CN112885085B - Confluence control strategy applied to reconstruction and extension of highway construction area - Google Patents
Confluence control strategy applied to reconstruction and extension of highway construction area Download PDFInfo
- Publication number
- CN112885085B CN112885085B CN202110054207.7A CN202110054207A CN112885085B CN 112885085 B CN112885085 B CN 112885085B CN 202110054207 A CN202110054207 A CN 202110054207A CN 112885085 B CN112885085 B CN 112885085B
- Authority
- CN
- China
- Prior art keywords
- traffic
- confluence
- lane
- vehicle
- clustering
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/07—Controlling traffic signals
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/09—Arrangements for giving variable traffic instructions
- G08G1/095—Traffic lights
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/16—Anti-collision systems
- G08G1/167—Driving aids for lane monitoring, lane changing, e.g. blind spot detection
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Evolutionary Biology (AREA)
- Theoretical Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Genetics & Genomics (AREA)
- Evolutionary Computation (AREA)
- Chemical & Material Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Physiology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Analytical Chemistry (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a confluence control strategy for rebuilding and expanding a highway construction area, which comprises functions of traffic state identification, confluence control and confluence strategy decision, can identify a dynamic traffic flow state, and respectively adopts an early confluence strategy, a late confluence strategy and a signal confluence strategy aiming at the problems of confluence conflict and congestion which may occur in different traffic states, so as to realize the functions of standardizing confluence behavior, avoiding vehicle congestion and guiding congestion traffic flow evacuation, and improve the traffic safety level and traffic efficiency of the rebuilding and expanding highway construction area.
Description
Technical Field
The invention belongs to the technical field of intelligent transportation and traffic safety, and particularly relates to a confluence control strategy for rebuilding and expanding a highway construction area.
Background
The highway vehicle flow for reconstruction and extension projects is large, and the reconstruction and extension projects have large influence range and long duration on traffic flow. When the roadways are trimmed, especially when a single road surface is constructed, lanes are occupied for construction, only part of the lanes are opened to form a bottleneck road section, congestion is formed at a peak time period, social influence is great, and secondary accidents are easily caused. In addition, compared with the conventional road section, the road section of the reconstruction and expansion construction area limits the vehicle speed, compresses the driving space and influences the lane changing behavior of the vehicle. When a vehicle runs on the bottleneck road section with a construction area, frequent acceleration, deceleration and lane change are easy, and mutual interference among the vehicles is large, so that traffic problems of large vehicle speed space-time distribution difference, multiple confluence conflicts and the like are caused.
Due to the lack of effective management on traffic organization and facility arrangement of highway reconstruction and extension construction area road sections at present, the highway reconstruction and extension construction area road sections only conform to national standards or local standards under most conditions, and the non-professionality of construction personnel is difficult to enable the construction personnel to implement effective guidance measures like traffic guidance personnel in European and American countries when traffic jam and other traffic safety problems exist. Therefore, a specific control scheme is often adopted for specific traffic problems in different traffic states due to the lack of related safety facilities in construction area road sections, and the vehicles cannot be subjected to confluence control and congestion evacuation, so that problems such as traffic conflicts and traffic congestion are induced, and the traffic safety and traffic efficiency of the highway are affected.
Therefore, in order to timely solve the problems of congestion, congestion and confliction which may occur in the reconstruction and extension construction area, it is necessary to adopt traffic flow acquisition and data mining technology on the road section of the construction area to identify the traffic flow state with dynamic and uncertain state, and meanwhile, to adopt corresponding effective control measures for different traffic problems occurring in different traffic states. The prior patent provides a solution for rebuilding and expanding a highway construction area. The method for calculating the dynamic speed limit recommended value in the construction area of the expressway, which is provided by the prior art, adjusts the speed limit value according to the road condition, and guides the upstream vehicle to decelerate so as to evacuate the traffic; the method is characterized in that traffic flow is collected in real time, and the traffic flow entering from the upstream is controlled by limiting speed when queuing occurs, so as to relieve congestion; in addition, the method compares the upstream traffic flow with a set threshold value to decide whether to remind an upstream driver to carry out confluence lane change or not so as to avoid the traffic jam or accidents caused by excessive and concentrated confluence. However, the reconstructed and expanded highway often has large flow, long queuing length when congestion occurs, low road section traffic capacity, frequent vehicle congestion, difficult function of a variable speed-limiting strategy, and incapability of solving the problem of vehicle confluence conflict caused by frequent lane change when traffic is small. Furthermore, the decision to set a threshold for merging is too empirical and cannot adapt to dynamically changing traffic flows.
Disclosure of Invention
Aiming at the problems that the traditional control measures are unscientific and poor in effect and constructors lack specialized training, the invention provides a method which can identify the state of a dynamic traffic flow and respectively adopt an early-stage confluence strategy, a late-stage confluence strategy and a signal confluence strategy to realize the functions of standardizing confluence behaviors and evacuating congested vehicles and improve the traffic safety level and the traffic efficiency of a reconstruction and extension highway construction area aiming at the problems of confluence conflict and congestion which may occur in different traffic states. The specific technical scheme of the invention is as follows:
a confluence control strategy for rebuilding and expanding a highway construction area comprises the following steps:
s1: collecting traffic information; collecting traffic characteristic parameters including traffic volume, speed and occupancy through a video detector arranged on a trip road section in a warning area, and carrying out normalization processing:
wherein, x'cdThe d traffic characteristic parameter, x, of the c sample after normalizationcdFor raw data before normalization, xd,maxIs the maximum value, x, in the d-th column datad,minIs the minimum value in the d-th column data;
when each period is finished, obtaining a traffic characteristic parameter matrix of the row c and the column d, carrying out normalization processing on the traffic characteristic parameter matrix, and calculating the average value of the row c and the column c of the matrix as a traffic characteristic parameter of each period to obtain a traffic parameter vector of the row 1 and the column d;
s2: when each period starts, calculating the Euclidean distance between the traffic parameter vector in the step S1 and a traffic state clustering center forming a traffic state identification standard library, performing mode matching, and identifying the current traffic state;
s3: forming a confluence strategy decision mechanism according to the current traffic state, selecting a corresponding confluence control measure, and controlling the combination of the variable information board and the signal lamp to be switched on and off;
s4: when the current cycle ends, the process returns to step S2.
