A method for dynamically dividing highway lane functions

Abstract

The application discloses a method for dynamically dividing expressway lane functions, comprising the following steps: in real time, receiving expressway lane traffic flow, vehicle type composition and lane changing frequency data, and clearly defining basic types and combination types of the lanes; constructing a lane function division model, setting constraint conditions to be followed by lane function division, and establishing an objective function aiming at maximizing traffic efficiency and minimizing safety risks; using a genetic algorithm to solve the lane function division model, and outputting an optimal or suboptimal lane function division scheme; and dynamically publishing the division scheme to roadside facilities through lane-level information publishing facilities for execution or reminding. The method solves the problems of lack of theoretical support and poor timeliness of existing expressway lane function division, breaks through the limitations of traditional experience-driven modes, promotes the transformation of lane function division from "qualitative judgment" to "quantitative decision", and improves the operation efficiency and safety level of expressways.

Description

Technical Field

[0001] This invention relates to the field of active control technology for road lane functions, and in particular to a method for dynamically classifying the functions of highway lanes. Background Technology

[0002] The study of lane function division can be traced back to the traffic management concept of "passenger-freight separation," the core of which is to improve road traffic efficiency and driving safety by spatially separating vehicles with significantly different driving behaviors (such as passenger cars and freight cars). Traditional lane function division methods mainly generate corresponding lane function division schemes based on relevant practical experience, and propose recommended schemes through traffic simulation evaluation. When the scale of the highway network and the number of lanes are relatively limited, this combination of experience and simulation can meet the management and control needs of general highways. However, with the rapid development of highway traffic infrastructure in my country, more and more multi-lane highways with ten or more lanes in both directions have emerged. These highways are characterized by a large number of lanes, complex traffic composition, significant tidal phenomena, and high dynamic control requirements. Traditional lane function division methods are difficult to adapt to the lane function control needs of multi-lane highways and have the following shortcomings:

[0003] (1) Lack of theoretical support, the generation of schemes relies on subjective experience. Most existing division schemes rely on expert experience or analogy with existing engineering cases, and lack a systematic and quantifiable theoretical analysis framework to guide the scientific configuration of lane functions.

[0004] (2) It is difficult to meet the needs of dynamic lane adjustment in traffic scenarios with multiple lanes. As the number of lanes increases, the possible combinations of lane function divisions grow exponentially. If we still use the method of modeling and simulating one by one and comparing one by one, it is time-consuming and laborious, and it is difficult to meet the needs of rapid response and dynamic adjustment in actual engineering. In particular, it lacks flexibility and adaptability when dealing with sudden traffic events or periodic traffic fluctuations.

[0005] In modeling the actual traffic capacity of roads, Wang Wei et al. proposed to consider three influencing factors—traffic composition within lanes, driver characteristics, and roadside interference—to correct the traffic capacity (see Wei Xueyan, Xu Chengcheng, Wang Wei, et al. Traffic capacity analysis and modeling of multi-lane expressways [J]. Transportation Systems Engineering and Information, 2017, 17(2): 105-111.). This calculation method integrates multiple influencing factors, but lacks consideration of the interaction between lanes. The interaction between vehicles on expressways is closely related to lane-changing behavior.

[0006] Therefore, there is an urgent need to construct a dynamic lane function allocation method that is theoretically sound, structurally clear, and quantifiable, in order to improve the operational efficiency and safety level of highways. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method for dynamically dividing highway lane functions. By constructing a highway lane function division model with the goal of maximizing operational efficiency and minimizing safety risks, the dynamic division of lane functions can be achieved through quantitative analysis based on real-time traffic flow.

[0008] The objective of this invention is achieved through the following technical solution:

[0009] A method for dynamically allocating highway lane functions includes:

[0010] S1. Receive real-time data on traffic flow, vehicle type composition, and lane change frequency for each lane on the highway, and identify the basic types and combinations of lanes. The basic types of lanes include dedicated passenger lanes, dedicated freight lanes, and mixed-traffic lanes.

[0011] S2. Based on the basic types and combinations of lanes, a lane function partitioning model is constructed. The lane function partitioning model consists of an objective function and constraints. The objective function aims to maximize traffic efficiency and minimize safety risk. Traffic efficiency is represented by a capacity coefficient, and safety risk is represented by a mixing coefficient. The capacity coefficient is the ratio of the sum of the actual traffic capacity of each lane to the sum of the basic traffic capacity of each lane. The mixing coefficient is obtained by multiplying the passenger vehicle mixing coefficient and the freight vehicle mixing coefficient by the maximum value of the product of the passenger vehicle mixing coefficient and the freight vehicle mixing coefficient. The objective function sums the capacity coefficient constraint term and the mixing coefficient constraint term in a weighted form and then performs numerical scaling.

[0012] S3. Use a genetic algorithm to solve the lane function partitioning model and output the optimal or suboptimal lane function partitioning scheme;

[0013] S4. The lane function division plan will be dynamically published to roadside facilities through lane-level information publishing facilities for implementation or reminders.

[0014] Furthermore, in step S2, the actual capacity of a single lane is obtained by correcting the basic capacity. The correction items include traffic composition correction, driver characteristic correction, roadside interference correction, and interaction influence correction from adjacent lanes. The traffic composition correction includes the proportion of each vehicle type and the corresponding vehicle type conversion factor. The interaction influence of adjacent lanes is characterized by a factor representing the frequency of lane changing from adjacent lanes to this lane, so as to reflect the impact of lane changing behavior on the capacity of this lane.

[0015] Furthermore, the actual traffic capacity of a single lane It is expressed as follows:

[0016] ; (1)

[0017] ; (2)

[0018] ; (3)

[0019] ; (4)

[0020] ; (5)

[0021] In the formula, This refers to the basic capacity, corresponding to the maximum capacity of Level 5 service under the given speed limits. This refers to the actual traffic capacity of a single lane. To achieve the minimum safe headway, This is the correction factor for road capacity based on traffic composition. This is a correction factor for road capacity based on driver characteristics, with values ​​ranging from 0.95 to 1. This is the roadside interference correction factor, which is set to 1 for highways. The percentage of each type of vehicle. For vehicle model The vehicle conversion factor, The right-hand interaction factor. The left-side interactive influencing factor, These represent the current passenger vehicle traffic flow in the right lane, the passenger vehicle traffic flow in the left lane, the freight vehicle traffic flow in the right lane, and the freight vehicle traffic flow in the left lane, respectively. These represent the average number of lane changes for passenger vehicles and the average number of lane changes for freight vehicles, respectively.

