A method for constructing an intelligent management and control plan for a highway with double emergency lanes

Abstract

The application discloses a kind of construction methods of expressway intelligent management and control preplan of double emergency lane, by determining the traffic management and control comprehensive condition index calculation formula of road to be managed and controlled, the data of road are collected in real time, traffic management and control comprehensive condition characteristics are constructed, production control preplan cluster, then by calculation and simulation verification will obtain target preplan and its preplan execution effect, execution condition are recorded into preplan library, so that the expressway of double emergency lane can quickly search and directly provide solution when coping with large flow traffic or accident in inner lane, emergency response speed is faster, and by index calculation and simulation simulation two angles respectively to a series of preplans in management and control preplan cluster are compared horizontally, avoid the error caused by unreasonable influence factor parameter value, and obtain the emergency target preplan with high traffic efficiency and low secondary accident risk, so that accident risk is fully reduced, and vehicle traffic efficiency is greatly improved.

Description

Technical Field

[0001] This invention specifically relates to a method for constructing an intelligent management and control plan for highways with dual emergency lanes. Background Technology

[0002] Highways serve as a link between urban nodes, and their traffic is saturated year-round, with a high proportion of trucks. After some highways are expanded, they are equipped with two emergency lanes in one direction. The combination of ultra-wide roads and ultra-high traffic volume can meet the dual challenges of emergency rescue work.

[0003] However, traditional highway management methods based on single emergency lanes are not suitable for the complex traffic environment along the extension, which may lead to slow emergency response, secondary accidents, and low traffic efficiency.

[0004] Therefore, it is necessary to invent a method for constructing an intelligent management and control plan for highways with dual emergency lanes to solve the above problems. Summary of the Invention

[0005] The purpose of this invention is to provide a method for constructing an intelligent management and control plan for highways with dual emergency lanes, in order to address the aforementioned shortcomings in the technology.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for constructing an intelligent management and control plan for highways with dual emergency lanes, comprising the following steps:

[0007] Step 1: Based on the historical control focus, obtain the weight of the plan indicators for the roads to be controlled, and determine the calculation formula for the comprehensive traffic control status indicator of the roads to be controlled;

[0008] Step 2: Collect road data in real time from multiple dimensions, and construct comprehensive traffic control status features based on comprehensive traffic control status indicators;

[0009] Step 3: Generate a cluster of traffic control plans based on relevant traffic control knowledge, and obtain the execution effect of the plans within the cluster of plans based on the comprehensive characteristics of traffic control.

[0010] Step 4: In the cluster of control plans and the corresponding implementation effects of the plans, simulation is introduced for comparison to obtain the target plan and verify whether the target plan is reasonable.

[0011] Step 5: Conduct reinforcement learning on reasonable target plans, dynamically adjust influencing factors, provide feedback on plan results, and update the comprehensive characteristics of traffic control.

[0012] The aforementioned method for constructing a smart traffic control plan for highways with dual emergency lanes involves, in step 1, obtaining the plan indicator weights for the roads to be controlled based on historical control priorities, and determining the calculation formula for the comprehensive traffic control status index of the roads to be controlled. The specific steps are as follows:

[0013] 1.1 By acquiring and analyzing historical traffic accident data, traffic flow data, and existing control measures, we extract control experience in typical scenarios, clarify the focus of control, and use this as a reference benchmark to construct the weights w1', w2', and w3' of the contingency plan indicators;

[0014] 1.2. In the dynamic traffic simulation platform, simulate traffic flow characteristics under different scenarios and compare them with each other, and compare them with the actual historical road traffic conditions and attributes of historical scenarios. Obtain the theoretical calculated values ​​w1”, w2”, and w3” of the index weights from the perspective of numerical calculation.

[0015] 1.3. By fine-tuning w1', w2', w3' and w1”, w2”, w3” through simulation, the comprehensive traffic control status indicators are confirmed, so as to construct the integrated plan indicator weights w1, w2, w3;

[0016] 1.4. Based on the weights w1, w2, and w3 of the contingency plan indicators, the calculation formula for the comprehensive traffic control status index J is determined as follows:

[0017] J = w1⋅E eff − w2⋅E risk – w3⋅E rt ;

[0018] Where w1 is E eff The weights;

[0019] w2 is E risk The weights;

[0020] w3 is E rt The weights are w1 + w2 + w3 = 1;

[0021] E eff As an indicator of traffic efficiency;

[0022] E risk As a safety risk indicator;

[0023] E rt This serves as an indicator for emergency response time.

[0024] In the aforementioned method for constructing a smart management and control plan for highways with dual emergency lanes, in step 1.4, the traffic efficiency index E... eff The calculation formula is as follows:

[0025] ;

[0026] Where C is the maximum design capacity, which is the design capacity of the road to be managed, reflecting the road's vehicle carrying capacity;

[0027] Given condition X, the traffic density on the road to be controlled at a certain time.

[0028] Given condition X, the traffic flow speed on the road to be controlled at a certain time.

[0029] X includes routine situations, traffic incidents, and the implementation of control measures, while J is used to compare the advantages and disadvantages of different combinations of control strategies.

[0030] Among them, safety risk indicator E risk The calculation formula is as follows:

[0031] ;

[0032] ;

[0033] ;

[0034] Among them, R acc Given condition X, the probability of an accident on the road segment to be controlled and during the time to be controlled.

[0035] This refers to the reference value for the probability of accidents on the road section to be controlled and the time to be controlled before the execution condition X.

[0036] Reference values ​​for the road curvature influencing factors of the road section to be controlled and the time to be controlled before implementing condition X;

[0037] Reference values ​​for slope impact factors for the road section to be controlled and the time to be controlled before implementing condition X;

[0038] Reference values ​​for the number of lanes influencing the road segment to be controlled and the time to be controlled before implementing condition X;

[0039] Reference values ​​for the visibility distance impact factor of the road section to be controlled and the time to be controlled before implementing condition X;

[0040] Reference values ​​for the obstacle coefficient influencing factors of the road section to be controlled and the time to be controlled before implementing condition X;

[0041] Before implementing condition X, the reference values ​​for traffic flow density of the road section to be controlled and the time to be controlled;

[0042] Reference values ​​for traffic flow speed on the road segment to be controlled and during the control period before condition X is implemented;

[0043] Among them, the emergency response time indicator E rt The calculation formula is as follows:

[0044] ;

[0045] in, This refers to the actual time taken for emergency response;

[0046] Distance between the rescue point and the accident site

[0047] For the time it takes for the rescue vehicle to travel;

[0048] for The minimum value;

[0049] For non-emergency rescue and control situations, w3 is 0.

[0050] The aforementioned method for constructing a smart traffic control plan for highways with dual emergency lanes involves, in step 2, collecting road data in real time from multiple dimensions and constructing comprehensive traffic control features based on comprehensive traffic control indicators. The specific steps are as follows:

[0051] 2.1 Obtain the curvature, elevation, and number of lanes of the road to be managed through road design documents or high-precision maps to obtain road attribute data;

[0052] 2.2. Real-time data acquisition of traffic flow, density, vehicle type, visibility, and obstacle coefficient of the road to be managed is obtained through actual road sensors to acquire real-time road data and historical multi-dimensional data;

[0053] 2.3. Based on road attribute data and historical multi-dimensional data, a comprehensive traffic control feature F is constructed, as follows:

[0054] F = [R c G, L, S, C, P st , α, β, γ, δ, η, θ, ζ];

[0055] Among them, R c This is the road curvature influencing factor, used to reflect the degree of road curvature;

[0056] G is the slope influencing factor, which reflects the steepness of the road's slope and affects the likelihood of accidents by influencing the difference in vehicle speed.