Further, in the step S2, the historical traffic status is divided into four categories of clear, general, crowded and crowded according to the service level to represent four clustering results, the traffic status recognition standard library for real-time traffic status pattern recognition, which is composed of four categories of centers, is obtained by fuzzy clustering analysis of historical traffic data in the reconstructed and expanded highway construction area, the traffic status recognition standard library is a 4-row and 3-column normalized matrix, 4-row data represents four traffic statuses of clear, general, crowded and crowded, 3-column data represents three traffic characteristics of traffic volume, speed and occupancy, the traffic status recognition standard library is constructed and realized by fuzzy C-means clustering analysis improved by genetic algorithm, and then real-time recognition of the reconstructed and expanded highway construction area is realized by adopting a pattern matching method, the specific process is as follows:
s2-1: optimizing an initial clustering center of a fuzzy C-means clustering algorithm through a genetic algorithm, namely, firstly solving a globally optimal clustering center by using the genetic algorithm as the initial clustering center of the fuzzy C-means clustering algorithm, specifically, selecting the initial clustering center as a population individual to carry out optimization solution, setting basic parameters of the genetic algorithm, wherein M is population scale, T is termination evolution algebra, P iscIs cross probability, PmFor the variation probability, the population fitness function takes the reciprocal of the objective function, Zb，qThe fitness of the qth individual of the b generation, b ∈ [1, T]，q∈[1，M]The objective function and the constraint condition are as follows:
wherein, the traffic data sample set x is matrix of n x 3 dimension, the cluster center set V is matrix of 4 x 3 dimension, and u(b)As a function of membership at the b-th iteration,
the Euclidean distance from the jth traffic data sample to the ith class center at the b-th iteration,
the Euclidean distance sum of the jth traffic data sample to all the clustering centers during the b-th iteration; i is the serial number of the clustering center set, and j is the serial number of the sample set;
set of cluster centers at b +1 iteration, uijIs a membership matrix of the sample points,
the membership degree of the jth traffic data sample relative to the ith class center; dijEuclidean distance, d, from jth traffic data sample to ith class centerkjThe Euclidean distance sum, x, of the jth traffic data sample to all the cluster centersjTraffic parameter vector, v, for the jth traffic data sampleiIs the cluster center of the ith class, xj、viIs a normalized 3-dimensional vector; v. ofkIs the cluster center of all classes;
s2-2: initializing a clustering center as a primary population, and updating membership and objective function values;
s2-3: encoding the population; binary coding is carried out on individuals of each population, and the number of coding bits is determined by setting coding precision and decoding space of each gene chain; after the coding is completed, the gene coding chain GN of the qth individual of the primary population is obtainedb，qWhen b is 0, q is ∈ [1, M ]]；
S2-4: carrying out roulette selection, single-point crossing and mutation operations on individual gene chains;
s2-5: updating an algebra b which is b +1, decoding the new population and updating the membership degree, the clustering center and the objective function value;
s2-6: judging the algebra, if b is less than T, returning to the step S2-3, otherwise, executing the step S2-7;
s2-7: decoding the individual with the maximum population fitness of the last generation to obtain the membership UbCluster center VbThe optimal solution of (2);
s2-8: clustering genetic algorithm to center VbThe optimal solution of the fuzzy C mean value clustering is used as an initial clustering center of the fuzzy C mean value clustering, the number C of the clustering centers is set to be 4, the threshold epsilon of iteration stop is set to be 0 at the moment;
s2-9: updating the membership function according to the cluster center and the formula (4), if a certain traffic data sample x is foundjWith a certain cluster centre viIf the Euclidean distance of the traffic data sample vector is 0, the membership degree from the traffic data sample vector to the clustering center is 1, and the other membership degrees are 0; otherwise, the updating process is as follows:
wherein the content of the first and second substances,
the membership degree of the jth traffic data sample relative to the ith class center;
for the jth traffic data sample to the ith category in the b-th iterationThe euclidean distance of the centers thereof,
the Euclidean distance sum of the jth traffic data sample to all the clustering centers during the b-th iteration;
s2-10: updating the clustering center V according to the updated membership degree by substituting formula (5)(b+1)；
S2-11: checking the iteration termination condition V(b+1)-V(b)If | < epsilon, executing step S2-12, otherwise, returning to step S2-9;
s2-12: outputting the optimal membership degree and the clustering center, completing four types of traffic state division of historical traffic data, and obtaining a traffic state identification standard library which is composed of the four types of clustering centers and used for real-time traffic state mode identification, wherein the traffic state identification standard library is a 4-row and 3-column normalized matrix, 4 rows of data respectively represent four types of traffic states, and 3 columns of data respectively represent three traffic characteristics of traffic volume, speed and occupancy;
s2-13: at the beginning of each period, completing pattern matching according to the Euclidean distance minimum principle of a traffic parameter vector of the previous period obtained by real-time acquisition and normalization processing and a cluster center of the current affiliated traffic state, and realizing real-time identification of the traffic state, wherein the formula of the Euclidean distance d (A, B) is as follows:
wherein, muA(μi) Is the ith element, mu, in the A point vectorB(μi) Is the ith element in the B point vector;
further, the confluence strategy decision mechanism in step S3 takes 2min as a cycle, and determines the traffic state at that time by calculating the euclidean distance between the traffic parameter vector acquired in the last cycle and obtained by normalization processing and each cluster center of the traffic state recognition standard library constructed in step S2 when each cycle starts; after the judgment of the traffic state is finished, an early confluence measure, a late confluence measure and a signal confluence measure are respectively and correspondingly selected according to three control scenes, namely unblocked control, general control, crowded control and crowded control; specifically, the method comprises the following steps:
when the traffic state is recognized as a clear, general state in step S2, it is decided to use an early merge control means that controls a variable information panel arranged in the middle section of the warning area to display information of "right lane closure" and "merge here", and guides the vehicle to complete the merge at that position; controlling a variable information board arranged at the tail end of the warning area to display 'right lane closed', and warning that the vehicle can not change lanes to the right lane any more; controlling a variable information board arranged at the starting point of the transition area to display 'please slow down' and 'close the right lane', guiding the vehicle to decelerate, and not changing the lane at will, so as to finish early guiding the vehicle converging behavior;
when the traffic state is identified as a more congested state in step S2, it is decided to use a late-stage confluence control measure, which controls a variable message board disposed at the middle section of the warning area to display information of "two lanes open" and "traveling to a confluence point" to prompt the driver not to finish confluence prematurely; controlling a variable information board arranged at the tail end of the warning area to display a 'lane keeping', guiding the vehicle to continuously run to a confluence point in the current lane, and warning the vehicle that the lane can not be changed randomly to jam a fleet of inner lanes; controlling a variable information board arranged at the starting point of the transition area to display 'confluence here' and 'right lane closure', guiding the vehicles to finish confluence here, and finishing late control on vehicle confluence behavior;
when the traffic state is identified as a congestion state in step S2, it is decided to use a signal confluence control measure, which controls a variable message board arranged at the middle section of the warning area to display information of "two lanes open" and "travel to confluence point" to prompt the driver not to finish confluence too early; controlling a variable information board arranged at the tail end of the warning area to display 'forbid lane changing', guiding the vehicle to continuously run to a confluence point in the current lane, and warning the vehicle that the lane can not be changed randomly to jam a fleet of the inner lane; and controlling a variable information board arranged at the starting point of the transition area to display the 'confluence here' and the 'compliance signal', prompting the vehicle to comply with the indication of the signal lamp, and simultaneously guiding the vehicle to complete the confluence here according to the right of way distributed by the signal lamp, so as to complete the signal control of the vehicle jammed on the bottleneck road section.
Further, in the case of a congestion state in step S3, the signal merge control means performs control in cooperation with the traffic light, the variable information panel, and the road temporary marking:
the temporary road marking is used for dividing a waiting area and a stop line, the waiting area refers to a queuing area of the vehicle during signal control, and the stop line is a signal control section of the road section;
the signal transition area is the distance between the position of the stop line and the starting point of the transition area, the length of the signal transition area is set to be 25m, and the layout position of the stop line is the position 25m away from the starting point of the transition area;
the waiting area refers to an area for a vehicle parking waiting signal separated by a white line, the length of which affects the traffic efficiency and the merging behavior of the vehicle, the length D of the waiting area passing through the maximum queuing length DLength of queueDetermining, according to the parking wave theory:
wherein v isfFor free flow velocity, kjFor the blocking density, t' is taken as the maximum red light duration, kfIs the actual density;
the signal timing scheme is a two-phase signal timing scheme, and aims to divide the right of way for vehicles in an open lane and a closed lane, and the calculation process is as follows:
s3-1: counting the peak hour flow q of the open lane of the road section of the construction area of the reconstruction and extension highway1And peak hourly flow q of closed lane2；
S3-2: setting the right of the open lane as phase one,The right of way of the closed lane is phase two, and phase one key flow rate ratio y is calculated1Phase two key flow rate ratio y2And sum of key flow rate ratios Y:
Y＝y1+y2
wherein S isTIs the straight-channel saturation flow rate;
s3-3: calculate the yellow light duration A of phase one1Full red duration r1And duration of green light interval I1Calculating the yellow light time length A of the phase two2Full red duration r2And duration of green light interval I2：
I1＝A1+r1
I2＝A2+r2
Where t is the driver reaction time, v85In order to reconstruct and expand the highest speed limit of a highway section in a highway construction area, a is automobile deceleration, g is gradient, w is the distance from a stop line to the tail end of a transition area, L' is automobile standard length, v1515% vehicle speed;
s3-4: calculating the signal loss time L of phase one1Phase two signal loss time L2Total signal loss time L and signal cycle duration C:
L1＝l1+r1
L2＝l2+r2
L＝L1+L2
wherein l1And l2The start-up losses for phase one and phase two, respectively;
s3-5: determining green light assignment:
g1＝gE1+l1-A1
g2＝gE2+l2-A2
wherein, gE1Effective green duration of phase one, gE2Effective green duration, g, for phase two1Green time duration of phase one, g2A green time period for phase two;
to sum up, the signal period duration C and the phase-one green duration g are obtained1And yellow light duration A1Green time duration g of phase two2And yellow light duration A2The two phases together form a signal timing scheme.
