CN115081662A

CN115081662A - Data processing method and device, electronic equipment and readable storage medium

Info

Publication number: CN115081662A
Application number: CN202110260898.6A
Authority: CN
Inventors: 宋洪正; 张法朝; 唐剑
Original assignee: Beijing Didi Infinity Technology and Development Co Ltd
Current assignee: Beijing Didi Infinity Technology and Development Co Ltd
Priority date: 2021-03-10
Filing date: 2021-03-10
Publication date: 2022-09-20

Abstract

The embodiment of the application provides a data processing method, a data processing device, electronic equipment and a readable storage medium, and relates to the technical field of computers. In the process, the accuracy of prediction can be effectively improved by the current road characteristics, the intersection to be predicted can be divided into finer granularity according to the type, and the process is free from manual participation.

Description

Data processing method and device, electronic equipment and readable storage medium

Technical Field

The present application relates to the field of computer technologies, and in particular, to a data processing method and apparatus, an electronic device, and a readable storage medium.

Background

At present, with the development of each region, the road condition of each region is more and more complex, and further, with the more and more complex road condition, the probability of occurrence of traffic accidents is increased.

The intersection is used as an intersection of a plurality of roads and is the most complex scene in a road network environment. As the place where people and vehicles densely meet, traffic accidents easily occur at the intersection.

In the related technology, the accidents are generally positioned and counted manually, and then the danger levels of the intersections are divided according to the frequency of the accidents at the intersections, so that the efficiency is low, the divided granularity is large, and the risk assessment on various turning types cannot be carried out.

Disclosure of Invention

In view of this, embodiments of the present application provide a data processing method, an apparatus, an electronic device, and a readable storage medium, so as to improve efficiency of predicting a probability of an accident occurring at an intersection, and at the same time, refine a type of the passing intersection, thereby predicting a risk more accurately.

In a first aspect, a data processing method is provided, where the method is applied to an electronic device, and the method includes:

determining the current road characteristics of a target track at an intersection to be predicted and the current passing type of the target track at the intersection to be predicted, wherein the current passing type is used for representing the driving direction of the target track at the intersection to be predicted; and

and inputting the current road characteristics and the current passing type into a pre-trained intersection risk prediction model, and determining the probability of a target accident of passing the target track through the intersection to be predicted.

In a second aspect, a data processing apparatus is provided, where the apparatus is applied to an electronic device, and the apparatus includes:

the system comprises a first determination module, a second determination module and a third determination module, wherein the first determination module is used for determining the current road characteristics of a target track at an intersection to be predicted and the current passing type of the target track at the intersection to be predicted, and the current passing type is used for representing the driving direction of the target track at the intersection to be predicted; and

and the prediction module is used for inputting the current road characteristics and the current passing type into a pre-trained intersection risk prediction model and determining the probability of a target accident that the target track passes through the intersection to be predicted.

In a third aspect, an embodiment of the present application provides an electronic device, including a memory and a processor, where the memory is used to store one or more computer program instructions, where the one or more computer program instructions are executed by the processor to implement the method according to the first aspect.

In a fourth aspect, embodiments of the present application provide a computer-readable storage medium on which computer program instructions are stored, which when executed by a processor implement the method according to the first aspect.

In a fifth aspect, embodiments of the present application provide a computer program product comprising computer programs/instructions which, when executed by a processor, implement the method according to the first aspect.

By the embodiment of the application, the target accident probability of the target track passing through the intersection to be predicted can be predicted at least based on the current road characteristics of the target track at the intersection to be predicted, the current passing type of the target track at the intersection to be predicted and a pre-trained intersection risk prediction model. In the process, the accuracy of prediction can be effectively improved by the current road characteristics, the intersection to be predicted can be divided into finer granularity according to the type, and the process is free from manual participation.

Drawings

The foregoing and other objects, features and advantages of the embodiments of the present application will be apparent from the following description of the embodiments of the present application with reference to the accompanying drawings in which:

fig. 1 is a schematic view of an intersection traffic condition provided in an embodiment of the present application;

fig. 2 is a flowchart of a data processing method according to an embodiment of the present application;

FIG. 3 is a schematic diagram of a target track provided in an embodiment of the present application;

fig. 4 is a schematic diagram of an incoming road section and an outgoing road section according to an embodiment of the present application;

fig. 5 is a schematic diagram of a road segment provided in an embodiment of the present application;

FIG. 6 is a schematic view of an intersection according to an embodiment of the present application;

FIG. 7 is a flow chart for determining dynamic characteristics according to an embodiment of the present application;

FIG. 8 is a flow chart of accident mining provided by an embodiment of the present application;

fig. 9 is an exemplary flowchart of an intersection risk prediction model in practical application according to an embodiment of the present application;

fig. 10 is a schematic structural diagram of a data processing apparatus according to an embodiment of the present application;

fig. 11 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed Description

The present application is described below based on examples, but the present application is not limited to only these examples. In the following detailed description of the present application, certain specific details are set forth in detail. It will be apparent to one skilled in the art that the present application may be practiced without these specific details. Well-known methods, procedures, components and circuits have not been described in detail so as not to obscure the present application.

