Urban road network left turn path coordination control method considering phase sequence optimization

Abstract

The invention belongs to the technical field of traffic engineering, and discloses an urban road network left-turn path coordination control method considering phase sequence optimization, which comprises the following steps: introducing a binary value variable, and optimizing the phase sequence of a left-turn intersection; the method comprises the following steps: improving a traditional AM-BAND variable bandwidth coordination control model, and constructing a green wave variable bandwidth coordination control model of a facing left turn path; and improving a classical fruit fly algorithm, and solving the green wave variable bandwidth coordination control model by using the improved fruit fly algorithm to obtain a left turn coordination control result. According to the method, the influence on road network background traffic is relatively small, compared with a classic AM-BAND model and a classical Synchro model, the vehicle average delays of all vehicles at the road network intersection are reduced by 2.43% and 10.35% respectively, and the effectiveness and applicability of the research model in left-turn path coordination control are verified.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of traffic engineering, and particularly relates to a left-turn path coordinated control method for urban road network considering phase sequence optimization. BACKGROUND

[0002] Due to objective factors such as urban industrial layout and road network distribution, a large number of traffic paths in the road network are composed of multiple rather than single trunks, especially the left-turn traffic flow in the path, which is prone to cause intersection congestion due to reasons such as turning radius and intersection widening length, greatly affecting the smoothness of urban traffic. By reasonably adjusting the phase sequence of the intersection, the green wave coordinated control effect of the left-turn path can be effectively improved.

[0003] Optimizing the signal phase sequence of the intersection is an effective method to improve green wave coordinated control and increase traffic capacity. Domestic and foreign scholars mainly focus on phase sequence adjustment and optimization methods in path signal control. By combining various strategies such as bidirectional green wave coordination and random signal system optimization, the path traffic efficiency is improved, the delay is reduced, and the green wave bandwidth is maximized. Xu Jieqiong et al. (Multi-period coordinated control optimization of urban trunk roads) designed a multi-period signal linkage scheme with the help of mixed clustering of trunk port flow, and used a multi-objective particle swarm algorithm to determine the switching time. Lin Li et al. (Ring-barrier phase-based arterial bus coordinated control) constructed an arterial phase sequence model considering bus operation based on the Ring-barrier double-ring structure. Zhang Chi (Bidirectional green wave setting method based on phase sequence adjustment) combined NENA phase and graphical means to propose a phase adjustment strategy to widen the bidirectional green wave bandwidth. Lu Shunda et al. (Optimization of bidirectional green wave coordinated control graphical method under asymmetric phase sequence mode) processed the bottleneck intersection, adjusted the phase sequence and difference, and gave a graphical coordination method to increase the green wave bandwidth.

[0004] Most of the above studies take straight traffic flow as the optimization object and focus on straight direction phase sequence optimization, ignoring the traffic scene when left-turn traffic flow dominates, which limits the adaptability of the model in actual application. In addition, most studies take straight traffic flow on arterial roads as the research object, and fail to consider the impact of the change of turning radius on path green wave coordinated control when vehicles turn at intersections. SUMMARY

[0005] The present application aims to solve the problems of the prior art and provides the following solutions:

[0006] A left-turn path coordinated control method for urban road network considering phase sequence optimization, comprising the following steps:

[0007] Introducing a binary-valued variable to optimize the phase sequence of the left-turn intersection;

[0008] The traditional AM-BAND variable bandwidth coordination control model is improved to construct a green wave variable bandwidth coordination control model for left-turn paths.

[0009] The classical fruit fly algorithm is improved, and the improved fruit fly algorithm is used to solve the green wave variable bandwidth coordination control model to obtain the left-turn coordination control result.

[0010] Preferably, the method for optimizing the phase sequence of the left-turn intersection comprises:

[0011] According to the left-turn phase mode, the uplink left-turn phase sequence is introduced in the inequality constraint of the coordination control model The uplink left-turn advance or lag is realized by solving the value of the uplink left-turn phase sequence in the bandwidth edge constraint.

[0012] For the intersection S n and the intersection S n+1 , the variable δ n and the variable The release mode of the two intersections is determined by solving the values of the variable δ n and the variable in the period constraint.

[0013] Preferably, the method for constructing the green wave variable bandwidth coordination control model comprises:

[0014] The objective function is constructed:

[0015]

[0016] b n =b n1 +b n2 ,

[0017]

[0018] Wherein, α n represents the weight coefficient of the uplink bandwidth of the intersection S n , represents the weight coefficient of the downlink bandwidth of the intersection S n , b n represents the green wave bandwidth of the uplink direction of the intersection S n , b n1 represents the first component of the green wave bandwidth of the uplink direction of the intersection S n , b n2 represents the second component of the green wave bandwidth of the uplink direction of the intersection S n , represents the green wave bandwidth of the downlink direction of the intersection S n , Indicates intersection S n The first component of the downlink green wave bandwidth, Indicates intersection S n The second component of the green wave bandwidth in the downlink direction, where n represents the nth intersection in the left-turn path and N represents the total number of intersections in the left-turn path as N;

[0019] Construct bandwidth edge constraints that consider phase sequence optimization, path speed constraints that consider left turns at intersections, acceleration constraints, path period constraints along the intersections under coordinated control, asymmetric bandwidth constraints, and other constraints.

[0020] The constructed constraints are sorted to obtain high-priority constraints, medium-priority constraints, and optional constraints.

[0021] Preferably, the bandwidth edge constraint considering phase sequence optimization includes:

[0022] Minimum bandwidth range constraint for processes at intersections in the path:

[0023]

[0024] ω i ≥b i2 i = n, n+2,

[0025]

[0026] Where, ω n+1 Indicates the intersection S in the uphill direction of the path. n+1 Green band b n+1 The interval between the left edge and the right edge of the adjacent red light. Indicates intersection S n+1 The red light is on for those going straight in the opposite direction. Indicates the intersection S in the downhill direction of the path. n+1 Green band The interval between the right edge and the left edge of the adjacent red light, ω i Indicates the intersection S in the uphill direction of the path. i Green band b i The interval between the left edge and the right edge of the adjacent red light, b i2 Indicates intersection S i The second component of the green wave bandwidth in the upward direction, Indicates the intersection S in the downhill direction of the path. i Green band The interval between the right edge and the adjacent left edge of the red light. Indicates intersection S i The first component of the green wave bandwidth in the downlink direction, where i represents the i-th intersection;

[0027] Maximum range constraint of process bandwidth edge value at straight intersections in the path:

[0028] ω n +b n1 ≤1-r Tn ,

[0029]

[0030] ω n+2 +b (n+1)1 ≤1-r Tn+2 -r n+2 ,

[0031]

[0032] Where, ω n Indicates the intersection S in the uphill direction of the path. n Green band b n The interval between the left edge and the right edge of the adjacent red light, rT n Indicates intersection S n During the red light period for those going straight uphill. Indicates the intersection S in the downhill direction of the path. n Green band The interval between the right edge and the adjacent left edge of the red light. Indicates intersection S n The red light is on for those going straight in the opposite direction. Indicates intersection S n Initial queue clearing time for the downlink, ω n+2 Indicates the intersection S in the uphill direction of the path. n+2 Green band b n+2 The interval between the left edge and the right edge of the adjacent red light, b (n+1)1 Indicates intersection S n+1 The first component of the green wave bandwidth in the uplink direction, r Tn+2 Indicates intersection S n+2 At the red light time for straight-ahead traffic, τ n+2 Indicates intersection S n+2 Initial queue clearing time for upward movement. Indicates the intersection S in the downhill direction of the path. n+2 Green band The interval between the right edge and the adjacent left edge of the red light. Indicates intersection S n+1 The second component of the downlink green wave bandwidth, Indicates intersection S n+2 The red light is on for those going straight in the opposite direction.