[0057] L is the lane number influence factor, which reflects the impact of the number of lanes on the probability of accidents and has an important influence on the dynamic characteristics of traffic flow.

[0058] S is the visibility distance impact factor, which measures the impact of visibility on road accidents;

[0059] P st The obstacle coefficient is a coefficient used to reflect the amount of obstacles on the road. The obstacle coefficient will increase when there are vehicles involved in unresolved traffic accidents on the road.

[0060] γ is R c The parameters are: δ is the parameter of G, η is the parameter of L, θ is the parameter of S, and ζ is the parameter of P. st The parameters are all greater than 0;

[0061] α is The parameters, β are The parameters.

[0062] The aforementioned method for constructing intelligent management and control plans for highways with dual emergency lanes involves, in step 3, generating a management and control plan cluster based on traffic management-related knowledge, and obtaining the plan execution effect from the management and control plan cluster based on the comprehensive characteristics of traffic management. The specific steps are as follows:

[0063] 3.1. Generate a cluster of control plans by using administrative, technical, and physical means to dynamically intervene in vehicles on the road. ;

[0064] 3.2. Based on the contingency plan For example, based on the comprehensive traffic control situation characteristic F, the corresponding contingency plan execution effect is obtained. .

[0065] The aforementioned method for constructing a smart control plan for highways with dual emergency lanes includes, in step 3.1, the methods for dynamically intervening in vehicles on the road, which include:

[0066] M1. Set the types of control and management plan clusters, specifically including speed limit, diversion, and lane closure;

[0067] Furthermore, control plans are generated for different types of control measures, including:

[0068] A. Only speed limit is applied, speed limit v limit Three speed limit schemes are proposed: 60, 70, and 80 km / h.

[0069] B. Only execute the traffic splitting ratio. Four diversion schemes with values ​​of 0.2, 0.4, 0.6, and 0.8 are proposed, and the influence on traffic flow density is achieved by affecting traffic volume.

[0070] at this time ∈ [0,1], Q = Q0 ⋅ (1 – );

[0071] Where Q represents the traffic after splitting;

[0072] Q0 represents the initial flow rate;

[0073] C. Only the emergency lane is open at this time. Satisfy the following formula:

[0074] ;

[0075] in, To determine the traffic density after opening i emergency lanes;

[0076] Original density;

[0077] The original number of lanes;

[0078] The number of lanes after opening i emergency lanes;

[0079] M2. For roads with two emergency lanes, there are four scenarios: emergency lane not open, emergency lane open on the left, emergency lane open on the right, and emergency lanes open on both sides. This results in a cluster of control plans. The details are as follows:

[0080] ;

[0081] ;

[0082] In step 3.2, according to the plan For example, based on the comprehensive traffic control situation characteristic F, the corresponding contingency plan execution effect is obtained. The specific steps are as follows:

[0083] 3.2.1 Calculate the initial comprehensive traffic control index using real-time road data and the comprehensive traffic control situation characteristic F. ;

[0084] 3.2.2. Contingency plans are directly derived from historical multi-dimensional data detection results. Under similar conditions (x) The actual effect of execution Changes;

[0085] Where x is a certain value;

[0086] Furthermore, when only speed limiting is applied, in most of the x cases:

[0087] ;

[0088] In rare cases:

[0089] ;

[0090] 3.2.3 Calculation Plan Under similar conditions (x) Multiple possible comprehensive traffic control indicators resulting from the implementation of the following measures ;

[0091] 3.2.4, Through and Receive control plan Representative comprehensive traffic control indicators The specific steps are as follows:

[0092] H1. Dataset Preparation: (Based on the contingency plan) Under similar conditions The results of multiple executions were organized into four sequences, which are as follows:

[0093] ;

[0094] ;

[0095] ;

[0096] Corresponding comprehensive traffic control indicators ;

[0097] H2. Calculate various statistical measures: Calculate the mean, median, mode, standard deviation, and interquartile range for each of the four sequences to characterize different statistical features;

[0098] H3. Compare the statistics and select representative values, as follows:

[0099] H31. Calculate the deviation between each statistic and its corresponding observations;

[0100] Among them, the statistics are mean, median, mode, standard deviation, and interquartile range;

[0101] The observed values ​​are the actual values ​​from the four sequences;

[0102] The deviation table includes: Mean Absolute Error (MAE), Root Mean Square Error (MSE), and Root Mean Square Deviation (RMSD).

[0103] H32. Select the statistic with the smallest bias as the representative value of each of the four sequences, denoted as H32. , , Subsequently, representative comprehensive traffic control indicators were calculated. ;

[0104] 3.2.5. Regarding the cluster of control and prevention plans The execution steps of all the contingency plans, from 3.2.2 to 3.2.4, yielded the execution effectiveness and representative comprehensive traffic control indicators for each of the ten contingency plans. The details are as follows:

[0105] .

[0106] The aforementioned method for constructing a smart control plan for highways with dual emergency lanes involves, in step 4, introducing simulation to compare the control plan cluster and its corresponding execution effects, deriving the target plan, and verifying its rationality. The specific steps are as follows:

[0107] 4.1 Constructing initial comprehensive traffic control indicators based on real-time road data and road attribute data. Execute the simulation control plan cluster The effectiveness of the simulated contingency plan was obtained by combining the results from the simulated detectors with the comprehensive traffic control indicators. Comprehensive indicators of simulated traffic control ;

[0108] 4.2 Comprehensive indicators of simulated traffic control Representative traffic control comprehensive status indicators Among them, the scheme with the maximum value is selected as the target contingency plan;

[0109] 4.3. Verify the consistency between the simulation execution effect of the target contingency plan and the historical execution effect of the corresponding contingency plan to check whether the target contingency plan has stability. After verification, store it in the contingency plan library.

[0110] 4.4 The reliability of the target plan is verified by simulation.

[0111] The aforementioned method for constructing a smart management and control plan for highways with dual emergency lanes involves, in step 4.3, verifying the consistency between the simulated execution effect of the target plan and the historical execution effect of the corresponding plan. This verifies the stability of the target plan. After verification, the plan is stored in the plan database. The specific steps are as follows:

[0112] 4.3.1 Obtain the simulation plan execution effect and historical traffic control comprehensive status indicators for different historical scenarios, specifically:

[0113] The target plan is set as follows: Select a control plan Subsequently, by analyzing historical multi-dimensional data, we identified contingency plans that had been executed with the same objectives. Different initial conditions y, obtain Corresponding historical contingency plan implementation effects ,as well as

[0114] Historical traffic conditions indicators ;

[0115] 4.3.2 In simulated fictional scenarios, obtain target plans for different fictional scenarios. The simulation plan's execution effectiveness and the comprehensive indicators of simulated traffic control are as follows:

[0116] The simulation simulates different initial scenarios to be controlled, obtains flow rate, density, and velocity through detectors, and then calculates the execution effect of the simulation plan. ,and

[0117] Simulated traffic control comprehensive status index ;

[0118] 4.3.3, Target contingency plan The effectiveness of historical contingency plans in different historical scenarios is combined with the effectiveness of simulated contingency plans in hypothetical scenarios to verify the target contingency plan. The stability under different conditions is as follows:

[0119] Use equivalence tests and distribution difference tests to... and Verification is required;

[0120] If stable, we will obtain the target plan, the effectiveness of the plan's implementation, and comprehensive indicators of traffic control. Stored in the contingency plan database;

[0121] If unstable, the obtained target plan, the effectiveness of the plan's implementation, and the overall traffic control indicators will be affected. Additional notes should be added to indicate that the condition for use is the initial comprehensive traffic control status index. The comprehensive characteristics of traffic control, F, are stored in the contingency plan database.

