behavior of the human driver reaches the preset condition;
and acquiring target data acquired within a preset time before and after the time point, and transmitting the target data back to a background server associated with the vehicle.
Further, the processor 1001 may be configured to invoke a traffic accident recognition prediction program stored in the memory 1005, and further perform the following operations:
classifying the target data according to the condition type of the preset condition;
and returning the classified target data to a preset data storage center so that a background server associated with the vehicle can download the target data from the data storage center.
Further, the processor 1001 may be configured to invoke a traffic accident recognition prediction program stored in the memory 1005, and further perform the following operations:
acquiring target data returned by the vehicle, and cleaning the target data to obtain an effective data set;
training a traffic accident recognition prediction algorithm according to the effective data set to obtain a trained algorithm;
and integrating according to the trained algorithm to obtain a software upgrading package, and transmitting the software upgrading package to the vehicle.
Further, the processor 1001 may be configured to invoke a traffic accident recognition prediction program stored in the memory 1005, and further perform the following operations:
acquiring the type of the target data, and respectively putting the target data into a first data set and a second data set according to the type of the target data;
and respectively cleaning the data in the first data set and the second data set to obtain the effective data set.
With the rapid development of new-generation information technology and artificial intelligence technology, the global automobile industry is in deep revolution, and the intelligent driving technology utilizes advanced devices such as vehicle-mounted sensors, controllers and actuators, and integrates the technologies such as artificial intelligence and computer vision, so that the automobile has the functions of complex environment perception, intelligent decision making, autonomous control and the like, and the final aim is to replace human to control the automobile in a safe and efficient manner. The method can identify complex traffic environment information through an algorithm and make a correct decision, and is a necessary function for realizing a final target. However, the current intelligent driving technology is a traffic accident recognition and prediction algorithm which is usually based on condition rules, and training data is difficult to obtain and the number of training samples is small. Therefore, when the complex and changeable actual road traffic condition is faced, the existing intelligent driving technology has the problem of low accuracy rate when the traffic accident is identified and predicted.
In order to solve the above problems, the present invention provides a traffic accident identification and prediction method, applied to a vehicle, including: judging whether the driving behavior of the virtual driver and/or the driving behavior of the human driver reach a preset condition or not; if yes, transmitting preset target data back to a background server associated with the vehicle; acquiring a software upgrading package transmitted by the background server, wherein the software upgrading package is integrated by a traffic accident algorithm obtained by training the background server according to the embedded point transmission data; and upgrading according to the software upgrading package to identify and predict the traffic accident.
According to the method, data return buried point design is carried out through respective driving behaviors of a virtual driver and/or a human driver of the vehicle, so that target data are returned to a background to train a traffic accident prediction algorithm, the algorithm for predicting the traffic accident of the vehicle is upgraded, the training mode of the traffic accident recognition prediction algorithm of the vehicle is effectively improved, more and richer sample data can be obtained to train the algorithm, and the mode for obtaining the sample data is more convenient and low in cost. Therefore, the problem that the accuracy of recognizing and predicting the traffic accident is low in the existing intelligent driving technology is solved, and the technical effect of improving the performance of predicting the traffic accident recognition in the intelligent driving technology is achieved.
An embodiment of the present invention provides a traffic accident identification and prediction method, and referring to fig. 2, fig. 2 is a schematic flow diagram of an embodiment of a traffic accident identification and prediction method according to the present invention.
In this embodiment, the traffic accident recognition and prediction method is applied to a vehicle, and the traffic accident recognition and prediction method includes:
and step S10, judging whether the driving behavior of the virtual driver and/or the driving behavior of the human driver reach preset conditions.
In the present embodiment, the execution subject is a vehicle equipped with an intelligent driving system. The virtual driver is a model which is different from a human driver and used for developing a traffic accident recognition and prediction function, operates in an intelligent driving system with a shadow mode function, and is essentially a piece of program code. The shadow mode refers to that when a human driver drives the vehicle, the virtual driver also runs in real time in the background, but does not send a control instruction, and only compares the control operation of the intelligent driving system and the control operation of the human driver on the vehicle.