[0022] A higher road capacity coefficient t indicates higher road traffic efficiency, expressed as:

[0023] ; (6)

[0024] In the formula, These represent the actual traffic capacity of a single lane for passenger vehicles, freight vehicles, and mixed traffic lanes, respectively. N represents the number of dedicated passenger vehicle lanes, the number of mixed-traffic lanes, the number of dedicated freight vehicle lanes, and the total number of lanes, respectively. The value can be 1, 0, or -1, representing respectively Section of highway lanes It belongs to the passenger vehicle lane, the mixed traffic lane, and the freight vehicle lane.

[0025] Furthermore, the mixing coefficient is determined through the following steps: calculating the actual proportion of passenger cars and freight cars in the mixed lanes, calculating the product of the passenger car mixing coefficient and the freight car mixing coefficient based on the proportion, and normalizing the product with the maximum possible product. The resulting value is the mixing coefficient, and the larger the value, the higher the degree of mixing and the greater the risk of conflict. In the process of setting the objective function, a positive exponential structure is adopted for the capacity coefficient to reward the improvement of capacity, and a negative exponential structure is adopted for the mixing coefficient to penalize the increase of mixing. The two terms are added according to a predetermined weight and then logarithmically transformed or other numerical scaling is performed to eliminate the difference in dimensions. The weights are set based on the traffic management objectives to achieve a trade-off between efficiency and safety.

[0026] Furthermore, , Let represent the mixing coefficients for passenger vehicles and freight vehicles in a mixed-traffic lane, respectively, as shown in the following formulas:

[0027] ; (7)

[0028] ; (8)

[0029] ; (9)

[0030] In the formula, For actual traffic demand; This indicates the proportion of passenger vehicles in actual traffic demand. This is the maximum value of the product of the passenger vehicle mixing coefficient and the freight vehicle mixing coefficient. At this value, the degree of mixed traffic is the highest, and the safety is the worst. These represent the actual traffic volume of passenger vehicles on the dedicated passenger vehicle lane and the actual traffic volume of freight vehicles on the dedicated freight vehicle lane, respectively.

[0031] The objective function expression is as follows:

[0032] (10)

[0033] In the formula, These represent the weights of the capacity constraint term and the hybrid coefficient constraint term, respectively, where e is a natural constant. This indicates that the road capacity coefficient t is exponentialized to serve as a weighting factor for the passage term. This indicates the mixing coefficient. The negative values ​​are exponentialized to serve as a suppression weight for the mixture term.

[0034] Furthermore, the constraints include:

[0035] (a) Lane number constraint: The total number of dedicated lanes shall not exceed the total number of lanes designed for the road segment, and at least one mixed traffic lane shall be provided;

[0036] (b) Lane position constraints: In the longitudinal cross section, the three lane types should follow the spatial order from the inside to the outside as dedicated passenger vehicle lane, mixed traffic lane, and dedicated freight vehicle lane;

[0037] (c) Capacity constraints: The actual capacity of each type of lane should be greater than the actual traffic demand allocated to that type of lane, and the total actual capacity of the road should be greater than the total traffic demand.

[0038] Furthermore, the solution process for step S3 is as follows:

[0039] S301. Randomly generate an initial lane function division scheme and calculate the objective function value of the initial scheme;

[0040] S302. Use a genetic algorithm to find an encoding scheme to encode individuals in the population, choosing either floating-point encoding or binary encoding;

[0041] S303. Using the multimodal function value as the fitness of an individual, calculate the fitness of each individual in the population; the population is a set containing several lane function partitioning schemes, and the individual is each lane function partitioning scheme in the corresponding set;

[0042] S304. Select individuals to participate in reproduction based on their fitness level;

[0043] S305. Perform crossover and mutation operations on the selected individuals to generate offspring, which will serve as a new lane function allocation scheme;

[0044] S306. Iterate the calculation of the objective function until the preset convergence condition or the preset maximum number of iterations is reached, and return the optimal or suboptimal lane function partitioning scheme.

[0045] Furthermore, in step S4, at least one set of lane-level electronic signs or variable message signs is installed every 2-3 kilometers along the highway to display the current lane function division scheme, open / closed status and lane type in graphic or text form to prompt the driver or roadside control unit.

[0046] The present invention also provides a management system for dynamically dividing highway lanes, comprising:

[0047] The data acquisition module is used to receive real-time data on traffic flow, vehicle type composition, and lane change frequency for each lane on the highway, and to identify the basic types and combinations of lane settings. The basic types of lanes include dedicated lanes for passenger vehicles, dedicated lanes for freight vehicles, and mixed-traffic lanes.

[0048] The lane function partitioning model module is used to construct a lane function partitioning model based on the basic types and combinations of lanes. The lane function partitioning model consists of an objective function and constraints. The objective function aims to maximize traffic efficiency and minimize safety risk. Traffic efficiency is represented by a capacity coefficient, and safety risk is represented by a mixing coefficient. The capacity coefficient is the ratio of the sum of the actual traffic capacity of each lane to the sum of the basic traffic capacity of each lane. The mixing coefficient is obtained by multiplying the passenger vehicle mixing coefficient and the freight vehicle mixing coefficient by the maximum value of the product of the passenger vehicle mixing coefficient and the freight vehicle mixing coefficient. The objective function sums the capacity coefficient constraint term and the mixing coefficient constraint term in a weighted form and then performs numerical scaling.

[0049] The optimization solution module is used to solve the lane function partitioning model using a genetic algorithm and output the optimal or suboptimal lane function partitioning scheme.

[0050] The publishing and execution module is used to dynamically publish the division scheme to roadside facilities through lane-level information publishing facilities for execution or reminder.

[0051] Furthermore, the lane function division model module and the optimization solution module are implemented in software and deployed on a server or edge computing device. The system also includes a database module for storing historical traffic data and lane function division scheme effect data to support online learning and offline verification of lane function division model parameters.

[0052] Compared with the prior art, the beneficial effects of the technical solution of the present invention are as follows:

[0053] 1. Real-time collection of traffic flow, vehicle type composition, and lane change frequency data for each lane: Utilizing roadside sensing devices or vehicle-to-infrastructure (V2I) platforms, real-time data is received on each lane's traffic flow, vehicle type composition (passenger / freight vehicle ratio), average lane change frequency, and other input variables. This real-time input enables the lane function partitioning model to recalculate the actual lane capacity and mixing coefficient based on the latest traffic flow, thereby generating an optimal lane function scheme that matches the current state. This significantly improves timeliness and responsiveness (e.g., it can quickly provide adjustment plans after peak hours or accidents, reducing congestion escalation). It solves the problems of traditional schemes relying on experience or historical static data, failing to respond to time-varying traffic conditions (peak hours, emergencies, tidal flow), and failing to accurately reflect dynamic interactions between lanes.

