Traffic accident risk management system and method based on road remaining capacity

Abstract

The present invention discloses a traffic accident risk management system and method based on the residual traffic capacity of a road, which relates to the technical field of traffic data analysis. The method comprises the following steps: collecting vehicle flow records and classifying them according to accident characteristics; selecting a flow record corresponding to an accident characteristic as a target flow record, setting a sampling section, obtaining vehicle types and flow rates, extracting characteristic values ​​of the target vehicle flow records and forming a characteristic sequence; drawing a two-dimensional scatter plot of the characteristic sequence in a plane coordinate system, and drawing a traffic capacity curve through function fitting; obtaining multiple traffic capacity curves, evaluating the goodness of fit and complexity, and determining the optimal fitting curve; taking the optimal fitting curve as the target curve of the accident characteristic, taking the flow rate at its extreme point as the residual traffic capacity of the sampling section, and calculating the residual capacity coefficient; finally, obtaining the current accident characteristics, matching the residual capacity coefficient, establishing a grading model in combination with actual vehicle traffic records, and quantitatively evaluating the impact of the accident.

Description

Technical Field

[0001] The present invention relates to the technical field of traffic data analysis, and in particular to a traffic accident risk management system and method based on road residual capacity. Background Art

[0002] The remaining traffic capacity of an accident scenario road refers to the maximum number of vehicles that can effectively pass through the remaining available lanes (including temporary lanes) within a unit time when some lanes of an expressway or road are closed or traffic conditions are restricted due to an accident. It is affected by a combination of factors such as the scope of the accident, temporary traffic organization measures, speed limits and driver behavior.

[0003] Existing methods for classifying road traffic accidents primarily base their classification on factors such as the number of lanes affected, the time of the accident, and the type and nature of the accident. For example, road traffic accidents are categorized into four types: minor, general, major, and extreme, based on the severity of casualties or the amount of property damage. Some regions also categorize traffic accidents based on the type and nature of the accident, such as rear-end collisions, fixed object collisions, skidding, and rollovers.

[0004] None of the aforementioned accident classification methods consider the actual traffic conditions at the time of the accident, lack a foresight into the scope and extent of the accident's impact, and therefore cannot directly guide road operators to take prompt measures to ensure smooth and safe traffic. In reality, traffic conditions are constantly changing. Failure to respond promptly and quickly to accidents can lead to significant secondary accidents, widespread congestion, and other traffic problems. Summary of the Invention

[0005] The purpose of the present invention is to provide a traffic accident risk management system and method based on road residual capacity to solve the problems raised in the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a traffic accident risk management method based on road remaining capacity, the method comprising:

[0007] Step S100: When a traffic accident occurs on a section of the highway, the flow records of vehicles on the section are collected, and the accident characteristics of the traffic accident are obtained to classify the vehicle flow records;

[0008] Step S200: Record the traffic flow records of vehicles corresponding to a certain accident characteristic as target traffic flow records, set a sampling section, obtain the types of vehicles passing through the sampling section and the traffic flow of each type of vehicle, extract the characteristic values ​​of the target traffic flow records, and aggregate the characteristic values ​​into a characteristic sequence;

[0009] Step S300: drawing a two-dimensional scatter plot of the characteristic sequence in a plane coordinate system, performing function fitting on the two-dimensional scatter plot, and drawing a capacity curve;

[0010] Step S400: obtaining a plurality of capacity curves, performing goodness of fit and complexity evaluation on the capacity curves, and obtaining the optimal fitting curve of the plurality of capacity curves;

[0011] Step S500: taking the optimal fitting curve as the target curve of a certain accident characteristic, taking the flow rate corresponding to the extreme point of the target curve as the remaining capacity of the sampling section, and calculating the remaining capacity coefficient of the remaining capacity;

[0012] Step S600: Obtain the accident characteristics of the current traffic accident, match the remaining traffic capacity coefficient, collect actual vehicle traffic records to establish a classification model, and conduct a quantitative assessment of the impact caused by the traffic accident.

[0013] Furthermore, step S100 includes:

[0014] The classification characteristics of traffic accidents include accident information and road section information;

[0015] Accident information includes the number of lanes occupied by the accident, accident location, accident type, and impact range;

[0016] The section information includes the composition of the cross section and the linear indicators of the section;

[0017] Each accident feature includes at least one classification feature, and the traffic records corresponding to each accident feature are classified and aggregated.