[0033] The maximum value range constraint of the left-turn intersection process bandwidth in the path:

[0034]

[0035] Wherein, τ n+1 represents the initial queuing emptying time of the uplink at the intersection S n+1 , r Ln+1 represents the red light time of the uplink left turn at the intersection S n+1 , g Ln+1 represents the green light time of the uplink left turn at the intersection S n+1 , represents the green light time of the downlink left turn at the intersection S n+1 .

[0036] Preferably, the path speed constraint considering the intersection left turn comprises:

[0037]

[0038]

[0039] Wherein, d n，n+1 represents the straight phase distance from the intersection S n to the intersection S n+1 , f n represents the lower limit of the uplink driving speed at the intersection S n , a represents the acceleration of the uplink vehicle when turning at the intersection, L n，n+1 represents the turning preparation phase distance from the intersection S n to the intersection S n+1 , z represents the reciprocal of the signal period length, t n，n+1 represents the uplink driving time of the vehicle from the intersection S n to the intersection S n+1 , e n represents the upper limit of the uplink driving speed at the intersection S n , represents the lower limit of the downlink driving speed at the intersection S n , represents the acceleration of the downlink vehicle when turning at the intersection, v represents the vehicle speed before entering the intersection, and r represents the turning radius when the vehicle turns, represents the downlink driving time of the vehicle from the intersection S n+1 to the intersection S n , represents the upper limit of the downlink driving speed at the intersection S n , n represents the upper limit of the uplink driving speed change at the intersection S n , d n+1，n+2Indicates intersection S n+1 To intersection S n+2 The straight-line phase distance, t n+1，n+2 Indicates that the vehicle is coming from intersection S n+1 To intersection S n+2 The uphill travel time, g n Indicates intersection S n The lower limit of the change in speed in the upward direction. Indicates intersection S n The upper limit of the change in speed in the downward direction. Indicates intersection S n+1 To intersection S n The distance of the straight-line phase, Indicates intersection S n+2 To intersection S n+1 The distance of the straight-line phase, Indicates that the vehicle is coming from intersection S n+2 To intersection S n+1 Downward travel time, Indicates intersection S n The lower limit of the change in speed in the downward direction.

[0040] Preferably, the acceleration constraint includes:

[0041]

[0042] Where μ represents the friction coefficient of the vehicle, g represents the gravitational acceleration, and sin(θ) represents the road slope.

[0043] Preferably, the periodic constraints at intersections along the path for coordinated control include:

[0044]

[0045] Where, r n+1 Indicates intersection S n Coordinate red light times on roads, g Ln Indicates intersection S n During the green light period for straight-ahead traffic. Indicates intersection S n At the time of the green light for going straight in the direction of traffic, δ i Indicates intersection S n Upward left turn 0-1 variable, g Li Indicates intersection S n The green light for turning left at the intersection. Indicates intersection S n Downward left turn 0-1 variable, Indicates intersection S n The green light time for turning left at the exit is m n,n+1integer adjustment factor satisfying the constraint condition between intersections S n and S n+1 .

[0046] Preferably, the asymmetric bandwidth constraint comprises:

[0047]

[0048] wherein N1 and N2 represent positive real numbers.

[0049] Preferably, the other constraint comprises:

[0050] Asymmetric constraint of unbalanced two-way traffic volume:

[0051]

[0052] wherein k n represents an unbalanced coefficient;

[0053] Period rationality constraint:

[0054]

[0055] wherein C max represents a maximum value of signal control period, C min represents a minimum value of signal control period;

[0056] Variable value constraint:

[0057]

[0058] m n,n+1 ,m n+1,n+2 integers,

[0059]

[0060] wherein m n+1，n+2 represents an integer adjustment factor satisfying the constraint condition between intersections S n+1 and S n+2 . n+2 δ n+2 represents an uplink left turn 0-1 variable, integers represent integers, and binary integers represent 0-1 integers.

[0061] Preferably, the method for solving the green wave variable bandwidth coordination control model comprises:

[0062] The geometric characteristics, road attributes, signal timing parameters and traffic demand of each intersection in the acquisition path are obtained, the Webster algorithm is used to solve the optimal cycle length of each independent intersection, and the maximum value among them is taken as the unified control cycle of the whole trunk;

[0063] According to the proportion of traffic volume of different directions in the intersection, the distribution proportion of green light time in the phase scheme design is determined;

[0064] Combined with the left-turn phase release mode of each intersection, the public signal cycle and the designed up-down green wave speed, the improved fruit fly algorithm is used to solve the green wave variable bandwidth coordination control model, and the phase sequence arrangement result of each intersection is obtained.

[0065] Based on the model solving result, the correlation between each variable and the phase offset in the time-distance diagram is analyzed, the phase difference between intersections is derived, and the corresponding time-distance diagram is drawn by using the phase difference, so that the left-turn coordination control result is obtained.

[0066] Compared with the prior art, the beneficial effects of the present application are:

[0067] The simulation results show that compared with the classic AM-BAND model and Synchro model, the green wave bandwidth of the path of the present application increases by 23.12% and 69.23% respectively; compared with the AM-BAND model, the path vehicle delay, average queue length and travel time of the present application are reduced by 9.74%, 18.46% and 5.66% respectively; compared with the classic Synchro model, the optimization effect is more significant, and the above indexes are reduced by 20.32%, 28.54% and 23.14% respectively. In addition, the influence of the present application on the road network background traffic is relatively small, and compared with the above two models, the vehicle delay of all vehicles at the intersection of the road network is reduced by 2.43% and 10.35% respectively. The effectiveness and applicability of the model in the left-turn path coordination control are verified. BRIEF DESCRIPTION OF DRAWINGS

[0068] In order to more clearly illustrate the technical solutions of the present application, the following briefly introduces the drawings needed to be used in the embodiments, and obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0069] Figure 1 The method flowchart of the embodiment of the present application;

[0070] Figure 2 The coordination flow control diagram of the embodiment of the present application;

[0071] Figure 3 The left-turn phase mode diagram of the embodiment of the present application;

[0072] Figure 4 Green wave time-distance diagram for left turn coordination control of the embodiment of the present application;

[0073] Figure 5 Intersection channelization diagram for the embodiment of the present application;

[0074] Figure 6 Time-space diagram for left turn path green wave coordination control of the present application in the embodiment of the present application;

[0075] Figure 7 Time-space diagram for left turn path green wave coordination control of the AM-BAND model in the embodiment of the present application;

[0076] Figure 8 Time-space diagram for left turn path green wave coordination control of the Synchro model in the embodiment of the present application;

[0077] Figure 9 Results diagram of simulation evaluation indexes of path coordination control of each model in the embodiment of the present application, wherein (a) is evaluation result of vehicle delay, (b) is evaluation result of path travel time, (c) is evaluation result of average number of stops of path, and (d) is evaluation result of average queue length of path.