[0054] 2. Divide lanes into three categories: dedicated for passenger vehicles, dedicated for freight vehicles, and mixed traffic, and limit the combination types: By clearly defining the three types of lanes and the allowed combinations (such as inner lane for passenger vehicles, outer lane for freight vehicles, and at least one mixed traffic lane), and writing them into the lane function division model constraints; the constraints ensure feasibility and safety (for example, forcing at least one mixed traffic lane to avoid the inability to adjust temporary needs under a completely dedicated system), improve the engineering feasibility and safety boundaries of the scheme; solve the problem that arbitrary division without constraints may lead to unreasonable or unsafe layout problems, such as a completely dedicated system without a mixed traffic lane or a chaotic spatial order.

[0055] 3. The actual capacity of a single lane is obtained by modifying the baseline capacity through multiple adjustments (see equations (1)-(6) and parameters in Table 2): Based on the basic maximum capacity (under Level 5 service), the actual capacity of each lane is calculated according to traffic composition (vehicle type ratio and conversion factor), driver characteristic correction factor, roadside interference factor, and the interaction influence factor of adjacent lanes. Through multi-factor correction, the capacity estimation is made closer to reality, so that the lane function division model will not overestimate / underestimate the capacity of a certain type of lane to carry traffic flow when allocating lane functions, thereby reducing congestion or safety risks caused by configuration errors. This solves the problem that a single "baseline capacity" cannot reflect the actual weakening or strengthening of capacity due to factors such as truck ratio, driver differences, and lane changing interference, which leads to misjudgment of capacity in the scheme.

[0056] 4. Introducing the interaction impact factor between left and right lanes and representing it with lane-changing frequency (interaction correction term): The capacity of each lane is corrected using the passenger / freight vehicle traffic flow and the average number of lane changes in the left and right lanes to reflect the impact of lane changes on traffic capacity. By explicitly modeling the interaction impact of lane-changing behavior, the model can reflect higher congestion / conflict levels in high-lane-changing scenarios (such as when freight vehicles and passenger vehicles occupy similar occupancy), prompting optimization schemes to prioritize specialization or layout adjustments to reduce conflicts, thereby improving safety and stability. Since vehicle lane changes (especially overtaking lane changes) significantly affect the local traffic capacity of the lane being changed, this technical solution solves the problem of traditional models ignoring or roughly estimating this impact.

[0057] 5. Overall definition of capacity coefficient t (sum of actual capacity of each lane / sum of basic capacity): Road operating efficiency is quantified into a capacity coefficient t (the larger the better), and this is incorporated into the objective function. t is a dimensionless index, facilitating trade-offs with the mixing coefficient u; quantifying the engineering index allows the optimizer (genetic algorithm) to directly search for higher efficiency solutions numerically. This addresses the problem in existing technologies of lacking a unified and comparable efficiency metric, which hinders the comparison and trade-off of multiple solutions.

[0058] 6. Definition and measurement of the mixing coefficient u: By calculating the proportion of passenger cars / freight vehicles in mixed lanes, their respective mixing coefficients are generated. The normalized mixing coefficient u is obtained by multiplying their products by the theoretical maximum product (the larger the value, the higher the degree of mixing and the greater the risk of conflict). After quantifying the degree of mixing, the objective function can directly reduce high-mixing schemes (i.e., more likely to choose lane specialization or change lane order to reduce mixing conflicts), thereby reducing the accident risk and speed fluctuations caused by lane changing / interaction conflicts.

[0059] 7. The objective function is calculated by weighting the positive exponent of the capacity term with the negative exponent of the mixing coefficient term and applying logarithmic scaling: The objective function is... ; and through weights and Adjusting efficiency / safety preferences. The exponential mapping amplifies improvements in the capacity coefficient *t* (encouraging schemes with significant speed increases or capacity expansion), while exponentially suppressing increases in the mixing coefficient *u* (strongly penalizing high mixing), thus more clearly distinguishing superior and inferior schemes in a high-dimensional discrete partition space. Using a logarithmically compressible scale improves numerical stability (avoiding the dominance of any single absolute value while maintaining comparability), facilitating convergence of the genetic algorithm. It addresses issues such as direct linear weighting potentially failing to emphasize the importance of increased capacity or effectively suppressing safety risks associated with high mixing, and the difficulty of directly adding different units.

[0060] 8. Employing Genetic Algorithms for Model Solving and Lane Combination Optimization: Genetic algorithms are used to encode and search lane types (multi-lane, multi-combination); addressing the problem that the combination space for lane function partitioning grows exponentially with the number of lanes, and exhaustive / simulation methods suffer from high computational complexity, long processing times, and difficulty in achieving real-time or near-real-time responses. Genetic algorithms excel at finding near-optimal solutions in large, discrete, non-convex, and nonlinear combination spaces; combined with appropriate encoding and a fitness function (objective function value), they can obtain engineering-acceptable optimal or suboptimal solutions within a finite number of iterations, thus meeting dynamic response requirements.

[0061] 9. Lane-level Information Dissemination and Execution: Lane-level electronic signs or variable message signs are deployed every 2-3 km along the route to disseminate the current lane function plan and distribute it to the roadside / vehicle terminals. Standardized dissemination frequency and expression improve the visibility and executability of the plan; connection with the vehicle-road cooperative platform enables command issuance and vehicle-side response, forming a closed loop and ensuring the actual traffic improvement effect of model optimization. This solves the problem that if model output cannot be promptly and clearly conveyed to the driver or lower-level control unit, it cannot be executed and the effects cannot be implemented.

[0062] 10. Systematic Hardware and Software Implementation: The lane function division model module and optimization solution module are implemented in software and deployed on servers or edge computing devices, including a historical data storage module for online learning and offline validation. Database-supported online learning can adjust correction coefficients (e.g., driver characteristics, interaction factor estimation) based on actual performance, improving model accuracy and robustness over the long term; edge deployment reduces latency and supports near real-time computation. This addresses issues such as the inability of a single model output to self-correct over long-term operation and the difficulty in parameter optimization due to a lack of historical data support.

[0063] 11. Ternary indicator variables represent lane type: using ternary variables with values ​​of 1 / 0 / −1. The type of each lane is clearly defined for constraint, calculation and coding; unified coding makes it easier to directly write physical constraints (number of lanes, position order) into the mathematical model and be processed by the optimizer, and also facilitates data docking with the simulation / execution system.

