A vehicle speed guidance method in highway tunnels based on dynamic road segmentation

Abstract

The present invention belongs to the field of intelligent transportation, and discloses a method for guiding vehicle speed in highway tunnels based on dynamic road segment division, including: obtaining CAVs information; dividing basic road segments according to the location distribution of CAVs; designing the number of clusters, that is, the number of speed-limited road segments; setting the optimal merging error function; speed-limited road segment division strategy based on hierarchical clustering; optimal speed-limited road segment division. The present invention can make the speed change of vehicles smoother and more fluid, can cope with real-time changing traffic flow, and can also reduce excessive repetition or similar calculation processes through reasonable road segment merging strategies. The idea of ​​dynamically dividing road segments in the present invention can effectively reduce the random effects brought about by the spatiotemporal variation characteristics of traffic flow. Moreover, the present invention takes into account the location distribution of CAVs when dividing road segments, which is conducive to the guiding role of CAVs on HVs. In summary, the present invention can improve the safety and efficiency of traffic in highway tunnels.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent transportation, and in particular relates to a vehicle speed guidance method in a highway tunnel based on dynamic road segment division. Background Art

[0002] In recent years, highway tunnels have continued to expand. However, the rapid growth in the number of vehicles has far outstripped the pace of highway tunnel construction, resulting in tunnel capacity being unable to meet the growing traffic demand. This, in turn, has led to increasingly prominent tunnel traffic congestion and safety issues. Furthermore, due to the unique traffic characteristics of tunnels, including limited capacity, restricted lane changes, and restricted entrances and exits, traffic congestion can trigger a series of traffic accidents, resulting in significant casualties and property damage. Therefore, considering how to improve highway tunnel traffic efficiency and reduce the occurrence of safety accidents is of great practical value.

[0003] To mitigate congestion and safety issues in highway tunnels, measures such as increasing road infrastructure construction and implementing advanced intelligent transportation technologies are commonly adopted. Compared to traditional infrastructure development, ITS can provide timely and rapid management of traffic participants based on real-time traffic conditions, offering advantages such as low cost, flexibility, efficiency, real-time capabilities, and intelligence. ITS, centered around intelligent connected vehicles (CAVs) and vehicle-road collaboration, improves road network utilization and road safety through real-time data exchange and intelligent algorithm optimization. CAVs integrate advanced sensors, control systems, actuators, and communication and networking technologies, effectively addressing traffic congestion and safety challenges in highway tunnels through information interaction. It has been demonstrated that when some vehicles in a tunnel area are equipped with network connectivity, tunnel efficiency is significantly improved. Real-time data from CAVs in highway tunnels provides effective data support and a foundation for intelligent decision-making for ITS to optimize tunnel traffic flow.

[0004] By reviewing relevant patents and papers, it is found that many scholars have conducted in-depth exploration of the vehicle speed guidance problem from different perspectives and have achieved many remarkable results. Despite this, most of the current research still focuses on fixed bottleneck scenarios (such as lane reduction areas), and there is still a lack of research on vehicle speed guidance under moving bottlenecks in highway tunnels. At the same time, existing vehicle speed guidance methods are usually limited to the management of a single speed limit segment or the management of static equal-length speed limit segments, and lack the dynamic adjustment of the length of the speed limit area to respond to changes in real-time road conditions. In order to more effectively alleviate the congestion problem caused by complex traffic scenarios, it is necessary to propose a more flexible and comprehensive method to optimize traffic flow control, in order to ensure smooth and safe traffic flow in a wider range of road conditions that change in real time. The literature search did not find any research on vehicle speed guidance methods that consider dynamic road segment division. Summary of the Invention

[0005] In light of this, the present invention aims to provide a method for speed guidance within highway tunnels based on dynamic road segmentation. This method aims to address the limitations of existing methods, which are unsuitable for speed guidance within bottleneck highway tunnels, and lack the ability to dynamically adjust the length of speed-limited areas to respond to real-time road conditions.

