Training method of risk radius prediction model and method of determining transportation path

Abstract

This disclosure provides a training method, apparatus, electronic device, and medium for a risk radius prediction model and a method for determining transportation routes, relating to the field of transportation technology. The specific implementation involves: determining accident information for each accident sample in a historical traffic accident sample set, wherein the accident information includes at least accident segment information; determining the accident impact radius of each accident sample and using the accident impact radius as the corresponding annotation information in the accident sample; training a risk radius prediction model based on the accident information and corresponding annotation information of each accident sample, wherein the risk radius prediction model is used to predict the transportation risk radius of transport vehicles on various road segments. The technical solution of this disclosure can train a risk radius prediction model capable of predicting the transportation risk radius of transport vehicles on various road segments, helping to quickly and efficiently determine safe transportation routes for vehicles and ensuring the transportation safety of oversized transport vehicles on highways.

Description

Technical Field

[0001] The present disclosure relates to the field of transportation technology, and in particular to a training method for a risk radius prediction model and a method, device, electronic device and medium for determining a transportation path. Background Art

[0002] Large-item transportation is an important part of daily social production and life, and large-item transportation accidents often have serious impacts on society. Therefore, for highway traffic management, it is very important to reduce the impact risk during large-item transportation. Determining the transportation risk radius of large-item transportation vehicles on each section of the highway can provide a reference for the travel route planning of large-item transportation vehicles and ensure the transportation safety of large-item transportation vehicles on the highway. Summary of the invention

[0003] The present disclosure provides a training method for a risk radius prediction model and a method, device, electronic device and medium for determining a transportation path.

[0004] According to a first aspect of the present disclosure, a training method for a risk radius prediction model is provided, comprising: determining accident information of each accident sample in a historical traffic accident sample set, wherein the accident information at least includes accident road section information;

[0005] Determine the accident impact radius of each accident sample, and use the accident impact radius as the annotation information corresponding to the accident information in the accident sample;

[0006] According to the accident information of each accident sample and the corresponding annotation information, a risk radius prediction model is trained, wherein the risk radius prediction model is used to predict the transportation risk radius of the transport vehicle on each section of the highway.

[0007] According to a second aspect of the present disclosure, a method for determining a transportation path is provided, comprising:

[0008] Determining the transportation risk radius of the target vehicle on each of the multiple road sections using a risk radius prediction model, wherein the multiple road sections are all road sections within the target area, and the risk radius prediction model is obtained by training based on the method of the first aspect of the present disclosure;

[0009] According to the transportation risk radius of the target vehicle on each road section, determine the local risk value of the target vehicle on each road section;

[0010] According to the local risk value, a transportation path of the target vehicle in the target area is determined, wherein the local risk value of the section with the largest local risk value in the transportation path is less than the risk threshold.

[0011] According to a third aspect of the present disclosure, a training device for a risk radius prediction model is provided, comprising:

[0012] An accident information determination module, used to determine the accident information of each accident sample in the historical traffic accident sample set, wherein the accident information at least includes the accident road section information;

[0013] A labeling information determination module is used to determine the accident impact radius of each accident sample, and use the accident impact radius as labeling information corresponding to the accident information in the accident sample;

[0014] The training module is used to train a risk radius prediction model based on the accident information of each accident sample and the corresponding annotation information, wherein the risk radius prediction model is used to predict the transportation risk radius of the transport vehicle on each section of the highway.

[0015] According to a fourth aspect of the present disclosure, a device for determining a transport path is provided, comprising:

[0016] A risk radius determination module, used to determine the transportation risk radius of the target vehicle on each of the multiple road sections using a risk radius prediction model, wherein the multiple road sections are all road sections within the target area, and the risk radius prediction model is obtained by training based on the method of the first aspect of the present disclosure;

[0017] A risk value determination module is used to determine the local risk value of the target vehicle on each road section according to the transportation risk radius of the target vehicle on each road section;

[0018] The path determination module is used to determine the transportation path of the target vehicle in the target area according to the local risk value, wherein the local risk value of the section with the largest local risk value in the transportation path is less than the risk threshold.

[0019] According to a fifth aspect of the present disclosure, there is provided an electronic device, including:

[0020] at least one processor; and

[0021] a memory communicatively connected to the at least one processor; wherein,

[0022] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method in any embodiment of the present disclosure.

[0023] According to a sixth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable a computer to execute the method in any embodiment of the present disclosure.

[0024] According to the training method of the risk radius prediction model disclosed in the present invention, historical accident data can be used as training samples to train the model, and a risk radius prediction model that can predict the transportation risk radius of transport vehicles on each section of the highway can be obtained through training. This makes it possible to evaluate the transportation risk of the transport path based on the transportation risk radius predicted by the risk radius prediction model, thereby helping to quickly and efficiently determine a safe transportation path for the vehicle and ensure the transportation safety of large-scale transport vehicles on the highway.

[0025] According to the method for determining the transport path disclosed in the present invention, the risk radius prediction model can be used to quickly predict the transport risk radius of the target vehicle on each road section in the target area, and then the transport risk of each road section can be quickly determined, providing a reference for the selection of the transport path for the target vehicle, thereby greatly improving the efficiency of decision-making for the selection of a safe transport path.

[0026] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0028] Figure 1 is a flow chart of a method for training a risk radius prediction model according to an embodiment of the present disclosure;

[0029] Figure 2 is a flow chart of a method for determining a transport path according to an embodiment of the present disclosure;

[0030] Figure 3 is a schematic diagram of a training device for a risk radius prediction model according to an embodiment of the present disclosure;

[0031] Figure 4 is a schematic diagram of a device for determining a transport path according to an embodiment of the present disclosure;

[0032] Figure 5 is a block diagram of an electronic device for implementing an embodiment of the present disclosure. DETAILED DESCRIPTION

[0033] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0034] Figure 1 : is a flow chart of a method for training a risk radius prediction model according to an embodiment of the present disclosure, comprising:

[0035] Step S110, determining the accident information of each accident sample in the historical traffic accident sample set, wherein the accident information at least includes the accident road section information;

[0036] Step S120, determining the accident impact radius of each accident sample, and using the accident impact radius as labeling information corresponding to the accident information in the accident sample;

[0037] Step S130, training a risk radius prediction model based on the accident information of each accident sample and the corresponding annotation information, wherein the risk radius prediction model is used to predict the transportation risk radius of the transport vehicle on each section of the highway.

[0038] For example, the historical traffic accident sample set includes record information of traffic accidents occurring on a highway within a period of time, the accident section information refers to information of the section where the traffic accident occurred, and the accident impact radius is used to indicate the scope affected by the traffic accident.

[0039] It is understandable that for transport vehicles, especially those transporting large items, since the transport vehicles themselves are large trucks with high loads, the risk of traffic accidents on the road is greater. Therefore, government management departments or freight carriers need to predict the transportation risks of the vehicle's optional transportation routes, so as to select appropriate transportation routes to minimize the possibility of accidents involving transport vehicles and ensure the transportation safety of large-scale transport vehicles on the road. The transportation risk radius represents the predicted impact range of a transportation accident involving a vehicle on a specific road section, which is crucial for the transportation risk assessment of the transportation route.

[0040] By adopting the method of the embodiment of the present application, historical accident data can be used as a training sample training model to obtain a risk radius prediction model that can predict the transportation risk radius of transport vehicles on each section of the highway. This makes it possible to evaluate the transportation risk of the transport path based on the transportation risk radius predicted by the risk radius prediction model, thereby helping to quickly and efficiently determine a safe transportation path for the vehicle and ensure the transportation safety of large-scale transport vehicles on the highway.

[0041] In one embodiment, in step S110, the accident information of each accident sample in the historical traffic accident sample set is determined, including: based on traffic survey data, obtaining the traffic volume information and the average vehicle speed of the road section where the accident occurred in each accident sample; wherein the traffic volume information includes the number of vehicles in the road section and the proportion of different types of vehicles; and using the traffic volume information and the average vehicle speed of the road section as the accident section information.

[0042] For example, vehicles have mobility, so for each road section, the traffic volume on the road section at different times tends to stabilize within a certain range. The average of the traffic volume of the road section at multiple different times can be used to represent the traffic volume of the road section under normal circumstances. Therefore, for each accident sample, the traffic survey data can be queried to obtain the traffic volume information and the average vehicle speed of the road section where the accident occurred in each accident sample as the accident section information.

[0043] It is understandable that traffic volume information and the average vehicle speed of a road section can to a certain extent reflect the possibility of a traffic accident occurring on the road section. For example, if there are too many vehicles or too many large vehicles on the road section, the possibility of a traffic accident occurring on the road section will increase, and the severity of the traffic accident will also increase, so the impact range of the traffic accident will be relatively increased. For another example, if the average speed of vehicles on the road section is high, it will also increase the possibility of a traffic accident and increase the impact range of a traffic accident.

[0044] Optionally, the accident road section information may also include basic attribute information of the road section, such as road section material, road section width, etc. The vehicle type classification can be based on the functionality of the vehicle, such as sedans, buses, trucks, etc., and can also be based on the vehicle model, axle type and wheel type. The specific classification method can be set according to actual needs, and this application is not limited to this.

[0045] In one embodiment, in step S110, the accident information also includes accident vehicle information, and determining the accident information of each accident sample in the historical traffic accident sample set also includes: determining the size of the accident vehicle in each accident sample as the accident vehicle information, wherein the size of the accident vehicle includes at least the length, width and height of the accident vehicle.

[0046] It is understandable that the impact range of a traffic accident is not only related to the road conditions, but also to the vehicle's own properties. The larger the vehicle, the larger the impact area will be after the traffic accident.

[0047] In one embodiment, in step S120, the accident impact radius of each accident sample is determined, including: determining the accident impact radius of each accident sample based on the accident impact area recorded in each accident sample; or, determining the accident impact radius of each accident sample using an expert investigation method based on the accident information of each accident sample.

[0048] For example, the accident sample may record the specific impact caused by the traffic accident, such as the time of suspension of traffic on the section of road where the accident occurred or other vehicles or buildings affected by the accident. The accident impact radius of the traffic accident can be determined based on the specific impact recorded in the accident sample. In addition, when the specific impact of the accident is not recorded in the accident sample, the accident impact radius can also be manually determined for the accident sample through expert survey method (also known as Delphi method), so that the accident impact radius can be used as the annotation information of the accident information to train the risk radius prediction model, so that the model predicts the transportation risk radius based on the input information.

[0049] In one embodiment, in step S130, a risk radius prediction model is trained based on the accident information of each accident sample and the corresponding annotation information, including: inputting the accident information of the i-th accident sample among the accident samples into a preset model to obtain the transportation risk radius output by the preset model; wherein i is an integer greater than or equal to 1; adjusting the parameters of the preset model based on the transportation risk radius output by the preset model and the annotation information corresponding to the accident information of the i-th road section; and using the preset model as the risk radius prediction model when the preset convergence conditions are met.

[0050] Exemplarily, the preset model may be a model designed based on a deep learning network architecture. For example, the preset model may include a convolutional network and a decoder. The convolutional network is used to extract feature information of the accident information (at least including the accident road section information) input into the model, and the decoder is used to decode the feature information to obtain the predicted transport risk radius of the transport vehicle on the road section.

[0051] It can be understood that the accident information of each accident sample among multiple accident samples (at least including the accident road section information) is used as the sample accident information of the risk radius prediction model. By inputting the traffic flow information of the i-th accident sample among the multiple accident samples into the preset model, the prediction model predicts and outputs the transportation risk radius corresponding to the i-th accident sample, and uses the accident information of each accident sample and the accident impact radius of the i-th road section actually obtained as the annotation information. According to the loss of the predicted information compared to the annotation information, the parameters of the preset model are adjusted.

[0052] Exemplarily, the preset convergence condition may be that the difference between the predicted information and the labeled information is less than a preset threshold, wherein the difference may be characterized by a loss function. The preset convergence condition may also be that the number of parameter adjustments reaches a preset threshold, for example, a sample data pair is randomly selected from multiple sample data pairs each time to perform the above-mentioned model parameter adjustment process, and the convergence condition is determined to be satisfied when the number of samplings reaches the preset threshold. It is understandable that the convergence condition can also be flexibly designed as other conditions according to the actual application scenario, and this application is not limited to this.

[0053] It can be seen that in the above method, the accident information of a large number of accident samples can be used to input the prediction model, and the above process of adjusting the preset model parameters can be repeated. If the preset convergence conditions are met, it can be considered that the trained preset model has a certain degree of ability to predict the transportation risk radius, and the preset model that meets the preset convergence conditions is used as the risk radius prediction model.

[0054] Optionally, the accident information of the accident sample may also include accident vehicle information and the road width and road material of the accident section, etc. The input during model training can be adjusted according to actual needs to improve the accuracy of the risk radius prediction model.

[0055] By adopting the method of the above embodiment, historical accident data is taken as samples to obtain accident information (accident vehicle information and accident road section information) in each accident in the sample and the actual accident impact radius of each accident. By using the accident impact radius as the annotation information corresponding to its accident information, after input training with a large number of samples, a risk radius prediction model can be obtained that only needs to obtain road section information and vehicle information to predict the transportation risk radius of a vehicle traveling on a road section.

[0056] As Figure 1 The application of the risk radius prediction model trained in the embodiment, the embodiment of the present disclosure also provides a method for determining a transportation path, such as Figure 2 As shown, the method includes:

[0057] Step S210, using a risk radius prediction model to determine the transport risk radius of the target vehicle on each of the multiple road sections, wherein the multiple road sections are all road sections within the target area, and the risk radius prediction model is based on Figure 1 The medium risk radius prediction model is trained using the training method;

[0058] Step S220, determining the local risk value of the target vehicle on each road section according to the transportation risk radius of the target vehicle on each road section;

[0059] Step S230, determining a transport path of the target vehicle within the target area according to the local risk value, wherein the local risk value of the section with the largest local risk value in the transport path is less than a risk threshold.

[0060] It can be understood that in step S210, the risk radius prediction model can be used to estimate the transport risk radius of the target vehicle on each road section, which is specifically determined according to the training process of the risk radius prediction model. For example, the input information of the risk radius prediction model is limited to the road section information (traffic volume information and average vehicle speed on the road section, etc.). In this case, it is only necessary to determine the road section information of each road section in the target area based on the traffic survey data to predict the transport risk radius of the target vehicle on each road section.

[0061] Similarly, if the input information of the risk radius prediction model is limited to road section information and vehicle information (such as vehicle size, vehicle mass, etc.), it is necessary to combine the traffic survey data and the target vehicle’s own parameters as input information to predict the transportation risk radius of the target vehicle on each road section.

[0062] It is understandable that government management departments or freight carriers need to consider the impact of the transportation of goods on population centers within the transportation area (including but not limited to residential areas, school areas, industrial parks, etc.), and keep transportation vehicles as far away from population centers as possible to reduce the number of people affected by the transportation process. In step S220, the local risk value of the target vehicle in each section of the road is determined. It can be based on the predicted transportation risk radius, and the number of population centers within the transportation risk radius and the number of people in the population centers are determined to determine the local risk value of the target in different sections of the road. The specific method of determining the local risk value of the section according to the transportation risk radius and the population center situation can be calculated according to existing risk assessment methods, and the present application is not limited to this.

[0063] In one embodiment, the target vehicle may have more than one transportation path in the target area that is determined based on the local risk value and the risk threshold in step S230. The government management department or the cargo carrier may select the path with the lowest transportation cost (e.g., the shortest time) from multiple transportation paths as the actual transportation path according to the needs, or may determine the transportation path with the lowest maximum local risk value from multiple transportation paths as the actual transportation path, wherein the maximum local risk value refers to the local risk value of the section with the largest local risk value in the path. Alternatively, a more balanced transportation path may be selected as the actual transportation path by combining transportation cost and risk control.

[0064] By adopting the method of the above embodiment, the risk radius prediction model can be used to quickly predict the transportation risk radius of the target vehicle on each road section in the target area, and then the transportation risk of each road section can be quickly determined, providing a reference for the selection of the transportation route of the target vehicle, greatly improving the efficiency of decision-making for the selection of safe transportation routes, and ensuring the transportation safety of large-scale transport vehicles on roads.

[0065] The above describes the specific settings and implementation methods of the embodiments of the present application from different perspectives. Using the method provided in the above embodiments, a transportation risk radius prediction model can be trained based on historical accident data of the highway, so that the trained model can predict the transportation risk radius of the vehicle in the road section based on vehicle information and the intersection data of the road section, thereby helping to quickly and efficiently determine a safe transportation path for the vehicle.

[0066] As the implementation of the above methods, Figure 3 As shown, the embodiment of the present disclosure also provides a training device for a risk radius prediction model, which may include:

[0067] The accident information determination module 310 is used to determine the accident information of each accident sample in the historical traffic accident sample set, wherein the accident information at least includes the accident road section information;

[0068] The annotation information determination module 320 is used to determine the accident impact radius of each accident sample, and use the accident impact radius as the annotation information corresponding to the accident information in the accident sample;

[0069] The training module 330 is used to train a risk radius prediction model based on the accident information of each accident sample and the corresponding annotation information, wherein the risk radius prediction model is used to predict the transportation risk radius of the transport vehicle on each section of the highway.

[0070] Exemplarily, the accident information determination module 310 is used to:

[0071] According to the traffic survey data, the traffic volume information and the average vehicle speed of the road section where the accident occurred in each accident sample are obtained; the traffic volume information includes the number of vehicles in the road section and the proportion of different types of vehicles; the traffic volume information and the average vehicle speed of the road section are used as the accident section information.

[0072] Exemplarily, the accident information also includes accident vehicle information, and the accident information determination module 310 is further used to determine the size of the accident vehicle in each accident sample as the accident vehicle information, wherein the size of the accident vehicle includes at least the length, width and height of the accident vehicle.

[0073] Exemplarily, the annotation information determination module 320 is used to:

[0074] The accident impact radius of each accident sample is determined based on the accident impact area recorded in each accident sample; or, based on the accident information of each accident sample, the accident impact radius of each accident sample is determined using an expert investigation method.

[0075] Exemplarily, the training module 330 is used to:

[0076] The accident information of the i-th accident sample in each accident sample is input into the preset model to obtain the transportation risk radius output by the preset model; wherein i is an integer greater than or equal to 1;

[0077] Adjust the parameters of the preset model according to the transport risk radius output by the preset model and the annotation information corresponding to the accident information of the i-th road section;

[0078] When the preset convergence conditions are met, the preset model is used as the risk radius prediction model.

[0079] like Figure 4As shown, the embodiment of the present disclosure also provides a device for determining a transportation path, which may include:

[0080] A risk radius determination module 410 is used to determine the transportation risk radius of the target vehicle on each of the multiple road sections using a risk radius prediction model, wherein the multiple road sections are all road sections within the target area, and the risk radius prediction model is obtained by training based on the risk radius prediction model training method in the above embodiment;

[0081] The risk value determination module 420 is used to determine the local risk value of the target vehicle on each road section according to the transportation risk radius of the target vehicle on each road section;

[0082] The path determination module 430 is used to determine the transportation path of the target vehicle in the target area according to the local risk value, wherein the local risk value of the section with the largest local risk value in the transportation path is less than the risk threshold.

[0083] The functions of each unit, module or sub-module in each device of the embodiments of the present disclosure can refer to the corresponding description in the above method embodiments, and have corresponding beneficial effects, which will not be repeated here.

[0084] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0085] Figure 5 FIG. 1 is a block diagram showing a structure of an electronic device according to an embodiment of the present application. Figure 5 As shown, the electronic device includes: a memory 510 and a processor 520, and the memory 510 stores instructions that can be run on the processor 520. When the processor 520 executes the instruction, the method in the above embodiment is implemented. The number of the memory 510 and the processor 520 can be one or more. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices and other similar computing devices. The components shown herein, their connections and relationships, and their functions are only examples, and are not intended to limit the implementation of the present application described and / or required herein.

[0086] The electronic device may also include a communication interface 530 for communicating with external devices and performing data exchange transmission. The various devices are interconnected using different buses and can be installed on a common motherboard or in other ways as needed. The processor 520 can process instructions executed in the electronic device, including instructions stored in or on the memory to display graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In other embodiments, if necessary, multiple processors and / or multiple buses can be used with multiple memories and multiple memories. Similarly, multiple electronic devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 5 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0087] Optionally, in a specific implementation, if the memory 510, the processor 520 and the communication interface 530 are integrated on a chip, the memory 510, the processor 520 and the communication interface 530 can communicate with each other through an internal interface.

[0088] It should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. It is worth noting that the processor may be a processor supporting the Advanced RISC Machines (ARM) architecture.

[0089] The embodiment of the present application provides a computer-readable storage medium (such as the memory 510 described above), which stores computer instructions. When the program is executed by a processor, the method provided in the embodiment of the present application is implemented.

[0090] Optionally, the memory 510 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and applications required for at least one function; the data storage area may store data created based on the use of the electronic device for identifying lane edges, etc. In addition, the memory 510 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory 510 may optionally include a memory generated remotely relative to the processor 520, and these remote memories may be connected to the electronic device for identifying lane edges via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0091] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other physical types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic disk storage or other magnetic storage media or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include non-transitory media such as modulated data signals and carrier waves.

[0092] This embodiment also provides a vehicle, including a controller, which can be used to execute the method of this embodiment, or the controller can include any device of this embodiment, or the controller can be any electronic device of this embodiment.

[0093] Exemplarily, the processor in the controller or electronic device may include at least one of an autonomous driving domain control module, a vehicle body domain control module, and an audio and video entertainment domain control module.

[0094] For example, the vehicle in this embodiment can be a fuel vehicle, an electric vehicle, a solar vehicle, or any other powered vehicle. For example, the vehicle in this embodiment can be an autonomous driving vehicle.

[0095] Other components of the vehicle of this embodiment, such as the specific structure of the frame and wheels and the connecting and fastening components, can adopt various technical solutions known to ordinary technicians in this field now and in the future, and will not be described in detail here.

[0096] In the description of this specification, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0097] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A training method for a risk radius prediction model, comprising: Determine the accident information of each accident sample in the historical traffic accident sample set, wherein the accident information includes accident road section information and accident vehicle information; Determine the accident impact radius of each accident sample, and use the accident impact radius as labeling information corresponding to the accident information in the accident sample; According to the accident information of each accident sample and the corresponding annotation information, a risk radius prediction model is trained, wherein the risk radius prediction model is used to predict the transportation risk radius of the transport vehicle on each section of the highway, so as to determine the safe transportation path of the transport vehicle on the highway according to the transportation risk radius; Among them, the accident information of each accident sample in the historical traffic accident sample set is determined, including: According to the traffic survey data, the traffic volume information and the average vehicle speed of the road section where the accident occurred in each accident sample are obtained; wherein the traffic volume information includes the number of vehicles in the road section and the proportion of different types of vehicles; Using the traffic volume information and the average vehicle speed of the road section as the accident road section information; Determining the size of the accident vehicle in each accident sample as the accident vehicle information, wherein the size of the accident vehicle at least includes the length, width and height of the accident vehicle; Determining the accident impact radius of each accident sample includes: Determine the accident impact radius of each accident sample according to the accident impact area recorded in each accident sample; or, According to the accident information of each accident sample, the accident impact radius of each accident sample is determined by utilizing the expert investigation method.

2. The method according to claim 1, wherein: According to the accident information of each accident sample and the corresponding annotation information, a risk radius prediction model is trained, including: Inputting the accident information of the i-th accident sample among the accident samples into the preset model, and obtaining the transportation risk radius output by the preset model; wherein i is an integer greater than or equal to 1; Adjusting the parameters of the preset model according to the transport risk radius output by the preset model and the annotation information corresponding to the accident information of the i-th road section; When the preset convergence conditions are met, the preset model is used as the risk radius prediction model.

3. A method for determining a transportation route, comprising: Determine the transport risk radius of the target vehicle on each of a plurality of road sections using a risk radius prediction model, wherein the plurality of road sections are all road sections within a target area, and the risk radius prediction model is obtained by training based on the method of claim 1 or 2; Determining the local risk value of the target vehicle at each road section according to the transportation risk radius of the target vehicle at each road section; A transport path of the target vehicle in a target area is determined according to the local risk value, wherein the local risk value of a section with the largest local risk value in the transport path is less than a risk threshold.

4. A training device for a risk radius prediction model, comprising: An accident information determination module, used to determine the accident information of each accident sample in the historical traffic accident sample set, wherein the accident information includes accident road section information and accident vehicle information; A labeling information determination module, used to determine the accident impact radius of each accident sample, and use the accident impact radius as labeling information corresponding to the accident information in the accident sample; A training module, used for training a risk radius prediction model according to the accident information of each accident sample and the corresponding annotation information, wherein the risk radius prediction model is used to predict the transportation risk radius of the transport vehicle on each section of the highway, so as to determine the safe transportation path of the transport vehicle on the highway according to the transportation risk radius; Wherein, the accident information determination module is also used for: According to the traffic survey data, the traffic volume information and the average vehicle speed of the road section where the accident occurred in each accident sample are obtained; wherein the traffic volume information includes the number of vehicles in the road section and the proportion of different types of vehicles; Using the traffic volume information and the average vehicle speed of the road section as the accident road section information; Determining the size of the accident vehicle in each accident sample as the accident vehicle information, wherein the size of the accident vehicle at least includes the length, width and height of the accident vehicle; The annotation information determination module is further used to: determine the accident impact radius of each accident sample according to the accident impact area recorded in each accident sample; or, According to the accident information of each accident sample, the accident impact radius of each accident sample is determined by utilizing the expert investigation method.

5. A device for determining a transport path, comprising: A risk radius determination module, used to determine the transport risk radius of the target vehicle on each of a plurality of road sections using a risk radius prediction model, wherein the plurality of road sections are all road sections within a target area, and the risk radius prediction model is obtained by training based on the method of claim 1 or 2; A risk value determination module, used to determine the local risk value of the target vehicle at each road section according to the transportation risk radius of the target vehicle at each road section; The path determination module is used to determine the transportation path of the target vehicle in the target area according to the local risk value, wherein the local risk value of the section with the largest local risk value in the transportation path is less than a risk threshold.

6. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 3.

7. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-3.

Images