The driving behavior of the virtual driver is generated by the virtual driver through observing the surrounding traffic environment in real time through an environment perception sensor of an intelligent driving system on the vehicle when the human driver controls the vehicle to run, identifying and predicting whether the surrounding vehicle has collision or not and making corresponding vehicle control decisions such as acceleration, braking, steering and the like. For example, its driving behavior includes: predicting the occurrence of an external traffic accident; observing the occurrence of external traffic accidents; generating decision intentions for accelerating, braking and steering the potential or occurring traffic accidents; the driving behaviors of stepping on an accelerator pedal, stepping on a brake pedal and steering a steering wheel.
The driving behavior of a human driver refers to the behavior of controlling the acceleration, deceleration, braking, steering and the like of a vehicle during driving. The preset condition is a preset buried point condition of the returned data, and is used for judging whether to collect the required data and return the data, and may also be referred to as a buried point condition.
It should be noted that the virtual driver operates in the background when the human driver does not turn on the automatic driving function, and the vehicle control decision made by the virtual driver while operating in the background does not directly control the vehicle.
And step S20, if yes, transmitting preset target data back to the background server associated with the vehicle.
In this embodiment, the preset target data refers to data that needs to be returned when a preset condition is reached, and includes: raw sensing signal data detected by an environment sensing device of the vehicle, such as a video stream detected by a visual camera, a point cloud detected by a laser radar, an energy map detected by a millimeter wave radar, an echo signal detected by ultrasonic waves, an imaging map detected by infrared rays, and the like, wherein the environment sensing device can be the visual camera, the laser radar, the millimeter wave radar, the ultrasonic waves, the infrared rays, and the like; vehicle motion state signal data, such as vehicle speed, acceleration, angular velocity, wheel speed, etc.; signal data of the virtual driver and the human driver operating the vehicle, such as a steering wheel angle, a torque, an accelerator pedal opening, a retarder pedal opening, and the like. After the target data is transmitted back to the background server, the background server can finally obtain effective data for algorithm training by cleaning the target data, namely the target data is the original sample data for algorithm training subsequently.
The background server is a server which performs algorithm training and can issue a software upgrade package integrated by the trained algorithm to the vehicle. The target data can be returned through a vehicle-mounted data return terminal of the vehicle, such as a 4G or 5G module. The association relationship between the vehicle and the background server is as follows: the vehicle can be directly connected with the background server, or the vehicle can also communicate with the background server through a cloud.
Specifically, for example, the preset target data is transmitted back to the background server associated with the vehicle through the vehicle-mounted data transmission terminal of the vehicle. The target data can be transmitted to the cloud or the data storage center through the vehicle-mounted data return terminal, and then can be acquired by the background server.
And step S30, acquiring the software upgrading package transmitted by the background server, wherein the software upgrading package is integrated by an algorithm obtained by the background server according to the target data.
In this embodiment, the software upgrade package is used to upgrade an intelligent driving system of a vehicle, so as to improve and optimize a traffic accident recognition prediction algorithm of the intelligent driving system. The software upgrading package is obtained by integrating an algorithm obtained by training the background server according to the target data.
And step S40, upgrading according to the software upgrading package to identify and predict traffic accidents.
In this embodiment, after the vehicle acquires the software upgrade package, the intelligent driving system of the vehicle may be upgraded, or the traffic accident recognition and prediction algorithm in the intelligent driving system of the vehicle may be upgraded, so that the upgraded intelligent driving system or traffic accident recognition and prediction algorithm is used to perform recognition and prediction on the traffic accident during the driving process of the vehicle. The identification and prediction of the traffic accident means that an intelligent driving system of the vehicle judges whether the surrounding vehicles have or are about to have the traffic accidents such as collision and the like based on the acquired surrounding environment information.
Specifically, for example, the intelligent driving system of the vehicle is upgraded according to the software upgrading package, and the upgraded intelligent driving system is used for recognizing and predicting the traffic accident.
According to the embodiment, when the respective driving behaviors of the virtual vehicle driver and/or the human driver reach the preset conditions, the target data are returned to the background for training of the traffic accident prediction algorithm, so that the algorithm for predicting the traffic accident of the vehicle is upgraded, and the training mode of the traffic accident recognition prediction algorithm of the vehicle is effectively improved. Therefore, the problem that the accuracy of recognizing and predicting the traffic accident is low in the existing intelligent driving technology is solved, and the technical effect of improving the performance of predicting the traffic accident recognition in the intelligent driving technology is achieved. In addition, because the training sample data is acquired in the driving process of the vehicle, more and richer sample data can be acquired to train the algorithm, and the mode of acquiring the sample data is more convenient and has low cost.