[0122] The aforementioned method for constructing a smart management and control plan for highways with dual emergency lanes involves verifying the reliability of the target plan through simulation in step 4.4. The specific steps are as follows:

[0123] 4.4.1 Based on the target plan For example, obtain the contingency plan The simulation execution effect;

[0124] Specifically: Obtaining the contingency plan The effectiveness of historical contingency plans under x similar conditions Corresponding historical traffic control comprehensive status indicators ,as well as

[0125] Under similar conditions X The simulation plan execution effect is then implemented. Corresponding simulated comprehensive traffic conditions indicators ;

[0126] 4.4.2 Calculate the comprehensive traffic condition index in the simulation Comprehensive indicators of historical traffic control The root mean square error is calculated using the following formula:

[0127] ;

[0128] in, = , ;

[0129] Among them, when Based on simulated comprehensive traffic conditions index Comprehensive indicators of historical traffic control Based on actual comparison, the target plan is considered to be If it's unreasonable, then it's reasonable; choose a reasonable target plan. .

[0130] The aforementioned method for constructing a smart traffic control plan for highways with dual emergency lanes involves, in step 5, performing reinforcement learning on a reasonable target plan, dynamically adjusting influencing factors, providing feedback on the plan results, and updating the comprehensive traffic control characteristics F. Specifically:

[0131] Comparing simulation results yields a comprehensive traffic condition index. Historical traffic control comprehensive status indicators obtained through numerical calculation The values ​​are compared to determine if they are equal, and the difference between the simulated values ​​is calculated. The specific formula is as follows:

[0132] ;

[0133] Among them, if If the value is greater than 0.0085, reinforcement learning is triggered;

[0134] The specific steps of reinforcement learning are as follows:

[0135] 5.1. Based on the control type, determine the main influencing factors causing the difference value, and calculate the new difference value through repeated iterations. until The value meets the requirements, and the dynamic adjustment of the impact factor is completed, specifically:

[0136] The main parameters involved in opening an emergency lane ,by For example, directly adjust or The value is used to reduce the error, and the process is iterated repeatedly until... Less than or equal to 0.0085;

[0137] If, after multiple adjustments to the count, the difference cannot be reduced to the standard level, then select the following in sequence: g、i、a、b、c , d Adjust the value until... Less than or equal to 0.0085;

[0138] 5.2 When the difference value calculation results are satisfactory, the original comprehensive traffic condition index is updated to the target comprehensive traffic condition index to complete the dynamic adjustment of the influencing factors and obtain the plan results;

[0139] 5.3 Update F by combining historical multi-dimensional data from the simulation with the target comprehensive traffic condition index.

[0140] Compared with the prior art, the beneficial effects of the present invention are:

[0141] This invention determines the calculation formula for the comprehensive traffic control status index of the road to be controlled, collects road data in real time, constructs the comprehensive traffic control status characteristics, and generates a cluster of control plans. Subsequently, through calculation and simulation verification, the obtained target plans, their execution effects, and execution conditions are recorded in the plan database. This allows highways with dual emergency lanes to quickly search for and directly provide solutions when dealing with high traffic volume or accidents in the inner lane, resulting in a faster emergency response speed. Furthermore, by comparing a series of plans in the control plan cluster from both the perspectives of index calculation and simulation, all control measures are fully considered, and control measures that may be implemented under various scenarios are fully verified, avoiding errors caused by unreasonable values ​​of influencing factor parameters. Simultaneously, it derives emergency target plans with high traffic efficiency and low risk of secondary accidents, effectively reducing traffic congestion, improving the overall traffic capacity of highways, significantly reducing accident risk, and greatly improving vehicle traffic efficiency. Attached Figure Description

[0142] Figure 1 This is a flowchart of the present invention;

[0143] Figure 2 Emergency lanes will not be opened;

[0144] Figure 3 To open the left emergency lane;

[0145] Figure 4 To open the right-hand emergency lane;

[0146] Figure 5 To open emergency lanes on both sides. Detailed Implementation

[0147] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0148] This invention provides, for example Figure 1-5 The method for constructing an intelligent management and control plan for a highway with dual emergency lanes, as shown, includes the following steps:

[0149] Step 1: Based on historical control priorities, obtain the weights of the planned indicators for the roads to be controlled, and determine the comprehensive traffic control status index J for the roads to be controlled. The specific steps are as follows:

[0150] 1.1 By acquiring and analyzing historical traffic accident data, traffic flow data, and existing control measures, we extract control experience in typical scenarios, clarify the focus of control, and use this as a reference benchmark to construct the weights w1', w2', and w3' of the contingency plan indicators;

[0151] 1.2. In the dynamic traffic simulation platform, simulate traffic flow characteristics under different scenarios and compare them with each other, and compare them with the actual historical road traffic conditions and attributes of historical scenarios. Obtain the theoretical calculated values ​​w1”, w2”, and w3” of the index weights from the perspective of numerical calculation.

[0152] 1.3. By fine-tuning w1', w2', w3' and w1”, w2”, w3” through simulation, the comprehensive traffic control status indicators are confirmed, so as to construct the integrated plan indicator weights w1, w2, w3;

[0153] 1.4. Based on the weights w1, w2, and w3 of the contingency plan indicators, the calculation formula for the comprehensive traffic control status index J is determined as follows:

[0154] J = w1⋅E eff − w2⋅E risk – w3⋅E rt ;

[0155] Where w1 is E eff The weights;

[0156] w2 is E risk The weights;

[0157] w3 is E rt The weights are w1 + w2 + w3 = 1;

[0158] E eff This is a traffic efficiency indicator; the higher the value, the higher the traffic efficiency.

[0159] E risk This is a safety risk indicator; the higher the value, the greater the safety risk.

[0160] E rt This is an indicator of emergency response time; the smaller the value, the shorter the rescue time.

[0161] Among them, traffic efficiency index E eff The calculation formula is as follows:

[0162] ;

[0163] Where C is the maximum design capacity, which is the design capacity of the road to be managed, reflecting the road's vehicle carrying capacity;

[0164] Given condition X, the traffic density on the road to be controlled at a certain time.

[0165] Given condition X, the traffic flow speed on the road to be controlled at a certain time.

[0166] X includes routine situations, traffic incidents, and the implementation of control measures, while J is used to compare the advantages and disadvantages of different combinations of control strategies.

[0167] Among them, safety risk indicator E risk The calculation formula is as follows:

[0168] ;

[0169] ;

[0170] ;

[0171] Among them, R acc Given condition X, the probability of an accident on the road segment to be controlled and during the time to be controlled.