[0064] 12. Engineering Examples and Parameterization: Based on the example scenario verification of Implementation Example 1, the examples show that the model can generate different and reasonable lane configurations in different scenarios (such as introducing a dedicated truck lane under high traffic volume and high truck ratio), providing reproducible engineering evidence and enhancing feasibility and scalability.

[0065] 13. This invention combines real-time data acquisition, interactive correction, and actual capacity calculation to achieve a more accurate match between real-time traffic demand and capacity on highways; significantly reduces congestion misjudgments or safety hazards caused by estimation errors; and improves the accuracy of objective function evaluation, thereby enabling the genetic algorithm to find more practical and executable solutions.

[0066] 14. This invention, by combining an exponential objective function with a genetic algorithm, is more conducive to finding "high-return, low-risk" solutions. Exponentialization highlights solutions that significantly improve traffic capacity while reducing mixed-traffic conflicts, and the genetic algorithm, at this fitness level, is more likely to converge to an engineering-feasible optimal solution. Real-world examples show that a significant improvement in the objective function value corresponds to a more reasonable lane layout (after switching from fully mixed-traffic to partially dedicated lanes, t increases and u decreases, resulting in an improved objective function).

[0067] 15. This invention uses an online database combined with historical performance storage and parameter correction to enhance the model's adaptability. Over time, the model can correct parameters such as left / right interaction factors and driver characteristic coefficients through historical feedback, so that the solution can continuously improve in long-term operation and reduce the sensitivity of the model to abnormal data in a single instance.

[0068] 16. Enhance decision-making transparency and interpretability: The mathematical modeling-based approach gives the entire partitioning process a clear logical structure and parameter input-output path, improving the traceability and interpretability of the decision-making process and helping to enhance the understanding and acceptance of traffic policies by managers and the public.

[0069] 17. Excellent flexibility and responsiveness: When faced with dynamically changing traffic demands (such as peak hours, emergencies, or severe weather), the model can quickly calculate and generate lane function allocation schemes that adapt to the current traffic conditions, providing traffic management departments with real-time or near-real-time control suggestions, which helps to achieve the transformation from passive response to proactive management.

[0070] 18. Facilitates integration with intelligent transportation systems: This model can be integrated with traffic sensing devices, dynamic sign control systems, vehicle-road cooperative platforms, etc., to form a closed-loop control system and promote the evolution of lane function division from static planning to dynamic intelligence.

[0071] In summary, compared with the traditional classification method that relies on experience-based judgment, the modeling approach proposed in this invention achieves a fundamental shift from "subjective experience-driven" to "data and model-driven," which not only improves the scientific rigor and efficiency of scheme generation but also provides new methodological support and technical pathways for the optimal allocation of lane resources in future intelligent transportation systems. Attached Figure Description

[0072] Figure 1 A flowchart is shown showing the method for setting the dynamic lane division function for highways according to the present invention.

[0073] Figure 2 A schematic diagram illustrating lane function division and lane-level information management according to an embodiment of the present invention is shown.

[0074] Figures 3a to 3c The diagrams show the recommended lane division schemes for scenarios with low traffic volume and low truck ratio, high traffic volume and low truck ratio, and high traffic volume and high truck ratio, respectively.

[0075] Figure 4 This is a schematic diagram of the iterative training curve of the lane function partitioning model. Detailed Implementation

[0076] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only for explaining the present invention and are not intended to limit the present invention.

[0077] The description in this section pertains only to typical embodiments, and the present invention is not limited to the scope of the embodiments described. Combinations of different embodiments, substitution of some technical features in different embodiments, and substitution of similar or identical prior art with some technical features in the embodiments are also within the scope of the description and protection of the present invention.

[0078] This invention addresses the shortcomings of quantitative analysis in highway lane function allocation methods by proposing a dynamic method for allocating highway lane functions, such as... Figure 1 As shown, it includes:

[0079] S1. Receive real-time data on traffic flow, vehicle type composition, and lane change frequency for each lane on the highway, and determine the basic types and combinations of lanes that need to be set up.

[0080] Based on the existing highway lane function classification, they are generally divided into three types: dedicated passenger vehicle lanes, dedicated freight vehicle lanes, and mixed-traffic lanes. Therefore, this study focuses on these three types of dedicated lanes. In practical application, there are four possible combinations of these three types of lanes, and their basic setup requirements are as follows:

[0081] (1) Mixed lanes. The most basic lane function division method, suitable for scenarios with relatively low traffic volume.

[0082] (2) Bus lane + mixed traffic lane. Suitable for scenarios with a large volume of bus traffic. Usually, the bus lane is on the inside and the mixed traffic lane is on the outside.

[0083] (3) Mixed traffic lane + truck lane. Suitable for scenarios with high truck traffic volume. Typically, the mixed traffic lane is on the inside and the truck lane is on the outside.

[0084] (4) Bus lane + mixed traffic lane + freight truck lane. Suitable for scenarios where both bus and freight traffic are relatively high. Typically, the bus lane is on the inside, the mixed traffic lane is in the middle, and the freight truck lane is on the outside.

[0085] In addition, theoretically, there is also a combination of dedicated bus lanes and dedicated freight truck lanes. However, this combination is rarely used in practice due to its weak ability to withstand traffic risks and poor flexibility, and therefore it is not considered for inclusion in the modeling and analysis.

[0086] S2. Construct a lane function division model.

[0087] First, the overall objective is determined and the objective function is established. Then, the constraints are proposed. Together, these two constitute the lane function partitioning model.

[0088] S2-1, Objective Function

[0089] The goal of the lane function partitioning model is to obtain a lane function partitioning scheme that balances efficiency and safety. Efficiency is mainly reflected in traffic capacity, which is represented by a traffic capacity coefficient. The value of this coefficient is the ratio of the sum of the actual traffic capacity of each lane to the sum of the basic traffic capacity of each lane, and the larger the value, the better. Safety is mainly reflected in the proportion of trucks and the vehicle lane-changing rate. After the adoption of dedicated lanes, the lane-changing rate has been controlled accordingly. Therefore, safety is mainly reflected in the degree of mixing of different types of vehicles in mixed lanes, which is represented by a mixing coefficient, and the smaller the value, the better.

[0090] (1) Calculation of traffic capacity coefficient

[0091] In addition to geometric constraints, road capacity must also consider the road's ability to handle the corresponding traffic load, that is, the relationship between the road's actual capacity and actual traffic demand. Road capacity is one of the indicators of road operating efficiency and is of great reference value for lane function classification.