[0078] Figure 10 Results diagram of simulation evaluation indexes of intersection coordination control of each model in the embodiment of the present application, wherein (a) is evaluation result of vehicle delay of each intersection, (b) is evaluation result of average queue length of each intersection, and (c) is evaluation result of average number of stops of each intersection. DETAILED DESCRIPTION

[0079] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0080] In order to make the above objectives, features and advantages of the present application more apparent, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0081] Current green wave signal coordination control models primarily optimize for through traffic on arterial roads, aiming to maximize their traffic efficiency. While through traffic on arterial roads typically constitutes a higher volume and should right-of-way with a green wave, at specific intersections where turning traffic demand is more significant, prioritizing these vehicles yields better overall benefits. In such cases, adhering to a straight-through priority strategy would reduce system efficiency, while coordinating turning traffic flow could significantly improve network efficiency. The following example illustrates a left-turn path formed by three signalized intersections (e.g.,...). Figure 2 The problem studied in this invention will be specifically illustrated as shown in the figure.

[0082] At intersection S n+1 Traditional green wave optimization strategies primarily coordinate traffic flow in the straight-ahead direction, enabling it to pass through the entire road continuously or with minimal stops, thereby improving traffic efficiency and achieving good traffic benefits. However, when S... n+1 The traffic demand for left turns is significantly higher than that for straight traffic, making the existing straight-through priority scheme no longer reasonable. In this case, adjusting the coordination target to left-turn traffic can significantly improve overall traffic efficiency. Therefore, this invention addresses the issue at intersection S... n+1 Left-turning traffic When implementing coordinated control, the left-turn release method at the intersection is considered, and phase sequence optimization is performed using a scheme selection approach to ensure maximum efficiency for left-turn traffic in the path. Currently, there are four overlapping phase release methods at intersections. This invention, based on these, considers scheme selection-based phase sequence optimization. By establishing a coordinated control model and introducing 0-1 variables, the phase sequence is optimized to determine the left-turn release method at the intersection, thereby maximizing the traffic efficiency of coordinated left-turn control in the path.

[0083] Furthermore, this invention takes the path in the road network as the research object. During the left turn at the intersection, there is a turning path, which results in a change in turning acceleration. This has a certain impact on the speed constraints in path coordination control, and there is a certain degree of difference from the speed constraints of traditional arterial roads.

[0084] Basic assumptions:

[0085] Due to the instability of traffic conditions along the route, the following basic assumptions are set for the established model:

[0086] (1) The traffic demand at each intersection along the route is in a stable, unsaturated state.

[0087] (2) The distance between adjacent intersections does not exceed 1000 meters, ensuring a high degree of correlation of traffic flow between road segments;

[0088] (3) At intersections where left turns are coordinated, the upstream left-turn traffic flow is significantly higher than the straight-through traffic flow, while the downstream right-turn traffic flow is controlled by the straight-through signal;

[0089] (4) The left-turn traffic volume of the non-coordinated left-turn intersection is obviously lower than the straight traffic volume, and the right-turn traffic volume has little impact on the main road traffic because it is not controlled by the signal light and has small flow;

[0090] (5) The study assumes that motor vehicles completely comply with traffic rules and does not consider the interference of bus stops, pedestrians and non-motor vehicles on path traffic flow;

[0091] (6) The intersection adopts a fixed-length signal timing scheme, and the next phase is started immediately after the end of the previous phase.

[0092] By reconstructing the traffic coordination optimization variable system between adjacent intersections, relying on the green wave time interval distribution diagram, the quantitative relationship between the coordinated control parameters and the vehicle running trajectory characteristics is systematically studied.

[0093] Embodiment one:

[0094] In this embodiment, as shown in Figure 1 , a left-turn path coordinated control method for urban road network considering phase sequence optimization includes the following steps:

[0095] S1. Introducing a binary-valued variable to optimize the phase sequence of the left-turn intersection.

[0096] In this embodiment, in order to improve the credibility of the model, the release mode of the left-turn phase needs to be studied first. According to the current green wave band coordinated control optimization theory, there are four direction modes for the left-turn signal of intersection coordinated flow, as shown in Figure 3 . By selecting the value of 0-1 variable, the optimization selection of different phase sequences at the intersection can be realized, so this embodiment adopts the phase sequence optimization mode of scheme selection, and the corresponding 0-1 variable value is obtained by solving the coordinated control model, so as to realize the optimization selection of the phase sequence of the intersection on the left-turn path, and provide a larger green wave passing range for the left-turn traffic flow on the path.

[0097] The method for optimizing the phase sequence of the left-turn intersection includes:

[0098] According to the left-turn phase mode, the uplink left-turn phase sequence is introduced in the inequality constraint of the coordinated control model. The value of the uplink left-turn phase sequence is solved in the bandwidth edge constraint to realize the advance or lag of the left-turn in the uplink direction; however, since the downlink direction is right-turn, it is usually not controlled by the signal light, and when there is a straight-right lane, the right-turn is controlled by the straight green light, so the variable has no effect on the right-turn signal phase. The change mode of the variable has an effect on the uplink left-turn signal phase, as shown in Table 1:

[0099] Table 1

[0100]

[0101] For intersection S n and intersection S n+1 , the release mode of the two intersections is determined by solving the values of variables δ n and variables In the cycle constraint, the release mode of the two intersections is determined by solving the values of variables δ n and variables When the values of variables δ n and variables are different, the release modes of the left-turn phases are different. The correspondence between the values of variables δ n and variables and the release modes is shown in Table 2:

[0102] Table 2

[0103]

[0104] By selecting the values of 0-1 variables, different phase sequences can be realized at the intersections, so the embodiment adopts a scheme selection type of phase sequence optimization, and the corresponding values of 0-1 variables are obtained by solving the model of coordinated control, so as to realize the optimization and selection of the phase sequence of the intersections on the left-turn path, and provide a larger green wave passing range for the left-turn traffic flow on the path.