Further, those of ordinary skill in the art will appreciate that the drawings provided herein are for illustrative purposes and are not necessarily drawn to scale.

Unless the context clearly requires otherwise, throughout the description, the words "comprise", "comprising", and the like are to be construed in an inclusive sense as opposed to an exclusive or exhaustive sense; that is, what is meant is "including, but not limited to".

In the description of the present application, it is to be understood that the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance. In addition, in the description of the present application, "a plurality" means two or more unless otherwise specified.

For example, as shown in fig. 1, fig. 1 is a schematic view of an intersection traffic condition provided by the embodiment of the present application, where the schematic view includes: vehicle 11, pedestrian 12, and vehicle 13.

When the vehicle 11 turns left at the intersection, there is a possibility that a traffic accident may occur with the turning vehicle (e.g., the vehicle 13) and, at the same time, with the passing pedestrian (e.g., the pedestrian 12).

Similarly, when the vehicle 13 turns around at the intersection, there is a possibility of a traffic accident with a left-turning vehicle (e.g., the vehicle 11) and, at the same time, with a passing pedestrian (e.g., the pedestrian 12).

Of course, fig. 1 is only an example provided in the embodiment of the present application, and the actual road conditions may be more complicated.

In order to solve the above problem, embodiments of the present application provide a data processing method to improve efficiency of predicting a probability of an accident occurring at an intersection, and at the same time, may further refine a type of the passing intersection, so as to perform risk prediction more accurately.

The data processing method may be applied to an electronic device, where the electronic device may be a terminal or a server, the terminal may be a smart phone, a tablet Computer, a Personal Computer (PC), or the like, and the server may be a single server, a server cluster configured in a distributed manner, or a cloud server.

A data processing method provided in the embodiments of the present application will be described in detail below with reference to specific embodiments, as shown in fig. 2, the specific steps are as follows:

in step 21, the current road characteristics of the target track at the intersection to be predicted and the current passing type of the target track at the intersection to be predicted are determined.

The current passing type is used for representing the driving direction of the target track at the intersection to be predicted.

In addition, in the embodiment of the present application, the target trajectory may be a planned trajectory in advance, for example, the target trajectory may be a recommended travel path recommended by the network car booking service platform to the network car booking driver, or a navigation path recommended by the map service platform to a general user. In an alternative embodiment, the target positioning information may be obtained first, and then the target positioning information is matched with the road network through a Map Matching (MM) algorithm, so as to determine the position of the vehicle in the Map. The target positioning information may be positioning information of a vehicle, the positioning information of the vehicle may receive a satellite signal and perform a positioning operation at a predetermined period (for example, every 30 seconds) through a positioning device on the vehicle or a terminal device arranged on the vehicle, and then the longitude and latitude information (i.e., the target positioning information) obtained by the positioning operation is reported, and further, the longitude and latitude information reported by the vehicle may be matched with a road network to determine the position of the vehicle in a map.

After the position of the vehicle in the map is determined, an optimal driving route can be planned for the vehicle according to the position of the vehicle in the map, wherein the optimal driving route is the target track.

It should be noted that the target location information refers to information obtained through a legal way, for example, in a car-booking scenario, the car-booking service platform needs to obtain the target location information of the user after obtaining the authorization of the user.

As shown in fig. 3, fig. 3 is a schematic diagram of a target track provided in the embodiment of the present application, where the schematic diagram includes: vehicle 31, target track 32, intersection a, and intersection B.

As shown in fig. 3, before the vehicle 31 enters the intersection a, the embodiment of the present application can predict the probability of an accident occurring when the vehicle 31 passes through the intersection a with the target track 32 according to the target track 32, the current road characteristics of the intersection a, and the passing type of the target track 32 passing through the intersection a, where the intersection a is the intersection to be predicted, and the current passing type is a right turn.

In addition, the probability of the vehicle 32 causing an accident when the vehicle 31 passes through the intersection B with the target track 32 can also be predicted through the target track 32, the current road characteristics of the intersection B and the passing type of the target track 32 passing through the intersection B, at this time, the intersection B is the intersection to be predicted, and the current passing type is left turn.

In an alternative embodiment, the current passing type of the target track at the intersection to be predicted may be determined based on the entering road segment and the exiting road segment of the target track at the intersection to be predicted, and specifically, the process may be performed as: the method comprises the steps of determining an entrance road section and an exit road section of a target track at an intersection to be predicted, and determining the current passing type of the target track based on an included angle between the entrance road section and the exit road section of the target track at the intersection to be predicted.

Specifically, as shown in fig. 4, fig. 4 is a schematic diagram of an incoming road segment and an outgoing road segment provided in an embodiment of the present application, where the schematic diagram includes: the target track 41, the entrance section 42 of the target track 41 at the intersection a, the exit section 43 of the target track 41 at the intersection a, the entrance section 44 of the target track 41 at the intersection B, and the exit section 45 of the target track 41 at the intersection B.

As can be seen from fig. 4, the included angle between the entering section 42 and the exiting section 43 is 90 °, and therefore, the current passing type of the target trajectory 41 at the intersection a can be determined as a right turn according to the included angle. The angle between the incoming route section 44 and the outgoing route section 45 is 270 °, and therefore, the current passing type of the target trajectory 41 at the intersection B can be judged as a left turn according to the angle.