[0006] The present invention provides a method for guiding vehicle speed in a highway tunnel based on dynamic road segment division, comprising the following steps:

[0007] S1. Obtain the location distribution information of CAVs within 0.3 to 2.8 km upstream of the mobile bottleneck at time t, and divide the speed-limited road section into l basic sections based on the CAV location distribution information;

[0008] S2. Calculate the average vehicle speed and traffic density within each basic road segment based on the collected CAVs information;

[0009] S3. Optimize the merging strategy using the optimal merging error function to merge adjacent basic road sections into speed-limited road sections;

[0010] S4. Input the traffic information of the new road section after the merger into the METANET model, set the update cycle T of the METANET model, and calibrate the parameters of the METANET model according to the real-time road section information;

[0011] S5. Set the prediction period to M p , use the calibrated METANET model to calculate M p * The traffic status prediction value at each step within time T, and the optimization objective function is designed based on the current traffic status and the predicted traffic status;

[0012] S6. Set the road section constraints and use the optimization algorithm to optimize the optimization objective function designed in step S5. When the minimum value of the objective function is obtained, the optimal speed limit combination upstream and downstream of the tunnel mobile bottleneck can be obtained;

[0013] S7. Send the optimal speed limit value to CAVs on each road section for speed guidance;

[0014] Set the control period to M c , from the current moment to the next M c * Within the time of T, the speed limit value is sent to the CAVs on each speed-limited road section.

[0015] Furthermore, in step S2, the calculation expressions for the average vehicle speed and traffic density in each basic road section are as follows:

[0016]

[0017]

[0018] Where, is the average speed of the basic road segment i; N is the number of vehicles on the road segment i; v n is the instantaneous speed of the vehicle; is the traffic density of road section i; X i is the length of the road segment.

[0019] Furthermore, in step S3, after the basic road sections are merged into speed-limited road sections, the number of speed-limited road sections is 5, and the length of each speed-limited road section is not less than 300m.

[0020] Furthermore, in step S3, the expression of the optimal combined error function is as follows:

[0021]

[0022] Where, k=5 represents the final divided control section; [m,n] represents the division from the mth basic section to the nth basic section; v j represents the average road speed of the jth basic road section; ρ j represents the traffic density of the jth basic road segment; The average speed of the final divided control section of the i-th segment, that is, the average speed of the section from the m-th basic section to the n-th basic section; It represents the traffic density of the control section finally divided in the i-th section.

[0023] Furthermore, in step S4, the METANET model is as follows:

[0024] q m (k) = ρ m (k).v m (k).λ m

[0025]

[0026]

[0027]

[0028] Where q m (k) is the average flow rate of the mth road at the kth time, veh / h; ρ m (k) is the average density of the mth road at the kth time, veh / km; v m (k) is the average speed of the mth road at the kth time, km / h; v freerepresents the free flow speed, km / h; ρ critical represents the critical density, veh / km; T represents the METANET model prediction update cycle time, h; τ is the traffic state change reaction time, h; L m is the length of the mth road section, km; λ m is the number of lanes in the mth road section; μ is the basic parameter in the expected speed equation of the METANET model; κ is the positive compensation coefficient; η is the sensitivity coefficient.

[0029] Furthermore, step S5 includes the following sub-steps:

[0030] S5.1 Set the prediction period to M p , use the calibrated METANET model to calculate M p * Traffic status prediction value at each step within time T;

[0031] S5.2 Design optimization objectives based on current and predicted traffic conditions;

[0032] The optimization objectives are the total travel time TTT, the total travel distance TTD, and the total speed difference TSD, and the calculation formulas are as follows:

[0033]

[0034]

[0035]

[0036] Where, M represents the total number of road sections; M p represents the improved METANET model prediction period; T represents the improved METANTE model update period; λ m Indicates the number of lanes in the mth section of road; L m represents the length of the mth road section; ρ METANET (m,k) represents the traffic density of the mth road section at the kth time step; v METANET (m,k) represents the traffic speed of the mth road at the kth time step predicted by the METANET model; v vsl (m,k) represents the speed limit of the mth road at the kth time; γ represents the road CAVs penetration rate; v free represents the free flow speed;

[0037] S5.3 setting the objective function according to the selected optimization goal;

[0038]

[0039] Where F(z) represents the overall objective function; α1 represents the weight coefficient of the total travel time TTT; α2 represents the weight coefficient of the total travel distance TTD; and α3 represents the weight coefficient of the total speed difference TSD.