Further, in the step S3, ST1650veh/h is taken, and t is taken as 1 s; l' is 5-6m, L1And l2Take 3 s.
The invention has the beneficial effects that:
1. the fuzzy C-means clustering algorithm improved by the genetic algorithm is used for traffic state identification, the algorithm is suitable for traffic parameters with boundary ambiguity, the reliability and the sensitivity of the identification algorithm are improved by selecting multi-dimensional traffic characteristic parameters, the convergence speed of the FCM algorithm is improved by improving the genetic algorithm, the defect that the FCM algorithm cannot find a global optimal solution is overcome, and the whole algorithm has great advantages in traffic identification.
2. According to the traffic state identification result, different confluence control measures are selected to solve the traffic problems of confluence conflict and congestion which may occur in different traffic states, and the traffic safety level and the traffic efficiency of the construction area of the reconstructed highway are improved.
3. The invention combines real-time traffic states, assists constructors in completing guidance and evacuation tasks of traffic flow in a construction area, provides basis for the management and control of constructors, and makes up the defects and non-professionality of domestic constructors in assisting traffic management and control.
Drawings
In order to illustrate embodiments of the present invention or technical solutions in the prior art more clearly, the drawings which are needed in the embodiments will be briefly described below, so that the features and advantages of the present invention can be understood more clearly by referring to the drawings, which are schematic and should not be construed as limiting the present invention in any way, and for a person skilled in the art, other drawings can be obtained on the basis of these drawings without any inventive effort. Wherein:
FIG. 1 is a schematic workflow diagram of a merge control strategy;
FIG. 2 is a flow chart of a fuzzy C-means clustering algorithm improved by a genetic algorithm;
FIG. 3 is the layout position of all the merge control facilities in the merge strategy;
FIG. 4 is an early-stage merge control action enforcement scenario;
FIG. 5 is a scenario for implementing late-stage confluence control measures;
FIG. 6 is a signal confluence control measure implementation scenario;
fig. 7 is a schematic diagram of a waiting area and a signal transition area.
The reference numbers illustrate:
1-signal lamp; 2-a stop line; 3-lane line; 4-variable information board; 5-a video detector; 6-static information board.
Detailed Description
In order that the above objects, features and advantages of the present invention can be more clearly understood, a more particular description of the invention will be rendered by reference to the appended drawings. It should be noted that the embodiments of the present invention and features of the embodiments may be combined with each other without conflict.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, however, the present invention may be practiced in other ways than those specifically described herein, and therefore the scope of the present invention is not limited by the specific embodiments disclosed below.
The existing traffic organization and traffic control technology for reconstructing and expanding a highway construction area depends on the guidance of traditional traffic sign marked lines and constructors, but the traffic sign marked lines are difficult to play an effective role in dynamic and uncertain traffic flow and complex traffic problems of the construction area. In addition, the variable speed limit control strategy proposed in recent years is often difficult to play a role in reconstructing and expanding the traffic characteristics of high traffic flow in a construction area of an expressway and often exceeding the road traffic capacity of the construction area. The method can accurately identify the dynamic and uncertain traffic flow states, and simultaneously, aiming at the problems of confluence conflict, congestion and the like which may occur in different traffic states, corresponding effective confluence control measures are adopted, so that the safety level and the traffic efficiency of the reconstruction and expansion highway construction area are improved.
Specifically, aiming at the characteristic of reconstructing and expanding the highway, the problem of low trafficability caused by large traffic flow and vehicle congestion can be effectively solved by indicating the vehicles to keep lane driving, distributing road rights through signals and indicating the vehicles to merge at the starting point of a transition area when congestion occurs; in a smooth and common traffic state, the vehicle confluence behavior is normalized by specifying the confluence position, so that the influence of vehicles in a closed lane on the traffic flow of an open lane is reduced, and confluence conflict is reduced; compared with a method for decision making by adopting a threshold value, the method disclosed by the invention has the advantages that the traffic state is identified in real time, the scientific decision making of the confluence control measure is carried out, the traffic problems of confluence conflict and congestion which may occur in different traffic states are solved in real time, the confluence behavior of the vehicles is normalized, the congestion of the vehicles in congestion is avoided, the congestion traffic flow is guided to be evacuated, and the method is more suitable for the traffic flow with dynamic and uncertain.
The method is applied to reconstruction and extension of a highway construction area, and has the functions of traffic state identification, confluence control and confluence strategy decision.
Firstly, traffic state recognition is realized through traffic information acquisition, a traffic state recognition standard library and pattern matching:
(1) traffic information collection
And collecting multi-dimensional traffic flow characteristic parameters including traffic volume, speed and occupancy in the reconstruction and expansion construction area in real time, and improving the reliability and sensitivity of traffic state identification through multi-dimensional traffic characteristic data. The traffic information acquisition equipment adopts a video detector which is convenient, reliable and convenient to detach and maintain, digitalizes image information and removes image noise data in a hardware processing and image processing mode, extracts required multi-dimensional traffic characteristic parameters, normalizes the parameters, and preprocesses data for traffic state identification.
(2) Traffic status recognition criteria library and pattern matching
The construction of the traffic state identification standard library uses a fuzzy analysis algorithm, the algorithm is suitable for analyzing traffic characteristic parameters with boundary ambiguity, specifically, a fuzzy C-means clustering algorithm improved by a genetic algorithm is used for carrying out clustering analysis on four traffic states of historical data divided according to service levels, and a traffic state identification standard library is constructed by each traffic state clustering center obtained through clustering results;
and the real-time identification of the traffic state of the reconstruction and extension highway construction area is realized by adopting a mode matching method for the real-time collected multidimensional traffic data through the traffic state identification standard library. And pattern matching is used for carrying out pattern recognition on the current traffic flow state by calculating Euclidean distances between multi-dimensional traffic flow characteristic parameters which are acquired in real time and subjected to normalization processing and various clustering centers in a traffic state recognition standard library.
Second, the merge control function includes early merge control, late merge control, and signal merge control measures. The three confluence control measures are respectively applied to three different traffic states of smooth general traffic, relatively congested traffic and congested traffic, and specifically comprise the following steps:
(1) early confluence control measures: aiming at the problem of irregular confluence of frequent acceleration and deceleration and lane change of the vehicle in a smooth general state, the information display of the variable information board is controlled to guide the vehicle to finish confluence at the appointed confluence position in the midstream of the warning area in advance, the vehicle is prohibited from driving back to a closed lane, the frequent lane change of the vehicle in the closed lane is avoided, and the confluence behavior of the vehicle is standardized.
(2) Late confluence control measures: aiming at the problems of low traffic capacity and vehicle lane change and congestion in a congestion state, the vehicle is guided to keep lane running by controlling the display of the variable information board until the vehicle reaches the starting point of the transition area and then is converged, so that the vehicle lane change and congestion are avoided, and the road space is fully used.
(3) Signal confluence control measures: aiming at the problems of low traffic capacity, queuing on bottleneck road sections and vehicle lane change and congestion in a congestion state, the variable information board and the signal lamp are used in a combined mode to indicate the vehicle to keep driving on a lane, confluence is completed at a stop line, the vehicle is prevented from changing lanes and congestion at will, and parking yielding and traffic confluence are performed according to the indication of the signal lamp.