[0092] The initial actual capacity of a single lane did not take into account the impact of lane interaction. This embodiment takes into account the differences in vehicle composition between adjacent lanes and introduces a lane interaction impact factor. This relationship is represented by formulas (1), (4), and (5). The right-side interaction factor mainly considers the impact of vehicles in the right lane changing lanes to the left on the driving status of vehicles in this lane. This impact is related to the average number of lane changes from the right to the left. Similarly, the left-side interaction factor mainly describes the impact of vehicles in the left lane changing lanes to the right on this lane. For the innermost and outermost lanes, only the right lane or the left lane interacts, while other lanes need to consider the interaction relationship between the two lanes simultaneously.

[0093] ; (1)

[0094] ; (2)

[0095] ; (3)

[0096] ; (4)

[0097] ; (5)

[0098] In the formula Basic capacity (corresponding to the maximum capacity of Level 5 service under the corresponding speed limit conditions). This refers to the actual traffic capacity of a single lane. To achieve the minimum safe headway, This is the correction factor for road capacity based on traffic composition. This is a correction factor for road capacity based on driver characteristics, with values ​​ranging from 0.95 to 1. This is the roadside interference correction factor, which is set to 1 for highways. The percentage of each type of vehicle. For vehicle model The vehicle conversion factor, The right-hand interaction factor. The left-side interactive influencing factor, These represent the current passenger vehicle traffic flow in the right lane, the passenger vehicle traffic flow in the left lane, the freight vehicle traffic flow in the right lane, and the freight vehicle traffic flow in the left lane, respectively. These represent the average number of lane changes for passenger cars and the average number of lane changes for freight cars, respectively.

[0099] The road capacity coefficient t is the ratio of the sum of the actual capacity of all lanes to the sum of the basic capacity of all lanes. A higher value indicates higher road efficiency. It is expressed as:

[0100] ; (6)

[0101] In the formula, These represent the actual traffic capacity of a single lane for passenger vehicles, freight vehicles, and mixed traffic lanes, respectively. N represents the number of dedicated passenger vehicle lanes, the number of mixed-traffic lanes, the number of dedicated freight vehicle lanes, and the total number of lanes, respectively.

[0102] (2) Calculation of mixing coefficient

[0103] A mixing coefficient *u* is introduced to characterize the relationship between traffic composition and safety level. Its value is equal to the product of the passenger vehicle mixing coefficient and the freight vehicle mixing coefficient, divided by the maximum value of the product of the passenger vehicle mixing coefficient and the freight vehicle mixing coefficient. It is mainly used to describe the impact of the proportion of passenger vehicles and freight vehicles in the traffic composition on safety in mixed-traffic lanes. For mixed-traffic lanes, where both freight vehicles and passenger vehicles travel together, when… ,along with The continuous increase in the proportion of passenger vehicles on mixed-traffic lanes leads to an increase in the mixing coefficient factor, increased conflicts in mixed-traffic lanes, and a decrease in safety levels. ,along with With the continuous increase in traffic volume, the proportion of trucks has decreased, the mixing factor has decreased, conflicts in mixed lanes have decreased, and the level of road safety has improved.

[0104] ; (7)

[0105] ; (8)

[0106] ; (9)

[0107] In the formula, For actual traffic demand; This indicates the proportion of passenger vehicles in actual traffic demand. This is the maximum value of the product of the passenger vehicle mixing coefficient and the freight vehicle mixing coefficient. At this value, the degree of mixed traffic is the highest, and the safety is the worst. These represent the actual traffic volume of passenger vehicles on the dedicated passenger vehicle lane and the actual traffic volume of freight vehicles on the dedicated freight vehicle lane, respectively.

[0108] (3) Objective function considering traffic efficiency and safety:

[0109] Considering the dimensional differences between the capacity coefficient and the mixing coefficient, the actual capacity needs to be normalized. A larger capacity coefficient indicates higher road operating efficiency, so a positive exponential structure is used; a larger mixing coefficient indicates a more uniform mix of passenger and freight vehicles, but also more interaction conflicts between vehicles, so a negative exponential structure is used. The capacity constraint and the mixing coefficient constraint are summed, and then numerically scaled using a logarithmic transformation. The objective function expression is as follows:

[0110] (10)

[0111] In the formula, These represent the weights of the capacity constraint term and the hybrid coefficient constraint term, respectively, where e is a natural constant. This indicates that the road capacity coefficient t is exponentialized to serve as a weighting factor for the passage term. This indicates the mixing coefficient. The negative values ​​are exponentialized to serve as a suppression weight for the mixture term;

[0112] S2-2, Constraints:

[0113] In order to characterize Section of highway lanes Which of the three lane types does it belong to? Introduce a ternary variable. (Values ​​range from 1, 0, -1), and its specific definition is as follows:

[0114] (11)

[0115] (1) Lane number and location constraints:

[0116] Regarding the number of lanes, the number of dedicated lanes cannot exceed the total number of lanes determined during the road planning phase. Based on the basic requirements for lane configuration, at least one mixed-traffic lane is required to accommodate changes in traffic flow for different vehicle types. The formula is as follows:

[0117] (12)

[0118] (13)

[0119] In the formula, It is the total number of lanes planned. This represents the total number of road segments.

[0120] In terms of lane location, according to the basic requirements for lane setting, the spatial relationship of the three lane types on highways from the inside out should be as follows: the innermost lane is for passenger vehicles, followed by mixed traffic lanes, and the outermost lane is for freight vehicles, as shown in the formula below:

[0121] (14)

[0122] (15)

[0123] In the formula, It is the total number of lanes planned. This represents the total number of road segments.

[0124] (2) Traffic capacity constraints:

[0125] The overall actual traffic capacity of the road should be greater than the total traffic demand allocated to the road, and the traffic capacity of each type of lane in all lanes should not be less than its corresponding actual traffic demand, as shown in formulas (16)-(21).

[0126] (16)

[0127] (17)

[0128] (18)

[0129] (19)

[0130] (20)

[0131] (twenty one)

[0132] In the formula, These represent the actual traffic capacity of a single lane for passenger vehicles, freight vehicles, and mixed traffic lanes, respectively. These represent the number of dedicated passenger vehicle lanes, the number of mixed-traffic lanes, and the number of dedicated freight vehicle lanes, respectively. For actual traffic demand; These represent the proportion of passenger vehicles in mixed-traffic lanes and the proportion of passenger vehicles in actual traffic demand, respectively.