[0018] Furthermore, step S200 includes:

[0019] Step S201: Obtain the location of the traffic accident, and from the location, obtain the nearest interchange entrance and exit in the road section in the opposite direction of the road traffic direction. The interchange entrance and exit may include a hub-type interchange entrance and exit or a service-type interchange entrance and exit;

[0020] Step S202: Setting a distance threshold d, recording a position that is d away from the position in the opposite direction of the road as the associated position of the traffic accident, and the sampling section is located between the interchange entrance and exit and the associated position;

[0021] Step S203: Record the sampled section as the upstream section of the traffic accident, obtain the record of vehicles passing through the upstream section per unit time, count the number of vehicles by vehicle type, and obtain the ratio of the number of vehicles of each type to the total number of vehicles passing through the upstream section per unit time;

[0022] Step S204: Calculate the correction coefficient f of traffic composition to traffic capacity per unit timeCB , , where P x The PCE represents the ratio of the number of vehicles of type x to the total number of vehicles passing through the upstream section. x represents the vehicle conversion coefficient for the xth vehicle type, and k represents the total number of vehicle types;

[0023] Step S205: Obtain the vehicle record of the t-th unit time period, record the total number of vehicles passing the upstream section in the t-th unit time period as Qt, and obtain the correction coefficient f of the t-th unit time period. t CB , calculate the i-th traffic volume Q i , , where τ represents the constant term coefficient, and T represents the length of the t-th unit time period;

[0024] Step S206: Obtain the total number of vehicles passing the location of the traffic accident in the t-th unit time period, and record it as q i , Q i and q i Form a data set R (Q i ,q i ), collect the data sets of w time periods to obtain the feature sequence.

[0025] Furthermore, step S300 includes:

[0026] Step S301: Setting a first coordinate axis and a second coordinate axis to establish a plane coordinate system, wherein the first coordinate axis represents: the traffic volume of the upstream section of the accident point per unit time, and the second coordinate axis represents the number of traffic flows passing the location of the traffic accident per unit time, obtaining all data groups of the feature sequence, mapping the values ​​in the data groups to coordinates on the plane coordinate system, and plotting the two-dimensional scatter points corresponding to all data groups in the plane coordinate system;

[0027] Step S302: Perform function fitting on the two-dimensional scattered points using a polynomial fitting function. Each fitting is performed to obtain a fitting function. After h fittings, the obtained h fitting functions are collected into a reference function set.

[0028] Furthermore, step S300 includes:

[0029] Step 3-1: Set the data group number threshold α. When w < α, collect the real value of vehicle traffic records. Use the VISSIM simulation model to obtain the simulated traffic volume of the upstream section per unit time and the simulated number of vehicles passing the accident location per unit time.

[0030] In order to ensure the accuracy of the fitting function, at least 15 sets of discrete points are required as reference data for the fitting function. When the number of discrete points is too small, it is necessary to supplement it with simulated data.

[0031] Step 3-2: Group the simulated traffic volume and the simulated number of vehicles into a simulated data group, and aggregate the simulated data group into a feature sequence.

[0032] Furthermore, step S400 includes:

[0033] Step S401: Calculate the Akaike Information Criterion value of each fitting function in the reference function set, where the Akaike Information Criterion value of the j-th fitting function is AIC j , , where p j Indicates the number of parameters of the fitting model when fitting the j-th fitting function, n j Indicates the number of observations of the fitting model when fitting the j-th fitting function, RSS j Represents the residual sum of squares when fitting the j-th fitting function;

[0034] In the formula for calculating the AIC value,

[0035] p represents the number of parameters of the model, that is, the power of the polynomial fitting function;

[0036] n represents the number of model observations, that is, the number of scattered samples;

[0037] RSS stands for residual sum of squares, which is the cumulative square difference between the predicted value of the fitting function and the true sample value;

[0038] For different polynomial fitting functions, the Akaike Information Criterion (AIC) value of each fitting function is calculated, and the model with the smallest value is selected to ensure the optimal balance between the degree of fit and complexity.

[0039] Step S402: Calculate the Akaike Information Criterion values ​​of all fitting functions in the reference function set, and take the fitting function corresponding to the minimum Akaike Information Criterion value as the optimal fitting function.

[0040] Furthermore, step S500 includes:

[0041] Step S501: Taking the traffic volume of the upstream section of the accident point per unit time as an independent variable, deriving the optimal fitting function to obtain the extreme value point of the optimal fitting function;

[0042] Step S502: Mark the coordinates of the extreme point as (Q ep ,q ep ), when the number of extreme points is 1, q epAs the residual capacity cv of the upstream section, when the number of extreme points exceeds 1, q ep The maximum value is taken as the residual capacity c of the upstream section v ;

[0043] Step S503: Calculate the residual capacity coefficient f for a certain accident characteristic r , f r =c v / c0, where c0 represents the basic traffic capacity coefficient of the road section;

[0044] Step S504: establishing a corresponding relationship between the accident characteristics, the remaining capacity and the remaining capacity coefficient, and collecting them into the road network traffic flow operation impact database.

[0045] Furthermore, step S600 includes:

[0046] Step S601: Record the upstream section of the current traffic accident as the target upstream section, and obtain the traffic volume Rt of the target upstream section per unit time;

[0047] Step S602: Obtain the accident characteristics of the current traffic accident and match the remaining traffic capacity cr in the road network traffic flow operation impact database;

[0048] Step S603: Calculate the supply-demand ratio SR, SR = Rt / cr, set three limit thresholds β1, β2 and β3, where β1 < β2 < β3, and determine the risk level of the traffic accident based on whether the supply-demand ratio is within the interval formed by the limit thresholds.

[0049] In order to better implement the above method, a traffic accident risk management system based on road remaining capacity is also proposed. The system includes: historical data management module, feature management module, fitting curve management module, remaining capacity assessment module and impact quantification module;

[0050] The historical data management module is used to store the vehicle flow records in the road section after the traffic accident occurs. The feature management module is used to extract the characteristic values ​​of the flow records. The fitting curve management module is used to draw the capacity curve and optimize the capacity curve to obtain the optimal fitting curve. The residual capacity assessment module is used to assess the residual capacity of the road. The impact quantification module is used to quantitatively assess the impact caused by the traffic accident.

[0051] Furthermore, the feature management module includes: a section management unit, a vehicle identification unit, a correction coefficient management unit, a traffic volume calculation unit, and a feature sequence management unit; wherein the section management unit is used to manage the upstream section corresponding to the traffic accident, the vehicle identification unit is used to identify the vehicle type and number in the vehicle record, the correction coefficient management unit is used to calculate the correction coefficient, the traffic volume calculation unit is used to calculate the traffic volume, and the feature sequence management unit is used to aggregate the data group to obtain the feature sequence;

[0052] Furthermore, the fitting curve management module includes: a numerical simulation unit, a function fitting unit, and a curve evaluation unit; wherein the numerical simulation unit is used to supplement the two-dimensional scatter plot by simulating data groups when the number of data groups is less than a threshold value; the function fitting unit is used to obtain the two-dimensional scatter plot and obtain the capacity curve by function fitting; and the curve evaluation unit is used to evaluate the capacity curve and obtain the optimal fitting function;

[0053] Furthermore, the remaining capacity assessment module includes: a remaining capacity management unit, a remaining capacity coefficient management unit, and a database management unit; wherein the remaining capacity management unit is used to obtain the extreme value point of the optimal fitting function to obtain the remaining capacity, the remaining capacity coefficient management unit is used to calculate the remaining capacity coefficient, and the database management unit is used to manage the road network traffic flow operation impact database;

[0054] Furthermore, the impact quantification module includes: an accident monitoring unit, a feature matching unit and a risk assessment unit; wherein, the accident monitoring unit is used to obtain the accident characteristics of the current traffic accident, the feature matching unit is used to match the remaining traffic capacity corresponding to the current traffic accident, and the risk assessment unit is used to calculate the supply-demand ratio and judge the risk level of the traffic accident.

[0055] Compared with the prior art, the beneficial effects of the present invention are: the present invention comprehensively considers core factors such as road characteristics, accident occupancy, traffic level, and accident duration, and can more accurately reflect the actual impact of accidents on traffic conditions, providing road operators with a behavioral benchmark for rapid response and disposal. When an accident occurs, road operators can quickly take corresponding measures based on the grading method of the present invention to ensure the smooth flow and safety of roads, reduce the risk of secondary accidents and traffic problems such as the spread of congestion, and protect the lives and property of drivers and passengers. For example, when faced with accidents of different levels, operators can carry out targeted traffic diversion, deploy rescue resources, etc., to improve the efficiency of accident handling, avoid the rapid spread of congestion, and reduce safety hazards. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 Schematic diagram of the structure of the traffic accident risk management system based on the remaining road capacity of the present invention;

[0057] Figure 2 Schematic diagram of the flow of the traffic accident risk management method based on the remaining road capacity of the present invention;

[0058] Figure 3 Schematic diagram of simulation parameters of the traffic accident risk management method based on the remaining road capacity of the present invention;

[0059] Figure 4 A two-dimensional scatter plot diagram of the traffic accident risk management method based on the remaining road capacity of the present invention;

[0060] Figure 5 Schematic diagram of function fitting of the traffic accident risk management method based on the remaining road capacity of the present invention;

[0061] Figure 6 The figure is a flow chart of an embodiment of a traffic accident risk management method based on the remaining road capacity of the present invention. DETAILED DESCRIPTION

[0062] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0063] Example: Figures 1-6 As shown, the present invention provides a technical solution, a traffic accident risk management method based on the remaining road capacity.

[0064] Step S100: When a traffic accident occurs on a section of the highway, the flow records of vehicles on the section are collected, and the accident characteristics of the traffic accident are obtained to classify the vehicle flow records;

[0065] Wherein, step S100 includes:

[0066] The classification characteristics of traffic accidents include accident information and road section information;

[0067] Accident information includes the number of lanes occupied by the accident, accident location, accident type, and impact range;

[0068] The section information includes the composition of the cross section and the linear indicators of the section;

[0069] Each accident feature includes at least one classification feature, and the traffic records corresponding to each accident feature are classified and aggregated.

[0070] Step S200: Record the traffic flow records of vehicles corresponding to a certain accident characteristic as target traffic flow records, set a sampling section, obtain the types of vehicles passing through the sampling section and the traffic flow of each type of vehicle, extract the characteristic values ​​of the target traffic flow records, and aggregate the characteristic values ​​into a characteristic sequence;

[0071] Wherein, step S200 includes:

[0072] Step S201: Obtain the location of the traffic accident, and from the location, obtain the nearest interchange entrance and exit in the road section in the opposite direction of the road traffic direction. The interchange entrance and exit may include a hub-type interchange entrance and exit or a service-type interchange entrance and exit;

[0073] Highway interchanges are the junctions between highways and other roads. Their main function is to connect traffic between different regions and improve transportation efficiency. Hub interchanges allow vehicles to transfer between different highways.

[0074] Step S202: Setting a distance threshold d, recording a location d away from the location in the opposite direction of the road as the associated location of the traffic accident, and the sampling section is located between the interchange entrance and exit and the associated location;

[0075] Step S203: Record the sampled section as the upstream section of the traffic accident, obtain the record of vehicles passing through the upstream section per unit time, count the number of vehicles by vehicle type, and obtain the ratio of the number of vehicles of each type to the total number of vehicles passing through the upstream section per unit time;

[0076] Preferably, the unit time is 1 hour.

[0077] In the embodiment, video surveillance information on the highway is collected, and the Yolov-5 algorithm is used to analyze the number of vehicles in the accident point section and the upstream traffic flow of real accident samples;

[0078] Step S204: Calculate the correction coefficient f of traffic composition to traffic capacity per unit time CB , , where P x The PCE represents the ratio of the number of vehicles of type x to the total number of vehicles passing through the upstream section. x represents the vehicle conversion coefficient of the xth vehicle type, and k represents the total number of vehicle types;

[0079] In the embodiment, the types of vehicles on the highway are divided into: small cars, medium-sized cars, large cars and car trains. According to the provisions of the "Highway Engineering Technical Standard" (JTG B01-2014), the vehicle conversion coefficients of small cars, medium-sized cars, large cars and car trains are set to 1.0, 1.5, 2.5 and 4.0 respectively;

[0080] Step S205: Obtain the vehicle record of the t-th unit time period, record the total number of vehicles passing the upstream section in the t-th unit time period as Qt, and obtain the correction coefficient f of the t-th unit time period. t CB , calculate the i-th traffic volume Q i , , where τ represents the constant term coefficient, and T represents the length of the t-th unit time period;

[0081] When the unit time is 1 hour, τ is 60;

[0082] Step S206: Obtain the total number of vehicles passing the location of the traffic accident in the t-th unit time period, and record it as q i , Q i and q i Form a data set R (Q i ,q i ), collect the data sets of w time periods to obtain the feature sequence.

[0083] Step S300: drawing a two-dimensional scatter plot of the characteristic sequence in a plane coordinate system, performing function fitting on the two-dimensional scatter plot, and drawing a capacity curve;

[0084] Step S300 includes:

[0085] Step 3-1: Set the data group number threshold α. When w < α, collect the real value of vehicle traffic records. Use the VISSIM simulation model to obtain the simulated traffic volume of the upstream section per unit time and the simulated number of vehicles passing the accident location per unit time.

[0086] Step 3-2: Group the simulated traffic volume and the simulated number of vehicles into a simulated data group, and aggregate the simulated data group into a feature sequence;

[0087] When α=15, and w<15, data supplementation is performed by VISSIM simulation;

[0088] Obtain information about the highway section to be simulated, such as road alignment, design speed, cross-section characteristics, etc., and build a matching VISSIM micro-simulation model.

[0089] The simulation model is calibrated and fine-tuned using indicators such as the hourly traffic volume Q of the section upstream of the accident point, the hourly number of vehicles passing the accident point q, and the queue length to restore the real accident scene.

[0090] The vehicle input, vehicle model ratio, and accident location in the VISSIM simulation model are consistent with data from real accident videos. By adjusting parameters such as the maximum deceleration, safety distance reduction factor, and minimum headway in driving behavior in VISSIM, the error between the hourly number of vehicles passing the accident point output by the simulation model and the hourly number of vehicles passing the accident point counted in real video data is kept within 5%. At the same time, the error of other indicators that can be obtained from the video, such as queue length, is also kept within 5% as much as possible.

[0091] By adjusting parameters such as the maximum deceleration, safety distance reduction coefficient, and minimum headway, the error between the number of vehicles passing the accident point per hour and the queue length output by the simulation model and the actual value is within 5%, ensuring the credibility of the simulation model.

[0092] Take 500n (n=4, 5, 6...) as the vehicle input in the VISSIM simulation model input parameters Simulate step by step and output the number of vehicles passing through the accident point per hour , It is the "Vehicle Input" parameter item in VISSIM;

[0093] When 2000≤V i ≤9000, by setting up data collection points at the accident sites, collecting the hourly vehicle count corresponding to different vehicle inputs , thus obtaining the simulation series .

[0094] Step S301: Setting a first coordinate axis and a second coordinate axis to establish a plane coordinate system, wherein the first coordinate axis represents: the traffic volume of the upstream section of the accident point per unit time, and the second coordinate axis represents the number of traffic flows passing the location of the traffic accident per unit time, obtaining all data groups of the feature sequence, mapping the values ​​in the data groups to coordinates on the plane coordinate system, and plotting the two-dimensional scatter points corresponding to all data groups in the plane coordinate system;

[0095] Step S302: Perform function fitting on the two-dimensional scattered points using a polynomial fitting function. Each fitting is performed to obtain a fitting function. After h fittings, the obtained h fitting functions are collected into a reference function set.

[0096] In the embodiment, a fitting function is selected to fit the scattered data and a capacity curve is drawn;

[0097] According to the distribution characteristics of the scatter plot, a polynomial fitting function is selected to fit the data set. Fitting is performed and the sum of squares of errors is minimized by the least squares method to find the best matching function for the data.

[0098] The set of scattered series is:

[0099] ;

[0100] in, The relationship satisfies the function:

[0101] ;

[0102] Then the polynomial fitting function is:

[0103] ;in, ——polynomial coefficients;

[0104] Solve the polynomial coefficients through the matrix, that is:

[0105] , ;

[0106] Then the sum of squared errors of the entire point set can be expressed as:

[0107] ;

[0108] Taking the derivative with respect to W, we get:

[0109] ; Finally, the polynomial coefficient W is obtained, namely:

[0110] .

[0111] Step S400: obtaining a plurality of capacity curves, performing goodness of fit and complexity evaluation on the capacity curves, and obtaining the optimal fitting curve of the plurality of capacity curves;

[0112] Wherein, step S400 includes:

[0113] Step S401: Calculate the Akaike Information Criterion value of each fitting function in the reference function set, where the Akaike Information Criterion value of the j-th fitting function is AIC j , , where p j Indicates the number of parameters of the fitting model when fitting the j-th fitting function, n j Indicates the number of observations of the fitting model when fitting the j-th fitting function, RSS j Represents the residual sum of squares when fitting the j-th fitting function;

[0114] Step S402: Calculate the Akaike Information Criterion values ​​of all fitting functions in the reference function set, and take the fitting function corresponding to the minimum Akaike Information Criterion value as the optimal fitting function.

[0115] In the embodiment, there are now 15 groups (Q, q), the number of model observations n=15, and a cubic polynomial is selected, so p=3, and the residual sum of squares RSS of the fitting function is calculated to be 753483.49, then AIC=2*3-15*ln(753483.49 / 15)=174.37;

[0116] Then choose the quartic polynomial, p=4, RSS=415205.57, then AIC=2*4-15*ln(415205.57 / 15)=170.19; the AIC value of the quartic polynomial is lower.

[0117] Step S500: taking the optimal fitting curve as the target curve of a certain accident characteristic, taking the flow rate corresponding to the extreme point of the target curve as the remaining capacity of the sampling section, and calculating the remaining capacity coefficient of the remaining capacity;

[0118] Wherein, step S500 includes:

[0119] Step S501: Taking the traffic volume of the upstream section of the accident point per unit time as an independent variable, deriving the optimal fitting function to obtain the extreme value point of the optimal fitting function;

[0120] Step S502: Mark the coordinates of the extreme point as (Q ep ,q ep ), when the number of extreme points is 1, q ep As the residual capacity cv of the upstream section, when the number of extreme points exceeds 1, q ep The maximum value is taken as the residual capacity c of the upstream section v ;

[0121] Step S503: Calculate the residual capacity coefficient f for a certain accident characteristic r , f r =c v / c0, where c0 represents the basic traffic capacity coefficient of the road section;

[0122] Step S504: establishing a corresponding relationship between the accident characteristics, the remaining capacity and the remaining capacity coefficient, and collecting them into the road network traffic flow operation impact database.

[0123] Step S600: Obtain the accident characteristics of the current traffic accident, match the residual capacity coefficient, collect actual vehicle traffic records to establish a classification model, and conduct a quantitative assessment of the impact caused by the traffic accident;

[0124] Step S600 includes:

[0125] Step S601: Record the upstream section of the current traffic accident as the target upstream section, and obtain the traffic volume Rt of the target upstream section per unit time;

[0126] Step S602: Obtain the accident characteristics of the current traffic accident and match the remaining traffic capacity cr in the road network traffic flow operation impact database;

[0127] Step S603: Calculate the supply-demand ratio SR, SR = Rt / cr, set three limit thresholds β1, β2 and β3, where β1 < β2 < β3, and determine the risk level of the traffic accident by the supply-demand ratio being within the interval formed by the limit thresholds.

[0128] In the embodiment, β1 is in the range of [0.4, 0.6] and is related to the queue length and average vehicle speed. When the traffic situation shows intermittent queues of vehicles, regular fluctuations in driving speed, and vehicles following each other, for example, when the opportunity for drivers to change lanes autonomously is significantly reduced, the ratio of actual traffic flow to remaining traffic capacity is β1;

[0129] The range of β2 is [0.9, 1.1], at which point the actual traffic flow is close to the remaining capacity of the road.

[0130] β3 comprehensively considers the requirements on queue length and handling time in the "Traffic Blockage Information Reporting System", and follows the principle that "sudden blockage information that causes congestion of more than 2km on expressways (including toll stations and service areas), and queues longer than 500 meters and lasting for more than 30 minutes at mainline toll stations must be reported to the competent authorities". When the congestion in the blockage state is close to 2km, and the queue length at the mainline toll station is close to 500 meters and lasts close to 30 minutes, the ratio of actual traffic flow to remaining traffic capacity is β3.

[0131] When SR < β1, the actual traffic flow demand is far lower than the remaining road capacity, and the traffic accident has little impact on traffic flow. Drivers and passengers can basically pass the accident point smoothly as they wish. The driving speed is smooth and there is enough space to change lanes. There is almost no risk of safe driving. The handling plan is: issue an accident warning message;

[0132] When β1<SR<β2, the actual traffic flow demand is still lower than or close to the remaining traffic capacity of the road. The impact of traffic accidents on traffic flow increases. Drivers and passengers need to consider the interference of neighboring vehicles when passing the accident site. The driving speed at the accident site may be reduced, and there may be local parking queues at the accident site, but they can dissipate naturally. The handling plan is to issue accident warning information and take speed limit measures.

[0133] When β2 < SR < β3, the actual traffic demand exceeds the remaining road capacity, and the accident duration does not exceed the requirements of the "Traffic Blockage Information Reporting System" and other documents (generally 30 minutes). Queues form upstream of the accident site and congestion continues to spread, significantly increasing the risk of road driving safety. The response plan is to issue accident warning information, implement speed limit measures, and take traffic control measures.

[0134] When SR>β3, the actual traffic flow demand is far higher than the remaining traffic capacity of the road, or the accident duration exceeds the requirements of the "Traffic Blockage Information Reporting System". If the control strategy is not immediately implemented, the congestion upstream of the accident point will spread rapidly, and the queue length will exceed the requirements of the "Traffic Blockage Information Reporting System" (generally 2km). Accident warning information will be issued, speed limit measures will be taken, and flow control measures will be taken.

[0135] The system includes: historical data management module, feature management module, fitting curve management module, remaining capacity assessment module and impact quantification module;

[0136] The historical data management module is used to store the traffic flow records of vehicles in the road section after the traffic accident occurs;

[0137] The feature management module is used to extract feature values ​​of traffic records, wherein the feature management module includes: a section management unit, a vehicle identification unit, a correction coefficient management unit, a traffic volume calculation unit, and a feature sequence management unit; wherein the section management unit is used to manage the upstream section corresponding to the traffic accident, the vehicle identification unit is used to identify the vehicle type and number in the vehicle record, the correction coefficient management unit is used to calculate the correction coefficient, the traffic volume calculation unit is used to calculate the traffic volume, and the feature sequence management unit is used to aggregate the data group to obtain the feature sequence;

[0138] The fitting curve management module is used to draw the capacity curve, optimize the capacity curve, and obtain the optimal fitting curve. The fitting curve management module includes: a numerical simulation unit, a function fitting unit, and a curve evaluation unit. The numerical simulation unit is used to supplement the two-dimensional scatter plot by simulating data groups when the number of data groups is less than a threshold. The function fitting unit is used to obtain the two-dimensional scatter plot and obtain the capacity curve by function fitting. The curve evaluation unit is used to evaluate the capacity curve and obtain the optimal fitting function.

[0139] The remaining capacity assessment module is used to assess the remaining capacity of the road, wherein the remaining capacity assessment module includes: a remaining capacity management unit, a remaining capacity coefficient management unit, and a database management unit; wherein the remaining capacity management unit is used to obtain the extreme value point of the optimal fitting function to obtain the remaining capacity, the remaining capacity coefficient management unit is used to calculate the remaining capacity coefficient, and the database management unit is used to manage the road network traffic flow operation impact database;

[0140] The impact quantification module is used to quantitatively evaluate the impact caused by traffic accidents, wherein the impact quantification module includes: an accident monitoring unit, a feature matching unit and a risk assessment unit; wherein the accident monitoring unit is used to obtain the accident characteristics of the current traffic accident, the feature matching unit is used to match the remaining traffic capacity corresponding to the current traffic accident, and the risk assessment unit is used to calculate the supply-demand ratio and determine the risk level of the traffic accident.

[0141] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims, not the foregoing description, and all variations within the meaning and range of equivalents of the claims are intended to be included therein. Any reference sign in a claim should not be construed as limiting the claim to which it relates.

Claims

1. A traffic accident risk management method based on road remaining capacity, characterized in that: Methods include: Step S100: When a traffic accident occurs on a section of the highway, the flow records of vehicles on the section are collected, and the accident characteristics of the traffic accident are obtained to classify the vehicle flow records; Step S200: Record the traffic flow records of vehicles corresponding to a certain accident characteristic as target traffic flow records, set a sampling section, obtain the types of vehicles passing through the sampling section and the traffic flow of each type of vehicle, extract the characteristic values ​​of the target traffic flow records, and aggregate the characteristic values ​​into a characteristic sequence; Step S300: drawing a two-dimensional scatter plot of the characteristic sequence in a plane coordinate system, performing function fitting on the two-dimensional scatter plot, and drawing a capacity curve; Step S400: obtaining a plurality of capacity curves, performing goodness of fit and complexity evaluation on the capacity curves, and obtaining the optimal fitting curve of the plurality of capacity curves; Step S500: taking the optimal fitting curve as the target curve of a certain accident characteristic, taking the flow rate corresponding to the extreme point of the target curve as the remaining capacity of the sampling section, and calculating the remaining capacity coefficient of the remaining capacity; Step S600: Obtain the accident characteristics of the current traffic accident, match the remaining traffic capacity coefficient, collect actual vehicle traffic records to establish a classification model, and conduct a quantitative assessment of the impact caused by the traffic accident.

2. The traffic accident risk management method based on road remaining capacity according to claim 1 is characterized by: The traffic accident classification method in step S100 includes: The classification characteristics of traffic accidents include accident information and road section information; The accident information includes the number of lanes occupied by the accident, the accident location, the accident type and the affected area; The road section information includes the composition of the cross section and the linear index of the road section; Each accident feature includes at least one classification feature, and the traffic records corresponding to each accident feature are classified and aggregated.

3. The traffic accident risk management method based on road remaining capacity according to claim 2 is characterized by: Step S200 includes: Step S201: Obtaining the location of the traffic accident, and obtaining the nearest interchange to the location in the road section from the location in the direction opposite to the road traffic direction, wherein the interchange includes a hub-type interchange or a service-type interchange; Step S202: Setting a distance threshold d, recording a position that is d away from the position in the opposite direction of the road as an associated position of the traffic accident, and the sampling section is located between the interchange entrance and the associated position; Step S203: Recording the sampled section as the upstream section of the traffic accident, obtaining a record of vehicles passing through the upstream section per unit time, and counting the number of vehicles by vehicle type to obtain the ratio of the number of each type of vehicles to the total number of vehicles passing through the upstream section per unit time; Step S204: Calculate the correction coefficient f of traffic composition to traffic capacity per unit time CB , , where P x The PCE represents the ratio of the number of vehicles of type x to the total number of vehicles passing through the upstream section. x represents the vehicle conversion coefficient of the xth vehicle type, and k represents the total number of vehicle types; Step S205: Obtain the vehicle record of the t-th unit time period, record the total number of vehicles passing the upstream section in the t-th unit time period as Qt, and obtain the correction coefficient f of the t-th unit time period. t CB , calculate the i-th traffic volume Q i , , where τ represents the constant term coefficient, and T represents the length of the t-th unit time period; Step S206: Obtain the total number of vehicles passing the location of the traffic accident in the t-th unit time period, and record it as q i , Q i and q i Form a data set R (Q i ,q i ), collect the data sets of w time periods to obtain the feature sequence.

4. The traffic accident risk management method based on road remaining capacity according to claim 3 is characterized by: Step S300 includes: Step S301: Setting a first coordinate axis and a second coordinate axis to establish a plane coordinate system, wherein the first coordinate axis represents: the traffic volume of the upstream section of the accident point per unit time, and the second coordinate axis represents the number of traffic flows passing the location of the traffic accident per unit time, obtaining all data groups of the feature sequence, mapping the values ​​in the data groups to coordinates on the plane coordinate system, and plotting the two-dimensional scatter points corresponding to all data groups in the plane coordinate system; Step S302: Perform function fitting on the two-dimensional scattered points using a polynomial fitting function. Each fitting is performed to obtain a fitting function. After h fittings, the obtained h fitting functions are collected into a reference function set.

5. The traffic accident risk management method based on road remaining capacity according to claim 4 is characterized in that: Step S300 includes: Step 3-1: Set the data group number threshold α. When w < α, collect the real value of vehicle traffic records. Use the VISSIM simulation model to obtain the simulated traffic volume of the upstream section per unit time and the simulated number of vehicles passing the accident location per unit time. Step 3-2: Group the simulated traffic volume and the simulated number of vehicles into a simulated data group, and aggregate the simulated data group into a feature sequence.

6. The traffic accident risk management method based on road remaining capacity according to claim 4 is characterized by: Step S400 includes: Step S401: Calculate the Akaike Information Criterion value of each fitting function in the reference function set, where the Akaike Information Criterion value of the j-th fitting function is AIC j , , where p j Indicates the number of parameters of the fitting model when fitting the j-th fitting function, n j Indicates the number of observations of the fitting model when fitting the j-th fitting function, RSS j Represents the residual sum of squares when fitting the j-th fitting function; Step S402: Calculate the Akaike Information Criterion values ​​of all fitting functions in the reference function set, and take the fitting function corresponding to the minimum Akaike Information Criterion value as the optimal fitting function.

7. The traffic accident risk management method based on road remaining capacity according to claim 6 is characterized by: Step S500 includes: Step S501: Taking the traffic volume of the upstream section of the accident point per unit time as an independent variable, deriving the optimal fitting function to obtain the extreme value point of the optimal fitting function; Step S502: Mark the coordinates of the extreme point as (Q ep ,q ep ), when the number of extreme points is 1, q ep As the residual capacity cv of the upstream section, when the number of extreme points exceeds 1, q ep The maximum value is taken as the residual capacity c of the upstream section v ; Step S503: Calculate the residual capacity coefficient f for a certain accident characteristic r , f r =c v / c0, where c0 represents the basic traffic capacity coefficient of the road section; Step S504: establishing a corresponding relationship between the accident characteristics, the remaining capacity and the remaining capacity coefficient, and collecting them into the road network traffic flow operation impact database.

8. The traffic accident risk management method based on road remaining capacity according to claim 7 is characterized by: Step S600 includes: Step S601: Record the upstream section of the current traffic accident as the target upstream section, and obtain the traffic volume Rt of the target upstream section per unit time; Step S602: Obtain the accident characteristics of the current traffic accident and match the remaining traffic capacity cr in the road network traffic flow operation impact database; Step S603: Calculate the supply-demand ratio SR, SR = Rt / cr, set three limit thresholds β1, β2 and β3, where β1 < β2 < β3, and determine the risk level of the traffic accident by the supply-demand ratio being within the interval formed by the limit thresholds.

9. A traffic accident risk management system based on road remaining capacity, configured to implement the traffic accident risk management method based on road remaining capacity according to any one of claims 1 to 8, characterized in that: The system includes: Historical data management module, feature management module, fitting curve management module, remaining capacity assessment module and impact quantification module; Among them, the historical data management module is used to store the vehicle flow records in the road section after the traffic accident occurs, the feature management module is used to extract the characteristic values ​​of the flow records, the fitting curve management module is used to draw the capacity curve, optimize the capacity curve to obtain the optimal fitting curve, the remaining capacity evaluation module is used to evaluate the remaining capacity of the road, and the impact quantification module is used to quantitatively evaluate the impact caused by the traffic accident.

10. The traffic accident risk management system based on road remaining capacity according to claim 9, characterized in that: The feature management module includes: a section management unit, a vehicle identification unit, a correction coefficient management unit, a traffic volume calculation unit, and a feature sequence management unit; wherein the section management unit is used to manage the upstream section corresponding to the traffic accident, the vehicle identification unit is used to identify the vehicle type and number in the vehicle record, the correction coefficient management unit is used to calculate the correction coefficient, the traffic volume calculation unit is used to calculate the traffic volume, and the feature sequence management unit is used to aggregate the data group to obtain the feature sequence; The fitting curve management module includes: a numerical simulation unit, a function fitting unit, and a curve evaluation unit; wherein the numerical simulation unit is used to supplement the two-dimensional scatter plot by simulating data groups when the number of data groups is less than a threshold; the function fitting unit is used to obtain the two-dimensional scatter plot and obtain the capacity curve by function fitting; the curve evaluation unit is used to evaluate the capacity curve and obtain the optimal fitting function; The remaining capacity assessment module includes: a remaining capacity management unit, a remaining capacity coefficient management unit, and a database management unit; wherein the remaining capacity management unit is used to obtain the extreme value point of the optimal fitting function to obtain the remaining capacity, the remaining capacity coefficient management unit is used to calculate the remaining capacity coefficient, and the database management unit is used to manage the road network traffic flow operation impact database; The impact quantification module includes: an accident monitoring unit, a feature matching unit and a risk assessment unit; among them, the accident monitoring unit is used to obtain the accident characteristics of the current traffic accident, the feature matching unit is used to match the remaining traffic capacity corresponding to the current traffic accident, and the risk assessment unit is used to calculate the supply-demand ratio and judge the risk level of the traffic accident.

Images