And finally, the decision of the confluence strategy is to select a confluence control measure according to the result of the traffic state recognition and control the combined opening and closing of confluence control facilities:
(1) early confluence control decision: when the vehicle is unblocked and in a common state, an early confluence control measure is used for decision, the variable information board arranged at the middle section of the warning area is controlled to display information of 'right lane closure' and 'confluence here', the variable information board arranged at the tail section of the warning area is controlled to display 'right lane closure', the variable information board arranged at the starting point of the transition area is controlled to display 'please walk' and 'right lane closure', the vehicle is guided to merge in advance at a specified position, the great influence of frequent lane changing of the vehicle in the closed lane on the traffic flow in the open lane is avoided, and confluence conflict is reduced.
(2) And (3) late confluence control decision: when the vehicle is crowded, a decision is made to use a late confluence control measure, a variable information board arranged at the middle section of an alarm area is controlled to display information of 'two lanes open' and 'driving to a confluence point', a variable information board arranged at the tail section of the alarm area is controlled to display a 'lane keeping', a variable information board arranged at the starting point of a transition area is controlled to display 'confluence at the position' and 'right lane closing', the vehicle is indicated to keep lanes until the vehicles are converged in the transition area, vehicle congestion is avoided, and lane driving space is fully utilized.
(3) And (3) signal confluence control decision: and when the vehicle is in a crowded state, a decision is made to use a signal confluence control measure, the variable information board arranged at the middle section of the warning area is controlled to display information of 'two lanes are open' and 'driving to confluence point', the variable information board arranged at the tail section of the warning area is controlled to display 'lane changing forbidding', the variable information board arranged at the starting point of the transition area is controlled to display 'here confluence' and 'compliance signal', the vehicle is instructed to keep lanes to avoid traffic jam, the right of way is distributed for the traffic flow of the open lane and the traffic flow of the closed lane, and the vehicle jammed at the bottleneck road section is evacuated.
Specifically, as shown in fig. 1, a confluence control strategy for rebuilding and expanding a construction zone of a highway includes the following steps:
s1: collecting traffic information; collecting traffic characteristic parameters including traffic volume, speed and occupancy through a video detector arranged on a trip road section in a warning area, and carrying out normalization processing:
wherein, x'cdThe d traffic characteristic parameter, x, of the c sample after normalizationcdFor raw data before normalization, xd,maxIs the maximum value, x, in the d-th column datad,minIs the minimum value in the d-th column data;
when each period is finished, obtaining a traffic characteristic parameter matrix of the row c and the column d, carrying out normalization processing on the traffic characteristic parameter matrix, and calculating the average value of the row c and the column c of the matrix as a traffic characteristic parameter of each period to obtain a traffic parameter vector of the row 1 and the column d;
s2: when each period starts, calculating the Euclidean distance between the traffic parameter vector in the step S1 and the traffic state clustering center forming the traffic state identification standard library, performing mode matching, and identifying the current traffic state, as shown in FIG. 2;
s3: forming a confluence strategy decision mechanism according to the current traffic state, selecting a corresponding confluence control measure, and controlling the combination of the variable information board and the signal lamp to be switched on and off;
s4: when the current cycle ends, the process returns to step S2.
In the step S2, the historical traffic state is divided into four categories of clear, general, crowded and crowded according to the service level to represent four clustering results respectively, as shown in table 1, a traffic state recognition standard library for real-time traffic state pattern recognition, which is composed of four categories of centers, is obtained by fuzzy clustering analysis of historical traffic data in the reconstructed and expanded highway construction area, the traffic state recognition standard library is a 4-row and 3-column normalized matrix, 4-row data represents four traffic states of clear, general, crowded and crowded respectively, 3-column data represents three traffic characteristics of traffic volume, speed and occupancy respectively, the constructed traffic state recognition standard library is realized by fuzzy C-means clustering analysis improved by a genetic algorithm, and then real-time recognition of the reconstructed and expanded highway construction area is realized by adopting a pattern matching method;
TABLE 1 traffic flow division Standard
The specific process of step S2 is:
s2-1: optimizing the initial clustering center of the fuzzy C-means clustering algorithm by a genetic algorithm, namely, firstly solving a global optimal clustering center by the genetic algorithm to be used as the initial clustering center of the fuzzy C-means clustering algorithm, selecting the initial clustering center as a population individual to carry out optimal solution, specifically, setting the genetic algorithmBasic parameters, M is population size, T is termination evolution algebra, PcIs cross probability, PmFor the variation probability, the population fitness function takes the reciprocal of the objective function, Zb,qThe fitness of the qth individual of the b generation, b ∈ [1, T]，q∈[1,M]The objective function and the constraint condition are as follows:
wherein, the traffic data sample set x is matrix of n x 3 dimension, the cluster center set V is matrix of 4 x 3 dimension, and u(b)As a function of membership at the b-th iteration,
the Euclidean distance from the jth traffic data sample to the ith class center at the b-th iteration,
the Euclidean distance sum of the jth traffic data sample to all the clustering centers during the b-th iteration; i is the serial number of the clustering center set, and j is the serial number of the sample set;
set of cluster centers at b +1 iteration, uijIs a membership matrix of the sample points,
the membership degree of the jth traffic data sample relative to the ith class center; dijEuclidean distance, d, from jth traffic data sample to ith class centerkjThe Euclidean distance sum, x, of the jth traffic data sample to all the cluster centersjTraffic parameter vector, v, for the jth traffic data sampleiIs the cluster center of the ith class, xj、viIs a normalized 3-dimensional vector; v. ofkIs the cluster center of all classes;
s2-2: initializing a clustering center as a primary population, and updating membership and objective function values;
s2-3: encoding the population; binary coding is carried out on individuals of each population, and the number of coding bits is determined by setting coding precision and decoding space of each gene chain; after the coding is completed, the gene coding chain GN of the qth individual of the primary population is obtainedb，qWhen b is 0, q is ∈ [1, M ]]；
S2-4: carrying out roulette selection, single-point crossing and mutation operations on individual gene chains;
s2-5: updating an algebra b which is b +1, decoding the new population and updating the membership degree, the clustering center and the objective function value;
s2-6: judging the algebra, if b is less than T, returning to the step S2-3, otherwise, executing the step S2-7;
s2-7: decoding the individual with the maximum population fitness of the last generation to obtain the membership UbCluster center VbThe optimal solution of (2);
s2-8: clustering genetic algorithm to center VbThe optimal solution of the fuzzy C mean value clustering is used as an initial clustering center of the fuzzy C mean value clustering, the number C of the clustering centers is set to be 4, the threshold value epsilon for stopping iteration is set to be 0.01, and the iteration frequency b is set to be 0;
s2-9: updating the membership function according to the cluster center and the formula (4), if a certain traffic data sample x is foundjWith a certain cluster centre viIf the Euclidean distance of the traffic data sample vector is 0, the membership degree from the traffic data sample vector to the clustering center is 1, and the other membership degrees are 0; otherwise, the updating process is as follows:
wherein the content of the first and second substances,
the membership degree of the jth traffic data sample relative to the ith class center;
the Euclidean distance from the jth traffic data sample to the ith class center at the b-th iteration,
the Euclidean distance sum of the jth traffic data sample to all the clustering centers during the b-th iteration;
s2-10: updating the clustering center V according to the updated membership degree by substituting formula (5)(b+1)；
S2-11: checking the iteration termination condition V(b+1)-V(b)If | < epsilon, executing step S2-12, otherwise, returning to step S2-9;
s2-12: outputting optimal membership and a clustering center, completing four types of traffic state division of historical traffic data, and obtaining a traffic state identification standard library which is composed of the four types of centers and used for real-time traffic state pattern identification, wherein the traffic state identification standard library is a 4-row and 3-column normalized matrix, 4 rows of data respectively represent four types of traffic states, and 3 columns of data respectively represent three traffic characteristics of traffic volume, speed and occupancy;