[0105] S2. The traditional AM-BAND variable bandwidth coordinated control model is improved, and a green wave variable bandwidth coordinated control model facing the left-turn path is constructed.

[0106] In the embodiment, the method for constructing the green wave variable bandwidth coordinated control model includes:

[0107] The bandwidth values between adjacent intersections on the path are optimized with the maximum green wave bandwidth as the target. By introducing the green wave progress line, the continuity of the green wave band at the intersections is ensured, and the symmetric bandwidth constraint is cancelled, so as to improve the available time of the green time. For the green wave band in the uplink direction, it is divided into two independent components b n = b n1 +b n2 , which respectively represent the components of the green wave band in the uplink direction on the left and right sides of the progress line. Similarly, the green wave band in the downlink direction is also divided into two components, i.e. which respectively represent the components of the green wave band in the downlink direction on the left and right sides of the progress line. The objective function can be expressed as:

[0108]

[0109] Where max represents the maximum of the objective function, and α n represents the intersection S na weight coefficient of the uplink bandwidth, denotes the intersection S n a weight coefficient of the downlink bandwidth, a n and the value of b is the ratio of the total traffic flow of each turning direction of the uplink (downlink) to the total saturated traffic flow of each turning direction of the uplink (downlink), n denotes the intersection S n the green wave bandwidth of the uplink direction, b n1 denotes the intersection S n the first component of the green wave bandwidth of the uplink direction, b n2 denotes the intersection S n the second component of the green wave bandwidth of the uplink direction, denotes the intersection S n the green wave bandwidth of the downlink direction, denotes the intersection S n the first component of the green wave bandwidth of the downlink direction, denotes the intersection S n the second component of the green wave bandwidth of the downlink direction, n denotes the nth intersection in the left-turn path, and N denotes the total number of intersections in the left-turn path as N;

[0110] The bandwidth edge constraint considering phase sequence optimization, the path speed constraint considering intersection left turn, the acceleration constraint, the path along the intersection cycle constraint of coordinated control, the asymmetric bandwidth constraint and other constraints are constructed.

[0111] According to the analysis of the left-turn phase sequence release mode, the intersection S n+1 has four release modes, and the model derivation is performed by taking the release mode I as an example, and the rest of the direction modes are changed according to the left-turn binary variable, and then the path green wave time-distance diagram under the change of the release mode I is as shown in Figure 4 . Considering different modeling objects (the AM-BAND model takes the straight traffic flow as the modeling object, and the model in the present embodiment takes the uplink left-turn traffic flow and the downlink right-turn traffic flow as the modeling objects), the integer equality constraint condition and the inequality constraint condition in the AM-BAND model cannot be directly applied to the model in the present embodiment. Therefore, in order to establish the related constraint conditions (integer equality constraint and inequality constraint) suitable for the model in the present embodiment, the AM-BAND model needs to be re-derived.

[0112] (1) To ensure that both the left and right edges of the green wave band are within the green light time and to prevent the green wave band from intersecting with the red light time, and to determine the left turn phase sequence optimization variable value, corresponding constraints are made for the maximum and minimum values ​​of the bandwidth edge variable to ensure the feasibility of the model. The constraints are as follows: to ensure that the process bandwidth part in the signal cycle does not conflict with the red light time, to adjust the allocation of process bandwidth according to the left turn phase sequence, and to determine its minimum value range constraint. Compared with the traditional green wave bandwidth control model where the edge bandwidth is greater than 0, this embodiment considers the bandwidth edge constraints for cross-password edge bandwidth greater than 0, including the following:

[0113] Minimum bandwidth range constraint for processes at intersections in the path:

[0114]

[0115] ω i ≥b i2 i = n, n+2,

[0116]

[0117] Where, ω n+1 Indicates the intersection S in the uphill direction of the path. n+1 Green band b n+1 The interval between the left edge and the right edge of the adjacent red light. Indicates intersection S n+1 The red light is on for those going straight in the opposite direction. Indicates the intersection S in the downhill direction of the path. n+1 Green band The interval between the right edge and the left edge of the adjacent red light, ω i Indicates the intersection S in the uphill direction of the path. i Green band b i The interval between the left edge and the right edge of the adjacent red light, b i2 Indicates intersection S i The second component of the green wave bandwidth in the upward direction, Indicates the intersection S in the downhill direction of the path. i Green band The interval between the right edge and the adjacent left edge of the red light. Indicates intersection S i The first component of the green wave bandwidth in the downlink direction, where i represents the i-th intersection;

[0118] Maximum range constraint of process bandwidth edge value at straight intersections in the path:

[0119] ω n +b n1 ≤1-r Tn ,

[0120]

[0121] ω n+2 +b (n+1)1 ≤1-r Tn+2 -τ n+2 ,

[0122]

[0123] Where, ω n Indicates the intersection S in the uphill direction of the path. n Green band b n The interval between the left edge and the right edge of the adjacent red light, rT n Indicates intersection S n During the red light period for those going straight uphill. Indicates the intersection S in the downhill direction of the path. n Green band The interval between the right edge and the adjacent left edge of the red light. Indicates intersection S n The red light is on for those going straight in the opposite direction. Indicates intersection S n Initial queue clearing time for downstream traffic, ω n+2 Indicates the intersection S in the uphill direction of the path. n+2 Green band b n+2 The interval between the left edge and the right edge of the adjacent red light, b( n+1 )1 represents intersection S n+1 The first component of the green wave bandwidth in the uplink direction, rT n+2 Indicates intersection S n+2 At the red light time for straight-ahead traffic, τ n+2 Indicates intersection S n+2 Initial queue clearing time for upward movement. Indicates the intersection S in the downhill direction of the path. n+2 Green band The interval between the right edge and the adjacent left edge of the red light. Indicates intersection S n+1 The second component of the downlink green wave bandwidth, Indicates intersection S n+2 The red light is on for those going straight in the opposite direction.

[0124] When determining the maximum range of bandwidth for processes at left-turn intersections in the path, the left-turn phase order is also considered, ensuring that its maximum value is less than:

[0125]

[0126] Where, τ n+1Indicates intersection S n+1 At the initial queue clearing time for upward movement, r Ln+1 Indicates intersection S n+1 At the red light time for turning left on the uphill route, g Ln+1 Indicates intersection S n+1 The green light for turning left at the intersection. Indicates intersection S n+1 The green light time for turning left when going downhill.

[0127] (2) Consider the path speed constraint for left turns at intersections:

[0128] At intersections, changes in acceleration lead to fluctuations in driving speed. Due to the influence of the turning radius, the green wave band needs to be widened at intersections to ensure a certain left-turn capacity. Therefore, this differs from the speed constraints on main roads.