[0133] S2-3, Lane Function Division Model Structure:

[0134] In summary, the lane function division model for port access expressways can be described as follows: And formulas (12)-(21).

[0135] S3. Solving the lane function partitioning model. A genetic algorithm is used to solve the model; the specific steps are as follows:

[0136] S301. Randomly generate an initial lane function division scheme and calculate the objective function value of the initial scheme;

[0137] S302. Use a genetic algorithm to find a suitable encoding scheme to encode the individuals in the population. Common encoding schemes such as floating-point encoding or binary encoding can be selected.

[0138] S303. Using the multimodal function value as the fitness of an individual, calculate the fitness of each individual in the population; the population is a set containing several lane function partitioning schemes, and the individual is each lane function partitioning scheme in the corresponding set;

[0139] S304. Select the parent and parent individuals to participate in reproduction based on their fitness levels. The principle of selection is that individuals with higher fitness levels are more likely to be selected.

[0140] S305. Perform genetic operations on the selected father and mother, that is, copy the genes of the father and mother, and use crossover, mutation and other operators to produce offspring as a new lane function division scheme;

[0141] S306. Calculate the objective function value of the new lane function partitioning scheme, and determine whether the objective function value has converged or reached the specified number of iterations. If not, continue iterating; if yes, end the calculation and return the optimal solution and the suboptimal solution.

[0142] S4. Dynamically publish the optimal lane division scheme to roadside facilities through lane-level information publishing facilities for execution or reminder.

[0143] Specifically, at least one set of lane-level electronic signs or variable message signs can be installed every 2-3 kilometers along the highway to display the current lane function division scheme, open / closed status and lane type in graphic or text form to prompt the driver or roadside control unit.

[0144] Preferably, the system further includes data integration between the lane function division model and road traffic sensing equipment, traffic flow detection system, vehicle-road cooperative platform, and dynamic sign control system. This allows the system to utilize real-time traffic flow, vehicle type ratio, and lane-changing statistics as model inputs and to release and execute the model outputs in a coordinated manner, forming a closed-loop control system. Figure 2 As shown in the diagram, this system can dynamically publish whether lane functions are enabled and the specific functions of each lane as needed.

[0145] In summary, this invention, based on a lane function division model and combined with real-time traffic volume on highways, dynamically divides lane functions, breaking through the limitations of the traditional experience-driven model, and promoting the transformation of lane function division from "qualitative judgment" to "quantitative decision-making," thereby improving the operational efficiency and safety level of highways.

[0146] Example 1

[0147] This embodiment is based on the above method and process, taking a two-way 10-lane expressway as an example, with a single-lane cross-section shown below. Figures 3a to 3c The highway is 11km long and has 3 pairs of ramps and exits. Three scenarios were designed: low traffic volume with low truck ratio, high traffic volume with low truck ratio, and high traffic volume with high truck ratio, serving as three typical scenarios for studying lane function division. In the low traffic volume scenario, the traffic volume is 6000 pcu / h one-way, with the maximum flow direction being 1 to 8, at 4200 pcu / h. In the high traffic volume scenario, the traffic volume is 9000 pcu / h one-way, with the maximum flow direction being 1 to 8, at 6300 pcu / h. Detailed traffic flow between each ramp and exit under different scenarios is shown in Table 1.

[0148] Table 1. Different traffic flow test scenarios

[0149]

[0150] The calculation process is illustrated below using a high-traffic scenario (with trucks accounting for 20%) as an example:

[0151] The first step is to set the initial plan, [0, 0, 0, 0, 0], which are all mixed traffic lanes.

[0152] The second step is to calculate the objective function value of the initial scheme.

[0153] (1) Calculate the actual traffic capacity of each lane.

[0154] Based on the capacity correction formula and the parameter settings in Table 2, the actual capacity of each lane can be obtained. In this embodiment, the capacity calculation correction considers not only the usual factors such as truck ratio, driver characteristics, and roadside interference, but also the impact of lane-changing behavior from the left and right lanes on the current lane. Statistical results from actual data show that the left-side interaction factor is slightly larger than the right-side interaction factor, which is closely related to drivers' habit of overtaking by changing lanes from the left. It should be noted that not all lanes require correction for both left and right-side influence factors; for example, the innermost lane only needs to consider the right-side influence factor. The average actual capacity of each lane is obtained by summing and averaging the capacity of all lanes. In subsequent iterative calculations, under a specific lane function division scheme, the vehicle distribution in different lanes can be statistically analyzed using a traffic flow model, thereby dynamically updating the left and right interaction factors of each lane.

[0155]

[0156]

[0157] To represent the average actual traffic capacity of each lane, Let represent the actual traffic capacity of the i-th lane.

[0158] Table 2. Parameters related to traffic capacity calculation

[0159]

[0160] (2) Calculate the objective function value

[0161] First, calculate: =2.171;

[0162] Next, calculate: =0.527;

[0163] With all weighting coefficients set to 0.5, the calculated objective function value is 0.29936.

[0164] The third step is to use a genetic algorithm to solve the lane function partitioning model, generate new schemes, and calculate new objective function values.

[0165] The parameter settings for the genetic algorithm are detailed in Table 3. The population size is 100, the crossover probability and mutation probability are 0.7 and 0.1 respectively, and the number of iterations depends on the initial scheme and the random seed settings. In most cases, convergence can be achieved in 200-243 iterations, so 300 iterations are chosen. The convergence criterion is that the objective function value changes less than the convergence threshold for five consecutive iterations. , The value is set to 0.005. The new scheme generated by the genetic function is [0, 0, 0, -1, -1], which means 3 mixed lanes and 2 truck lanes, with a corresponding objective function value of 0.30598.

[0166] Table 3. Parameter settings for genetic algorithm solution

[0167]

[0168] The fourth step is the iterative training and calculation of the lane function division model. According to the convergence rules, the new lane function division scheme did not meet the convergence criteria and required repeated iterative calculations. The model converged on the 225th iteration, with the corresponding objective function value being [1,1, 0, 0, 0]. The corresponding lane function scheme, from the inside out, consists of 2 dedicated bus lanes and 3 mixed-traffic lanes. See the table and figure below for details of the calculation process.

[0169] Table 4 Calculation Process Table

[0170]

[0171] The iterative training curve of the lane function partitioning model in this embodiment is shown in the figure. Figure 4 .