s2-13: at the beginning of each period, completing pattern matching according to the Euclidean distance minimum principle of a traffic parameter vector of the previous period obtained by real-time acquisition and normalization processing and a cluster center of the current affiliated traffic state, and realizing real-time identification of the traffic state, wherein the formula of the Euclidean distance d (A, B) is as follows:
wherein, muA(μi) Is the ith element, mu, in the A point vectorB(μi) Is the ith element in the B point vector;
the confluence strategy decision mechanism in the step S3 takes 2min as a period, and when each period starts, the Euclidean distance between the traffic parameter vector acquired in the last period and obtained through normalization processing and each cluster center of the traffic state recognition standard library constructed in the step S2 is calculated, and the traffic state at the moment is judged; after the judgment of the traffic state is completed, according to three control scenes, namely unblocked control scene, general control scene, more crowded control scene and crowded control scene, an early confluence measure, a late confluence measure and a signal confluence measure are respectively and correspondingly selected, as shown in fig. 3; specifically, the method comprises the following steps:
when the traffic state is identified to be a smooth and general state in the step S2, the problem of irregular merging of frequent acceleration and deceleration and lane change of the vehicle occurs, and it is necessary to guide the vehicle to finish merging in advance at the designated merging position in the midstream of the warning area, and prohibit the vehicle from driving back to the closed lane, so as to avoid frequent lane change of the vehicle in the closed lane and standardize the merging behavior of the vehicle; after the state judgment is completed, as shown in fig. 4, an early confluence control measure is used for decision, and the early confluence control measure controls a variable information board arranged at the middle section of the warning area to display information of 'right lane closing' and 'confluence here', and guides the vehicle to complete confluence at the position; controlling a variable information board arranged at the tail end of the warning area to display 'right lane closed', and warning that the vehicle can not change lanes to the right lane any more; controlling a variable information board arranged at the starting point of the transition area to display 'please slow down' and 'close the right lane', guiding the vehicle to decelerate, and not changing the lane at will, so as to finish early guiding the vehicle converging behavior;
when the traffic state is identified to be a crowded state in the step S2, the problems of low traffic capacity and lane change and traffic jam of the vehicle occur, and the vehicle needs to be guided to keep driving in a lane until the vehicle reaches the starting point of the transition area and then confluence is carried out, so that lane change and traffic jam of the vehicle are avoided, and the road space is fully used; after the state judgment is completed, as shown in fig. 5, a decision is made to use a late-stage confluence control measure, and the late-stage confluence control measure controls a variable information board arranged at the middle section of the warning area to display information of 'two lanes open' and 'driving to a confluence point', so as to prompt a driver not to finish confluence too early; controlling a variable information board arranged at the tail end of the warning area to display a 'lane keeping', guiding the vehicle to continuously run to a confluence point in the current lane, and warning the vehicle that the lane can not be changed randomly to jam a fleet of inner lanes; controlling a variable information board arranged at the starting point of the transition area to display 'confluence here' and 'right lane closure', guiding the vehicles to finish confluence here, and finishing late control on vehicle confluence behavior;
when the traffic state is identified as a congestion state in the step S2, the problems of low traffic capacity, queuing on the bottleneck road section, and lane changing and congestion adding of the vehicle occur, and the vehicle needs to be instructed to keep driving on the lane, merge at the stop line, avoid the vehicle from changing lane and congestion randomly, guide the vehicle on the bottleneck road section to pass in order, and evacuate congestion; after the state judgment is completed, as shown in fig. 6, a decision is made to use a signal confluence control measure, and the signal confluence control measure controls a variable information board arranged at the middle section of the warning area to display information of 'two lanes open' and 'driving to a confluence point', so as to prompt a driver not to finish confluence too early; controlling a variable information board arranged at the tail end of the warning area to display 'forbid lane changing', guiding the vehicle to continuously run to a confluence point in the current lane, and warning the vehicle that the lane can not be changed randomly to jam a fleet of the inner lane; and controlling a variable information board arranged at the starting point of the transition area to display the 'confluence here' and the 'compliance signal', prompting the vehicle to comply with the indication of the signal lamp, and simultaneously guiding the vehicle to complete the confluence here according to the right of way distributed by the signal lamp, so as to complete the signal control of the vehicle jammed on the bottleneck road section.
In step S3, in the case of a congestion state, the signal merge control measure is controlled in cooperation with a traffic light, a variable information panel, and a road temporary marking:
the temporary road marking is used for dividing a waiting area and a stop line, the waiting area refers to a queuing area of the vehicle during signal control, and the stop line is a signal control section of the road section;
in order to scientifically and reasonably arrange temporary road marking lines and signal lamps, two concepts of a signal transition area and a waiting area are added, as shown in fig. 7, the signal transition area refers to the distance between the position of a stop line and the starting point of the transition area, the length of the signal transition area is set to be 25m, and the arrangement position of the stop line is the position 25m away from the starting point of the transition area;
the waiting area refers to an area for a vehicle parking waiting signal separated by a white line, the length of which affects the traffic efficiency and the merging behavior of the vehicle, the length D of the waiting area passing through the maximum queuing length DLength of queueDetermining, according to the parking wave theory:
wherein v isfFor free flow velocity, kjFor the blocking density, t' is taken as the maximum red light duration, kfThe actual density is determined by the upstream traffic volume and the upstream speed limit;
the signal timing scheme is a two-phase signal timing scheme, and aims to divide the right of way for vehicles in an open lane and a closed lane, and the calculation process is as follows:
s3-1: counting the peak hour flow q of the open lane of the road section of the construction area of the reconstruction and extension highway1And peak hourly flow q of closed lane2；
S3-2: setting the right of the open lane as phase one and the right of the closed lane as phase two, and calculating the phase-critical flow rate ratio y1Phase two key flow rate ratio y2And sum of key flow rate ratios Y:
Y＝y1+y2
wherein S isTFor straight-path saturated flow rate, ST1650veh/h is taken;
s3-3: calculate the yellow light duration A of phase one1Full red duration r1And duration of green light interval I1Calculating the yellow light time length A of the phase two2Full red duration r2And duration of green light interval I2：
I1＝A1+r1
I2＝A2+r2
Wherein t is the reaction time of the driver, and t is 1s, v85In order to reconstruct and expand the highest speed limit of a highway section in a highway construction area, a is automobile deceleration, g is gradient, w is the distance from a stop line to the tail end of a transition area, L 'is automobile standard length, L' is 5-6m, v1515% vehicle speed;
s3-4: calculating the signal loss time L of phase one1Phase two signal loss time L2Total signal loss time L and signal cycle duration C:
L1＝l1+r1
L2＝l2+r2
L＝L1+L2
wherein l1And l2Startup losses of phase one and phase two, respectively,/1And l2Taking for 3 s;
s3-5: determining green light assignment:
g1＝gE1+l1-A1
g2＝gE2+l2-A2
wherein, gE1Effective green duration of phase one, gE2Effective green duration, g, for phase two1Green time duration of phase one, g2A green time period for phase two;
to sum up, the signal period duration C and the phase-one green duration g are obtained1And yellow light duration A1Green time duration g of phase two2And yellow light duration A2The two phases together form a signal timing scheme.
For the convenience of understanding the above technical aspects of the present invention, the following detailed description will be given of the above technical aspects of the present invention by way of specific examples.
Example 1
The method comprises the steps of establishing a traffic state identification standard library containing four traffic state clustering centers for actually measured historical traffic data of an expressway reconstruction and expansion construction area of Kaiyang highway Yingping-Yangjiang section in Guangdong province, processing 60 groups of traffic flow data including traffic volume, speed and occupancy in a cycle of 2min, and obtaining a standard library result through a fuzzy C mean value clustering algorithm improved by a genetic algorithm:
wherein, V1，kThe cluster center is the traffic volume, speed and occupancy rate after normalization in a more crowded state; r is2，kThe clustering center of the traffic volume, the speed and the occupancy rate after normalization in the unblocked state; v3，kThe cluster center is the traffic volume, speed and occupancy rate after normalization in the congestion state; v4，kThe cluster center of the traffic volume, the speed and the occupancy rate after normalization in a common state.