[0040] Furthermore, in step S6, the road section constraint condition includes a speed limit constraint condition and a tunnel constraint condition;

[0041] Speed ​​limit constraints:

[0042] V min,diff ≤|v vsl (m,k)-v vsl (m,k+1)|≤V max,diff

[0043] V min,diff ≤|v vsl (m,k)-v vsl (m+1,k)|≤V max,diff

[0044] Where, v vsl (m,k) represents the speed limit calculated at time k on the mth road; v vsl (m+1,k) represents the speed limit calculated at time k on the m+1th road; V max,diff Indicates the maximum speed limit difference; V min,diff Indicates the minimum speed limit difference;

[0045] Tunnel constraints:

[0046] v low ≤v vsl (m,k)≤v high

[0047] 0≤Q(m,k)≤Q max

[0048] Where Q(m,k) represents the traffic flow of the mth road section at the kth moment; Q max Indicates the maximum throughput; v low Indicates the minimum speed limit; v high Indicates the maximum speed limit.

[0049] Furthermore, the optimization algorithm in step S6 is an artificial fish swarm algorithm.

[0050] Beneficial effects:

[0051] The present invention is different from the traditional variable speed limit control of a single speed limit section. The control strategy of multiple sections can make the speed change of the vehicle smoother. Furthermore, the present invention is also different from the variable speed limit control of static equal-length multiple speed limit sections. The dynamic division of the section length can not only cope with the real-time changing traffic flow, but also reduce excessive repetition or similar calculation processes through a reasonable section merging strategy. The idea of ​​dynamically dividing the sections in the present invention can effectively reduce the random effects brought by the spatiotemporal variation characteristics of the traffic flow. Moreover, the location distribution of CAVs is taken into account when dividing the sections. Such a section division makes it more conducive to the guiding role of CAVs on HVs when dealing with mixed traffic problems. In summary, the present invention can improve the safety and efficiency of traffic in highway tunnels.

[0052] Other advantages, objects, and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art upon examination of the following description or may be learned from practice of the present invention. The objects and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 A brief process for dynamically dividing road sections;

[0054] Figure 2 Demonstration results for road segmentation;

[0055] Figure 3 Schematic diagram of the spatiotemporal changes of the METANET model for two-lane tunnel traffic;

[0056] Figure 4 This is the flow chart of the AFSA-based tunnel multi-section speed guidance optimization strategy. DETAILED DESCRIPTION

[0057] To make the technical solutions, advantages, and purposes of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the described embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0058] This embodiment provides a method for guiding vehicle speed in a highway tunnel based on dynamic road segmentation. The method employs the dynamic road segmentation method of the present invention, uses a METANET model for vehicle speed prediction, uses an MPC control strategy for control, and uses an artificial fish swarm algorithm for target optimization. Specifically, the method may include the following steps:

[0059] S1. Obtain the location distribution information of CAVs within 0.3 to 2.8 km upstream of the mobile bottleneck at time t, and divide the speed-limited road section into l basic sections based on the CAV location distribution information;

[0060] The road segment set is: R = {r1, r2, r3, ... r l};

[0061] S2. Calculate the average vehicle speed and traffic density within each basic road segment based on the collected CAVs information;

[0062] The calculation expressions for the average vehicle speed and traffic density in each basic road section are as follows:

[0063]

[0064]

[0065] Where, is the average speed of the basic road segment i; N is the number of vehicles on the road segment i; v n is the instantaneous speed of the vehicle; is the traffic density of road section i; X i is the length of the road segment.