[0172] Through iterative training and solving, we obtained vehicle-to-truck (VTL) function partitioning schemes for three scenarios: low traffic volume with low truck ratio, high traffic volume with low truck ratio, and high traffic volume with high truck ratio, as shown in Figure 3. The specific schemes are as follows:

[0173] The lane function division scheme for low traffic volume and low truck ratio scenarios is: 5 mixed traffic lanes.

[0174] The lane function division scheme for high traffic volume and low truck ratio scenarios is as follows: the functions of the 5 lanes from the inside out are passenger vehicle lane, passenger vehicle lane, mixed traffic lane, mixed traffic lane, and mixed traffic lane.

[0175] The lane function division scheme for high traffic volume and high truck proportion scenarios is as follows: from the inside out, the functions of the 5 lanes are 1 dedicated passenger vehicle lane + 2 mixed traffic lanes + 2 dedicated truck lanes.

[0176] Example 2

[0177] Based on the same inventive concept, embodiments of this application also provide a management system for dynamically dividing highway lanes, which can be used to implement the method for dynamically dividing highway lanes as described above, as illustrated in the following embodiments. Since the principle of the management system for dynamically dividing highway lanes is similar to that of the method for dynamically dividing highway lanes, the implementation of the management system for dynamically dividing highway lanes can refer to the implementation of the method for dynamically dividing highway lanes, and repeated details will not be elaborated further. As used below, the terms "unit" or "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the system described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0178] The present invention provides a specific implementation of a management system for dynamically dividing highway lanes, which is capable of realizing the function of dynamically dividing highway lanes. The management system for dynamically dividing highway lanes specifically includes the following:

[0179] The data acquisition module is used to receive real-time data on traffic flow, vehicle type composition, and lane change frequency for each lane on the highway, and to identify the basic types and combinations of lane settings. The basic types of lanes include dedicated lanes for passenger vehicles, dedicated lanes for freight vehicles, and mixed-traffic lanes.

[0180] The lane function partitioning model module is used to construct a lane function partitioning model based on the basic types and combinations of lanes. The lane function partitioning model consists of an objective function and constraints. The objective function aims to maximize traffic efficiency and minimize safety risk. Traffic efficiency is represented by a capacity coefficient, and safety risk is represented by a mixing coefficient. The capacity coefficient is the ratio of the sum of the actual traffic capacity of each lane to the sum of the basic traffic capacity of each lane. The mixing coefficient is obtained by multiplying the passenger vehicle mixing coefficient and the freight vehicle mixing coefficient by the maximum value of the product of the passenger vehicle mixing coefficient and the freight vehicle mixing coefficient. The objective function sums the capacity coefficient constraint term and the mixing coefficient constraint term in a weighted form and then performs numerical scaling.

[0181] The optimization solution module is used to solve the lane function partitioning model using a genetic algorithm and output the optimal or suboptimal lane function partitioning scheme.

[0182] The publishing and execution module is used to dynamically publish the division scheme to roadside facilities through lane-level information publishing facilities for execution or reminder.

[0183] The lane function partitioning model module and the optimization solution module are implemented in software and deployed on servers or edge computing devices.

[0184] In one embodiment, the management system for dynamically dividing highway lane functions also includes a database module for storing historical traffic data and lane function division scheme effect data, so as to support online learning and offline verification of lane function division model parameters.

[0185] Preferably, embodiments of this application also provide a specific implementation of an electronic device capable of implementing all steps of the method for dynamically dividing highway lanes in the above embodiments. The electronic device specifically includes the following:

[0186] Processor, memory, communications interface, and bus;

[0187] The processor, memory, and communication interface communicate with each other via a bus; the communication interface is used to realize information transmission between server-side devices, metering devices, and user-side devices.

[0188] The processor is used to call a computer program in memory, and when the processor executes the computer program, it implements all the steps in the method for dynamically dividing highway lanes in the above embodiments.

[0189] Preferably, embodiments of this application also provide a computer-readable storage medium capable of implementing all steps of the method for dynamically dividing highway lanes in the above embodiments. The computer-readable storage medium stores a computer program that, when executed by a processor, implements all steps of the method for dynamically dividing highway lanes in the above embodiments.

[0190] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on its differences from other embodiments. In particular, hardware + program embodiments are relatively simple in description because they are fundamentally similar to method embodiments; relevant parts can be referred to the descriptions in the method embodiments.

[0191] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0192] While this application provides method operation steps as shown in the embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-inventive labor. The order of steps listed in the embodiments is merely one possible execution order among many and does not represent the only execution order. In actual device or client product execution, the method can be executed in the order shown in the embodiments or drawings or in parallel (e.g., in a parallel processor or multi-threaded processing environment).

[0193] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0194] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0195] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0196] This invention is not limited to the embodiments described above. The above description of specific embodiments is intended to illustrate and explain the technical solutions of this invention. The specific embodiments described above are merely illustrative and not restrictive. Without departing from the spirit and scope of the claims, those skilled in the art can make many specific modifications based on the teachings of this invention, and these modifications all fall within the scope of protection of this invention.

Claims

1. A method of dynamically dividing highway lane functions, characterized by, Comprise: S1. Real-time receive highway each lane traffic flow, vehicle type composition and lane changing frequency data, and clear lane basic type and combination type, lane basic type includes passenger car lane, truck lane and mixed lane; S2. Based on the basic type and combination type of the lane, a lane function division model is constructed, the lane function division model is composed of an objective function and a constraint condition, the objective function takes maximizing traffic efficiency and minimizing safety risk as the goal; the traffic efficiency is represented by the traffic capacity coefficient, and the safety risk is represented by the mixed coefficient, the traffic capacity coefficient is the ratio of the sum of the actual traffic capacity of each lane to the sum of the basic traffic capacity of each lane; The mixed coefficient is obtained by multiplying the passenger car mixed coefficient and the truck mixed coefficient, and dividing the product by the maximum value of the product of the passenger car mixed coefficient and the truck mixed coefficient; the objective function sums and numerically scales the traffic capacity coefficient constraint term and the mixed coefficient constraint term in a weighted form; Single lane actual traffic capacity is represented as follows: ； （1） ； （2） ； （3） ； （4） ； （5） In the formula, is the basic traffic capacity, corresponding to the maximum traffic capacity of the five-level service level under the corresponding speed limit condition, is the actual traffic capacity of a single lane, is the minimum safe headway, is the traffic composition correction factor for road traffic capacity, is the driver characteristic correction factor for road traffic capacity, with a value between 0.95 and 1, is the roadside interference correction factor, with a value of 1 for expressways, is the proportion of various vehicle types, is the vehicle conversion coefficient of the vehicle type , is the right side interaction factor, is the left side interaction factor, respectively represent the passenger car flow of the right lane, the passenger car flow of the left lane, the truck flow of the right lane, and the truck flow of the left lane, respectively represent the average lane changing frequency of passenger cars and the average lane changing frequency of trucks; The greater the value of the road traffic capacity coefficient t, the higher the road traffic efficiency, which is represented as: ； （6） In the formula, respectively represent the actual capacity of a single lane for passenger cars, freight cars, and mixed lanes; N respectively represent the number of passenger car lanes, the number of mixed lanes, the number of freight car lanes, and the total number of lanes, take the value of 1, 0 or -1, respectively indicating Section highway lane Belongs to passenger car lane, mixed lane and freight car lane; , respectively represent the passenger vehicle mixing coefficient and the truck mixing coefficient on the mixed lane, and the formulas are as follows: ； （7） ； （8） ； （9） In the formula, is the actual traffic demand; represents the passenger car ratio in the actual traffic demand, is the maximum value of the product of the passenger car mixing coefficient and the truck mixing coefficient, at which the mixing degree is the highest and the safety is the worst; respectively represent the actual traffic volume of passenger cars on the passenger car-only lane and the actual traffic volume of trucks on the truck-only lane; represents the mixing coefficient; The objective function expression is as follows: ； (10) In the formula, respectively represent the weight of the traffic capacity constraint term and the weight of the mixing coefficient constraint term, e is a natural constant, represents the exponentialization of the negative value of the road traffic capacity coefficient t as the inhibition weight of the mixing term, represents the exponentialization of the negative value of the mixing coefficient as the inhibition weight of the mixing term. S3. The lane function division model is solved by using a genetic algorithm, and an optimal or suboptimal lane function division scheme is output; S4. The lane function division scheme is dynamically published to the roadside facility through the lane-level information publishing facility for execution or reminder.

2. The method of claim 1, wherein, In step S2, the actual traffic capacity of a single lane is obtained by modifying the basic traffic capacity, and the modification term includes traffic composition modification, driver characteristic modification, roadside disturbance modification, and interaction influence modification from left and right adjacent lanes, the traffic composition modification includes the proportion of each vehicle type and the corresponding conversion coefficient, and the interaction influence of the left and right adjacent lanes is represented by a factor representing the lane changing frequency of the vehicles in the adjacent lanes to the lane, to reflect the influence of lane changing behavior on the traffic capacity of the lane.

3. The method of claim 1, wherein, The mixed coefficient is determined by the following steps: calculating the actual proportion of passenger cars and trucks on the mixed lane, calculating the product of the passenger car mixed coefficient and the truck mixed coefficient based on the actual proportion of passenger cars and trucks, and normalizing the product with the maximum product that can occur, the obtained value is the mixed coefficient, the larger the value, the higher the mixing degree and the greater the conflict risk; In the target function setting process, a positive exponential structure is adopted for the traffic capacity coefficient term to reward the increase of traffic capacity, and a negative exponential structure is adopted for the mixed coefficient term to punish the increase of mixed degree, and after the two terms are added with a predetermined weight, logarithmic transformation or other numerical scaling is performed to eliminate the dimension difference, and the weight is set based on the traffic management goal to realize the trade-off between efficiency and safety.

4. The method of claim 1, wherein the dynamic division of the highway lane function is characterized by, The constraint conditions include: (a) Lane number constraint: the total number of special lanes should not exceed the total number of lanes designed for the road section, and at least one mixed lane should be set; (b) Lane position constraint: in the longitudinal cross section, the spatial order of the three types of lanes from inside to outside should be passenger car lane, mixed lane and truck lane; (c) Traffic capacity constraint: the actual traffic capacity of each type of lane should be greater than the actual traffic demand allocated to the lane, and the actual total traffic capacity of the road should be greater than the total traffic demand.

5. The method of claim 1, wherein the dynamic division of the highway lane function is characterized by, The solving process of step S3 is as follows: S301. Randomly generate an initial scheme of lane function division, and calculate an initial scheme objective function value; S302. Use a genetic algorithm to find a coding scheme to code individuals in a population, and select floating-point number coding or binary coding; S303. Use a multi-peak function value as the fitness of an individual, and calculate the fitness of each individual in the population; the population is a set containing several lane function division schemes, and the individual is each lane function division scheme in the set; S304. Select individuals participating in reproduction according to the fitness; S305. Perform crossover and mutation operations on the selected individuals to generate offspring as new lane function division schemes; S306. Iteratively calculate the objective function until a preset convergence condition is reached or a preset maximum number of iterations is reached, and return the optimal or suboptimal lane function division scheme.

6. The method of claim 1, wherein the dynamic division of the highway lane function is characterized by, In step S4, at least one set of lane-level electronic sign or variable message board is set every 2-3 kilometers along the expressway, and the current lane function division scheme, open / close state and lane type are broadcast in the form of graphics or text to prompt the driver or roadside control unit.

7. A management system for dynamically dividing a function of a highway lane based on the method for dynamically dividing a function of a highway lane according to claim 1, characterized by, It comprises: a data acquisition module for receiving real-time expressway lane traffic flow, vehicle composition and lane changing frequency data, and determining the basic type and combination type of the lane setting, the basic type of the lane including passenger car exclusive lane, truck exclusive lane and mixed lane; a lane function division model module for constructing a lane function division model based on the basic type and combination type of the lane, the lane function division model being composed of an objective function and a constraint condition, the objective function aiming to maximize traffic efficiency and minimize safety risk; the traffic efficiency being represented by a traffic capacity coefficient, and the safety risk being represented by a mixing coefficient; the traffic capacity coefficient being the ratio of the sum of the actual traffic capacity of each lane to the sum of the basic traffic capacity of each lane; the mixing coefficient being obtained by multiplying the passenger car mixing coefficient and the truck mixing coefficient, and dividing the product by the maximum value of the product of the passenger car mixing coefficient and the truck mixing coefficient; the objective function summing and scaling the traffic capacity coefficient constraint term and the mixing coefficient constraint term in a weighted form; an optimization solving module for solving the lane function division model using a genetic algorithm, and outputting the optimal or suboptimal lane function division scheme; a publishing and executing module for dynamically publishing the division scheme to roadside facilities through lane-level information publishing facilities for execution or prompting.

8. The management system for dynamically dividing a highway lane function according to claim 7, wherein The lane function division model module and the optimization solving module are realized in software form and deployed on a server or an edge computing device, and the system further comprises a database module for storing historical traffic data and lane function division scheme effect data to support online learning and offline verification of lane function division model parameters.

Images