Further, in another embodiment of the traffic accident identification and prediction method according to the present invention, the step S10 of determining whether the driving behavior of the virtual driver meets the preset condition includes:
step S11, acquiring first behavior information generated by the driving behavior of the virtual driver, and determining whether the first behavior information meets a first preset condition.
In this embodiment, the first behavior information refers to decision behavior information generated during the driving process of the virtual driver vehicle and used for recognizing and predicting the traffic accident. The first preset condition is that the virtual driver predicts that a traffic accident is about to occur or recognizes that the traffic accident occurs.
Specifically, whether the first behavior information reaches a first preset condition is judged by judging whether the virtual driver generates decision information for predicting the imminent occurrence of the traffic accident or identifying the occurrence of the traffic accident in the driving process of the vehicle.
Step S12, if the first behavior information reaches a first preset condition, determining that the driving behavior of the virtual driver reaches the preset condition.
In this embodiment, when the virtual driver generates decision information for predicting that a traffic accident is about to occur or recognizing that a traffic accident occurs during the driving of the vehicle, it is determined that the first behavior information reaches the first preset condition, so that it is determined that the driving behavior of the virtual driver reaches the first preset condition.
In the embodiment, the first behavior information of the virtual driver is acquired, whether the first behavior information reaches the first preset condition is judged, and whether the first behavior information reaches the preset condition is further judged, so that the returning of the target data is triggered, namely, the acquisition of the original sample data for algorithm training is triggered, and the sample data for training is enriched.
Alternatively, the step S10, determining whether the driving behavior of the human driver meets the preset condition, includes:
step S13, obtaining second behavior information generated by the driving behavior of the human driver, and determining whether the second behavior information meets a second preset condition.
In this embodiment, the second behavior information is driving behavior information of a human driver stepping on a brake or a steering wheel. The second preset condition is that the brake pedal opening change rate of the vehicle is greater than a certain threshold value
And at the deep pedal for a time threshold t 6 Second, and the vehicle speed during this time is not zero (V) vehicle > 0); or the rate of rotation of the steering wheel being above a certain threshold
And for a time threshold t 7 Second, and the vehicle speed during this time is not zero (V) vehicle > 0). Wherein, the specific threshold value can be determined by a developer according to the specific application boundary.
Specifically, for example, whether the second behavior information reaches the second preset condition is determined by vehicle brake pedal opening information, time information, and vehicle speed information when the human driver steps on the brake, or by steering wheel turning angle rate information, time information, and vehicle speed information when the human driver turns the steering wheel.
Step S14, if the second behavior information reaches a second preset condition, determining that the driving behavior of the human driver reaches the preset condition.
In this embodiment, when the human driver steps on the brake, the brake pedal opening change rate of the vehicle is greater than a certain threshold, and continues for a period of time at the deep pedal, and the vehicle speed of the period of time is not zero, or when the human driver rotates the steering wheel, the steering wheel rotation angle rate of the vehicle is greater than a certain threshold, and continues for a period of time at a threshold, and the vehicle speed of the period of time is not zero, it is determined that the second behavior information reaches the second preset condition, and it is determined that the driving behavior of the human driver reaches the preset condition.
Optionally, in step S10, the determining whether the driving behavior of the virtual driver and the driving behavior of the human driver meet the preset condition includes:
step S15, behavior difference information of the driving behavior of the virtual driver and the driving behavior of the human driver is obtained, and whether the behavior difference information reaches a third preset condition is judged;
in the present embodiment, the behavior difference information refers to difference information when the virtual driver and the human driver perform the same driving behavior, such as controlling vehicle acceleration, controlling vehicle deceleration, or controlling vehicle steering. Specifically, for example, when both the human driver and the virtual driver control the acceleration of the vehicle, the difference between the accelerator pedal opening degrees of the human driver and the virtual driver is the difference between the intended acceleration. The virtual driver realizes the behavior of stepping on an accelerator pedal by sending an acceleration instruction to the engine, realizes the behavior of stepping on a brake pedal by sending a deceleration instruction to the chassis ESP system, and realizes the behavior of steering by sending a steering angle instruction or a torque instruction to the steering wheel EPS system.