[0066] S3. Optimize the merging strategy using the optimal merging error function to merge adjacent basic road sections into speed-limited road sections;

[0067] In step S3, after the basic road sections are merged into speed-limited road sections, the number of speed-limited road sections is 5, and the length of each speed-limited road section is not less than 300m;

[0068] The traffic parameters of the merged road section are judged and the merging strategy is optimized to ensure that the speed and density errors are minimized. The expression of the optimal merging error function is as follows:

[0069]

[0070] Where, k=5 represents the final divided control section; [m,n] represents the division from the mth basic section to the nth basic section; v j represents the average road speed of the jth basic road section; ρ j represents the traffic density of the jth basic road segment; The average speed of the final divided control section of the i-th segment, that is, the average speed of the section from the m-th basic section to the n-th basic section; represents the traffic density of the control section finally divided in the i-th section;

[0071] The final optimal road segment division method is obtained. The brief division process and result demonstration are as follows: Figure 1and Figure 2 shown.

[0072] S4. Input the traffic information of the new road section after merging into the METANET model, set the update cycle T of the METANET model, and calibrate the parameters of the METANET model according to the real-time road section information. The spatiotemporal changes of the METANET model are as follows: Figure 3 As shown;

[0073] The METANET model is as follows:

[0074] q m (k) = ρ m (k).v m (k).λ m

[0075]

[0076]

[0077]

[0078] The meanings of the various parameters in the METANET model are shown in the following table:

[0079]

[0080] S5. Set the prediction period to M p , use the calibrated METANET model to calculate M p * The traffic status prediction value at each step within time T, and the optimization objective function is designed based on the current traffic status and the predicted traffic status;

[0081] S5.1 Set the prediction period to M p , use the calibrated METANET model to calculate M p * Traffic status prediction value at each step within time T;

[0082] S5.2 Design optimization objectives based on current and predicted traffic conditions;

[0083] The optimization objectives are total travel time (TTT), total travel distance (TTD), and total speed difference (TSD), and the calculation formulas are as follows:

[0084]

[0085]

[0086]

[0087] Where, M represents the total number of road sections; M p represents the improved METANET model prediction period; T represents the improved METANTE model update period; λ m Indicates the number of lanes in the mth section of road; L m represents the length of the mth road section; ρ METANET (m,k) represents the traffic density of the mth road section at the kth time step; v METANET (m,k) represents the traffic speed of the mth road at the kth time step predicted by the METANET model; v vsl (m,k) represents the speed limit of the mth road at the kth time; γ represents the road CAVs penetration rate; v free represents the free flow speed;

[0088] S5.3 Set a reasonable objective function based on the selected optimization goal and its practical meaning. The present invention sets it as follows:

[0089]

[0090] Where F(z) represents the overall objective function; α1 represents the weight coefficient for the total travel time (TTT); α2 represents the weight coefficient for the total travel distance (TTD); and α3 represents the weight coefficient for the total speed difference (TSD). Generally, when α2 = -1, α1 = [65, 80] achieves saturation in the optimization rates for both total travel time and throughput. Therefore, in this paper, α1 = 65 and α2 = -1 are used. Furthermore, since traffic efficiency is the primary consideration, α3 is set to -1.

[0091] S6. Set the road section constraints and use the optimization algorithm to optimize the optimization objective function designed in step S5. When the minimum value of the objective function is obtained, the optimal speed limit combination upstream and downstream of the tunnel mobile bottleneck can be obtained;

[0092] Road section constraints include speed limit constraints and tunnel constraints;

[0093] If the speed limit differs too much from the current road speed, it will lead to increased traffic congestion and easily cause safety accidents. If the speed limit differs too little from the current road speed, the speed limit will not be effective and will not help alleviate traffic congestion. Therefore, the calculated speed limit must meet certain constraints, as shown below:

[0094] V min,diff ≤|v vsl (m,k)-v vsl (m,k+1)|≤V max,diff

[0095] V min,diff ≤|v vsl (m,k)-v vsl (m+1,k)|≤V max,diff

[0096] Where, v vsl (m,k) represents the speed limit calculated at time k on the mth road; v vsl (m+1,k) represents the speed limit calculated at time k on the m+1th road; V max,diff Indicates the maximum speed limit difference, and its absolute difference should not be too large; V min,diff The minimum speed limit difference should not be too small. Therefore, in order to ensure the stability of traffic flow, V is set here. min,diff =5, V max,diff =20.