That is to say, according to the embodiment of the application, the current passing type corresponding to the included angle between the incoming road section and the outgoing road section can be determined according to the included angle between the incoming road section and the outgoing road section and the corresponding relationship between the pre-stored included angle and the passing type.

For example, as shown in the following table one, the table one is a correspondence table of an included angle and a passing type provided in the embodiment of the present application, and specifically, the following table is provided:

watch 1

Included angle	Type of passage
		[0°，90°]	Right turn
(90°，270°)	Straight going
		[270°，360°)	Left turn
360°	Turning round

It should be noted that the table i is only an example of the embodiment of the present application, and in practical application, the corresponding relationship between the included angle and the passing type may be appropriately adjusted according to practical situations.

In an alternative embodiment, the current road characteristics include a current intersection static characteristic, a current intersection dynamic characteristic, a current road section static characteristic, and a current road section dynamic characteristic;

the static characteristics of the current road junction are used for representing the objective attributes of the road junction, the dynamic characteristics of the current road junction are used for representing the driving rules of vehicles in the road junction, the static characteristics of the current road section are used for representing the objective attributes of the road section, and the dynamic characteristics of the current road section are used for representing the driving rules of the vehicles in the road section.

For example, the current intersection static characteristics may include: width, traffic direction, toll set, start and end point type, length, crossing limit, speed limit, type of intersection, traffic lights, pedestrian crosswalk, number of connected roads and the like. The current road segment static characteristics may include: the type, width, passing direction, toll arrangement, start and end point type, length, crossing limitation, speed limit, traffic lights, pedestrian crossing, number of connecting roads and the like of the road. It should be noted that the above example is only one example provided in the embodiments of the present application, and may be appropriately adjusted in practical applications.

In an implementation manner, the static features may be preprocessed by an encoding (e.g., one-hot encoding) technique, specifically, the electronic device may encode data of the static features to obtain a feature vector of the static features, and then the electronic device may perform calculation based on the feature vector of the static features.

One-hot (one-hot) coding, also known as one-bit-efficient coding, is a method of using an n-bit status register to encode n states, each state having its own independent register bit and only one of which is active at any one time. In the embodiment of the present application, the one-hot coding may be used to pre-process data of a category, for example, three types of intersections, namely, an intersection, a t-intersection, and other intersections are included in the category of an intersection, and after the one-hot coding is performed on the category of an intersection, three types of data 100 (an intersection), 010 (a t-intersection), and 001 (other intersections) may be obtained.

As shown in fig. 5, fig. 5 is a schematic diagram of a road segment provided in an embodiment of the present application, where the current road segment static characteristics of the road segment may include: two-way traffic lane, width 50 meters, length 200 meters, no charge and speed limit 60 kilometers/hour.

As shown in fig. 6, fig. 6 is a schematic diagram of an intersection provided by the embodiment of the present application, where the current intersection static characteristics of the intersection may include: crossroad, traffic lights, pedestrian crosswalk and 4 connecting roads.

It should be noted that fig. 5 and fig. 6 are only schematic diagrams provided in the present embodiment, and since the actual road segment or intersection is more complex than fig. 5 and fig. 6, the actual road segment or intersection may include more static features than fig. 5 and fig. 6.

The current intersection dynamic characteristics or the current road section dynamic characteristics can include: the track point density of the vehicles in the intersection/road section, the speed distribution of the vehicles in the intersection/road section, and the overspeed distribution information in the intersection/road section, etc. It should be noted that the above example is only one example provided in the embodiments of the present application, and may be appropriately adjusted in practical applications.

Wherein, the track point density of vehicle can be used for the representation to pass through the vehicle flowrate in crossing or highway section, and in addition, the track point density of vehicle can also be divided into the track point density of vehicle during peak period, the track point density of vehicle during flat peak period and the track point density of vehicle during night period.

The speed distribution of the vehicles can be used for representing the congestion condition of an intersection or a road section, and in addition, the speed distribution of the vehicles can be further divided into the speed distribution of the vehicles in a peak period, the speed distribution of the vehicles in a flat period and the speed distribution of the vehicles in a night period.

The overspeed distribution information can be used for representing overspeed violation conditions of intersections or road sections, and in addition, the overspeed distribution information can also be divided into peak overspeed distribution information, flat peak overspeed distribution information and night overspeed distribution information.

In combination with the above, the current intersection dynamic feature or the current road section dynamic feature may be predetermined based on historical positioning information of a plurality of vehicles, specifically, as shown in fig. 7, fig. 7 is a flowchart for determining the dynamic feature provided in this embodiment of the present application.

The positioning information of the vehicles 1 to n in fig. 7 is: positioning information in the process that the vehicles 1 to n pass through the intersection or the road section.