The third preset condition is that the absolute value of the difference value of the intention accelerations (integrated accelerations) of the human driver and the virtual driver is larger than a certain threshold (| ia) H -ia V |＞ia TH ) And for a time threshold t 1 Second, or the absolute value of the difference between the accelerator pedal opening (accelerator pedal opening) of the human driver and the accelerator pedal opening (accelerator pedal opening) of the virtual driver is larger than a certain threshold (| apo) H -apo V |＞apo TH ) And for a time threshold t 2 Second; or the absolute value of the difference between the intended decelerations of the human driver and the virtual driver is greater than a certain threshold (| id) H -id V |＞id TH ) And for a time threshold t 3 Second, or the absolute value of the difference between the opening degrees of the brake speed reduction pedals of the human driver and the virtual driver is larger than a certain threshold (| dpo) H -dpo V |＞dpo TH ) And for a time threshold t 4 Second; or the absolute value of the difference between the intended steering wheel angle of the human driver and the intended steering wheel angle of the virtual driver is greater than a threshold (| isa) H -isa V |＞isa TH ) And for a period of time threshold t 5 And second. Wherein, the specific threshold value can be determined by a developer according to the specific application boundary.
Specifically, for example, it is determined whether the behavior difference information reaches the third preset condition by the intention acceleration difference absolute value information of the human driver and the virtual driver and the duration information thereof. The step of determining whether the behavior difference information meets the third preset condition through the other information is the same as the previous step, and is not listed here.
Step S16, if the behavior difference information reaches a third preset condition, determining that the driving behavior of the virtual driver and the driving behavior of the human driver reach the preset condition.
In this embodiment, when the absolute value of the difference between any one of the intended acceleration, the accelerator pedal opening, the intended deceleration, the brake/deceleration pedal opening, and the intended steering wheel angle of the human driver and the virtual driver is greater than a threshold value and continues for a certain period of time, it is determined that the behavior difference information reaches the third preset condition, and it is determined that the driving behavior of the intended driver and the driving behavior of the human driver reach the preset condition.
In the embodiment, whether different specific preset conditions are met or not is judged by obtaining the driving behavior information or the behavior difference information of various virtual drivers and human drivers, so that whether the preset conditions are met or not is judged, namely whether the target data is triggered to be returned or not is judged, algorithm training development of returning various types of data is realized, and the accuracy of traffic accident recognition and prediction can be improved due to more sample data.
Further, in another embodiment of the traffic accident identification and prediction method of the present invention, step S20, the step of transmitting the preset target data back to the background server associated with the vehicle includes:
step S21, determining a point in time at which the driving behavior of the virtual driver and/or the driving behavior of the human driver reaches the preset condition.
In this embodiment, the time point when the preset condition is reached is determined by determining the current time when the driving behavior of the virtual driver reaches the preset condition, or determining the current time when the driving behavior of the human driver reaches the preset condition, or determining the current time when the driving behaviors of the virtual driver and the human driver reach the preset condition. For determining when to start acquiring target data.
And step S22, acquiring target data acquired within a preset time length before and after the time point, and transmitting the target data back to a background server associated with the vehicle.
In this embodiment, the preset time duration is used to determine the time duration of the target data to be acquired, and may be determined according to the actual demand, for example, if the preset time duration is set to 15 seconds, the target data 15 seconds before and after the time point of the preset condition is determined will be acquired. The target data are cached in a cache area in advance in the vehicle, so that a controller of the vehicle can acquire the target data when a preset condition is reached.
Optionally, in step S22, the step of returning the target data to the vehicle-associated background server includes:
and step S221, classifying the target data according to the condition type of the preset condition.
In this embodiment, the condition type of the preset condition refers to a preset condition specifically for triggering data backhaul, and is divided into the first preset condition, the second preset condition, and the third preset condition. Specifically, the target data returned under different preset conditions may be labeled to indicate which preset condition the target data is returned under, so as to classify the target data.
Step S222, the classified target data is transmitted back to a preset data storage center, so that the background server associated with the vehicle downloads the target data from the data storage center.