[0129]

[0130] Where, d n，n+1 Indicates intersection S n To intersection S n+1 The straight-line phase distance, f n Indicates intersection S n The lower limit of the speed limit for vehicles traveling in the uphill direction, 'a' represents the acceleration of a vehicle turning at an intersection in the uphill direction, and 'L' represents the speed limit for vehicles traveling in the uphill direction. n,n+1 Indicates intersection S n To intersection S n+1 The distance during the pre-turn preparation phase, z represents the reciprocal of the signal period duration, t n，n+1 Indicates that the vehicle is coming from intersection S n To intersection S n+1 The uphill travel time, e n Indicates intersection S n The maximum speed limit for travel in the upward direction. Indicates intersection S n The lower limit of the speed limit for travel in the downward direction. This represents the acceleration of a vehicle turning at the intersection in the downhill direction, where v represents the vehicle's speed before entering the intersection, and r represents the turning radius. Indicates that the vehicle is coming from intersection S n+1 To intersection S n Downward travel time, Indicates intersection S n Downward speed limit, h n Indicates intersection S n The upper limit of the change in speed in the upward direction, d n+1，n+2 Indicates intersection S n+1 To intersection S n+2The straight-line phase distance, t n+1，n+2 Indicates that the vehicle is coming from intersection S n+1 To intersection S n+2 The uphill travel time, g n Indicates intersection S n The lower limit of the change in speed in the upward direction. Indicates intersection S n The upper limit of the change in speed in the downward direction. Indicates intersection S n+1 To intersection S n The distance of the straight-line phase, Indicates intersection S n+2 To intersection S n+1 The distance of the straight-line phase, Indicates that the vehicle is coming from intersection S n+2 To intersection S n+1 Downward travel time, Indicates intersection S n The lower limit of the change in speed in the downward direction.

[0131] (3) Acceleration constraints:

[0132]

[0133] Where μ represents the friction coefficient of the vehicle, g represents the gravitational acceleration, and sin(θ) represents the road slope.

[0134] (4) Periodic constraints at intersections along the route for coordinated control:

[0135] Assuming all intersections on the path use a common cycle time, then intersection S n and S n+1 Phase difference between variables They are respectively expressed by the following formulas:

[0136] φ n，n+1 =0.5r Tn +ω n +t n,n+1 -τ n,n+1 -ω n+1 -0.5r n+1 ,

[0137]

[0138] Compared to traditional green wave coordination control methods, the phase node difference between the upper and lower intersections in this embodiment is 1.5 times the period, unlike the traditional model where the period is an integer multiple. Therefore, the following relationship must be satisfied:

[0139]

[0140] △ n = δ n g Ln -0.5g Ln - δ n g Ln +0.5g Ln ,

[0141] S n and S n+1 between the intersection S

[0142]

[0143] where r n+1 denotes the red light time on the coordinated road at intersection S n , g Ln denotes the uplink straight green light time at intersection S n , g n denotes the downlink straight green light time at intersection S i , δ n denotes the uplink left turn 0-1 variable at intersection S Li , g n denotes the uplink left turn green light time at intersection S n , δ n denotes the downlink left turn 0-1 variable at intersection S n，n+1 , m n denotes the downlink left turn green light time at intersection S n+1 , Δ n denotes the integer adjustment factor satisfying the constraint condition between intersection S n+1 and S n+2 , and Δ n+1,n+2 denotes the time interval between the uplink red light midpoint and the closest downlink red light midpoint.

[0144] Similarly, the variable phase difference between intersection S n+1 and S n+1 is calculated, and since the downlink direction is the straight direction of the coordinated intersection, the variable phase difference is respectively represented by the following formula:

[0145] φ n+1,n+2 =0.5r n+2 +ω n+2 +t Tn+1 -τ n+1 -ω n+2 -0.5r Ln+2 ,

[0146]

[0147] The above two formulas are combined with Then the intersection S can be calculated. n+1 and S n+2 The periodic constraints between them are as follows:

[0148]

[0149] Among them, g Ln+2 Indicates intersection S n+2 The green light for turning left at the intersection. Indicates intersection S n+2 At the time of the green light for turning left while going downhill, δ n+2 Indicates intersection S n+2 Upward left turn 0-1 variable, Indicates intersection S n+2 Downward left turn 0-1 variable, m n+2，n+1 Indicates intersection S n+1 and S n+2 Integer adjustment coefficients that satisfy the constraints between them.

[0150] (5) Asymmetric bandwidth constraint:

[0151] To ensure that the left and right components of the uplink and downlink green wave bands along the development path are not zero, i.e. to achieve asymmetric multi-bandwidth optimization, the following constraints are set:

[0152]

[0153] Where N1 and N2 represent any positive real numbers. The upper and lower limits of the green band uplink / downlink ratio are adjusted based on the values ​​of N1 and N2.

[0154] (6) Other constraints include:

[0155] Due to the asymmetric problem of uneven two-way traffic volume in this embodiment, an imbalance coefficient k is set. n To ensure that the asymmetric constraints for the imbalance of two-way traffic volume are as follows:

[0156]

[0157] Where, k n Indicates the imbalance coefficient;

[0158] To ensure the reasonableness of the model's cycle, cycle reasonableness constraints are set to determine the upper and lower limits of the cycle for each intersection:

[0159]

[0160] Among them, C max C represents the maximum value of the signal control period. min This indicates the minimum value of the signal control cycle;

[0161] Variable value constraints, indicating that these variables take values of non-negative numbers, integers, and 0 and 1, respectively:

[0162]

[0163] m n,n+1 , m n+1，n+2 integers,

[0164]

[0165] wherein m n+1,n+2 represents an integer adjustment coefficient satisfying a constraint condition between intersections S n+1 and S n+2 , δ n+2 represents an uplink left-turn 0-1 variable at intersection S n+2 , integers represent integers, and binary integers represent 0-1 integers.

[0166] The constructed constraint conditions are sorted to obtain high-priority constraints, medium-priority constraints, and optional constraints.

[0167] In this embodiment, since the coordination control model established in this embodiment involves many constraint variables, the constraint conditions are prioritized according to the problem objective (maximizing green wave bandwidth), physical feasibility (such as signal period), and actual traffic demand (such as asymmetric bandwidth and unbalanced upstream and downstream flow), and the model is dynamically solved according to the sorting order of the constraint conditions to avoid the model falling into a no-solution state. The constraint conditions of the model are divided into three parts: high-priority constraints, medium-priority constraints, and optional constraints.

[0168] (1) High-priority constraints:

[0169] The bandwidth edge constraint considering phase sequence optimization is the core guarantee for the continuity and effectiveness of the green wave band and plays a role in optimizing the left-turn phase sequence. The period constraint of the coordinated control path along the intersection plays a role in coordinating the phase difference between intersections. In addition, other constraints involve the determination of model variable parameters and need to strictly satisfy the constraint conditions. These constraints are directly related to the objective function (maximizing bandwidth) and key coordination logic.