Specifically, in the embodiment of the present application, the positioning information of the vehicles 1 to n passing through a certain road or a certain intersection may be obtained, then the positioning information of the vehicles 1 to n is matched with the road network through the MM algorithm, so as to obtain the tracks 1 to n of the vehicles 1 to n in the road network, and then the track point density of the vehicles at the peak/flat peak/night time, the speed distribution of the vehicles at the peak/flat peak/night time, and the overspeed distribution information at the peak/flat peak/night time of the road or the intersection may be calculated according to the tracks 1 to n.

Furthermore, in the process of determining the target accident probability, the dynamic characteristics of the current intersection or the dynamic characteristics of the current road section can be preprocessed, the characteristic vector of the dynamic characteristics of the current intersection or the current road section is determined, and then the electronic device can calculate based on the characteristic vector of the dynamic characteristics of the current intersection or the current road section.

Specifically, the preprocessing of the dynamic features of the current intersection or the current road section can be performed in a numerical normalization mode, the normalization processing is generally used for preprocessing numerical data (such as track point density of a vehicle), and specifically, the numerical data can be mapped between (0 and 1) through the normalization processing, so that the numerical data can be processed more conveniently and rapidly.

In step 22, the current road characteristics and the current passing type are input into a pre-trained intersection risk prediction model, and the probability of a target accident that a target track passes through the intersection to be predicted is determined.

In an alternative embodiment, the intersection risk prediction model is obtained by type training based on at least historical road characteristics and history of a plurality of historical tracks at the corresponding accident intersections.

In practical applications, the intersection risk prediction model may adopt a Gradient Boosting iterative Decision Tree (GBDT) model.

The GBDT model is composed of a series of decision trees, complex interaction relations among features can be captured, the probability of traffic accidents can be accurately predicted based on the GBDT model, in addition, terms with regularization effects can be introduced in the calculation process, and overfitting of the model can be effectively prevented.

It should be noted that, for a specific implementation framework, the electronic device may adopt any applicable implementation framework (for example, an XGBoost framework, etc.), and the embodiment of the present application does not specifically limit the selection of the implementation framework.

In an alternative embodiment, in addition to the current road characteristics and the current passing type, parameters for characterizing the state attributes of the driver may be input into a pre-trained intersection risk prediction model, specifically, step 22 may be performed as: and inputting the current road characteristics, the current passing type and the current driving state information into a pre-trained intersection risk prediction model, and determining the probability of a target accident of passing through the intersection to be predicted by using a target track.

Wherein the current driving state information is used to characterize a state attribute of the driver. For example, the current driving state information may include: whether the driver has experienced dangerous driving behavior, the driver's age, the driver's driving age, the driver's fatigue information, and the like. Taking the network car booking scene as an example, the dangerous driving behaviors can include the behaviors of overspeed of a network car booking driver in a past period, call receiving and making during driving, safety belt fastening and the like, and the fatigue information of the driver can include the continuous working time of the network car booking driver and the like.

In an alternative embodiment, the intersection risk prediction model may be trained based on a training set, and specifically, the intersection risk prediction model may be trained based on the following steps: acquiring a training set, and training an intersection risk prediction model based on historical road characteristics, historical passing types and accident labels.

The training set comprises historical road characteristics of each historical track at the corresponding accident intersection, historical passing types of each historical track at the corresponding accident intersection and accident marks corresponding to each historical track at the corresponding accident intersection.

In this embodiment of the present application, the historical track may be determined by historical positioning information, and specifically, the process may be executed as: and determining historical positioning information of the historical passing events, and performing map matching on the historical positioning information and road network data to determine historical tracks of the historical passing events.

In the embodiment of the application, the historical road characteristics can comprise historical intersection static characteristics, historical intersection dynamic characteristics, historical road section static characteristics and historical road section dynamic characteristics.

The static characteristics of the historical road junction are used for representing the objective attributes of the road junction, the dynamic characteristics of the historical road junction are used for representing the driving rules of vehicles in the road junction, the static characteristics of the historical road section are used for representing the objective attributes of the road section, and the dynamic characteristics of the historical road section are used for representing the driving rules of the vehicles in the road section.

For example, the historical intersection static characteristics may include: width, traffic direction, toll set, start and end point type, length, crossing limit, speed limit, type of intersection, traffic lights, pedestrian crosswalk, number of connected roads and the like. The historical road segment static characteristics may include: the type, width, passing direction, toll arrangement, start and end point type, length, crossing limitation, speed limit, traffic lights, pedestrian crossing, number of connecting roads and the like of the road. The current intersection dynamic characteristics or current road segment dynamic characteristics may include: the track point density of the vehicles in the intersection/road section, the speed distribution of the vehicles in the intersection/road section, and the overspeed distribution information in the intersection/road section, etc.

It should be noted that the above example is only one example provided in the embodiments of the present application, and may be appropriately adjusted in practical applications.

For the historical passing type of each historical track at the corresponding accident intersection, the historical passing type can be determined based on the driving-in road section and the driving-out road section of the historical track at the corresponding accident intersection, and specifically, the process can be executed as follows: and determining the driving-in road section and the driving-out road section of the historical track at the corresponding accident intersection according to each historical track, and determining the historical passing type of the historical track based on the included angle between the driving-in road section and the driving-out road section of the historical track at the corresponding accident intersection.

The process of determining the history passing type may refer to the content shown in fig. 4 and table one, which is not described herein again in this embodiment of the present application.