In this embodiment, since the amount of the returned data is large, the classified target data can be returned to a preset data storage center or a cloud, and then downloaded and acquired by the background server.
According to the embodiment, the target data are classified according to the types of different triggered preset conditions, and the fact that the target data are triggered and returned by specific preset conditions is noted, so that the method is beneficial to distinguishing and identifying a large amount of data.
Further, in another embodiment of the traffic accident identification and prediction method of the present invention, referring to fig. 3, fig. 3 is a schematic flow chart of another embodiment of the traffic accident identification and prediction method of the present invention, where the method is applied to a vehicle-associated background server, and includes:
step A10, acquiring target data returned by the vehicle, and cleaning the target data to obtain an effective data set;
in this embodiment, the background server may obtain the target data returned by the vehicle from the cloud or the data storage center. The effective data set refers to a data set which can be finally used for training a traffic accident recognition and prediction algorithm in target data returned by the vehicle.
The cleaning of the target data refers to cleaning of data which does not contain traffic accidents, namely data which do not contribute to algorithm training. For example, data including traffic accidents in target data can be extracted through a machine search method, namely an artificial intelligence algorithm, and the data is put into a data set to serve as an effective data set; and the rest data is classified into another data set for manually judging whether the data containing the traffic accident still exist, receiving the effective data manually screened out and putting the effective data into the effective data set.
Step A20, training a traffic accident recognition prediction algorithm according to the effective data set to obtain a trained algorithm;
in this embodiment, after obtaining the valid data set, the server trains the traffic accident recognition and prediction algorithm according to the valid data set, that is, sample data, so as to obtain a trained algorithm, and the trained algorithm is used for a software upgrade package of the integrated vehicle-mounted automatic driving system.
And A30, integrating according to the trained algorithm to obtain a software upgrading package, and transmitting the software upgrading package to the vehicle.
In this embodiment, the server integrates the trained algorithm into a software upgrade package of the vehicle-mounted intelligent driving system, and the software upgrade package is used for upgrading the intelligent driving system. The vehicle can be sent to the vehicle through Over-the-Air Technology (Over-the-Air Technology) at regular or irregular intervals, so that the vehicle can download and upgrade the intelligent driving system.
Optionally, in step a10, the step of cleansing the target data to obtain a valid data set includes:
step A11, acquiring the type of the target data, and respectively putting the target data into a first data set and a second data set according to the type of the target data;
in this embodiment, when the vehicle returns the target data, the target data is classified or labeled according to a specific preset condition triggered during returning, so that the server can acquire the type of the target data. The type of the acquired target data, that is, what kind of preset condition is specifically triggered to transmit the acquired target data back, may be divided into three types of target data transmitted back under the first, second, and third preset conditions.
Different types of target data are respectively classified into different data sets, and specifically, the target data triggered and returned by second and third preset conditions can be put into the first data set; and the target data which is triggered to return by the first preset condition is put into a second data set, namely the traffic accident data identified by the virtual driver or the predicted traffic accident data.
Step a12, cleansing the data in the first data set and the second data set respectively to obtain the valid data set.
In this embodiment, different cleansing methods are required for different types of target data, and thus are separately cleansed to obtain valid data sets. The valid data set may include a correct data set for forward training of the algorithm and an incorrect data set for reverse training of the algorithm.
Specifically, for example, the data in the first data set may include data of an external traffic accident or may not have data of a traffic accident due to a large amount of data. Therefore, the data including the traffic accident in the target data can be extracted by a machine searching method, namely an artificial intelligence algorithm, and then the data is put into a correct data set, the data left after the machine searching in the first data set is provided for the manual searching, the data screened by the manual searching is received, and the data is also put into the correct data set, so that an effective data set is obtained.
Aiming at the data in the second data set, because the data volume is small, the data can be selected and provided for manual searching, the data which is manually screened and contains the traffic accident is received, and the data is put into a correct data set; and if the data of the traffic accident is not contained, the judgment of the virtual driver is wrong, so that the data are put into a wrong data set for carrying out reverse training of the algorithm.