[0170] (2) Medium-priority constraints:

[0171] The path speed constraint considering intersection left-turn considers the upper and lower limits of vehicle speed and the acceleration change of vehicles when turning, enhancing the authenticity of the model, but allowing for parameter adjustment and relaxation; the asymmetric bandwidth constraint ensures the proportion of uplink and downlink green wave bandwidth, but since the uplink left-turn is the main research object in this embodiment, the downlink direction constraint proportion parameter can be adjusted appropriately.

[0172] (3) Optional constraints:

[0173] Since the turning speed at the intersection is considered in this embodiment to affect the coordinated control, the acceleration constraint can be appropriately relaxed in the constraint condition.

[0174] S3. The classical fruit fly algorithm is improved, and the improved fruit fly algorithm is used to solve the green wave variable bandwidth coordinated control model to obtain the left turn coordinated control result.

[0175] The left turn path coordinated control model established in this embodiment is improved on the basis of the traditional AM-BAND trunk coordinated control model. In the speed constraint part, the dynamic change of the acceleration variable at the intersection turning point is considered, and the established coordinated control model is a mixed integer nonlinear programming MINLP model.

[0176] Since there are different types of variables such as integer discrete variables, continuous optimization variables and nonlinear constraints in the model, the classical fruit fly algorithm is selected, and two improvements are made: (1) different strategies are adopted for updating different types of variables; (2) a periodic step function is used to guide the optimization of fruit flies.

[0177] The fruit fly strategy is derived from the collective foraging behavior of this species, and then evolved into a global optimization search method. Fruit flies are superior to other species in vision and olfaction, and their foraging search process features include: ① using olfactory organs to distinguish the source of target odor and flying towards it; ② using vision to find the target path during flight, and the specific calculation process is as follows:

[0178] 1. Initialize parameters, fruit fly population size (m), iteration number (p), parameters of step function ρ, σ, T, flight range of fruit fly in variable j (dom j ∈ [dom L，j , dom u，j ]) where L represents the lower bound of the jth variable (Lower bound), and u represents the upper bound of the jth variable (Upper bound).

[0179] 2. Randomly generate the initial position of the fruit fly group, where x axi，j is a continuous variable, y axi，j is a discrete or binary integer variable:

[0180] x axe,j = rand(dom j ),

[0181] y axe,j = rand(dom j ).

[0182] 3. The update rule of each variable in a fruit fly individual, i.e. its use of olfactory and visual discovery of food, such as x e，j The random represents the jth variable in the ith fruit fly. In the traditional fruit fly algorithm, the update of all variables usually uses a unified random disturbance method (such as random displacement in the olfactory search stage), without considering the differences in variable types (such as continuous variables, discrete variables, 0,1 integer variables, etc.). The improved algorithm in this paper distinguishes the update methods for different types of variables. The algorithm dynamically adjusts the update strategy according to the variable type (for example, continuous variables follow the traditional random disturbance, discrete variables use probability selection or neighborhood jumping), solving the problem of mismatch between variable type and update method, as follows:

[0183] x i，j = x axe，j + p j * random,

[0184] y i，j = y axe，j + s j * random.

[0185] 4. For the approximation process of the optimal solution, the embodiment designs a dynamic step adjustment mechanism:

[0186] p j = |dom u，j -dom L，j || * p p％T ,

[0187] s j = |dom u，j -dom L，j | * s p％T ,

[0188] Where % represents the remainder of p divided by T, and its range is [0, T-1]. Compared with the traditional random step strategy, it has higher precision. This mechanism effectively eliminates random factors in the search process through the synergistic effect of continuous variable parameter p (0.5≤p≤0.95) and discrete variable parameter s (0.2≤s≤0.5). The periodic function design gives the algorithm the ability to jump out of local extrema. The period parameter T (20≤T≤40) controls the convergence characteristics: smaller T value enhances the shock effect, and larger T value improves the convergence precision, where p represents the iteration number.

[0189] 5. Substitute x i，j , y i，j into the objective function to determine the smell concentration (smell i ) of the fruit fly at its position at this time:

[0190] smelli = Function(x i,j ,y i,j ).

[0191] 6. Determine the optimal odor concentration individual in the fruit fly population by searching:

[0192] b-smell = max(smell i ),

[0193] Wherein, b-smell represents the optimal odor concentration value, max represents the maximum value of the objective function;

[0194] 7. Keep the optimal odor concentration value and position of fruit flies, and all fruit flies fly to the position:

[0195] x axi，j = x[b-set],

[0196] y axi,j = y[b-set],

[0197] Wherein, set represents the position of fruit flies with optimal odor concentration value.

[0198] 8. Check whether the termination condition of the algorithm is met, if met, terminate the calculation process; if not met, loop the operation process of steps 3 to 7. In this process, the best odor intensity perceived by the current fruit fly population is compared with the global optimal record, if there is a better solution, go to step 8 to continue optimization.

[0199] The method for solving the green wave variable bandwidth coordination control model by using the improved fruit fly algorithm comprises:

[0200] Obtain the geometric characteristics, road attributes, signal timing parameters and traffic demand of each intersection in the path, solve the optimal cycle length of each independent intersection by using Webster algorithm, and take the maximum value thereof as the unified control cycle of the whole trunk;

[0201] According to the proportion of traffic volume of different directions in the intersection, the distribution proportion of green time in the phase scheme design is determined;

[0202] Combined with the left turn phase release mode of each intersection, the public signal cycle and the designed up-down green wave speed, the improved fruit fly algorithm is used to solve the green wave variable bandwidth coordination control model, and the phase sequence arrangement result of each intersection is obtained;

[0203] Based on the model solving result, the correlation between each variable and the phase offset in the time-distance diagram is analyzed, the phase difference between intersections is derived, and the corresponding time-distance diagram is drawn by using the phase difference, and the left turn coordination control result is obtained.

[0204] Embodiment two:

[0205] In the embodiment, the left-turn path composed of three intersections of l-m intersection, l-n intersection and n-o intersection formed by l section of road, m section of road, n section of road and o section of road in a core area of a certain area in A city is selected as a research object for analysis.

[0206] The total length of the path is about 1.55 / km, the intersection spacing is 822 / m and 702 / m respectively, and the channelization state of each intersection is as shown in Figure 5 The solid arrow represents the uplink left-turn coordination direction, and the dotted arrow represents the downlink direction.

[0207] The peak hour flow of each intersection along the line is collected, and the survey period is the morning peak on May 15, 2025: 7:00-8:00, and the flow of each intersection is shown in Table 3.