In the training process, in order to ensure the effectiveness of training, a training set is generally constructed by adopting data of real traffic accident events, so how to effectively mine accidents occurring at intersections is a problem which needs to be solved urgently.

In order to effectively mine accidents occurring at intersections, the embodiment of the application accurately judges the time of the accidents according to the sound generated when the accidents occur, and further judges the places of the accidents according to the accurate time of the accidents. Specifically, the process may be performed as: the method comprises the steps of determining audio data corresponding to historical tracks aiming at each historical track, responding to the fact that collision sound is detected in the audio data, determining accident occurrence time points corresponding to the collision sound, determining accident occurrence position points corresponding to the accident occurrence time points according to the accident occurrence time points and the historical tracks, responding to the fact that the distance between the accident occurrence position points and an intersection is smaller than a preset distance, and outputting accident intersections, driving-in road sections, driving-out road sections and accident marks corresponding to the historical tracks.

In the embodiment of the application, the audio data is legally recorded audio data, for example, in the process of the network car booking service, in order to ensure the personal safety of the user, whether the user receives the whole-course audio and video recording is firstly inquired, and if the user receives the whole-course audio and video recording in the process of the network car booking service, the audio and video recording equipment installed inside the network car booking will be started in the whole course in the process of the network car booking running. Wherein, this recording equipment generally is the recording equipment that the network car appointment platform was equipped with in unison.

For example, as shown in fig. 8, fig. 8 is a flowchart of accident mining provided in an embodiment of the present application.

In step 81, audio data corresponding to the historical track is determined.

Wherein the audio data is recorded by legal means.

In step 82, it is determined whether or not a collision sound is detected in the audio data, and if a collision sound is detected in the audio data, step 83 is executed, and if a collision sound is not detected in the audio data, step 81 is executed.

In practical applications, traffic accidents are generally accompanied by sounds which do not appear during normal driving, such as collisions, that is, after the audio data corresponding to the historical track is determined, the embodiment of the application can detect whether the audio data contains abnormal sounds.

In step 83, the accident occurrence time point corresponding to the impact sound is determined.

In practical applications, because the sound generated when an accident occurs is synchronized with the accident itself, the time when the accident occurs can be accurately determined by the time when the collision sound occurs, that is, the time when the collision sound occurs is the time when the accident occurs.

At step 84, an accident location point corresponding to the accident time point is determined.

Since the time determined in step 83 is the exact time of the accident occurrence, the exact accident occurrence location point can be determined at the location corresponding to the time of the accident occurrence in the history track according to the time of the accident occurrence.

In step 85, it is determined whether the accident position is less than 50 meters from the intersection, if the accident position is less than 50 meters from the intersection, step 86 is executed, and if the accident position is greater than or equal to 50 meters from the intersection, step 81 is executed.

The "50 meters" in step 85 is an example of the embodiment of the present application, and may be appropriately adjusted according to actual situations in practical applications.

At step 86, the accident intersection, the driving-in road section, the driving-out road section and the accident sign corresponding to the historical track are output.

In the embodiment of the application, the accurate time of the accident can be determined based on the collision sound in the audio data, then the accurate accident position point can be determined according to the accurate accident time, and then whether the accident occurs at the intersection can be judged according to the accurate accident position point. Therefore, through the embodiment of the application, accidents occurring at the intersection can be accurately mined, and the effectiveness of training data is improved.

During the training process, the training set may further include historical driving state information corresponding to each historical track, where the historical driving state information is used to characterize a state attribute of the driver, and for example, the historical driving state information may include: whether the driver has experienced dangerous driving behavior, the driver's age, the driver's driving age, the driver's fatigue information, and the like.

Further, the process of training based on the training set including the historical driving state information may be: and training an intersection risk prediction model based on the historical road characteristics, the historical passing type, the historical driving state information and the accident signs.

Through tests in practical application, when the trained intersection risk prediction model is used for predicting the probability of traffic accidents at an intersection within 30 days, the Area Under the working characteristic Curve (AUC) of a subject is 88%.

The more the value of AUC is close to 100%, the more accurate the prediction probability of the representation model is, and therefore the accident probability determined by the intersection risk prediction model obtained through training in the embodiment of the application is more accurate.

After the model training is finished, the trained intersection risk prediction model can be applied to an online platform and used for predicting the probability of accidents of the target track at the intersection to be predicted.

For example, as shown in fig. 9, fig. 9 is an exemplary flowchart of an intersection risk prediction model provided in the embodiment of the present application in practical application.

In step 91, target location information is determined.

The target positioning information may be real-time longitude and latitude information of the vehicle.

In step 92, the target location information is bound to the road network to determine the location of the vehicle in the map.

In step 93, it is determined whether the position of the vehicle on the map is located on the accident-prone link, and if the position of the vehicle on the map is located on the accident-prone link, step 94 is performed, and if the position of the vehicle on the map is not located on the accident-prone link, step 91 is performed.

At step 94, a target trajectory is determined.

The target track can be a preferred driving route planned in real time according to the position of the vehicle in the map.