The normalized standard library results were:
wherein, V1,kThe cluster center of the traffic volume, the speed and the occupancy rate in a more crowded state; v2,kA clustering center of traffic volume, speed and occupancy rate in a smooth state; v3,kThe cluster center is the traffic volume, speed and occupancy rate in the crowded state; v4,kThe cluster center of the traffic volume, speed and occupancy in the normal state.
At the beginning of each period, the euclidean distance between the traffic parameter vector normalized in step S1 and the standard library is calculated, pattern matching is performed, and the current traffic state is identified. Four traffic states are explained below:
when the traffic state identification result is in a smooth state, the average vehicle speed of the vehicle as the state clustering center is 82.05km/h, the traffic volume is 300veh/h, the occupancy is 0.021, and the Euclidean distance between the current traffic flow parameter vector and the clustering center is shortest. At the moment, an early confluence control measure is used for controlling a variable information board arranged in the middle section of the warning area to display information of 'right lane closing' and 'confluence here', and guiding the vehicle to finish confluence at the position; controlling a variable information board arranged at the tail end of the warning area to display 'right lane closed', and warning that the vehicle can not change lanes to the right lane any more; and controlling a variable information board arranged at the starting point of the transition area to display 'please walk slowly' and 'seal the right lane'. When the traffic state identification result is in a general state, the average vehicle speed of the vehicle as the state cluster center is 68.38km/h, the traffic volume is 570veh/h, the occupancy is 0.049, and the Euclidean distance between the current traffic flow parameter vector and the cluster center is shortest. At the moment, the decision uses the early confluence control measure, and the specific measure is the same as the above.
When the traffic state identification result is in a more crowded state, the average vehicle speed of the vehicle as the state clustering center is 38.56km/h, the traffic volume is 720veh/h, the occupancy is 0.135, and the Euclidean distance between the current traffic flow parameter vector and the clustering center is shortest. At the moment, a decision is made to use a late-stage confluence control measure, namely, the late-stage confluence control measure controls a variable information board arranged at the middle section of the warning area to display information of 'two lanes open' and 'driving to a confluence point', and prompts a driver not to finish confluence too early; controlling a variable information board arranged at the tail end of the warning area to display a 'lane keeping', guiding the vehicle to continuously run to a confluence point in the current lane, and warning the vehicle that the lane can not be changed randomly to jam a fleet of inner lanes; and controlling a variable information board arranged at the starting point of the transition area to display the 'confluence here' and the 'right lane closure'.
When the traffic state identification result is a congestion state, the average speed of the vehicles as the state clustering center is 23.04km/h, the traffic volume is 1050veh/h, the occupancy is 0.180, and the Euclidean distance between the current traffic flow parameter vector and the clustering center is shortest. At the moment, a decision is made to use a signal confluence control measure, namely, the signal confluence control measure controls a variable information board arranged at the middle section of the warning area to display information of 'two lanes open' and 'driving to confluence point', and prompts a driver not to finish confluence too early; controlling a variable information board arranged at the tail end of the warning area to display 'forbid lane changing', guiding the vehicle to continuously run to a confluence point in the current lane, and warning the vehicle that the lane can not be changed randomly to jam a fleet of the inner lane; and controlling a variable information board arranged at the starting point of the transition area to display a 'confluence here' and a 'compliance signal', prompting the vehicle to comply with the indication of the signal lamp, and guiding the vehicle to finish the confluence here according to the right-of-way traffic distributed by the signal lamp. The signaling scheme is as follows:
setting the free flow speed as v by actually measuring the average speed of the time span of the Yangyang high-speed Enping-Yangjiang section traffic flow in Guangdong province greater than 8 secondsf94km/h and a measured jam density k in the presence of a vehicle with a vehicle speed of 0j90pcu/km, calculated as D40 m; the signal timing scheme is a simple two-phase signal timing scheme and aims to divide the right of way for vehicles in an open lane and a closed lane. By actually measuring the peak hour flow of an open lane and a closed lane of an expansion construction area reconstructed at a Kaiyang high-speed Yingping-Yangjiang section in Guangdong province, the peak hour flow is q respectively1＝546veh/h，q2637veh/h, the calculation result is obtained by the calculation of step S3:
TABLE 2 Signal design
Finally, VISSIM simulation proves the effect of the confluence control decision, and the experimental result is as follows:
table 3 VISSIM simulation experimental results
In the present invention, the terms "first", "second", "third" and "fourth" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance. The term "plurality" means two or more unless expressly limited otherwise.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (2)
1. A confluence control strategy for rebuilding and expanding a highway construction area is characterized by comprising the following steps:
s1: collecting traffic information; collecting traffic characteristic parameters including traffic volume, speed and occupancy through a video detector arranged on a trip road section in a warning area, and carrying out normalization processing:
wherein, x'cdD traffic of c sample after normalizationCharacteristic parameter, xcdFor raw data before normalization, xd，maxIs the maximum value, x, in the d-th column datad，minIs the minimum value in the d-th column data;
when each period is finished, obtaining a traffic characteristic parameter matrix of the row c and the column d, carrying out normalization processing on the traffic characteristic parameter matrix, and calculating the average value of the row c and the column c of the matrix as a traffic characteristic parameter of each period to obtain a traffic parameter vector of the row 1 and the column d;
s2: when each period starts, calculating the Euclidean distance between the traffic parameter vector in the step S1 and a traffic state clustering center forming a traffic state identification standard library, performing mode matching, and identifying the current traffic state;
specifically, historical traffic states are divided into four types of smooth, general, crowded and crowded according to service levels to respectively represent four clustering results, a traffic state recognition standard library for real-time traffic state pattern recognition, which is composed of four types of centers, is obtained through fuzzy clustering analysis on historical traffic data of a reconstructed highway construction area, the traffic state recognition standard library is a 4-row and 3-column normalized matrix, 4-row data respectively represent four traffic states of smooth, general, crowded and crowded, 3-column data respectively represent three traffic characteristics of traffic volume, speed and occupancy, the constructed traffic state recognition standard library is specifically realized through fuzzy C mean value clustering analysis improved by a genetic algorithm, then real-time recognition on the traffic states of the reconstructed highway construction area is realized by adopting a pattern matching method, and the specific process is as follows:
s2-1: optimizing an initial clustering center of a fuzzy C-means clustering algorithm through a genetic algorithm, namely, firstly solving a globally optimal clustering center by using the genetic algorithm as the initial clustering center of the fuzzy C-means clustering algorithm, specifically, selecting the initial clustering center as a population individual to carry out optimization solution, setting basic parameters of the genetic algorithm, wherein M is population scale, T is termination evolution algebra, P iscIs cross probability, PmFor the variation probability, the population fitness function takes the reciprocal of the objective function, Zb，qThe fitness of the qth individual of the b generation, b ∈ [1, T]，q∈[1，M]The objective function and the constraint condition are as follows:
wherein, the traffic data sample set x is matrix of n x 3 dimension, the cluster center set V is matrix of 4 x 3 dimension, and u(b)As a function of membership at the b-th iteration,
the Euclidean distance from the jth traffic data sample to the ith class center at the b-th iteration,
the Euclidean distance sum of the jth traffic data sample to all the clustering centers during the b-th iteration; i is the serial number of the clustering center set, and j is the serial number of the sample set;
set of cluster centers at b +1 iteration, uijIs a membership matrix of the sample points,
the membership degree of the jth traffic data sample relative to the ith class center; dijEuclidean distance, d, from jth traffic data sample to ith class centerkjThe Euclidean distance sum, x, of the jth traffic data sample to all the cluster centersjTraffic parameter vector, v, for the jth traffic data sampleiIs the cluster center of the ith class, xj、viIs a normalized 3-dimensional vector; v. ofkIs the cluster center of all classes;
s2-2: initializing a clustering center as a primary population, and updating membership and objective function values;