[0172] This refers to the reference value for the probability of accidents on the road section to be controlled and the time to be controlled before the execution condition X.

[0173] Reference values ​​for the road curvature influencing factors of the road section to be controlled and the time to be controlled before implementing condition X;

[0174] Reference values ​​for slope impact factors for the road section to be controlled and the time to be controlled before implementing condition X;

[0175] Reference values ​​for the number of lanes influencing the road segment to be controlled and the time to be controlled before implementing condition X;

[0176] Reference values ​​for the visibility distance impact factor of the road section to be controlled and the time to be controlled before implementing condition X;

[0177] Reference values ​​for the obstacle coefficient influencing factors of the road section to be controlled and the time to be controlled before implementing condition X;

[0178] Before implementing condition X, the reference values ​​for traffic flow density of the road section to be controlled and the time to be controlled;

[0179] Reference values ​​for traffic flow speed on the road segment to be controlled and during the control period before condition X is implemented;

[0180] Among them, the emergency response time indicator E rt The calculation formula is as follows:

[0181]

[0182] in, This refers to the actual time taken for emergency response;

[0183] Distance between the rescue point and the accident site

[0184] For the time it takes for the rescue vehicle to travel;

[0185] for The minimum value;

[0186] For non-emergency rescue and control situations, w3 is 0;

[0187] In step 1, by importing historical traffic accident data, traffic flow data, and existing control measures, the intentions of the controllers are determined and quantified, resulting in the weights w1 for traffic efficiency indicators, w2 for safety risk indicators, and w3 for emergency response time indicators. This achieves the scientific quantification of the controllers' control objectives and provides strong support for the subsequent generation and optimization of contingency plans.

[0188] Step 2: Collect road data in real time from multiple dimensions, and construct the comprehensive traffic control status feature F based on the comprehensive traffic control status index J. The specific steps are as follows:

[0189] 2.1 Obtain the curvature, elevation, and number of lanes of the road to be managed through road design documents or high-precision maps to obtain road attribute data;

[0190] 2.2. Real-time data acquisition of traffic flow, density, vehicle type, visibility, and obstacle coefficient of the road to be managed is obtained through actual road sensors to acquire real-time road data and historical multi-dimensional data;

[0191] 2.3. Based on road attribute data and historical multi-dimensional data, a comprehensive traffic control feature F is constructed, as follows:

[0192] F = [R c G, L, S, C, P st , α, β, γ, δ, η, θ, ζ];

[0193] Among them, R c The road curvature factor reflects the degree of road curvature. Areas with greater curvature have reduced visibility, increased operational difficulty, increased centrifugal force, and a higher probability of accidents.

[0194] G is the slope influencing factor, which reflects the steepness of the road's slope and affects the likelihood of accidents by influencing the difference in vehicle speed.

[0195] L is the lane number influence factor, which reflects the impact of the number of lanes on the probability of accidents and has an important influence on the dynamic characteristics of traffic flow.

[0196] S is the visibility distance impact factor, which measures the impact of visibility on road accidents, such as abnormal weather like rain, snow, and fog. Different vehicle types have different driving characteristics and impacts on traffic flow.

[0197] P st The obstacle coefficient is a coefficient used to reflect the amount of obstacles on the road. The obstacle coefficient will increase when there are vehicles involved in unresolved traffic accidents on the road.

[0198] γ is R c The parameters are: δ is the parameter of G, η is the parameter of L, θ is the parameter of S, and ζ is the parameter of P. st The parameters are all greater than 0;

[0199] α is The parameters, β are Parameters;

[0200] In step 2, road attributes, traffic flow attributes, and environmental attributes are abstracted and quantified numerically to enable subsequent scientific and quantifiable analysis.

[0201] Step 3: Generate a cluster of traffic control plans based on relevant traffic control knowledge. Based on the comprehensive traffic control situation characteristic F, obtain the execution effect of the plans within the cluster. Specific steps are as follows:

[0202] 3.1. Generate a cluster of control plans by using administrative, technical, and physical means to dynamically intervene in vehicles on the road. ;

[0203] Methods for dynamically intervening in vehicles on the road include:

[0204] M1. Set the types of control and management plan clusters, specifically including speed limit, diversion, and lane closure;

[0205] Furthermore, control plans are generated for different types of control measures, including:

[0206] A. Only speed limit is applied, speed limit v limit Three speed limit schemes are proposed: 60, 70, and 80 km / h.

[0207] B. Only execute the traffic splitting ratio. Four diversion schemes with values ​​of 0.2, 0.4, 0.6, and 0.8 are proposed, and the influence on traffic flow density is achieved by affecting traffic volume.

[0208] at this time ∈ [0,1], Q = Q0 ⋅ (1 – );

[0209] Where Q represents the traffic after splitting;

[0210] Q0 represents the initial flow rate;

[0211] C. Only the emergency lane is open at this time. Satisfy the following formula:

[0212] ;

[0213] in, To determine the traffic density after opening i emergency lanes;

[0214] Original density;

[0215] The original number of lanes;

[0216] The number of lanes after opening i emergency lanes;

[0217] M2. For roads with two emergency lanes, there are four scenarios: emergency lane not open, emergency lane open on the left, emergency lane open on the right, and emergency lanes open on both sides. This results in a cluster of control plans. The details are as follows:

[0218] ;

[0219] ;

[0220] 3.2. Based on the contingency plan For example, based on the comprehensive traffic control situation characteristic F, the corresponding contingency plan execution effect is obtained. The specific steps are as follows:

[0221] 3.2.1 Calculate the initial comprehensive traffic control index using real-time road data and the comprehensive traffic control situation characteristic F. ;

[0222] 3.2.2. Contingency plans are directly derived from historical multi-dimensional data detection results. Under similar conditions (x) The actual effect of execution Changes;

[0223] Where x is a certain value;

[0224] Furthermore, when only speed limiting is applied, in most of the x cases:

[0225] ;

[0226] In rare cases:

[0227] ;

[0228] 3.2.3 Calculation Plan Under similar conditions (x) Multiple possible comprehensive traffic control indicators resulting from the implementation of the following measures ;

[0229] 3.2.4, Through and Receive control plan Representative comprehensive traffic control indicators The specific steps are as follows:

[0230] H1. Dataset Preparation: (Based on the contingency plan) Under similar conditions The results of multiple executions were organized into four sequences, which are as follows:

[0231] ;

[0232] ;

[0233] ;

[0234] Corresponding comprehensive traffic control indicators ;

[0235] H2. Calculate various statistical measures: Calculate the mean, median, mode, standard deviation, and interquartile range for each of the four sequences to characterize different statistical features;

[0236] H3. Compare the statistics and select representative values, as follows:

[0237] H31. Calculate the deviation between each statistic and its corresponding observations;

[0238] Among them, the statistics are mean, median, mode, standard deviation, and interquartile range;

[0239] The observed values ​​are the actual values ​​from the four sequences;

[0240] The deviation table includes: Mean Absolute Error (MAE), Root Mean Square Error (MSE), and Root Mean Square Deviation (RMSD).

[0241] H32. Select the statistic with the smallest bias as the representative value of each of the four sequences, denoted as H32. , , Subsequently, representative comprehensive traffic control indicators were calculated. ;

[0242] 3.2.5. Regarding the cluster of control and prevention plans The execution steps of all the contingency plans, from 3.2.2 to 3.2.4, yielded the execution effectiveness and representative comprehensive traffic control indicators for each of the ten contingency plans. The details are as follows:

[0243]

[0244] In step 3, by parameterizing a series of control measures, the transformation between specific control measures and abstract indicator changes is realized. Furthermore, the rationality of the control plan cluster generation results is ensured through the simulation of a large number of control measures. After parameterization, the parameters can be transformed into the influence on specific elements in the indicators, which facilitates subsequent optimization and deduction.

[0245] Step 4: In the cluster of control plans and their corresponding implementation effects, simulation is introduced for comparison to derive the target plan and verify its rationality. The specific steps are as follows:

[0246] 4.1 Constructing initial comprehensive traffic control indicators based on real-time road data and road attribute data. Execute the simulation control plan cluster The effectiveness of the simulated contingency plan was obtained by combining the results from the simulated detectors with the comprehensive traffic control indicators. Comprehensive indicators of simulated traffic control ;

[0247] In order to facilitate the distinction of the initial state in the simulation, the initial state is used. replace ;

[0248] 4.2 Comprehensive indicators of simulated traffic control Representative traffic control comprehensive status indicators Among them, the scheme with the maximum value is selected as the target contingency plan;

[0249] 4.3. Perform a consistency check between the simulation execution effect of the target contingency plan and the historical execution effect of the corresponding contingency plan to verify the stability of the target contingency plan. After verification, store it in the contingency plan library. The specific steps are as follows:

[0250] 4.3.1 Obtain the simulation plan execution effect and historical traffic control comprehensive status indicators for different historical scenarios, specifically:

[0251] The target plan is set as follows: Select a control plan Subsequently, by analyzing historical multi-dimensional data, we identified contingency plans that had been executed with the same objectives. Different initial conditions y, obtain Corresponding historical contingency plan implementation effects ,as well as

[0252] Compared with historical traffic conditions indicators ;

[0253] 4.3.2 In simulated fictional scenarios, obtain target plans for different fictional scenarios. The simulation plan's execution effectiveness and the comprehensive indicators of simulated traffic control are as follows:

[0254] The simulation simulates different initial scenarios to be controlled, obtains flow rate, density, and velocity through detectors, and then calculates the execution effect of the simulation plan. ,and

[0255] Simulated traffic control comprehensive status index ;

[0256] 4.3.3, Target contingency plan The effectiveness of historical contingency plans in different historical scenarios is combined with the effectiveness of simulated contingency plans in hypothetical scenarios to verify the target contingency plan. The stability under different conditions is as follows:

[0257] Use equivalence tests and distribution difference tests to... and Verification is required;

[0258] If stable, we will obtain the target plan, the effectiveness of the plan's implementation, and comprehensive indicators of traffic control. Stored in the contingency plan database;

[0259] If unstable, the obtained target plan, the effectiveness of the plan's implementation, and the overall traffic control indicators will be affected. Additional notes should be added to indicate that the condition for use is the initial comprehensive traffic control status index. And the comprehensive traffic control situation characteristics F, stored in the contingency plan database;

[0260] In step 4.3, the stability of the target plan under different conditions is verified through simulation, and the applicability, robustness and effectiveness of the target plan in different traffic scenarios are known.

[0261] 4.4 The reliability of the target plan is verified by simulation. The specific steps are as follows:

[0262] 4.4.1 Based on the target plan For example, obtain the contingency plan The simulation execution effect;

[0263] Specifically: Obtaining the contingency plan The effectiveness of historical contingency plans under x similar conditions Corresponding historical traffic control comprehensive status indicators ,as well as

[0264] Under similar conditions X The simulation plan execution effect is then implemented. Corresponding simulated comprehensive traffic conditions indicators ;

[0265] 4.4.2 Calculate the comprehensive traffic condition index in the simulation Comprehensive indicators of historical traffic control The root mean square error is calculated using the following formula:

[0266] ;

[0267] in, = , ;

[0268] Among them, when Based on simulated comprehensive traffic conditions index Comprehensive indicators of historical traffic control Based on actual comparison, the target plan is considered to be If it's unreasonable, then it's reasonable; choose a reasonable target plan. .

[0269] In step 4, a series of contingency plans in the control plan cluster are compared horizontally from two perspectives: indicator calculation and simulation. All control measures are fully considered, and the control measures that may be implemented in various scenarios are fully verified, avoiding errors that may be caused by unreasonable values ​​of influencing factor parameters.

[0270] Furthermore, the target contingency plan, execution conditions, and corresponding execution effects are obtained through calculation and simulation verification and recorded in the contingency plan library. In the event of an accident, a solution can be quickly searched and provided directly. By comparing a series of contingency plans in the control contingency plan cluster from the perspectives of indicator calculation and simulation, all control measures are fully considered, and control measures that may be implemented in various scenarios are fully verified. This avoids errors caused by unreasonable values ​​of influencing factor parameters. At the same time, an emergency plan with high traffic efficiency and low risk of secondary accidents is obtained, which significantly reduces the risk of accidents and improves the efficiency of vehicle traffic.

[0271] Step 5: Conduct reinforcement learning on the reasonable target contingency plan, dynamically adjust the influencing factors, provide feedback on the plan results, and update the comprehensive traffic control situation characteristic F, specifically as follows:

[0272] Comparing simulation results yields a comprehensive traffic condition index. Historical traffic control comprehensive status indicators obtained through numerical calculation The values ​​are compared to determine if they are equal, and the difference between the simulated values ​​is calculated. The specific formula is as follows:

[0273] ;

[0274] Among them, if If the value is greater than 0.0085, reinforcement learning is triggered;

[0275] The specific steps of reinforcement learning are as follows:

[0276] 5.1. Based on the control type, determine the main influencing factors causing the difference value, and calculate the new difference value through repeated iterations. until The value meets the requirements, and the dynamic adjustment of the impact factor is completed, specifically:

[0277] The main parameters involved in opening an emergency lane ,by For example, directly adjust or The value is used to reduce the error, and the process is iterated repeatedly until... Less than or equal to 0.0085;

[0278] If, after multiple adjustments to the count, the difference cannot be reduced to the standard level, then select the following in sequence: g、i、a、b、c , dAdjust the value until... Less than or equal to 0.0085;

[0279] 5.2 When the difference value calculation results are satisfactory, the original comprehensive traffic condition index is updated to the target comprehensive traffic condition index to complete the dynamic adjustment of the influencing factors and obtain the plan results;

[0280] 5.3 Update F by combining historical multi-dimensional data from the simulation with the target comprehensive traffic condition index;

[0281] In step 5, parameters are adjusted through simulation to further optimize the rationality of the calculation of the comprehensive traffic condition index, thus resolving the irrationality that may be introduced by setting influencing factors based on experience.

[0282] Furthermore, by updating the target plan with each update and optimizing the influencing factors, the optimality of the target plan strategy is continuously ensured, effectively improving the accuracy of the plan.

[0283] Verification Experiment

[0284] I. Example of contingency plan construction, the specific steps are as follows:

[0285] G. Acquire data on roads within the study area to obtain road curvature, elevation, and number of lanes. Use simulated sensors to acquire traffic flow, density, vehicle type, visibility, and obstacle coefficient. Model road features using existing information.

[0286] Construct a multi-traffic condition attribute feature that includes road attribute features:

[0287] F = [R c ,G,L,S,C,P st ,α,β,γ,δ,η,θ,ζ]

[0288] H. Clarify the weighting of the contingency plan experience by combining historical experience with simulation analysis. The specific steps are as follows:

[0289] H1. For the spatial unit of the road leaving the accident site, calculate... The specific formula is as follows:

[0290] ;

[0291] For the spatial units of roads leading away from the accident site, the movement of rescue vehicles does not need to be considered; only the comprehensive consideration of traffic efficiency and accident risk is required, with α = 1.2 and β = −1.5.

[0292] And R c G, S, P st It does not change with the opening or closing of the emergency lane;

[0293] H2, The Underwood model of velocity-density relationship is used for J2, and the specific formula is as follows:

[0294] ;

[0295] ;

[0296] ;

[0297] H3. For the road leading to the accident site and the spatial unit one of the accident site, calculate... 1. The specific formula is as follows: ;

[0298] Where α = 1.2, β = −1.5, and R c The values ​​G and S do not change with the opening or closing of the emergency lane or in the event of an accident;

[0299] H4. The comprehensive traffic control index J1 for spatial unit 1 adopts the Greenberg model of speed-density relationship, and the specific formula is as follows:

[0300] ;

[0301] ;

[0302] Among them, here It has no specific meaning and is only used for mathematical substitution.

[0303] H5. Based on historical experience and simulation, the weights of each influencing factor in this scenario are determined as follows:

[0304] E eff The weight is 0.652;

[0305] E risk The weight is 0.341;

[0306] E rt The weight is 0.007;

[0307] Among them, a two-way ten-lane roadway can ensure that traffic capacity does not decrease and the probability of accidents does not increase before rescue. Therefore, the impact of rescue time is relatively low, and the main focus is on ensuring traffic efficiency.

[0308] Since there is no upstream regulation, the number of vehicles passing through these two spatial units per hour is the same and does not change with the increase or decrease of the number of lanes;

[0309] K. Contingency plan parameterization, as detailed below:

[0310] Based on the hypothetical scenario and the nature of dual emergency lanes, the following conclusions are drawn. The calculation formula is as follows:

[0311] ;

[0312] Where i represents the i-th spatial unit, and k represents the time step. The emergency lane switch status at time k and position i;

[0313] Considering the road leading to the accident site and the accident site as Unit 1, and the road leading away from the accident site as Unit 2, the emergency lane switch states for these two spatial units are as follows: ,but:

[0314] ;

[0315] To study the emergency lane switching status of two spatial units in the road section;

[0316] The system identifies 16 possible states for emergency lane access. The controlled road section is divided into two segments based on the accident location. Each segment has four possible states for the dual emergency lanes. 4. A total of 16 types;

[0317] Considering that the emergency lane switch status will not change frequently in actual execution, we will not consider the status switching during execution for the time being, but will consider it in subsequent feedback iterations.

[0318] There are currently 16 provisional emergency lane control plans: X1, X2, ..., X 16 The parameters affected by the emergency lane switch include the lane number factor L;

[0319] L. Indicator calculation and multi-scenario evaluation, as detailed below:

[0320] Calculate the comprehensive traffic control status index of the spatial unit of the road leading away from the accident site. The value of the comprehensive traffic control index is obtained when both emergency lanes on both sides are open in each scheme.

[0321] Further calculations were performed on the road leading to the accident site and the comprehensive indicators of the accident site as a spatial unit. ,and The values ​​are used to create Table 1:

[0322] Table 1

[0323]

[0324] In Table 1, X1 indicates that the emergency lane is not open, X2 indicates that the left emergency lane is open, X3 indicates that the right emergency lane is open, and X4 indicates that both emergency lanes are open.

[0325] Analyze separately , and each of the following schemes X i Performance under various speed and density scenarios can be obtained, when When both emergency lanes of the two spatial units are fully open, the optimal strategy is to not consider the mid-way state switching.

[0326] M. Reinforcement learning, dynamically adjusting parameters, as detailed below:

[0327] If the initial evaluation results are consistent with the simulation demonstration results, the contingency plan feedback is directly added to the contingency plan library. After the second evaluation, the contingency plan results are compared and the better result is selected.

[0328] Based on the following steps: road condition numerical quantification → target numerical quantification → control numerical quantification → comparison of various control effects → model feedback optimization → result output, the above steps have been organized. The contingency plan evaluation system determines the framework at the very beginning of the process. Each time the above steps are completed, the contingency plan library is updated.

[0329] Since the simulation and numerical results are highly consistent, this result was directly adopted. However, the second simulation requires multiple results to be fitted, which necessitates the actual calibration of the mathematical model. Then, the above steps are repeated to determine whether the current plan is reasonable.

[0330] N. Feedback on the results of the contingency plan, as detailed below:

[0331] This step is executed when the simulation results match the calculated control indicators. Specifically:

[0332] Based on traffic flow indicators obtained from simulations or detected during actual execution, fine-tune the parameters and use comparative reinforcement learning to further update the parameters.

[0333] As shown in the example above, the multi-indicator evaluation results are consistent with the simulation results. The comprehensive traffic control status feature F is updated by combining multi-dimensional simulation data and traffic flow indicators during actual execution.

[0334] O. Evaluation system calibration

[0335] This step is executed when the simulation results differ from the calculated control indicators. Specifically:

[0336] If the actual effect of the simulation results is X4 > X2 > X1 > X3, that is, the simulation results are worse after opening the right emergency lane, and the control index should be lower, it means that the weights w1, w2, and w3 are set unreasonably. The weights w1, w2, and w3 need to be updated and the above steps GL need to be repeated until the simulation results are consistent with the index results.

[0337] The above is a simple example of constructing a type of contingency plan. When the contingency plan library is updated to improve accuracy, the road units will be further divided, and state changes will be incorporated into the calculation for further iteration.

[0338] Comparison of simulation verification and control effects:

[0339] Figure 2-5 This is a schematic diagram of the experiment. Figure 2-5 As can be seen, the simulation results are consistent with the index calculation results, both showing X4> X2> X3> X1, indicating that the effect is best when all lanes are open;

[0340] By comparing a series of control measures from the perspectives of indicator calculation and simulation, the system fully considers all control measures and verifies the control measures that may be implemented in various scenarios. The two perspectives of indicator calculation and simulation avoid errors that may be caused by unreasonable values ​​of influencing factor parameters.

[0341] In summary, this invention determines the calculation formula for the comprehensive traffic control status index of the road to be controlled, collects road data in real time, constructs the comprehensive traffic control status characteristics, and generates a cluster of control plans. Subsequently, through calculation and simulation verification, the obtained target plans, their execution effects, and execution conditions are recorded in the plan database. This allows highways with dual emergency lanes to quickly search for and directly provide solutions when dealing with high traffic volumes or accidents in the inner lane, resulting in a faster emergency response speed. Furthermore, by comparing a series of plans in the control plan cluster from both the perspectives of index calculation and simulation, all control measures are fully considered, and control measures that may be implemented in various scenarios are fully verified, avoiding errors caused by unreasonable values ​​of influencing factor parameters. Simultaneously, it yields emergency target plans with high traffic efficiency and low risk of secondary accidents, effectively reducing traffic congestion, improving the overall traffic capacity of highways, significantly reducing accident risks, and greatly improving vehicle traffic efficiency.

[0342] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used above are only some embodiments described in this invention. Obviously, those skilled in the art can obtain other drawings based on these drawings.

[0343] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for constructing an intelligent management and control plan for highways with dual emergency lanes, characterized by the following steps: Step 1: Based on historical control priorities, obtain the weights of the planned indicators for the roads to be controlled, and determine the calculation formula for the comprehensive traffic control status index of the roads to be controlled. The specific steps are as follows: 1.1 By acquiring and analyzing historical traffic accident data, traffic flow data, and existing control measures, we extract control experience in typical scenarios, clarify the focus of control, and use this as a reference benchmark to construct the weights w1', w2', and w3' of the contingency plan indicators; 1.

2. In the dynamic traffic simulation platform, simulate traffic flow characteristics under different scenarios and compare them with each other, and compare them with the actual historical road traffic conditions and attributes of historical scenarios. Obtain the theoretical calculated values ​​w1”, w2”, and w3” of the index weights from the perspective of numerical calculation. 1.

3. By fine-tuning w1', w2', w3' and w1”, w2”, w3” through simulation, the comprehensive traffic control status indicators are confirmed, so as to construct the integrated plan indicator weights w1, w2, w3; 1.

4. Based on the weights w1, w2, and w3 of the contingency plan indicators, the calculation formula for the comprehensive traffic control status index J is determined as follows: J = w1⋅E eff − w2⋅E risk – w3⋅E rt ; in, w1 is E eff The weights; w2 is E risk The weights; w3 is E rt The weights are given by w1 + w2 + w3 = 1; E eff E is a traffic efficiency indicator. risk For safety risk indicators; E rt As an indicator of emergency response time; Step 2: Collect road data in real time from multiple dimensions, and construct comprehensive traffic control status features based on comprehensive traffic control status indicators; Step 3: Generate a cluster of traffic control plans based on relevant traffic control knowledge. Then, based on the comprehensive characteristics of the traffic control situation, obtain the execution effect of the plans within this cluster. The specific steps are as follows: 3.

1. Generate a cluster of control plans by using administrative, technical, and physical means to dynamically intervene in vehicles on the road. ; The methods for dynamically intervening in vehicles on the road include: M1. Set the types of control and management plan clusters, specifically including speed limit, diversion, and lane closure; Furthermore, control plans are generated for different types of control measures, including: A. Only speed limit is applied, speed limit v limit Three speed limit schemes are proposed: 60, 70, and 80 km / h. B. Only execute the traffic splitting ratio. Four traffic diversion schemes with values ​​of 0.2, 0.4, 0.6, and 0.8 are proposed, and their impact on traffic density is determined by influencing traffic flow. ∈ [0,1], Q = Q0 ⋅ (1 – ); where Q is the traffic after splitting; Q0 is the initial traffic; C. Only the emergency lane is open at this time. Satisfy the following formula: ； in, To determine the traffic density after opening i emergency lanes; Original density; The original number of lanes; The number of lanes after opening i emergency lanes; M2. For roads with two emergency lanes, there are four scenarios: emergency lane not open, emergency lane open on the left, emergency lane open on the right, and emergency lanes open on both sides. This results in a cluster of control plans. The details are as follows: ； ； 3.

2. Based on the contingency plan For example, based on the comprehensive traffic control situation characteristic F, the corresponding contingency plan execution effect is obtained. ; Step 4: In the cluster of control plans and the corresponding implementation effects of the plans, simulation is introduced for comparison to obtain the target plan and verify whether the target plan is reasonable. Step 5: Conduct reinforcement learning on reasonable target plans, dynamically adjust influencing factors, provide feedback on plan results, and update the comprehensive characteristics of traffic control.

2. The method for constructing a smart control plan for highways with dual emergency lanes according to claim 1, characterized in that: In step 1.4, the traffic efficiency index E eff The calculation formula is as follows: ； Where C is the maximum design capacity, which is the design capacity of the road to be managed, reflecting the road's vehicle carrying capacity. Given condition X, the traffic density on the road to be controlled at a certain time. Given condition X, the traffic flow speed on the road to be controlled at a certain time. X includes routine situations, traffic incidents, and the implementation of control measures, while J is used to compare the advantages and disadvantages of different combinations of control strategies. Among them, safety risk indicator E risk The calculation formula is as follows: ； ； ； Among them, R acc Given condition X, the probability of an accident on the road segment and during the time to be controlled. This refers to the reference value for the probability of accidents on the road section to be controlled and the time to be controlled before the execution condition X. Reference values ​​for the road curvature influencing factors of the road section to be controlled and the time to be controlled before implementing condition X; Reference values ​​for slope impact factors for the road section to be controlled and the time to be controlled before implementing condition X; Reference values ​​for the number of lanes influencing the road segment to be controlled and the time to be controlled before implementing condition X; Reference values ​​for the visibility distance impact factor of the road section to be controlled and the time to be controlled before implementing condition X; Reference values ​​for the obstacle coefficient influencing factors of the road section to be controlled and the time to be controlled before implementing condition X; Before implementing condition X, the reference values ​​for traffic flow density of the road section to be controlled and the time to be controlled; Reference values ​​for traffic flow speed on the road segment to be controlled and during the control period before condition X is implemented; Among them, the emergency response time indicator E rt The calculation formula is as follows: ； in, This refers to the actual time taken for emergency response; Distance between the rescue point and the accident site For the time it takes for the rescue vehicle to travel; for The minimum value; For non-emergency rescue and control situations, w3 is 0.

3. The method for constructing a smart control plan for highways with dual emergency lanes according to claim 2, characterized in that: In step 2, road data is collected in real time from multiple dimensions. Based on the comprehensive traffic control status index, comprehensive traffic control status features are constructed. The specific steps are as follows: 2.1 Obtain the curvature, elevation, and number of lanes of the road to be managed through road design documents or high-precision maps to obtain road attribute data; 2.

2. Real-time data acquisition of traffic flow, density, vehicle type, visibility, and obstacle coefficient of the road to be managed is obtained through actual road sensors to acquire real-time road data and historical multi-dimensional data; 2.

3. Based on road attribute data and historical multi-dimensional data, a comprehensive traffic control feature F is constructed, as follows: F = [R c ,G,L,S,C,P st ,a,b,c,d,h,the,z]; Among them, R c This is the road curvature influencing factor, used to reflect the degree of road curvature; G is the slope influencing factor, which reflects the steepness of the road's slope and affects the likelihood of accidents by influencing the difference in vehicle speed. L is the lane number influence factor, which reflects the impact of the number of lanes on the probability of accidents and has an important influence on the dynamic characteristics of traffic flow. S is the visibility distance impact factor, which measures the impact of visibility on road accidents; P st The obstacle coefficient is a coefficient used to reflect the amount of obstacles on the road. The obstacle coefficient will increase when there are vehicles involved in unresolved traffic accidents on the road. γ is R c The parameters are: δ is the parameter of G, η is the parameter of L, θ is the parameter of S, and ζ is the parameter of P. st The parameters are all greater than 0; α is The parameters, β are The parameters.

4. The method for constructing a smart control plan for highways with dual emergency lanes according to claim 3, characterized in that: In step 3.2, according to the plan For example, based on the comprehensive traffic control situation characteristic F, the corresponding contingency plan execution effect is obtained. The specific steps are as follows: 3.2.1 Calculate the initial comprehensive traffic control index using real-time road data and the comprehensive traffic control situation characteristic F. ; 3.2.

2. The contingency plan is directly obtained from the results of historical multi-dimensional data detection. Under similar conditions (x) The actual effect of execution Changes; Where x is a certain value; Furthermore, when only speed limiting is applied, in most of the x cases: ； In rare cases: ； 3.2.3 Calculation Plan Under similar conditions (x) Multiple possible comprehensive traffic control indicators resulting from the implementation of the following measures ; 3.2.4, Through and Receive control plan Representative comprehensive traffic control indicators The specific steps are as follows: H1. Dataset Preparation: (Based on the contingency plan) Under similar conditions The results of multiple executions were organized into four sequences, which are as follows: ； ； ； Corresponding comprehensive traffic control indicators ; H2. Calculate various statistical measures: Calculate the mean, median, mode, standard deviation, and interquartile range for each of the four sequences to characterize different statistical features; H3. Compare the statistics and select representative values, as follows: H31. Calculate the deviation between each statistic and its corresponding observations; Among them, the statistics are mean, median, mode, standard deviation, and interquartile range; The observed values ​​are the actual values ​​from the four sequences; The deviation table includes: Mean Absolute Error (MAE), Root Mean Square Error (MSE), and Root Mean Square Deviation (RMSD). H32. The statistic with the smallest bias is selected as the representative value of each of the four sequences, denoted as H32. , , Subsequently, representative comprehensive traffic control indicators were calculated. ; 3.2.

5. Regarding the cluster of control and prevention plans The execution steps of all the contingency plans, from 3.2.2 to 3.2.4, yielded the execution effectiveness and representative comprehensive traffic control indicators for each of the ten contingency plans. The details are as follows: 。 5. The method for constructing a smart control plan for highways with dual emergency lanes according to claim 4, characterized in that: In step 4, simulation is introduced to compare the control plan cluster and its corresponding execution effect, derive the target plan, and verify whether the target plan is reasonable. The specific steps are as follows: 4.1 Constructing initial comprehensive traffic control indicators based on real-time road data and road attribute data. Execute the simulation control plan cluster The effectiveness of the simulated contingency plan was obtained by combining the results from the simulated detectors with the comprehensive traffic control indicators. Comprehensive indicators of simulated traffic control ; 4.2 Comprehensive indicators of simulated traffic control Representative traffic control comprehensive status indicators Among them, the scheme with the maximum value is selected as the target contingency plan; 4.

3. Verify the consistency between the simulation execution effect of the target contingency plan and the historical execution effect of the corresponding contingency plan to check whether the target contingency plan has stability. After verification, store it in the contingency plan library. 4.4 The reliability of the target plan is verified by simulation.

6. The method for constructing a smart control plan for highways with dual emergency lanes according to claim 5, characterized in that: In step 4.3, the consistency between the simulation execution effect of the target contingency plan and the historical execution effect of the corresponding contingency plan is checked to verify whether the target contingency plan has stability. After verification, it is stored in the contingency plan library. The specific steps are as follows: 4.3.1 Obtain the simulation plan execution effect and historical traffic control comprehensive status indicators for different historical scenarios, specifically: The target plan is set as follows: Select a control plan Subsequently, contingency plans with the same objectives were found in historical multi-dimensional data. Different initial conditions y, obtain Corresponding historical contingency plan implementation effects ,as well as Historical traffic conditions indicators ; 4.3.2 In simulated fictional scenarios, obtain target plans for different fictional scenarios. The simulation plan's execution effectiveness and the comprehensive indicators of simulated traffic control are as follows: The simulation simulates different initial scenarios to be controlled, obtains flow rate, density, and velocity through detectors, and then calculates the execution effect of the simulation plan. ,and Simulated traffic control comprehensive status index ; 4.3.3, Target contingency plan The effectiveness of historical contingency plans in different historical scenarios is combined with the effectiveness of simulated contingency plans in hypothetical scenarios to verify the target contingency plan. The stability under different conditions is as follows: Use equivalence tests and distribution difference tests to... and Verification is required; If stable, we will obtain the target plan, the effectiveness of the plan's implementation, and comprehensive indicators of traffic control. Stored in the contingency plan database; If unstable, the obtained target plan, the effectiveness of the plan's implementation, and the overall traffic control indicators will be affected. Additional notes should be added to indicate that the condition for use is the initial comprehensive traffic control status index. The comprehensive characteristics of traffic control, F, are stored in the contingency plan database.

7. The method for constructing a smart control plan for highways with dual emergency lanes according to claim 6, characterized in that: In step 4.4, the reliability of the target plan is verified through simulation. The specific steps are as follows: 4.4.1 Based on the target plan For example, obtain the contingency plan The simulation execution effect; Specifically: Obtaining the contingency plan The effectiveness of historical contingency plans under x similar conditions Corresponding historical traffic control comprehensive status indicators ,as well as Under similar conditions X The simulation plan execution effect is then implemented. Corresponding simulated comprehensive traffic conditions indicators ; 4.4.2 Calculate the comprehensive traffic condition index in the simulation Comprehensive indicators of historical traffic control The root mean square error is calculated using the following formula: ； in, = , ; Among them, when Based on simulated comprehensive traffic conditions index Comprehensive indicators of historical traffic control Based on actual comparison, the target plan is considered to be If it's unreasonable, then it's reasonable; choose a reasonable target plan. .

8. The method for constructing a smart control plan for highways with dual emergency lanes according to claim 7, characterized in that: In step 5, reinforcement learning is performed on the reasonable target plan, influencing factors are dynamically adjusted, the plan results are fed back, and the comprehensive traffic control situation feature F is updated, specifically as follows: Comparing simulation results yields a comprehensive traffic condition index. Historical traffic control comprehensive status indicators obtained through numerical calculation The values ​​are compared to determine if they are equal, and the difference between the simulated values ​​is calculated. The specific formula is as follows: ； Among them, if If the value is greater than 0.0085, reinforcement learning is triggered; The specific steps of reinforcement learning are as follows: 5.

1. Based on the control type, determine the main influencing factors causing the difference value, and calculate the new difference value through repeated iterations. until The value meets the requirements, and the dynamic adjustment of the impact factor is completed, specifically: The main parameters involved in opening an emergency lane ,by For example, directly adjust η The value is used to reduce the error, and the process is iterated repeatedly until... Less than or equal to 0.0085; If, after multiple adjustments to the count, the difference cannot be reduced to the standard level, then select the following in sequence: ζ, θ, α, β, γ , δ Adjust the value until... Less than or equal to 0.0085; 5.2 When the difference value calculation results are satisfactory, the original comprehensive traffic condition index is updated to the target comprehensive traffic condition index to complete the dynamic adjustment of the influencing factors and obtain the plan results; 5.3 Update F by combining historical multi-dimensional data from the simulation with the target comprehensive traffic condition index.

Images