[0097] At the same time, considering that there are speed limits inside and outside the highway tunnel, the speed limit value within the control period should not exceed the maximum speed limit value, nor should it be lower than the minimum speed limit value. Therefore, the tunnel constraints are set as follows:

[0098] v low ≤v vsl (m,k)≤v high

[0099] 0≤Q(m,k)≤Q max

[0100] Where Q(m,k) represents the traffic flow of the mth road section at the kth moment; Q max Indicates the maximum throughput; v low Indicates the minimum speed limit, here the minimum speed limit is 30km / h; v high Indicates the maximum speed limit, which is 80km / h here.

[0101] like Figure 4 As shown, the optimization algorithm in step S6 of this embodiment is the artificial fish swarm algorithm, and the steps are as follows:

[0102] Step 1: Define the parameters of the artificial fish swarm algorithm, including the number of fish; the fish field of view, which is set to 10 km / h here; the trial step size is 5 km / h; the number of attempts is 5; the maximum number of iterations is 100; and the crowding factor is designed to be a certain proportion of the number of fish.

[0103] Step 2: Initialize the artificial fish swarm position, which represents a multi-section speed limit combination. The absolute speed difference between adjacent sections must be greater than 5km / h and less than 20km / h; the absolute speed difference between adjacent time periods must be greater than 5km / h and less than 20km / h.

[0104] Step 3: Calculate the objective function value of each initialized artificial fish based on the overall objective function;

[0105] Step 4: Execute an iterative process, moving based on foraging, flocking, and tailgating behaviors. It can also determine congestion to find a new location and then update the location, finding a better solution in each iteration.

[0106] Step 5: Perform multiple iterations until the maximum number of iterations is reached and the global optimal solution of the final iteration is used as the optimal speed limit combination for multiple sections.

[0107] S7. Send the optimal speed limit value to CAVs on each road section for speed guidance;

[0108] Set the control period to M c , from the current moment to the next M c * Within the time of T, the speed limit value is sent to the CAVs on each speed-limited road section.

[0109] It is hereby stated that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the purpose and scope of the technical solutions, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for guiding vehicle speed in a highway tunnel based on dynamic road segment division, characterized in that: The following steps are involved: S1. Obtain the location distribution information of CAVs within 0.3 to 2.8 km upstream of the mobile bottleneck at time t, and divide the speed-limited road section into l basic sections based on the CAV location distribution information; S2. Calculate the average vehicle speed and traffic density within each basic road segment based on the collected CAVs information; S3. Optimize the merging strategy using the optimal merging error function to merge adjacent basic road sections into speed-limited road sections; In step S3, after the basic road sections are merged into speed-limited road sections, the number of speed-limited road sections is 5, and the length of each speed-limited road section is not less than 300m; In step S3, the expression of the optimal combined error function is as follows: Where, k=5 represents the number of speed-limited sections finally divided; [m,n] represents the division from the mth basic section to the nth basic section; v j represents the average road speed of the jth basic road segment; ρ j represents the traffic density of the jth basic road segment; The average speed of the final speed-limited road section of the i-th segment, that is, the average speed of the road section from the m-th basic road section to the n-th basic road section; represents the traffic density of the speed limit section finally divided in the i-th section; S4. Input the traffic information of the new road section after the merger into the METANET model, set the update cycle T of the METANET model, and calibrate the parameters of the METANET model according to the real-time road section information; S5. Set the prediction period to M p , use the calibrated METANET model to calculate M p * The traffic status prediction value at each step within time T, and the optimization objective function is designed based on the current traffic status and the predicted traffic status; S6. Set the road section constraints and use the optimization algorithm to optimize the optimization objective function designed in step S5. When the minimum value of the objective function is obtained, the optimal speed limit combination upstream and downstream of the tunnel mobile bottleneck can be obtained; S7. Send the optimal speed limit value to CAVs on each road section for speed guidance; Set the control period to M c , from the current moment to the next M c * Within the time of T, the speed limit value is sent to the CAVs on each speed-limited road section.

2. The method for guiding vehicle speed in a highway tunnel based on dynamic road segment division according to claim 1, characterized in that: In step S2, the calculation expressions of the average vehicle speed and traffic density in each basic road section are as follows: Where, is the average speed of the basic road segment i′; N is the number of vehicles on the road segment i′; v n is the instantaneous speed of the vehicle; is the traffic density of road section i′; X i′ is the length of the road segment.

3. The method for guiding vehicle speed in a highway tunnel based on dynamic road segment division according to claim 2, characterized in that: In step S4, the METANET model is as follows: q m (k)=ρ m (k).v m (k).λ m Where q m (k) is the average flow rate of the mth road at the kth time, veh / h; ρ m (k) is the average density of the mth road at the kth time, veh / km; v m (k) is the average speed of the mth road at the kth time, km / h; v free represents the free flow speed, km / h; ρ critical represents the critical density, veh / km; T represents the METANET model prediction update cycle time, h; τ is the traffic state change reaction time, h; L m is the length of the mth road section, km; λ m is the number of lanes in the mth road section; μ is the basic parameter in the expected speed equation of the METANET model; κ is the positive compensation coefficient; η is the sensitivity coefficient.

4. The method for guiding vehicle speed in a highway tunnel based on dynamic road segment division according to claim 3, characterized in that: The step S5 includes the following sub-steps: S5.1 Set the prediction period to M p , use the calibrated METANET model to calculate M p * Traffic status prediction value at each step within time T; S5.2 Design optimization objectives based on current and predicted traffic conditions; The optimization objectives are the total travel time TTT, the total travel distance TTD, and the total speed difference TSD, and the calculation formulas are as follows: Where, M represents the total number of road sections; M p represents the improved METANET model prediction period; T represents the improved METANTE model update period; λ m Indicates the number of lanes in the mth section of road; L m represents the length of the mth road section; ρ METANET (m,k) represents the traffic density of the mth road section at the kth time step; v METANET (m,k) represents the traffic speed of the mth road at the kth time step predicted by the METANET model; v vsl (m,k) represents the speed limit of the mth road segment at the kth time; γ represents the road CAVs penetration rate; S5.3 setting the objective function according to the selected optimization goal; Where F(z) represents the overall objective function; α1 represents the weight coefficient of the total travel time TTT; α2 represents the weight coefficient of the total travel distance TTD; and α3 represents the weight coefficient of the total speed difference TSD.

5. The method for guiding vehicle speed in a highway tunnel based on dynamic road segment division according to claim 4, characterized in that: In step S6, the road section constraint conditions include speed limit constraint conditions and tunnel constraint conditions; Speed ​​limit constraints: V min,diff ≤|v vsl (m,k)-v vsl (m,k+1)|≤V max,diff V min,diff ≤|v vsl (m,k)-v vsl (m+1,k)|≤V max,diff Where, v vsl (m,k) represents the speed limit calculated at time k on the mth road; v vsl (m+1,k) represents the speed limit calculated at time k on the m+1th road; V max,diff Indicates the maximum speed limit difference; V min,diff Indicates the minimum speed limit difference; Tunnel constraints: in low ≤in vsl (m,k)≤v high 0≤Q(m,k)≤Q max Where Q(m,k) represents the traffic flow of the mth road section at the kth moment; Q max Indicates the maximum throughput; v low Indicates the minimum speed limit; v high Indicates the maximum speed limit.

6. The method for guiding vehicle speed in a highway tunnel based on dynamic road segment division according to claim 5, characterized in that: The optimization algorithm in step S6 is an artificial fish swarm algorithm.

Images