In step 95, it is determined whether the target track includes an intersection, and if the target track includes an intersection, step 96 is executed, and if the target track does not include an intersection, step 91 is executed.

At step 96, current road characteristics and current pass type are determined.

Step 96 is the same as that described in step 21, and repeated contents are not described herein in this embodiment of the application.

At step 97, a target accident probability is determined.

The target accident probability may be determined based on a pre-trained intersection risk prediction model, that is, step 22, and repeated content is not described herein in this embodiment of the present application.

In addition, in the embodiment of the application, because the current passing type is introduced when the target accident probability is determined, the target accident probability corresponding to the current passing type can be determined according to the current passing type. For example, if the current pass type is "turn left", the target accident probability corresponding to "turn left" may be determined in step 97.

In step 98, it is determined whether the target probability of the accident is greater than a predetermined threshold, and if the target probability of the accident is greater than the predetermined threshold, step 99 is performed, and if the target probability of the accident is less than or equal to the predetermined threshold, step 91 is performed.

The predetermined threshold may be a probability threshold set according to actual situations, for example, the predetermined threshold may be 5%, 6%, 10%, and so on.

In step 99, the reminder is announced.

In the embodiment of the application, the broadcast prompt is used for prompting that the intersection which the driver is going to pass through is the accident-prone intersection, and if the broadcast prompt is needed to be performed on the driver, a prompt instruction can be sent to the terminal device of the driver (for example, a smartphone of the driver) or the vehicle-mounted terminal inside the vehicle driven by the driver, so that the terminal device of the driver or the vehicle-mounted terminal inside the vehicle driven by the driver can play a prompt audio, for example, "the front is the accident-prone intersection, and please drive carefully".

In addition, in the embodiment of the application, when the target accident probability is determined, the target accident probability corresponding to the current passing type can be determined according to the current passing type. Further, in step 99, a reminder may be announced for the current pass type. For example, "the front is the left turn accident high-incidence intersection, please drive carefully".

By the embodiment of the application, the target accident probability of the target track passing through the intersection to be predicted can be predicted at least based on the current road characteristics of the target track at the intersection to be predicted, the current passing type of the target track at the intersection to be predicted and a pre-trained intersection risk prediction model. In the process, the accuracy of prediction can be effectively improved by the current road characteristics, the intersection to be predicted can be divided into finer granularity according to the type, and in addition, manual participation is not needed in the process, namely, the efficiency and the accuracy of predicting the accident probability at the intersection can be improved through the embodiment of the method and the device, meanwhile, the type of the intersection can be refined, and further, the risk prediction can be more accurately carried out.

Based on the same technical concept, an embodiment of the present application further provides a data processing apparatus, as shown in fig. 10, the apparatus includes: a first determination module 101 and a prediction module 102.

The first determining module 101 is configured to determine a current road characteristic of a target track at an intersection to be predicted and a current passing type of the target track at the intersection to be predicted, where the current passing type is used to represent a driving direction of the target track at the intersection to be predicted.

And the prediction module 102 is configured to input the current road characteristics and the current passing type into a pre-trained intersection risk prediction model, and determine a target accident probability of passing the target trajectory through the intersection to be predicted.

Optionally, the intersection risk prediction model is obtained by type training based on at least historical road characteristics and history of a plurality of historical tracks at the corresponding accident intersections.

Optionally, the intersection risk prediction model is trained based on the following modules:

the acquisition module is used for acquiring a training set, wherein the training set comprises historical road characteristics of each historical track at the corresponding accident intersection, historical passing types of each historical track at the corresponding accident intersection and accident marks corresponding to each historical track at the corresponding accident intersection.

And the training module is used for training the intersection risk prediction model based on the historical road characteristics, the historical passing type and the accident labels.

Optionally, the apparatus further comprises:

and the second determination module is used for determining an entrance road section and an exit road section of each historical track at the corresponding accident intersection.

And the third determining module is used for determining the historical passing type of the historical track based on the included angle between the driving-in road section and the driving-out road section of the historical track at the corresponding accident intersection.

Optionally, the apparatus further comprises:

and the audio data module is used for determining the audio data corresponding to each historical track.

And the accident occurrence time point module is used for responding to the detection of the collision sound in the audio data and determining the accident occurrence time point corresponding to the collision sound.

And the accident occurrence position point module is used for determining an accident occurrence position point corresponding to the accident occurrence time point according to the accident occurrence time point and the historical track.

And the output module is used for responding to the situation that the distance between the accident occurrence position point and the intersection is smaller than the preset distance, and outputting the accident intersection, the driving-in road section, the driving-out road section and the accident mark corresponding to the historical track.

Optionally, the apparatus further comprises:

and the historical positioning information module is used for determining the historical positioning information of the historical passing event.

And the map matching module is used for performing map matching on the historical positioning information and the road network data to determine the historical track of the historical passing event.

Optionally, the historical road characteristics include historical intersection static characteristics, historical intersection dynamic characteristics, historical road section static characteristics and historical road section dynamic characteristics.

The static characteristics of the historical road junction are used for representing objective attributes of the road junction, the dynamic characteristics of the historical road junction are used for representing driving rules of vehicles in the road junction, the static characteristics of the historical road section are used for representing objective attributes of the road section, and the dynamic characteristics of the historical road section are used for representing driving rules of the vehicles in the road section.

Optionally, the training set further includes historical driving state information corresponding to each historical track, and the historical driving state information is used to represent a state attribute of the driver.

The training module is specifically configured to:

training the intersection risk prediction model based on the historical road characteristics, the historical passing type, the historical driving state information and the accident sign.

Optionally, the apparatus further comprises:

and the fourth determination module is used for determining the driving-in road section and the driving-out road section of the target track at the intersection to be predicted.

And the fifth determining module is used for determining the current passing type of the target track based on the included angle between the driving-in road section and the driving-out road section of the target track at the intersection to be predicted.

Optionally, the current road characteristic includes a current intersection static characteristic, a current intersection dynamic characteristic, a current road section static characteristic, and a current road section dynamic characteristic.

Optionally, the prediction module 102 is specifically configured to:

and inputting the current road characteristics, the current passing type and the current driving state information into a pre-trained intersection risk prediction model, and determining the probability of a target accident when the target track passes through the intersection to be predicted, wherein the current driving state information is used for representing the state attribute of a driver.

Fig. 11 is a schematic diagram of an electronic device according to an embodiment of the present application. As shown in fig. 11, the electronic device shown in fig. 11 is a general address query device, which includes a general computer hardware structure, which includes at least a processor 111 and a memory 112. The processor 111 and the memory 112 are connected by a bus 113. The memory 112 is adapted to store instructions or programs executable by the processor 111. Processor 111 may be a stand-alone microprocessor or may be a collection of one or more microprocessors. Thus, the processor 111 implements the processing of data and the control of other devices by executing the instructions stored by the memory 112 to perform the method flows of the embodiments of the present application as described above. The bus 113 connects the above components together, and also connects the above components to a display controller 114 and a display device and an input/output (I/O) device 115. Input/output (I/O) device 115 may be a mouse, keyboard, modem, network interface, touch input device, motion sensing input device, printer, and other devices known in the art. Typically, the input/output device 115 is connected to the system through an input/output (I/O) controller 116.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, apparatus (device) or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may employ a computer program product embodied on one or more computer-readable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations of methods, apparatus (devices) and computer program products according to embodiments of the application. It will be understood that each flow in the flow diagrams can be implemented by computer program instructions.

These computer program instructions may be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows.

These computer program instructions may also be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows.

Another embodiment of the present application is directed to a non-transitory storage medium storing a computer-readable program for causing a computer to perform some or all of the above-described method embodiments.

That is, as can be understood by those skilled in the art, all or part of the steps in the method for implementing the embodiments described above may be accomplished by specifying the relevant hardware through a program, where the program is stored in a storage medium and includes several instructions to enable a device (which may be a single chip, a chip, or the like) or a processor (processor) to execute all or part of the steps of the method described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.

Another embodiment of the present application relates to a computer program product comprising computer programs/instructions which, when executed by a processor, may implement some or all of the above-described method embodiments.

That is, as can be understood by those skilled in the art, the embodiments of the present application may specify related hardware (including the processor itself) by the processor executing the computer program product (computer program/instruction), so as to implement all or part of the steps in the method of the above embodiments.

The above description is only a preferred embodiment of the present application and is not intended to limit the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

The embodiment of the application discloses a TS1 and a data processing method, wherein the method comprises the following steps:

and inputting the current road characteristics and the current passing type into a pre-trained intersection risk prediction model, and determining the probability of a target accident that the target track passes through the intersection to be predicted.

The TS2, the method as in TS1, wherein the intersection risk prediction model is trained by type based on at least historical road characteristics and history of a plurality of historical tracks at respective corresponding accident intersections.

TS3, the method of TS1 or TS2, wherein the intersection risk prediction model is trained based on the steps of:

acquiring a training set, wherein the training set comprises historical road characteristics of each historical track at a corresponding accident intersection, historical passing types of each historical track at the corresponding accident intersection and accident marks of each historical track at the corresponding accident intersection; and

and training the intersection risk prediction model based on the historical road characteristics, the historical passing type and the accident signs.

TS4, the method of TS3, wherein the method further comprises:

aiming at each historical track, determining an entrance road section and an exit road section of the historical track at the corresponding accident intersection; and

and determining the historical passing type of the historical track based on the included angle between the driving-in road section and the driving-out road section of the historical track at the corresponding accident intersection.

TS5, the method of TS3, wherein the method further comprises:

for each historical track, determining audio data corresponding to the historical track;

in response to detecting a collision sound in the audio data, determining an accident occurrence time point corresponding to the collision sound;

determining an accident occurrence position point corresponding to the accident occurrence time point according to the accident occurrence time point and the historical track; and

and responding to the situation that the distance between the accident occurrence position point and the intersection is less than the preset distance, and outputting an accident intersection, an entrance road section, an exit road section and an accident mark corresponding to the historical track.

TS6, the method of TS3, wherein the method further comprises:

determining historical positioning information of the historical passing event; and

and performing map matching on the historical positioning information and road network data to determine the historical track of the historical passing event.

The TS7, the method as TS3, wherein the historical road characteristics comprise historical intersection static characteristics, historical intersection dynamic characteristics, historical road section static characteristics and historical road section dynamic characteristics;

TS8, the method of TS3, wherein the training set further includes historical driving state information corresponding to each historical trajectory, the historical driving state information being used to characterize a state attribute of the driver;

the training of the intersection risk prediction model based on the historical road characteristics, the historical passage type and the accident sign comprises:

TS9, the method of TS1, wherein the method further comprises:

determining an entrance road section and an exit road section of the target track at the intersection to be predicted; and

and determining the current passing type of the target track based on the included angle between the driving-in road section and the driving-out road section of the target track at the intersection to be predicted.

TS10, the method of TS1, wherein the current road characteristics include a current intersection static characteristic, a current intersection dynamic characteristic, a current road segment static characteristic, and a current road segment dynamic characteristic;

The TS11 method of TS1, wherein the inputting the current road characteristic and the current passing type into a pre-trained intersection risk prediction model to determine a target accident probability of passing the intersection to be predicted with the target track, comprises:

TS12, a data processing apparatus, wherein the apparatus comprises:

the device comprises a first determining module, a second determining module and a judging module, wherein the first determining module is used for determining the current road characteristics of a target track at an intersection to be predicted and the current passing type of the target track at the intersection to be predicted, and the current passing type is used for representing the driving direction of the target track at the intersection to be predicted; and

TS13, the apparatus of TS12, wherein the intersection risk prediction model is obtained by type training based on at least historical road characteristics and history of a plurality of historical tracks at each corresponding accident intersection.

TS14, the apparatus as described in TS12 or TS13, wherein the intersection risk prediction model is trained based on the following modules:

the acquisition module is used for acquiring a training set, wherein the training set comprises historical road characteristics of each historical track at the corresponding accident intersection, historical passing types of each historical track at the corresponding accident intersection and accident marks corresponding to each historical track at the corresponding accident intersection; and

TS15, the apparatus of TS14, wherein the apparatus further comprises:

the second determination module is used for determining an entrance road section and an exit road section of each historical track at the corresponding accident intersection; and

TS16, the apparatus of TS14, wherein the apparatus further comprises:

the audio data module is used for determining audio data corresponding to each historical track;

an accident occurrence time point module, configured to determine an accident occurrence time point corresponding to a collision sound in response to detecting the collision sound in the audio data;

the accident occurrence position point module is used for determining an accident occurrence position point corresponding to the accident occurrence time point according to the accident occurrence time point and the historical track; and

and the output module is used for responding that the distance between the accident occurrence position point and the intersection is less than the preset distance, and outputting the accident intersection, the driving-in road section, the driving-out road section and the accident mark corresponding to the historical track.

TS17, the apparatus of TS14, wherein the apparatus further comprises:

the historical positioning information module is used for determining the historical positioning information of the historical passing event; and

TS18, the apparatus of TS14, wherein the historical road characteristics include historical intersection static characteristics, historical intersection dynamic characteristics, historical road segment static characteristics and historical road segment dynamic characteristics;

The TS19, the apparatus as in TS14, wherein the training set further includes historical driving state information corresponding to each historical track, the historical driving state information being used to characterize a state attribute of a driver;

the training module is specifically configured to:

and training the intersection risk prediction model based on the historical road characteristics, the historical passing type, the historical driving state information and the accident sign.

TS20, the apparatus of TS12, wherein the apparatus further comprises:

the fourth determination module is used for determining the driving-in road section and the driving-out road section of the target track at the intersection to be predicted; and

TS21, the device as in TS12, wherein the current road characteristics include a current intersection static characteristic, a current intersection dynamic characteristic, a current road section static characteristic and a current road section dynamic characteristic;

TS22, the apparatus of TS12, wherein the prediction module is specifically configured to:

TS23, an electronic device comprising a memory and a processor, wherein the memory is configured to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement a method as set forth in any one of TS1-TS 11.

TS24, a computer readable storage medium, wherein the computer readable storage medium has stored therein a computer program which, when executed by a processor, implements the method of any one of TS1-TS 11.

TS25, a computer program product comprising computer programs/instructions, wherein the computer programs/instructions, when executed by a processor, implement the method of any one of TS1-TS 11.

Claims

1. A method of data processing, the method comprising:

2. The method of claim 1, wherein the intersection risk prediction model is trained based on the steps of:

3. The method of claim 2, further comprising:

4. The method of claim 2, further comprising:

5. The method of claim 2, wherein the training set further comprises historical driving state information corresponding to each historical track, the historical driving state information being used to characterize a state attribute of the driver;

6. The method of claim 1, further comprising:

7. The method of claim 1, wherein the inputting the current road characteristic and the current passing type into a pre-trained intersection risk prediction model to determine a target accident probability of passing the target track through the intersection to be predicted comprises:

8. An electronic device comprising a memory and a processor, wherein the memory is configured to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method of any of claims 1-7.

9. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the method of any one of claims 1 to 7.

10. A computer program product comprising computer programs/instructions, characterized in that the computer programs/instructions, when executed by a processor, implement the method of any of claims 1-7.