[0208] Table 3

[0209]

[0210] According to the flow and the current lane function division, the optimal cycle of each intersection is calculated by Webster, the overall coordination cycle of the path is selected as the maximum cycle value 124s in the three intersections, and the cycles of other intersections are also adjusted to 124s, and the phase green time of each intersection is calculated according to the vehicle flow of each intersection and is shown in Table 4:

[0211] Table 4

[0212]

[0213] Since the main focus of the application is the uplink left-turn flow, the results of the model solution mainly analyze the uplink green wave bandwidth and the left-turn phase sequence in the uplink direction, and the related parameters required for the model solution are shown in Table 5. The results of different models for left-turn coordination control path are obtained by solving the model, AM-BAND model and Synchro model, and the signal timing scheme of each intersection under different coordination control models is calculated and shown in Table 7.

[0214] Table 5

[0215]

[0216] Table 6

[0217]

[0218] Table 7

[0219]

[0220] In order to compare the differences between the green wave bandwidths of various road sections more intuitively, the time-distance diagrams of the path green wave coordinated control schemes optimized by the model, the AM-BAND model and the Synchro model are shown in Figure 6 、 Figure 7 and Figure 8 .

[0221] Simulation is performed by Vissim, and simulation parameters are set as follows: simulation period is 3600s, simulation accuracy is 5 steps / s, and random seed is 50. Considering the instability in the initial stage of simulation, the first 300s is a simulation warm-up period, and the simulation output data from 300s to 3600s is used for evaluation, and the average value of 10 simulations is taken as the final evaluation result.

[0222] In the case of traffic simulation, the performance of the model and the AM-BAND model is compared and evaluated from two aspects, (1) the left-turn path traffic flow composed of three intersections is taken as the simulation evaluation object, to show that it is feasible to coordinate the left-turn traffic flow and a larger traffic benefit can be obtained, but attention should be paid to that the traffic benefit of other directions at the intersection cannot be sacrificed when the left-turn traffic flow is coordinated. Therefore, (2) the simulation effect of each intersection along the left-turn path is evaluated.

[0223] (1) Simulation effect evaluation of left-turn path coordinated control

[0224] The vehicle delay, average queue length, average number of stops and travel time are taken as the evaluation indexes of the model, the AM-BAND model and the Synchro model, and the evaluation index output results are shown in Table 8, Figure 9 .

[0225] Table 8

[0226]

[0227] It can be seen from the table that the model is better than the other two models in each index. According to the analysis in Table 9, compared with the AM-BAND model, the vehicle delay of the model along the path is reduced by 9.74%, the average queue length is shortened by 18.46%, the average number of stops is reduced by 8.14%, and the travel time is reduced by 5.66%; compared with the Synchro model, the vehicle delay is reduced by 20.32%, the average queue length is shortened by 28.54%, the average number of stops is reduced by 37.30%, and the travel time is reduced by 23.14%.

[0228] Table 9

[0229]

[0230] (2) Simulation effect evaluation of coordinated control of intersections along the path

[0231] The evaluation indexes of the model, the AM-BAND model and the Synchro model are vehicle delay, average queue length and average stopping number, and the output results are shown in Table 10. Figure 10

[0232] Table 10

[0233]

[0234] It can be seen that the model has certain optimization in the evaluation indexes of each intersection along the left-turn path, but the traffic efficiency of non-coordinated directions is affected in the coordination of the left-turn path, resulting in low overall optimization of the intersection. According to the analysis in Table 11, compared with the AM-BAND model, the average vehicle delay of each intersection along the path is reduced by 2.43%, the average queue length is increased by 0.36%, and the average stopping number is reduced by 5.96%; compared with the Synchro model, the average vehicle delay is reduced by 10.35%, the average queue length is reduced by 14.55%, and the average stopping number is reduced by 15.18%.

[0235] Table 11

[0236]

[0237] In summary, the control effect of the path and the control effect of the intersection along the path, it can be found that compared with the AM-BAND model, the green wave signal coordination control scheme of the model can reduce the vehicle delay, reduce the queue length, reduce the stopping number, and shorten the travel time of the path in the coordination and optimization of the left-turn path, and the queue length is slightly increased in the coordination control of the intersection due to the influence of some non-coordinated directions, but the vehicle delay and the stopping number can be reduced; compared with the Synchro model, the green wave signal coordination control scheme generated by the model can reduce the vehicle delay, reduce the queue length, reduce the stopping number, and shorten the travel time of the path in the coordination and optimization of the left-turn path and the intersection. This verifies that the model built in the text provides green wave bandwidth for left-turn traffic, which can obtain better traffic efficiency.

[0238] The above-described embodiments are only descriptions of the preferred modes of the present application, and do not limit the scope of the present application. Without departing from the design spirit of the present application, various modifications and improvements to the technical solutions of the present application made by those skilled in the art shall fall within the protection scope determined by the claims of the present application.​

Claims

1. A left-turn path coordinated control method for urban road networks considering phase sequence optimization, characterized in that, The method comprises the following steps: A binary-valued variable is introduced to optimize the phase sequence of a left-turn intersection; A green wave variable bandwidth coordination control model oriented to a left-turn path is constructed by improving a traditional AM-BAND variable bandwidth coordination control model; The classical fruit fly algorithm is improved, and the green wave variable bandwidth coordination control model is solved by using the improved fruit fly algorithm to obtain a left-turn coordination control result.

2. The method of claim 1, wherein the method further comprises: The method for optimizing the phase sequence of a left-turn intersection comprises: According to the left-turn phase mode, the uplink left-turn phase sequence is introduced by 0-1 in the inequality constraint of the coordination control model The left-turn advance or lag in the uplink direction is realized by solving the value of the uplink left-turn phase sequence in the bandwidth edge constraint; For intersection S n and intersection S n+1 By introducing the variable δ n and variables In a periodic constraint, the variable δ is solved. n and the variables The values ​​are used to determine the passage method for these two intersections.

3. The method of claim 2, wherein the method further comprises: The method for constructing the green wave variable bandwidth coordination control model comprises: A target function is constructed: b n = b n1 + b n2 , wherein, a n represents the intersection S n the weight coefficient of the uplink bandwidth, represents the intersection S n the weight coefficient of the downlink bandwidth, b n represents the intersection S n the green wave bandwidth in the uplink direction, b m1 represents the intersection S n the first component of the green wave bandwidth in the uplink direction, b n2 represents the intersection S n the second component of the green wave bandwidth in the uplink direction, represents the intersection S n the green wave bandwidth in the downlink direction, represents the intersection S n the first component of the green wave bandwidth in the downlink direction, represents the intersection S n the second component of the green wave bandwidth in the downlink direction, n represents the nth intersection in the left-turn path, and N represents the total number of intersections in the left-turn path as N. Bandwidth edge constraints considering phase sequence optimization, path speed constraints considering left turns at intersections, acceleration constraints, path along-line intersection cycle constraints of coordination control, asymmetric bandwidth constraints, and other constraints are constructed. The constructed constraints are sorted to obtain high-priority constraints, medium-priority constraints, and optional constraints.

4. The method of claim 3, wherein, The bandwidth edge constraints considering phase sequence optimization comprise: A path-intersection process bandwidth edge minimum value range constraint: ω i ≥b i2 i = n, n + 2, Where, ω n+1 Indicates the intersection S in the uphill direction of the path. n+1 Green band b n+1 The interval between the left edge and the right edge of the adjacent red light. Indicates intersection S n+1 The red light is on for those going straight in the opposite direction. Indicates the intersection S in the downhill direction of the path. n+1 Green band The interval between the right edge and the left edge of the adjacent red light, ω i Indicates the intersection S in the uphill direction of the path. i Green band b i The interval between the left edge and the right edge of the adjacent red light, b i2 Indicates intersection S i The second component of the green wave bandwidth in the upward direction, Indicates the intersection S in the downhill direction of the path. i Green band The interval between the right edge and the adjacent left edge of the red light. Indicates intersection S i The first component of the green wave bandwidth in the downlink direction, where i represents the i-th intersection; A path-intersection process bandwidth edge maximum value range constraint: ω n +b n1 ≤1-r Tn , ω n+2 +b (n+1)1 ≤1-r Tn+2 -τ n+2 , Where, ω n Indicates the intersection S in the uphill direction of the path. n Green band b n The interval between the left edge and the right edge of the adjacent red light, r Tn Indicates intersection S n During the red light period for those going straight uphill. Indicates the intersection S in the downhill direction of the path. n Green band The interval between the right edge and the adjacent left edge of the red light. Indicates intersection S n The red light is on for those going straight in the opposite direction. Indicates intersection S n Initial queue clearing time for downstream traffic, ω n+2 Indicates the intersection S in the uphill direction of the path. n+2 Green band b n+2 The interval between the left edge and the right edge of the adjacent red light, b (n+1)1 Indicates intersection S n+1 The first component of the green wave bandwidth in the uplink direction, r Tn+2 Indicates intersection S n+2 At the red light time for straight-ahead traffic, τ n+2 Indicates intersection S n+2 Initial queue clearing time for upward movement. Indicates the intersection S in the downhill direction of the path. n+2 Green band The interval between the right edge and the adjacent left edge of the red light. Indicates intersection S n+1 The second component of the downlink green wave bandwidth, Indicates intersection S n+2 The red light is on for those going straight in the opposite direction. A path-intersection process bandwidth maximum value range constraint: where τ n+1 represents the initial queue clearance time at the intersection S n+1 , r Ln+1 represents the red light time for left turn at the intersection S n+1 , g Ln+1 represents the green light time for left turn at the intersection S n+1 , and represents the green light time for left turn at the intersection S n+1 .

5. The method of claim 4, wherein, The path speed constraints considering left turns at intersections comprise: wherein d n，n+1 represents a straight stage distance from the intersection S n to the intersection S n+1 , f n represents a lower limit of an uplink direction travel speed at the intersection S n , a represents an acceleration when a vehicle in the uplink direction turns at the intersection, L n,n+1 represents a turning preparation stage distance from the intersection S n to the intersection S n+1 , z represents an inverse of a signal cycle length, t n，n+1 represents an uplink travel time of the vehicle from the intersection S n to the intersection S n+1 , e n represents an upper limit of the uplink direction travel speed at the intersection S n , represents a lower limit of a downlink direction travel speed at the intersection S n , represents an acceleration when a vehicle in the downlink direction turns at the intersection, v represents a vehicle speed before entering the intersection, r represents a turning radius when the vehicle turns, represents a downlink travel time of the vehicle from the intersection S n+1 to the intersection S n , represents an upper limit of the downlink direction travel speed at the intersection S n , n represents an upper limit of a change in the uplink direction travel speed at the intersection S n , n+1，n+2 represents a straight stage distance from the intersection S n+1 to the intersection S n+2 , t n+1，n+2 represents an uplink travel time of the vehicle from the intersection S n+1 to the intersection S n+2 , n represents a lower limit of a change in the uplink direction travel speed at the intersection S n , represents an upper limit of a change in the downlink direction travel speed at the intersection S n , represents a straight stage distance from the intersection S n+1 to the intersection S n , represents a straight stage distance from the intersection S n+2 to the intersection S n+1 , represents a downlink travel time of the vehicle from the intersection S n+2 to the intersection S n+1 , represents a straight stage distance from the intersection S n Lower limit of the downlink travel speed change.

6. The method of claim 5, wherein the method further comprises: The acceleration constraints comprise: Wherein, μ represents the friction coefficient of a vehicle, g represents the acceleration of gravity, and sin(θ) represents the road slope.

7. The method of claim 6, wherein the method further comprises: The path along-line intersection cycle constraints of coordination control comprise: where r n+1 denotes the red light time on the road coordinated at intersection S n where g Ln denotes the green light time for going straight up at intersection S n where g n denotes the green light time for going straight down at intersection S i where δ n denotes the 0-1 variable for left turn up at intersection S Li where g n denotes the green light time for left turn up at intersection S n where δ n denotes the 0-1 variable for left turn down at intersection S n,n+1 where g n denotes the green light time for left turn down at intersection S n+1 and S n denotes the integer adjustment factor satisfying the constraint condition between intersection S max and S min .

8. The method of claim 3, wherein the method further comprises: The asymmetric bandwidth constraints comprise: Wherein, N1 and N2 represent positive real numbers.

9. The method of claim 7, wherein the method further comprises: The other constraints comprise: An asymmetric constraint for unbalanced two-way traffic volume: wherein k n represents an unbalance coefficient; A cycle rationality constraint: where C max represents the maximum value of the signal control period, C min represents the minimum value of the signal control period; A variable value constraint: m n,n+1 , m n+1，n+2 integers, where m n+1，n+2 denotes the integer adjustment factor satisfying the constraint condition between the intersections S n+1 and S n+2 denotes the integer adjustment factor satisfying the constraint condition between the intersections S n+2 denotes the integer adjustment factor satisfying the constraint condition between the intersections S n+2 denotes the integer adjustment factor satisfying the constraint condition between the intersections S 10. The method of claim 1, wherein, The method for solving the green wave variable bandwidth coordination control model by using the improved fruit fly algorithm comprises: The geometric characteristics, road attributes, signal timing parameters, and traffic demand of each intersection in the path are obtained, the Webster algorithm is used to solve the optimal cycle length of each independent intersection, and the maximum value thereof is taken as the unified control cycle of the entire trunk; According to the proportion of traffic volume in different directions in the intersection, the green time allocation proportion in the phase scheme design is determined; The left-turn phase release mode of each intersection, the public signal cycle, and the designed up-down green wave speed are combined, the green wave variable bandwidth coordination control model is solved by using the improved fruit fly algorithm, and the phase sequence arrangement result of each intersection is obtained; Based on the model solving result, the correlation between each variable and the phase offset in the time-distance diagram is analyzed, the phase difference between intersections is derived, and the corresponding time-distance diagram is drawn by using the phase difference to obtain the left-turn coordination control result.

Images