According to the embodiment, the target data is divided into two data sets according to the type of the target data, effective data sets are obtained by using different modes for cleaning, and the efficiency and the accuracy of effective data extraction are improved. In addition, the correct data set and the error data set are obtained to carry out forward and reverse training on the algorithm, so that the accuracy rate of the traffic accident recognition and prediction of the trained algorithm is improved.
Further, an embodiment of the present invention further provides a traffic accident identification and prediction device, where the device includes: the system comprises a memory, a processor and a traffic accident identification and prediction program which is stored on the memory and can be operated on the processor, wherein the traffic accident identification and prediction program is configured to implement the steps of the traffic accident identification and prediction method provided by the embodiment, and specific implementation steps can refer to the embodiment and are not described in detail herein.
Further, an embodiment of the present invention further provides a computer-readable storage medium, where a traffic accident identification prediction program is stored on the computer-readable storage medium, and when the traffic accident identification prediction program is executed by a processor, the steps of the traffic accident identification prediction method provided in the foregoing embodiment are implemented, and specific implementation steps may refer to the foregoing embodiment, and are not described in detail herein.
The traffic accident recognition and prediction device and the computer-readable storage medium provided by the embodiments of the present invention are used for implementing the touch terminal remote control method provided by the above embodiments, and solve the problem that the accuracy of the recognition and prediction of a traffic accident is low in the existing intelligent driving technology, compared with the prior art, the traffic accident recognition and prediction device and the computer-readable storage medium provided by the embodiments of the present invention have the same beneficial effects as the traffic accident recognition and prediction method of the above embodiments, and are not described herein again.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium (e.g., ROM/RAM, magnetic disk, optical disk) as described above and includes instructions for enabling a terminal device (e.g., a mobile phone, a computer, a server, or a network device) to execute the method according to the embodiments of the present invention.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (10)
1. A traffic accident recognition and prediction method is applied to a vehicle, and comprises the following steps:
judging whether the driving behavior of the virtual driver and/or the driving behavior of the human driver reach a preset condition or not;
if yes, transmitting preset target data back to a background server associated with the vehicle;
acquiring a software upgrading package transmitted by the background server, wherein the software upgrading package is integrated by an algorithm obtained by training the background server according to the target data;
and upgrading according to the software upgrading package so as to identify and predict the traffic accident.
2. The traffic accident recognition and prediction method according to claim 1, wherein the step of determining whether the driving behavior of the virtual driver meets the preset condition comprises:
acquiring first behavior information generated by the driving behavior of the virtual driver, and judging whether the first behavior information reaches a first preset condition;
and if the first behavior information reaches a first preset condition, judging that the driving behavior of the virtual driver reaches the preset condition.
3. The traffic accident recognition and prediction method according to claim 1, wherein the step of determining whether the driving behavior of the human driver meets a preset condition comprises:
acquiring second behavior information generated by the driving behavior of the human driver, and judging whether the second behavior information reaches a second preset condition;
and if the second behavior information reaches a second preset condition, judging that the driving behavior of the human driver reaches the preset condition.
4. The traffic accident recognition and prediction method according to claim 1, wherein the step of determining whether the driving behavior of the virtual driver and the driving behavior of the human driver meet the preset conditions comprises:
acquiring behavior difference information of the driving behavior of the virtual driver and the driving behavior of the human driver, and judging whether the behavior difference information reaches a third preset condition;
and if the behavior difference information reaches a third preset condition, judging that the driving behavior of the virtual driver and the driving behavior of the human driver reach the preset condition.
5. The traffic accident recognition and prediction method of claim 1, wherein the step of transmitting the preset target data back to the vehicle-associated back-office server comprises:
determining a point in time at which the driving behavior of the virtual driver and/or the driving behavior of the human driver reaches the preset condition;
and acquiring target data acquired within a preset time before and after the time point, and transmitting the target data back to a background server associated with the vehicle.
6. The traffic accident recognition and prediction method of claim 5, wherein the step of transmitting the target data back to a backend server associated with the vehicle comprises:
classifying the target data according to the condition type of the preset condition;
and returning the classified target data to a preset data storage center so that a background server associated with the vehicle can download the target data from the data storage center.
7. A traffic accident recognition and prediction method is applied to a vehicle-associated background server and comprises the following steps:
acquiring target data returned by the vehicle, and cleaning the target data to obtain an effective data set;
training a traffic accident recognition prediction algorithm according to the effective data set to obtain a trained algorithm;
and integrating according to the trained algorithm to obtain a software upgrading package, and transmitting the software upgrading package to the vehicle.
8. The traffic accident recognition and prediction method of claim 7, wherein the step of cleansing the target data to obtain a valid data set comprises:
acquiring the type of the target data, and respectively putting the target data into a first data set and a second data set according to the type of the target data;
and respectively cleaning the data in the first data set and the second data set to obtain the effective data set.
9. A traffic accident recognition and prediction apparatus, characterized in that the apparatus comprises: a memory, a processor, and a traffic accident recognition prediction program stored on the memory and executable on the processor, the traffic accident recognition prediction program configured to implement the steps of the traffic accident recognition prediction method of any of claims 1-6 or 7-8.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a traffic accident recognition prediction program which, when executed by a processor, implements the steps of the traffic accident recognition prediction method according to any one of claims 1-6 or 7-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210559958.9A CN114911501A (en) | 2022-05-20 | 2022-05-20 | Traffic accident recognition prediction method, device and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210559958.9A CN114911501A (en) | 2022-05-20 | 2022-05-20 | Traffic accident recognition prediction method, device and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114911501A true CN114911501A (en) | 2022-08-16 |
Family
ID=82768823
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210559958.9A Pending CN114911501A (en) | 2022-05-20 | 2022-05-20 | Traffic accident recognition prediction method, device and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114911501A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115938128A (en) * | 2023-03-15 | 2023-04-07 | 天津所托瑞安汽车科技有限公司 | Traffic accident prediction method, device, terminal and storage medium |
-
2022
- 2022-05-20 CN CN202210559958.9A patent/CN114911501A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115938128A (en) * | 2023-03-15 | 2023-04-07 | 天津所托瑞安汽车科技有限公司 | Traffic accident prediction method, device, terminal and storage medium |
| CN115938128B (en) * | 2023-03-15 | 2023-10-03 | 天津所托瑞安汽车科技有限公司 | Traffic accident prediction method, device, terminal and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20200331476A1 (en) | Automatic lane change with minimum gap distance | |
| US20200209874A1 (en) | Combined virtual and real environment for autonomous vehicle planning and control testing | |
| CN113366497B (en) | Agent Prioritization for Autonomous Vehicles | |
| US9304515B2 (en) | Regional operation modes for autonomous vehicles | |
| US10850739B2 (en) | Automatic lane change with lane-biased strategy | |
| CN108248609B (en) | Hybrid vehicle and method of predicting driving pattern in hybrid vehicle | |
| CN110473310B (en) | Automobile driving data recording method, system, equipment and storage medium | |
| US11518380B2 (en) | System and method for predicted vehicle incident warning and evasion | |
| CN111354187B (en) | Method for assisting a driver of a vehicle and driver assistance system | |
| US20200307589A1 (en) | Automatic lane merge with tunable merge behaviors | |
| US20200377087A1 (en) | Lane keep control of autonomous vehicle | |
| JP6418574B2 (en) | Risk estimation device, risk estimation method, and computer program for risk estimation | |
| US20210097780A1 (en) | Method and system for determining advanced driver assistance systems (adas) features | |
| CN113734203B (en) | Control method, device and system for intelligent driving and storage medium | |
| US20220089171A1 (en) | Method, device, computer program and computer program product for operating a driver assistance function of a vehicle | |
| CN111532267A (en) | Vehicle, and control device and control method thereof | |
| CN112937520A (en) | Emergency braking method and device for vehicle, commercial vehicle and storage medium | |
| JP2021026720A (en) | Driving support device, method for controlling vehicle, and program | |
| US20230415753A1 (en) | On-Vehicle Driving Behavior Modelling | |
| CN116438583A (en) | Available parking space recognition device, available parking space recognition method, and program | |
| CN114911501A (en) | Traffic accident recognition prediction method, device and computer readable storage medium | |
| US20200317119A1 (en) | Autonomous vehicle active interaction with surrounding environment | |
| CN110239539B (en) | Vehicle control device and vehicle | |
| CN111278708B (en) | Method and device for assisting driving | |
| CN115431995B (en) | Equipment control method and device based on different-level auxiliary driving |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |