Expressway bypass path shunting method and system in accident scene

Abstract

The invention discloses a highway bypass path shunting method and system in an accident scene, and the method comprises the steps: collecting a plurality of variables which affect the accident handling time, such as an accident attribute factor, a road attribute factor and an environment attribute factor, and constructing an accident duration prediction model through a random forest algorithm; constructing a microscopic simulation model under the accident condition by using the road data around the accident and the accident information; performing traffic flow dynamic loading in combination with historical OD information and the current road flow, estimating the maximum influence range of an accident, and obtaining the maximum queuing length of a sub-time window; according to the maximum queuing length under different time windows obtained through simulation, determining the starting and closing time of a path shunting strategy, carrying out accident influence range traceability analysis under different time windows, and determining shunting nodes based on a traceability analysis result; and carrying out dynamic traffic distribution based on the shunting nodes to obtain a final shunting scheme. The reasonable degree of the highway bypass path shunting scheme in the accident scene is improved, and the highway bypass path shunting scheme has high rationality.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent transportation technology, and in particular relates to a method and system for diverting detour routes on expressways in accident scenarios. Background Art

[0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art.

[0003] With the annual increase in car ownership, traffic pressure is increasing, and the incidence of road traffic accidents is also showing a rising trend. In this situation, if effective measures are not taken in a timely manner once a traffic accident occurs, it will often lead to increased traffic congestion and even secondary accidents, posing a huge challenge to urban traffic management and public safety.

[0004] After an accident, it's crucial to implement swift and effective diversion measures. By accurately assessing traffic conditions at the accident site and developing a scientifically sound diversion plan, vehicles that might have otherwise flooded the scene can be redirected to alternative roads, effectively alleviating traffic pressure at the accident site and its surrounding areas. Current methods for diverting traffic in accident scenarios suffer from a number of issues, including low accident duration prediction accuracy and poor assessment of the impact area, making them less effective. Summary of the Invention

[0005] In order to overcome the above-mentioned deficiencies of the prior art, the present invention provides a scientific and effective method and system for highway detour route diversion in accident scenarios, providing technical support for highway management to formulate reasonable diversion strategies.

[0006] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions:

[0007] In a first aspect, the present invention provides a method for diverting traffic on a highway detour in an accident scenario, comprising:

[0008] Obtain explanatory variables for the current accident, perform correlation analysis on the explanatory variables, and construct an input sample set; input the input sample set into a pre-trained accident duration prediction model to predict the current accident duration; compare the current accident duration with a diversion time threshold; if the current accident duration is less than the diversion time threshold, there is no need to divert vehicles within the accident impact area; otherwise, proceed to the subsequent steps;

[0009] Acquire the road data of the current accident to construct a microscopic simulation model of the accident section, and perform dynamic traffic flow loading on the microscopic simulation model of the accident section based on historical vehicle passing data to obtain the microscopic simulation model of the accident section after dynamic traffic flow loading;

[0010] Obtaining current accident report information, and reproducing the accident in a loaded accident section micro-simulation model based on the accident report information and the accident duration, to obtain the maximum impact range of the accident and the maximum queue length in each time window;

[0011] Based on the maximum impact range of the accident, it is determined whether to divert vehicles within the accident impact range. If diversion is to be performed, the start and close time of the path diversion strategy is determined based on the maximum queue length of the time window. A traceability analysis of the accident impact range under different time windows is carried out, and the diversion nodes are determined based on the traceability analysis results; dynamic traffic allocation is performed based on the diversion nodes to form a highway path diversion plan for accident scenarios.

[0012] According to a further technical solution, the explanatory variables include accident attribute factors, road attribute factors, environmental attribute factors, and traffic attribute factors.

[0013] According to a further technical solution, the accident duration prediction model is established using a random forest algorithm, and a grid search method is used to determine the model's hyperparameters for model training.

[0014] A further technical solution is to load dynamic traffic flow into the micro-simulation model of the accident section based on historical vehicle passing data. Specifically, according to the historical vehicle passing data, the historical OD conditions of vehicles around the accident section before the accident and during the accident period are obtained, and the historical OD is reduced according to the flow ratio based on the actual vehicle passing conditions at the accident scene, and loaded into the micro-simulation model to complete the dynamic traffic flow loading work.

[0015] A further technical solution is to obtain the maximum impact range of the accident and the maximum queue length in each time window as follows:

[0016] Based on the accident report information, the parking position, occupied lanes and number of accident vehicles of the accident vehicle are reproduced in the microscopic simulation model of the accident section;

[0017] According to the duration of the accident, the parking time is set to complete the reproduction of the accident in the microscopic simulation model of the accident section;

[0018] Run the micro-simulation model and set up a queue length detector upstream of the accident to obtain the maximum impact range of the accident;

[0019] Taking a fixed time as a time window, the maximum queue length of vehicles within the time window is output to obtain the maximum queue length of the divided time window.

[0020] A further technical solution is to determine the specific diversion nodes:

[0021] Determine the duration of the path diversion strategy based on the maximum queue length in each time window;

[0022] Analyze the road sections and traffic volume involved in the accident impact based on the duration of the path diversion strategy and the maximum queue length within the corresponding time;

[0023] Based on the current vehicle passing data, trace back all the sections in the section set, find the set number of source sets, proportions, and the length of the source sets from the sections before each section;

[0024] Use the more concentrated source nodes as diversion nodes.

[0025] Further technical solutions to form a highway route diversion plan for accident scenarios are as follows:

[0026] Based on the diversion nodes, analyze the previously set number of destination sets and their proportions for each diversion node through historical data;

[0027] Based on the source set and destination set, the OD set and its contribution rate to be routed are constructed. After sorting in descending order of contribution rate, the OD pairs are evenly divided into two levels.

[0028] The k-short circuits between two level OD pairs are calculated respectively, and traffic assignment is performed based on the user random equilibrium algorithm to obtain the highway path diversion plan under the accident scenario.

[0029] In a second aspect, the present invention provides a highway detour route diversion system in an accident scenario, comprising:

[0030] The time prediction module is configured to: obtain explanatory variables of the current accident, perform correlation analysis on the explanatory variables, and construct an input sample set; input the input sample set into a pre-trained accident duration prediction model to predict the current accident duration; compare the current accident duration with a diversion time threshold; if the current accident duration is less than the diversion time threshold, there is no need to divert vehicles within the accident impact area; otherwise, proceed to the subsequent steps;

[0031] An accident simulation module is configured to: obtain road data of a current accident to construct a microscopic simulation model of the accident section, and perform dynamic traffic flow loading on the microscopic simulation model of the accident section based on historical vehicle passing data to obtain the microscopic simulation model of the accident section after dynamic traffic flow loading;

[0032] An accident reproduction module is configured to: obtain current accident report information, reproduce the accident in a loaded accident section micro-simulation model based on the accident report information and the accident duration, and obtain the maximum impact range of the accident and the maximum queue length in each time window;

[0033] The solution generation module is configured to: determine whether to divert vehicles within the accident impact range based on the maximum impact range of the accident; if diversion is to be performed, determine the start and end time of the path diversion strategy based on the maximum queue length of the time window; conduct a source tracing analysis of the accident impact range under different time windows; determine the diversion node based on the source tracing analysis results; perform dynamic traffic allocation based on the diversion node to form a highway path diversion plan for accident scenarios.

[0034] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a method for diverting traffic through a detour path on a highway in an accident scenario as described in the first aspect.

[0035] In a fourth aspect, the present invention provides a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps of the method for diverting detour routes on a highway in an accident scenario as described in the first aspect are implemented.

[0036] One or more of the above technical solutions have the following beneficial effects:

[0037] The present invention collects multiple variables that affect accident handling time, such as accident attribute factors, road attribute factors, and environmental attribute factors, and uses a random forest algorithm to build an accident duration prediction model; uses road data and accident information around the accident to build a micro-simulation model under accident conditions; combines historical OD information with current road traffic flow to dynamically load traffic flow, estimate the maximum impact range of the accident and obtain the maximum queue length in each time window; determines the start and end time of the path diversion strategy based on the maximum queue length in different time windows obtained by simulation, conducts source tracing analysis of the accident impact range in different time windows, and determines the diversion nodes based on the source tracing analysis results; and performs dynamic traffic allocation based on the diversion nodes to obtain the final diversion plan.

[0038] The present invention improves the rationality of highway detour path diversion schemes in accident scenarios, comprehensively considers various factors affecting accident handling, makes full use of existing data, constructs multiple prediction models, and uses dynamic traffic flow distribution algorithms to determine diversion schemes, making highway detour path diversion schemes in accident scenarios highly rational.

[0039] This paper integrates multiple factors and uses a random forest algorithm to predict accident duration. Since there are many factors that affect accident duration, using a random forest algorithm to integrate multiple factors such as accident attributes, road attributes, and environmental attributes to comprehensively estimate accident duration is beneficial for achieving highly reasonable and accurate results.

[0040] The present invention builds an accident impact range prediction model based on micro-simulation. Micro-simulation can simulate the traffic flow operation state of vehicles in a real environment. By accurately reproducing the accident scene in micro-simulation, it can achieve an accurate prediction of the accident impact range.

[0041] The present invention generates a diversion strategy based on micro-simulation results. The starting conditions of the diversion strategy are obtained according to the micro-simulation results, and a traceability analysis is carried out to determine the diversion nodes. Finally, a dynamic traffic flow allocation algorithm is applied to obtain a diversion plan. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0043] Figure 1 The figure is a flow chart of a highway detour path diversion method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0044] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0045] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0046] In the absence of conflict, the embodiments of the present invention and the features thereof may be combined with each other.

[0047] Example 1

[0048] like Figure 1 As shown, this embodiment discloses a method for diverting traffic on a highway detour in an accident scenario, including:

[0049] S1: Obtain explanatory variables of the current accident, perform correlation analysis on the explanatory variables and construct an input sample set; input the input sample set into a pre-trained accident duration prediction model to predict the current accident duration.

[0050] S101: Collect and organize candidate explanatory variables, including accident attribute factors, road attribute factors, environmental attribute factors, and traffic attribute factors. Accident attribute factors include accident time, accident location, and lane occupancy; environmental attribute factors include weather conditions, temperature and humidity at the time of the accident; road attributes include the total number of lanes on the road where the accident occurred and whether it is located on the main line; and traffic attribute factors include traffic volume and passenger-to-freight ratio on the accident section.

[0051] S102: Perform correlation analysis on the candidate explanatory variables using the maximum information coefficient method, and construct the input sample set after removing the highly correlated variables:

[0052] {X=[x i1 ,x i2 ,...,x iM ],Υ=[y i1 ,y i2 ],i=1,2,...,N}

[0053] Among them, M and N represent the number of feature variables and samples respectively; X represents the feature variable, y i1 、y i2 Represents the time interval value and distance interval value of the i-th sample.

[0054] S103: Apply the random forest algorithm to establish an accident duration prediction model, and use the grid search method to determine the model's hyperparameters for model training.

[0055] S104: Use mean square error (MSE) and root mean square error (RMSE) as evaluation indicators to verify the prediction effect of the model. If any error is greater than the threshold, adjust the model hyperparameters and retrain. The error calculation formula is as follows:

[0056]

[0057] Among them, u is the estimated result of the accident duration model, is the actual value of the accident duration.

[0058] S105: Input the input sample set of the current accident into the trained accident duration prediction model, predict the current accident duration, compare the accident duration with the diversion time threshold, and if the accident duration prediction result is less than the diversion time threshold T min There is no need to divert vehicles within the accident impact area, otherwise proceed to the subsequent steps.

[0059] S2: Obtain the road data of the current accident to construct a microscopic simulation model of the accident section, and perform dynamic traffic flow loading on the microscopic simulation model of the accident section based on historical vehicle passing data to obtain the microscopic simulation model of the accident section after dynamic traffic flow loading.

[0060] S201: Road data includes road alignment, number of lanes, road connectivity, and speed limit signs. Specifically, based on the design data of the surrounding roads where the accident occurred, the road alignment data, number of lanes, and other information required for simulation are extracted and imported into micro-simulation software to create a basic road model. This basic model is then adjusted based on real-world road connectivity, and key road information such as speed limit signs is reproduced to create a micro-simulation model of the accident section.

[0061] S202: Based on historical traffic data, the historical OD of vehicles around the accident section before and during the accident is obtained. Based on the actual traffic at the accident site, this historical OD is reduced by the traffic volume ratio and loaded into the micro-simulation model to complete the dynamic traffic flow loading process. OD traffic volume refers to the number of trips between the starting and ending points. The micro-simulation model of the accident section after dynamic traffic flow loading is the constructed accident impact range prediction model.

[0062] S3: Obtain current accident report information, and reproduce the accident in the loaded accident section micro-simulation model according to the accident report information and the accident duration to obtain the maximum impact range of the accident and the maximum queue length in each time window.

[0063] S301: Obtain the current accident report information. Based on the accident report information, reproduce the parking position, occupied lanes, and number of accident vehicles in the loaded accident section micro-simulation model. Set the parking time based on the accident duration model prediction result in S1, i.e., the accident duration, to complete the reproduction of the accident in the micro-simulation model.

[0064] S302: Run the micro-simulation model, set a queue length detector upstream of the accident, and obtain the maximum queue length of the vehicles involved in the accident, that is, the maximum impact range of the accident.

[0065] S303: Taking a fixed time as a time window (eg, 5 minutes), output the maximum queue length of vehicles involved in the accident within the time window to obtain the maximum queue length of each time window.

[0066] S4: Based on the maximum impact range of the accident, determine whether to divert vehicles within the accident impact range. If diversion is to be performed, determine the start and close time of the path diversion strategy based on the maximum queue length of the time window, conduct a source tracing analysis of the accident impact range under different time windows, and determine the diversion nodes based on the source tracing analysis results; perform dynamic traffic allocation based on the diversion nodes to form a highway path diversion plan for accident scenarios.

[0067] S401: Determine whether to divert vehicles within the accident impact range based on the maximum impact range of the accident. Compare the maximum impact range of the accident with the diversion length threshold. If the maximum impact range of the accident is greater than the diversion length threshold L min Then proceed to the next step, otherwise no diversion is required.

[0068] S402: Based on the traffic operation status change trend under different time windows obtained by micro-simulation, determine the start and end time of the path diversion strategy, conduct source tracing analysis of the accident impact range under different time windows, and determine the diversion node based on the source tracing analysis results. Specifically, find the maximum queue length greater than the threshold L min For example, a single time period is used. The starting time window of the time period is the starting time t1 of the diversion strategy, and the ending time window is the ending time t2 of the diversion strategy. t=t2-t1, where t is the duration of the path diversion strategy. According to the duration of the path diversion strategy and the maximum queue length within the corresponding time, the set of road sections involved in the accident impact range R={r1,r2,…,r n} and flow N={n1,n2,…,n n Based on the current vehicle passing data, trace back all the sections in the section set and find the source set S of the set number (TOP20 in this embodiment) before each section. i , proportion B i The source set and the distance from the road segment are calculated. Tracing the source means obtaining the source of vehicles on a specific path (for example, which toll station the vehicle enters the highway from), the path, and the traffic volume, and using the more concentrated source nodes as diversion nodes.

[0069] Use the centralized traceability of all sources as a diversion node.

[0070] Source Set S i Expressed as: S i ={s i1 ,s i2 ,…,s i20}(i=1,2,3,…,n).

[0071] Proportion B i Expressed as: B i ={b i1 ,b i2 ,…,b i20}(i=1,2,3,…,n).

[0072] S403: Based on the diversion nodes, historical data is analyzed to determine the destination sets of a previously set number (in this example, the TOP 10) of each diversion node (source node, such as the XX toll station), as well as their proportions. Based on the source and destination sets, an OD set is constructed to be routed, along with its contribution to the road segment. After sorting in descending order of contribution, the OD pairs are evenly divided into two levels. O represents the source, and D represents the destination. One source and one destination constitute an OD pair, or OD, and multiple ODs constitute an OD set.

[0073] S404: Dynamically distribute traffic based on the determined diversion nodes to obtain a path diversion solution. Specifically, the length of each road segment within the accident impact area is first assigned to infinity. Based on the road network topology, the k-short circuits between each OD pair of level one are calculated, and traffic distribution is performed based on a user-randomized equilibrium algorithm. Then, the length of each road segment within the accident impact area is assigned to twice the actual length. Based on the road network topology, the k-short circuits between each OD pair of level two are calculated, and traffic distribution is performed based on a user-randomized equilibrium algorithm. Finally, a path diversion solution is obtained for the accident scenario.

[0074] The present invention performs traffic distribution through the k-short circuit method and the user random equilibrium algorithm. The k-short circuit problem refers to finding the k-shortest path from the source to the destination in the road network topology. When k=1, the k-short circuit problem is the shortest path problem, which is used to find the shortest path and alternative shortest path between ODs.

[0075] S405: The path diversion strategy for the time period t1 to t2 is completed, and S401 to S404 are repeated until the diversion plans for all time periods are generated.

[0076] Example 2

[0077] This embodiment discloses a highway detour route diversion system in an accident scenario, including:

[0078] The time prediction module is configured to: obtain explanatory variables of the current accident, perform correlation analysis on the explanatory variables, and construct an input sample set; input the input sample set into a pre-trained accident duration prediction model to predict the current accident duration; compare the current accident duration with a diversion time threshold; if the current accident duration is less than the diversion time threshold, there is no need to divert vehicles within the accident impact area; otherwise, proceed to the subsequent steps;

[0079] An accident simulation module is configured to: obtain road data of a current accident to construct a microscopic simulation model of the accident section, and perform dynamic traffic flow loading on the microscopic simulation model of the accident section based on historical vehicle passing data to obtain the microscopic simulation model of the accident section after dynamic traffic flow loading;

[0080] An accident reproduction module is configured to: obtain current accident report information, reproduce the accident in a loaded accident section micro-simulation model based on the accident report information and the accident duration, and obtain the maximum impact range of the accident and the maximum queue length in each time window;

[0081] The solution generation module is configured to: determine whether to divert vehicles within the accident impact range based on the maximum impact range of the accident; if diversion is to be performed, determine the start and end time of the path diversion strategy based on the maximum queue length of the time window; conduct a source tracing analysis of the accident impact range under different time windows; determine the diversion node based on the source tracing analysis results; perform dynamic traffic allocation based on the diversion node to form a highway path diversion plan for accident scenarios.

[0082] Example 3

[0083] The purpose of this embodiment is to provide a computing device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method of embodiment 1 when executing the program.

[0084] Example 4

[0085] The purpose of this embodiment is to provide a computer-readable storage medium, a computer-readable storage medium having a computer program stored thereon, which performs the steps of the method of embodiment 1 when executed by a processor.

[0086] The steps involved in the apparatuses of Examples 3 and 4 above correspond to those of Method Example 1. For detailed implementation, please refer to the relevant description of Example 1. The term "computer-readable storage medium" should be understood to mean a single medium or multiple media containing one or more instruction sets; it should also be understood to include any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and causing the processor to perform any of the methods of the present invention.

[0087] Those skilled in the art will appreciate that the modules or steps of the present invention described above can be implemented using a general-purpose computer device. Alternatively, they can be implemented using program code executable by a computing device, which can then be stored in a storage device and executed by the computing device. Alternatively, they can be fabricated into separate integrated circuit modules, or multiple modules or steps can be fabricated into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.

[0088] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

[0089] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without any creative work are still within the scope of protection of the present invention.

Claims

1. A method for diverting traffic on a highway detour route in an accident scenario, characterized in that: include: Obtaining explanatory variables of the current accident, and constructing an input sample set after performing correlation analysis on the explanatory variables; Inputting the input sample set into a pre-trained accident duration prediction model to predict the current accident duration; Compare the current accident duration with the diversion time threshold, if the current accident duration is less than the diversion time threshold, then there is no need to divert vehicles within the accident impact range, otherwise proceed to the subsequent steps; Acquire the road data of the current accident to build a microscopic simulation model of the accident section, and perform dynamic traffic flow loading on the microscopic simulation model of the accident section based on historical vehicle passing data to obtain the microscopic simulation model of the accident section after dynamic traffic flow loading; Obtaining current accident reporting information, reproducing the accident in the loaded accident section micro-simulation model according to the accident reporting information and the accident duration, and obtaining the maximum impact range of the accident and the maximum queue length in the time window; Based on the maximum impact range of the accident, determine whether to divert vehicles within the accident impact range. If diversion is performed, determine the start and close time of the path diversion strategy based on the maximum queue length of the divided time windows, conduct source tracing analysis of the accident impact range under different time windows, and determine the diversion node based on the source tracing analysis results; Dynamic traffic allocation is carried out based on diversion nodes to form a highway path diversion plan for accident scenarios.

2. A method for diverting traffic on a highway detour route in an accident scenario as claimed in claim 1, characterized in that: The explanatory variables include accident attribute factors, road attribute factors, environmental attribute factors, and traffic attribute factors.

3. The method for diverting traffic on a highway detour route in an accident scenario according to claim 1, characterized in that: The accident duration prediction model is established using a random forest algorithm, and a grid search method is used to determine the model's hyperparameters for model training.

4. The method for diverting traffic on a highway detour route in an accident scenario according to claim 1, characterized in that: The dynamic traffic flow loading of the accident section micro-simulation model based on the historical vehicle passing data is specifically as follows: according to the historical vehicle passing data, the historical OD of vehicles around the accident section before the accident and during the accident period is obtained, and according to the actual vehicle passing situation at the accident scene, the historical OD is reduced according to the flow ratio, and loaded into the micro-simulation model to complete the dynamic traffic flow loading work.

5. The method for diverting traffic on a highway detour route in an accident scenario according to claim 1, characterized in that: The maximum impact range of the accident and the maximum queue length in the time window are obtained as follows: According to the accident report information, the parking position, occupied lane and number of accident vehicles of the accident vehicle are reproduced in the microscopic simulation model of the accident road section; According to the duration of the accident, the parking time is set to complete the reproduction of the accident in the microscopic simulation model of the accident section; Run the micro-simulation model, set up a queue length detector upstream of the accident, and obtain the maximum impact range of the accident; Take a fixed time as a time window, output the maximum queue length of vehicles in the time window, and get the maximum queue length of the divided time windows.

6. The method for diverting traffic on a highway detour route in an accident scenario according to claim 1, characterized in that: Determine the specific diversion nodes: Determine the duration of the path diversion strategy based on the maximum queue length in the time window; Analyze the road sections and traffic volume involved in the accident impact range based on the duration of the path diversion strategy and the maximum queue length within the corresponding time; Based on the current vehicle passing data, all sections in the section set are traced back to find the set number of source sets, proportions and the length of the source sets from the sections before each section; Use the more concentrated source nodes as diversion nodes.

7. The method for diverting traffic on a highway detour route in an accident scenario as claimed in claim 1, characterized in that: The specific highway route diversion plan for accident scenarios is as follows: According to the diversion nodes, the previously set number of destination sets and their proportions of each diversion node are analyzed through historical data; Based on the source set and the destination set, the OD set and its contribution rate that need to be routed are constructed. After sorting in descending order of contribution rate, the OD pairs are evenly divided into two levels. The k short circuits between two levels of OD pairs are calculated respectively, and traffic allocation is performed based on the user random equilibrium algorithm to obtain the highway path diversion plan under the accident scenario.

8. A highway detour route diversion system in an accident scenario, characterized in that: include: A time prediction module is configured to: obtain explanatory variables of the current accident, and construct an input sample set after performing correlation analysis on the explanatory variables; Inputting the input sample set into a pre-trained accident duration prediction model to predict the current accident duration; Compare the current accident duration with the diversion time threshold, if the current accident duration is less than the diversion time threshold, then there is no need to divert vehicles within the accident impact range, otherwise proceed to the subsequent steps; The accident simulation module is configured to: obtain the road data of the current accident to construct a microscopic simulation model of the accident section, and perform dynamic traffic flow loading on the microscopic simulation model of the accident section based on historical vehicle passing data to obtain the microscopic simulation model of the accident section after dynamic traffic flow loading; The accident reproduction module is configured to: obtain current accident reporting information, reproduce the accident in the loaded accident section micro-simulation model according to the accident reporting information and the accident duration, and obtain the maximum impact range of the accident and the maximum queue length in the time window; A solution generation module is configured to: determine whether to divert vehicles within the accident impact range based on the maximum impact range of the accident, and if diversion is performed, determine the start and close time of the path diversion strategy based on the maximum queue length of the divided time windows, conduct source tracing analysis of the accident impact range under different time windows, and determine the diversion node based on the source tracing analysis results; Dynamic traffic allocation is carried out based on diversion nodes to form a highway path diversion plan for accident scenarios.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps in a highway detour path diversion method in an accident scenario are implemented as described in any one of claims 1-7.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps in the highway detour path diversion method in an accident scenario are implemented as described in any one of claims 1-7.

Images