s2-3: encoding the population; binary coding is carried out on individuals of each population, and the number of coding bits is determined by setting coding precision and decoding space of each gene chain; after the coding is completed, the gene coding chain GN of the qth individual of the primary population is obtainedb，qWhen b is 0, q is ∈ [1, M ]]；
S2-4: carrying out roulette selection, single-point crossing and mutation operations on individual gene chains;
s2-5: updating an algebra b which is b +1, decoding the new population and updating the membership degree, the clustering center and the objective function value;
s2-6: judging the algebra, if b is less than T, returning to the step S2-3, otherwise, executing the step S2-7;
s2-7: decoding the individual with the maximum population fitness of the last generation to obtain the membership UbCluster center VbThe optimal solution of (2);
s2-8: clustering genetic algorithm to center VbThe optimal solution of the fuzzy C mean value clustering is used as an initial clustering center of the fuzzy C mean value clustering, the number C of the clustering centers is set to be 4, the threshold epsilon of iteration stop is set to be 0 at the moment;
s2-9: updating the membership function according to the cluster center and the formula (4), if a certain traffic data sample x is foundjWith a certain cluster centre viIf the Euclidean distance of the traffic data sample vector is 0, the membership degree from the traffic data sample vector to the clustering center is 1, and the other membership degrees are 0; otherwise, the updating process is as follows:
wherein the content of the first and second substances,
the membership degree of the jth traffic data sample relative to the ith class center;
is the b th iterationThe Euclidean distance from j traffic data samples to the class i center,
the Euclidean distance sum of the jth traffic data sample to all the clustering centers during the b-th iteration;
s2-10: updating the clustering center V according to the updated membership degree by substituting formula (5)(b+1)；
S2-11: checking the iteration termination condition V(b+1)-V(b)If | < epsilon, executing step S2-12, otherwise, returning to step S2-9;
s2-12: outputting the optimal membership degree and the clustering center, completing four types of traffic state division of historical traffic data, and obtaining a traffic state identification standard library which is composed of the four types of clustering centers and used for real-time traffic state mode identification, wherein the traffic state identification standard library is a 4-row and 3-column normalized matrix, 4 rows of data respectively represent four types of traffic states, and 3 columns of data respectively represent three traffic characteristics of traffic volume, speed and occupancy;
s2-13: at the beginning of each period, completing pattern matching according to the Euclidean distance minimum principle of a traffic parameter vector of the previous period obtained by real-time acquisition and normalization processing and a cluster center of the current affiliated traffic state, and realizing real-time identification of the traffic state, wherein the formula of the Euclidean distance d (A, B) is as follows:
wherein, muA(μi) Is the ith element, mu, in the A point vectorB(μi) Is the ith element in the B point vector;
s3: forming a confluence strategy decision mechanism according to the current traffic state, selecting a corresponding confluence control measure, and controlling the combination of the variable information board and the signal lamp to be switched on and off;
the confluence strategy decision mechanism takes 2min as a period, and judges the traffic state at the moment by calculating the Euclidean distance between the traffic parameter vector acquired in the last period and obtained by normalization processing and each clustering center of the traffic state identification standard library constructed in the step S2 when each period starts; after the judgment of the traffic state is finished, an early confluence measure, a late confluence measure and a signal confluence measure are respectively and correspondingly selected according to three control scenes, namely unblocked control, general control, crowded control and crowded control; specifically, the method comprises the following steps:
when the traffic state is recognized as a clear, general state in step S2, it is decided to use an early merge control means that controls a variable information panel arranged in the middle section of the warning area to display information of "right lane closure" and "merge here", and guides the vehicle to complete the merge at that position; controlling a variable information board arranged at the tail end of the warning area to display 'right lane closed', and warning that the vehicle can not change lanes to the right lane any more; controlling a variable information board arranged at the starting point of the transition area to display 'please slow down' and 'close the right lane', guiding the vehicle to decelerate, and not changing the lane at will, so as to finish early guiding the vehicle converging behavior;
when the traffic state is identified as a more congested state in step S2, it is decided to use a late-stage confluence control measure, which controls a variable message board disposed at the middle section of the warning area to display information of "two lanes open" and "traveling to a confluence point" to prompt the driver not to finish confluence prematurely; controlling a variable information board arranged at the tail end of the warning area to display a 'lane keeping', guiding the vehicle to continuously run to a confluence point in the current lane, and warning the vehicle that the lane can not be changed randomly to jam a fleet of inner lanes; controlling a variable information board arranged at the starting point of the transition area to display 'confluence here' and 'right lane closure', guiding the vehicles to finish confluence here, and finishing late control on vehicle confluence behavior;
when the traffic state is identified as a congestion state in step S2, it is decided to use a signal confluence control measure, which controls a variable message board arranged at the middle section of the warning area to display information of "two lanes open" and "travel to confluence point" to prompt the driver not to finish confluence too early; controlling a variable information board arranged at the tail end of the warning area to display 'forbid lane changing', guiding the vehicle to continuously run to a confluence point in the current lane, and warning the vehicle that the lane can not be changed randomly to jam a fleet of the inner lane; controlling a variable information board arranged at the starting point of the transition area to display 'confluence here' and 'compliance signal', prompting the vehicle to comply with the indication of a signal lamp, and simultaneously guiding the vehicle to finish confluence here according to the right of way distributed by the signal lamp, so as to finish signal control on the vehicle jammed on the bottleneck road section;
in a crowded state, the signal confluence control measures cooperate with the signal lights, the variable information board and the road temporary marking to control:
the temporary road marking is used for dividing a waiting area and a stop line, the waiting area refers to a queuing area of the vehicle during signal control, and the stop line is a signal control section of the road section;
the signal transition area is the distance between the position of the stop line and the starting point of the transition area, the length of the signal transition area is set to be 25m, and the layout position of the stop line is the position 25m away from the starting point of the transition area;
the waiting area refers to an area for a vehicle parking waiting signal separated by a white line, the length of which affects the traffic efficiency and the merging behavior of the vehicle, the length D of the waiting area passing through the maximum queuing length DLength of queueDetermining, according to the parking wave theory:
wherein v isfFor free flow velocity, kjFor the blocking density, t' is taken as the maximum red light duration, kfIs the actual density;
the signal timing scheme is a two-phase signal timing scheme, and aims to divide the right of way for vehicles in an open lane and a closed lane, and the calculation process is as follows:
s3-1: statistics of the reconstruction and expansion highwayConstruction area road section open lane peak hour flow q1And peak hourly flow q of closed lane2；
S3-2: setting the right of the open lane as phase one and the right of the closed lane as phase two, and calculating the phase-critical flow rate ratio y1Phase two key flow rate ratio y2And sum of key flow rate ratios Y:
Y＝y1+y2
wherein S isTIs the straight-channel saturation flow rate;
s3-3: calculate the yellow light duration A of phase one1Full red duration r1And duration of green light interval I1Calculating the yellow light time length A of the phase two2Full red duration r2And duration of green light interval I2：
I1＝A1+r1
I2＝A2+r2
Where t is the driver reaction time, v85In order to reconstruct and expand the highest speed limit of a highway section in a highway construction area, a is automobile deceleration, g is gradient, w is the distance from a stop line to the tail end of a transition area, L' is automobile standard length, v1515% vehicle speed;
s3-4: meterSignal loss time L for phase one calculation1Phase two signal loss time L2Total signal loss time L and signal cycle duration C:
L1＝l1+r1
L2＝l2+r2
L＝L1+L2
wherein l1And l2The start-up losses for phase one and phase two, respectively;
s3-5: determining green light assignment:
g1＝gE1+l1-A1
g2＝gE2+l2-A2
wherein, gE1Effective green duration of phase one, gE2Effective green duration, g, for phase two1Green time duration of phase one, g2A green time period for phase two;
to sum up, the signal period duration C and the phase-one green duration g are obtained1And yellow light duration A1Green time duration g of phase two2And yellow light duration A2The two phases jointly form a signal timing scheme;
s4: when the current cycle ends, the process returns to step S2.
2. A method for rebuilding and extending height according to claim 1The merging control strategy of the construction area of the expressway, wherein S is performed in step S3T1650veh/h is taken, and t is taken as 1 s; l' is 5-6m, L1And l2Take 3 s.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110054207.7A CN112885085B (en) | 2021-01-15 | 2021-01-15 | Confluence control strategy applied to reconstruction and extension of highway construction area |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110054207.7A CN112885085B (en) | 2021-01-15 | 2021-01-15 | Confluence control strategy applied to reconstruction and extension of highway construction area |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112885085A CN112885085A (en) | 2021-06-01 |
| CN112885085B true CN112885085B (en) | 2022-04-29 |
Family
ID=76048095
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110054207.7A Active CN112885085B (en) | 2021-01-15 | 2021-01-15 | Confluence control strategy applied to reconstruction and extension of highway construction area |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112885085B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113643534B (en) * | 2021-07-29 | 2023-04-18 | 北京万集科技股份有限公司 | Traffic control method and equipment |
| CN116311940B (en) * | 2023-03-23 | 2024-03-26 | 东南大学 | Dynamic traffic guidance system and method for expressway reconstruction and expansion operation area |
| CN116663964B (en) * | 2023-04-28 | 2023-11-03 | 长电新能有限责任公司 | Engineering unit price rapid calculation method and system based on characteristic values of list items |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104103180A (en) * | 2014-07-15 | 2014-10-15 | 同济大学 | Urban expressway on-ramp and main line collaborative signal control system and method |
| CN104778847A (en) * | 2015-04-09 | 2015-07-15 | 南京全司达交通科技有限公司 | Method for controlling traffic signals under closed condition of lane on outermost side of multi-lane highway |
| CN107945544A (en) * | 2017-12-12 | 2018-04-20 | 常州大学 | A kind of artificial intelligence system based on Dynamic Vehicle assignment of traffic traffic lights |
| CN108510758A (en) * | 2018-05-24 | 2018-09-07 | 金陵科技学院 | City expressway Entrance ramp signal control method based on video real time information |
| CN108765956A (en) * | 2018-06-14 | 2018-11-06 | 广东工业大学 | A kind of traffic status of express way comprehensive estimation method |
| CN109377785A (en) * | 2018-10-22 | 2019-02-22 | 北京航空航天大学 | A kind of highway merging area vehicle collision method for early warning based on deep learning |
| CN110570675A (en) * | 2019-10-17 | 2019-12-13 | 中国公路工程咨询集团有限公司 | roadside control system of highway construction area under vehicle-road collaborative environment |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100444210C (en) * | 2007-04-20 | 2008-12-17 | 东南大学 | Mixed controlling method of single dot signal controlling crossing |
| US9830813B2 (en) * | 2013-06-18 | 2017-11-28 | Carnegie Mellon University, A Pennsylvania Non-Profit Corporation | Smart and scalable urban signal networks: methods and systems for adaptive traffic signal control |
| CN106548633A (en) * | 2016-10-20 | 2017-03-29 | 中国科学院深圳先进技术研究院 | A kind of variable guided vehicle road control method of road network tide flow stream |
| CN108842652B (en) * | 2018-07-31 | 2020-11-27 | 东南大学 | Variable confluence marking line of ramp confluence area of expressway and setting method thereof |
| CN109191873B (en) * | 2018-10-15 | 2020-02-21 | 深圳市市政设计研究院有限公司 | Method and device for setting pre-signal lamp at traffic confluence point |
| CN109859500B (en) * | 2018-12-17 | 2021-02-02 | 浙江海康智联科技有限公司 | High-speed confluence area safety early warning method based on vehicle-road cooperation |
| CN110634299B (en) * | 2019-10-25 | 2022-03-25 | 福州大学 | Urban traffic state fine division and identification method based on multi-source track data |
| CN111462491B (en) * | 2019-12-10 | 2022-03-01 | 北京航空航天大学 | Expressway confluence area traffic conflict early warning method based on ramp control |
| CN111292534A (en) * | 2020-02-13 | 2020-06-16 | 北京工业大学 | Traffic state estimation method based on clustering and deep sequence learning |
| CN112085947B (en) * | 2020-07-31 | 2023-10-24 | 浙江工业大学 | Traffic jam prediction method based on deep learning and fuzzy clustering |
-
2021
- 2021-01-15 CN CN202110054207.7A patent/CN112885085B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104103180A (en) * | 2014-07-15 | 2014-10-15 | 同济大学 | Urban expressway on-ramp and main line collaborative signal control system and method |
| CN104778847A (en) * | 2015-04-09 | 2015-07-15 | 南京全司达交通科技有限公司 | Method for controlling traffic signals under closed condition of lane on outermost side of multi-lane highway |
| CN107945544A (en) * | 2017-12-12 | 2018-04-20 | 常州大学 | A kind of artificial intelligence system based on Dynamic Vehicle assignment of traffic traffic lights |
| CN108510758A (en) * | 2018-05-24 | 2018-09-07 | 金陵科技学院 | City expressway Entrance ramp signal control method based on video real time information |
| CN108765956A (en) * | 2018-06-14 | 2018-11-06 | 广东工业大学 | A kind of traffic status of express way comprehensive estimation method |
| CN109377785A (en) * | 2018-10-22 | 2019-02-22 | 北京航空航天大学 | A kind of highway merging area vehicle collision method for early warning based on deep learning |
| CN110570675A (en) * | 2019-10-17 | 2019-12-13 | 中国公路工程咨询集团有限公司 | roadside control system of highway construction area under vehicle-road collaborative environment |
Non-Patent Citations (2)
| Title |
|---|
| 基于改进FCM聚类的交通控制时段划分;于德新 等;《华南理工大学学报(自然科学版)》;20161231;第44卷(第12期);第53-59页 * |
| 高速公路施工区交通组织与作业段长度优化研究;朱永光;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20110315;第16-21页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112885085A (en) | 2021-06-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112885085B (en) | Confluence control strategy applied to reconstruction and extension of highway construction area | |
| CN105913661B (en) | A kind of express highway section traffic state judging method based on charge data | |
| CN109410577B (en) | Self-adaptive traffic control subarea division method based on space data mining | |
| CN110728842A (en) | Abnormal driving early warning method based on reasonable driving range of vehicles at intersection | |
| CN104778834B (en) | Urban road traffic jam judging method based on vehicle GPS data | |
| CN105825669B (en) | A kind of system and method for identifying urban expressway traffic bottleneck | |
| CN109523786A (en) | A kind of traffic accidents genetic analysis method | |
| CN107492251A (en) | It is a kind of to be identified and driving condition supervision method based on the driver identity of machine learning and deep learning | |
| CN105355049B (en) | A kind of highway evaluation of running status method based on macroscopical parent map | |
| CN108848462A (en) | Real-time vehicle trajectory predictions method suitable for signalized crossing | |
| CN105225500A (en) | A kind of traffic control aid decision-making method and device | |
| CN106205156A (en) | A kind of crossing self-healing control method for the sudden change of part lane flow | |
| CN104951764A (en) | Identification method for behaviors of high-speed vehicle based on secondary spectrum clustering and HMM (Hidden Markov Model)-RF (Random Forest) hybrid model | |
| CN109345832B (en) | Urban road overtaking prediction method based on deep recurrent neural network | |
| CN108922168B (en) | A kind of mid-scale view Frequent Accidents road sentences method for distinguishing | |
| CN110363997A (en) | One kind having construction area intersection signal timing designing method | |
| CN113516854B (en) | Multi-interface coordination self-adaptive control method based on police card and video detector | |
| CN109118787A (en) | A kind of car speed prediction technique based on deep neural network | |
| CN113436432A (en) | Method for predicting short-term traffic risk of road section by using road side observation data | |
| CN108961788A (en) | Traffic lights wisdom transform method | |
| CN106683442A (en) | Multi-index based intersection signal timing plan evaluation method | |
| CN111754771B (en) | Individual travel time prediction method based on traffic signals and density delay | |
| CN111882858A (en) | Method for predicting queuing length of expressway abnormal events based on multi-source data | |
| CN112990544A (en) | Traffic accident prediction method for expressway intersection area | |
| Prontri et al. | Traffic signal control using fuzzy logic |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |