CN103280113B - Self-adaptive intersection signal control method - Google Patents

Abstract

The invention belongs to the field of urban traffic control and provides an adaptive intersection signal control method. The method detects the number of vehicles by installing a vehicle monitor, obtains the vehicle queuing length, and determines the next cycle through the intersection vehicle queuing length in the previous period. Periodic signal control strategy. Use the vehicle detectors that have been laid in the city to obtain traffic flow data and calculate the queuing length of each lane, and dynamically adjust the signal timing scheme for the next cycle according to the different proportions of queuing length. The method calculates the timing scheme with the period as the unit, and the calculation amount is small, and it can achieve the purpose of adaptive control of the intersection signal according to the traffic flow density. In addition, the algorithm of this method is simple, and the requirements for hardware and engineering quantity are low, which is convenient for promotion.

Description

A self-adaptive intersection signal control method

technical field

The invention belongs to the field of urban traffic control, and in particular relates to an adaptive intersection signal control method.

Background technique

Traffic signal control takes the microscopic traffic flow at the intersection as the control object, allocates the right of way of the vehicle at the intersection with the phase and time as the scale, separates the conflicting traffic flow from the time, and conducts reasonable guidance and scheduling of the traffic flow, so as to reduce traffic accidents and improve traffic safety. The purpose of traffic safety and transportation efficiency.

At present, there are mainly three types of signal control methods at intersections at home and abroad: timing control, induction control and adaptive control. Timing control is suitable for intersections with large traffic volume and uniform traffic distribution; induction control is suitable for small traffic volume and variable traffic. Uneven intersections, but when the traffic volume is heavy, it is impossible to adjust and optimize the signal state according to traffic demand to alleviate traffic congestion; adaptive control is based on the traffic volume data sent by the detector, real-time generation of the best performance under a certain performance index The optimal timing scheme automatically adjusts the signal timing parameters such as green signal ratio, cycle length and green time difference, so that the signal timing can adapt to the changing traffic flow. At present, SCOOT and SCATS are two mainstream adaptive control methods for urban traffic lights method. The SCATS system adjusts the signal control parameters in real time according to the traffic flow changes, but the number of options is limited, and it takes a long time to execute each plan (change every ten minutes), and it does not have a feedback function, and the detection equipment is placed in front of the stop line. Unable to detect the queue length of vehicles; the SCOOT system is highly practical and can continuously fine-tune the timing parameters, but its related control models are all simulated through mathematical models. The more complex the mathematical model, the longer the simulation time required , there will be a contradiction between reliability and real-time performance. In addition, both SCOOT and SCATS systems need to collect a large amount of data, and the construction costs are high.

The queuing length of vehicles at an intersection is an important indicator to characterize the traffic state. At present, the optimal real-time signal control is mainly solved based on the queuing length of each entrance lane at the intersection. However, calculating the queue length is complicated, usually relying on multiple detectors or using video image data, real-time image processing requires a lot of calculation, is expensive, and is easily affected by weather, and the reliability of the system is poor. At present, most cities have laid induction coils and other equipment to obtain real-time traffic information such as traffic volume, vehicle occupancy time, road occupancy rate, congestion level and queue length, but the traffic demand in all directions at the intersection is not necessarily balanced. The queue lengths are not necessarily equal, especially in the case of heavy traffic, when the detector is occupied by a vehicle in the queue for a long time, it is impossible to estimate the number of vehicles arriving in real time and the queue length and other information, which leads to vehicle queuing The length calculation accuracy is not high, and the signal timing scheme is not accurate. If multiple detectors are installed on each lane, it will not only cause multiple damages to the original road surface, but also cause a large amount of engineering, inconvenient installation, and low engineering feasibility and implementation operability.

Contents of the invention

Aiming at the defects and deficiencies in the prior art, the object of the present invention is an adaptive intersection signal control method, which realizes intersection signal control by the vehicle queuing length at the intersection in the previous cycle.

An adaptive intersection signal control method, the method is carried out according to the following steps:

Step 1, install a vehicle detector on each lane of the intersection, the distance between the vehicle detector and the stop line is denoted as L, and each vehicle detector is used to collect the red light turning-on moment of each phase nth cycle The moment when the green light turns on in the nth cycle The moment when the red light turns on in the n+1th cycle The occupancy time t _o of each vehicle passing the vehicle detector, the idle time interval t _f of the vehicle detector and the number n of vehicles passing the detector;

The phases include the straight phase of the straight queue and the left phase of the left queue;

A cycle is from the moment when the red light turns on to the time when the green light goes out, record G _min as the minimum green light time, record g _lay as the unit green light extension or shortening time, record G _main as the maximum total green light time of the main road;

Step 2: At the end of a period of effective green light time, the vehicle detector transmits the detection data to the background data processing system, analyzes the changes in the traffic state in this period, and finds out the transition time T _A , T _B and T _C ;

T _A is the state transition time when vehicle v1 starts to occupy the vehicle detector for a long time after the red light is turned on. After T _A , the vehicle detector is occupied by vehicle v1 for a long time. Time T _A can be used to determine whether there is a long vehicle queue;

T _B is the transition time when vehicle v1 leaves the vehicle detector after the green light is turned on. After the green light is turned on and before time T _B , the vehicles in front of the stop line leave the intersection one after another at a saturated flow rate, but the vehicles passing the detector The number is 0, after _TB , the vehicle passes the vehicle detector at the maximum flow rate, and the occupancy time and interval time of the vehicle detector decrease;

T _C is the state transition moment when the last vehicle v2 in the vehicle queuing queue passes the vehicle detector. The basis for determining the last vehicle _v2 in the queuing queue is that the time interval between consecutive vehicles passing the detector in the queuing queue is relatively small. The time interval between the car and its following car passing the detector is greater than the threshold A, indicating that the following car is not in this queue;

According to the statistical results of the current traffic survey data, the value of the threshold A is 3s;

Step three, adopt from Calculate the traffic flow density k _j in the saturated condition from the number of vehicles n _L passing through the detector from time T to time T _A , then k _j = n _L /L; Calculate the dissipation speed v of the vehicle queuing queue from the time period from time to _TB and the distance L between the stop line and the vehicle detector, then Then calculate the maximum queuing length Q in this cycle by the vehicle number n _j of the detector in the T _B and T _C time periods;

(1) If there is no time T _A , then Q<L;

(2) If there is a time T _A , then further determine the time T _C ;

(a) If there is T _C moment, then Q=L+n _j /k _j ;

(b) If there is no T _C time, the time when the red light will be turned on in the next cycle As the transition time T _C , statistics T _B and The number n _j of vehicles passing the detector within the time period, then Q＞L+n _j /k _j ;

Step 4, adjust the signal control strategy of the main road, the specific process is as follows:

Calculate the straight line length Q _S and the left turn queue length Q _L according to the method of step 3 respectively;

(1) If the lengths of the two queues are both less than L, the minimum green light time is assigned to both the left-turn phase time and the straight-going phase time, that is, T _g1 =T _g2 =G _min ;

(2) If there is only one queue whose length is less than L, then the phase of the queue is assigned the minimum green light time T _g1 =G _min , and another queue whose length is greater than or equal to L needs to determine whether there is a transition time T _C in this queue:

(a) If there is a transition time T _C , then according to T _C and Calculate the minimum speed v _m required by the vehicle to pass L in this time interval, then there is Determine if there is a green light slack:

a) If v _m < v, there is a surplus time for the green light, and the original green light duration is subtracted from the unit green light change time g _lay , that is, T _g2 =T _g2 -g _lay ;

b) If v _m ≥ v, there is no surplus time for the green light, that is, there is a remaining queue. If T _g2 <G _main -T _g1 , then add g _lay to the original green light duration, that is, T _g2 =T _g2 +g _lay , if T _g2 = G _main - T _g1 , then T _g2 = T _g2 ;

(b) If there is no transition time T _C , it means that at the end of the green light time, there is still a queuing queue upstream of the detector, and the green light duration of this phase is T _g2 =G _main -T _g1 ;

(3) If the queue lengths of the two queues are both greater than or equal to L, then determine whether there is a transition time T _C in each queue, then:

(a) If time T _C exists in both, determine whether each phase has a surplus green light time, and follow the strategy (a) of step (2) to extend or reduce the green light duration T _g1 and T _g2 , where: T _g1 +T _g2 ≤ G _main ;

(b) If there is a time T _C in one queue, and there is no time T _C in the other queue, then the phase of the time T _C exists according to the strategy (a) of step (2) to extend or reduce the green light duration T _g1 , and there is no time The green light time of the phase of T _C is T _g2 =G _main -T _g1 ;

(c) If there is no transition time T _C in both, the green light time of the left-turn phase and the straight-go phase is allocated according to the length ratio of Q _S and Q _L , that is $\left\{\begin{array}{c}{T}_L=G_{\mathrm{main}}\&times;\frac{Q_L}{Q_L+Q_S}\\ T_S=G_{\mathrm{main}}\&times;\frac{Q_S}{Q_L+Q_S}\end{array}\right.$

Among them: T _L is the green light duration of the left-turning phase, T _S is the green light duration of the straight-going phase, Q _S is the queue length of the straight-going phase, and Q _L is the queue length of the left-turning phase;

Step five, adjust the signal control strategy of the secondary trunk road, the specific process is as follows:

Going straight and turning left on the secondary road are controlled by a phase. Record G _minor as the maximum green light time of the secondary road, G _min as the minimum green light time, and g _lay as the unit of green light extension or shortening time. First, calculate the secondary road according to the method of step 3 Queue length Q, T _G is the green light time of the secondary trunk road;

(1) If the queue length Q<L, assign the shortest green light time T _G =G _min ;

(2) If the queue length Q≥L, determine whether there is a transition time T _C , then:

(a) If there is a transition time T _C , then according to T _C and Calculate the minimum speed v _m required by the vehicle to pass L in this time interval, then there is Determine if there is a green light slack:

a) If v _m < v, there is a green light surplus time, then T _G =T _G -g _lay ;

b) If v _m ≥ v, there is no surplus time for the green light, which means that there is a remaining queue. If T _G <G _minor , then T _G =T _G +g _lay , if T _G =G _minor , then T _G =T _G ;

(b) If there is no transition time T _C , then T _G =G _minor .

Compared with the prior art, the present invention has the beneficial effects of:

The method of the invention utilizes the laid vehicle detectors in the city to obtain traffic flow data, calculates the queue length of each lane, and dynamically adjusts the signal timing scheme of the next cycle according to the different ratios of the queue length. The method calculates the timing scheme with the period as the unit, and the calculation amount is small, and it can achieve the purpose of adaptive control of the intersection signal according to the traffic flow density. In addition, the algorithm of this method is simple, and the requirements for hardware and engineering quantity are low, and the existing equipment does not increase the construction cost, which is convenient for promotion.

Description of drawings

Figure 1 is a schematic diagram of the layout of vehicle detectors at the intersection of major and minor arterial roads.

Fig. 2 is a schematic diagram of the vehicle queuing queue changing with time in a lane.

Fig. 3 is a schematic diagram of the time-varying change of the queue with a transition time T _C and a surplus green light time during the dissipating process of the vehicle queuing queue.

Fig. 4 is a schematic diagram of the time-varying change of a queue with a transition moment T _C but no spare green light time during the dissipating process of the vehicle queue.

Fig. 5 is a schematic diagram of the time-varying change of the queue without the transition moment T _C in the process of dissipating the vehicle queuing queue.

Fig. 6 is a flow chart of trunk road signal adaptive control.

Fig. 7 is a flow chart of adaptive control of secondary trunk road signals.

The specific content of the present invention will be described in further detail below in conjunction with the accompanying drawings and embodiments.

Detailed ways

It should be noted that the dimensions of the physical quantities given in this embodiment are unified standard dimensions in the field, and those skilled in the art can know the dimensions according to the physical quantities, so the following content will not be given one by one.

It should be noted that a cycle is from the time when the red light is turned on to the time when the green light is extinguished. According to the current traffic rules, vehicles within the stop line cannot pass when the yellow light is flashing. Therefore, the flashing time of the yellow light in the present invention is included in the red light In the duration, the yellow light will no longer be considered in the following embodiments.

According to the above technical scheme, as shown in Figure 1, it is a schematic diagram of the layout of vehicle detectors at the intersection of main and secondary arterial roads. The direction is a secondary road, and a vehicle detector is installed at a distance Ls from the intersection. The main road has a left-turn phase and a straight-go phase, and a secondary road is controlled by a phase for straight-going and left-turning. There is a signal controller at the intersection, and all vehicle detectors detect the traffic volume, detector occupancy time, and idle time and transmit them to the background data processing system. The real-time data processing system in the background dynamically determines the signal timing scheme for the next cycle according to the calculation results, and transmits it to the signal controller.

Since the calculation method of vehicle queuing length on the main road and the secondary road is the same, the following scheme takes a main road as an example for control, and the distance between the detector and the stop line is L.

As shown in Figure 2, it is a schematic diagram of a lane queuing queue changing with time. When the red light is turned on, the vehicles start to stop and queue until the green light is on, and the queuing queue begins to dissipate, and the cycle repeats like this. When there is no transition time point T _A within the time period of the red light, it indicates that the queue length Q<L, otherwise, it indicates that the queue overflows the detector. according to The number of queued vehicles n _L in the time period from T to _A calculates the traffic density under saturation:

k _j =n _L /L

From the moment of green light At time T _B , the queue leaves the intersection at a saturated flow rate, and the dissipation speed of the queue is:

$\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; L</span>}<span\; class="notranslate">\; /</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; B</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; G</span>}}^{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span>$

As shown in Figure 3, there is a transition time T _C in the queuing queue dissipation process and there is a schematic diagram of the queue with a surplus green light time changing with time; Figure 4 shows a transition time T _C but no surplus green light time in the queuing queue dissipation process. Schematic representation of queue evolution over time. According to T _C and interval calculation If v _m <v, there is a surplus green light time as shown in Fig. 3; if v _m ≥ v, there is no surplus green light time as shown in Fig. 4 . Use the number n of vehicles passing the vehicle detector during the period T _B and T _C to calculate the maximum queue length in this period:

Q＝L+n/k _j

As shown in Figure 5, it is a schematic diagram of the queue changing with time when there is no transition time T _C in the queue dissipation process. The absence of time T _C indicates that when the effective green light time ends, there is still a waiting queue behind the detector. See T _C , count T _B and The number n _j of vehicles passing the detector in the time period, the queue length is:

Q＞L+n _j /k _j

Next, determine the parameters in the control scheme. Since the period of an intersection signal cannot be too long or too short, if it is too short, the switching of each phase will be frequent, the number of vehicles parked will increase and the vehicles will start and stop frequently. If it is too long, then The waiting time of the vehicles in the queue is too long, so in this scheme, the minimum green light time G _min (generally 15s), the unit green light change time g _lay (3s), and the maximum total green light time G _main of the main road (generally 15s) are set. 80s), the maximum green light time G _minor of secondary roads (generally 45s).

As shown in Figure 6, it is the flow chart of signal adaptive control on the main road. First, calculate the length of the straight queue Q _S and the length of the left turn queue Q _L , and then carry out signal control according to the following steps:

(1) If the lengths of the two queues are both less than L, then the left-turn phase and the straight-going phase are assigned the minimum green light time, that is, T _g1 =T _g2 =G _min ;

(2) If the length of a certain queue is less than L, then the phase allocates the minimum green light time T _g1 =G _min , and the length of another queue is greater than or equal to L, then determine whether there is T _C in this queue:

(a) If there is a transition time T _C , then according to T _C and Calculate the minimum speed required for the vehicle to pass through L in the time interval Determining if there is excess green light time

a) If v _m < v, there is a surplus time for the green light, then subtract the unit green light change time g _lay from the original green light duration, that is, T _g2 = T _g2 -g _lay ;

b) If v _m ≥ v, there is no surplus time for the green light, it means that there is a remaining queue. If T _g2 <G _main -T _g1 , then add the unit green light change time g _lay to the original green light duration, that is, T _g2 =T _g2 +g _lay , if T _g2 =G _main -T _g1 , then T _g2 =T _g2 ;

(b) If there is no transition time T _C , it means that at the end of the green light time, there is still a queuing queue upstream of the detector, and the green light duration of this phase is T _g2 =G _main -T _g1 ;

(3) If the queue lengths of the two queues are both greater than or equal to L, then determine whether there is a state transition time T _C in each queue:

(a) If there is a transition time T _C in both, determine whether each phase has a surplus green light time, and implement the green light extension or reduction according to the strategy (a) of step (2);

(b) If one has a transition time T _C and the other does not exist, then the phase with a transition time follows the strategy (a) of step (2) to extend or reduce the green light duration T _g1 , and the green light time of the other phase is T _g2 = G _main -T _g1 ;

(c) If there is no transition time T _C in both, the green light time of the left-turn phase and the straight-go phase is allocated according to the length ratio of Q _S and Q _L , that is $\left\{\begin{array}{c}{T}_L=G_{\mathrm{main}}\&times;\frac{Q_L}{Q_L+Q_S}\\ T_S=G_{\mathrm{main}}\&times;\frac{Q_S}{Q_L+Q_S}\end{array}\right..$

Among them: T _L is the green light duration of the left-turn phase, T _S is the green light duration of the straight-going phase, Q _S is the queue length of the straight-go phase, and Q _L is the queue length of the left-turn phase.

As shown in Figure 7, it is the flow chart of adaptive control of the secondary trunk road signal. The straight line and left turn of the secondary trunk road are controlled by one phase, so the queue length Q of the secondary trunk road is calculated first,

(1) If the queue length Q<L, assign the shortest green light time T _G =G _min ;

(2) If the queue length Q≥L, determine whether there is a transition time T _C , then:

(a) If there is a transition time T _C , then according to T _C and Calculate the minimum speed v _m required by the vehicle to pass L in this time interval, then there is Determine if there is a green light slack:

a) If v _m < v, there is a green light surplus time, then T _G =T _G -g _lay ;

b) If v _m ≥ v, there is no surplus time for the green light, which means that there is a remaining queue. If T _G <G _minor , then T _G =T _G +g _lay , if T _G =G _minor , then T _G =T _G ;

(b) If there is no transition time T _C , then T _G =G _minor .

Specific embodiments of the present invention are provided below, and it should be noted that the present invention is not limited to the following specific embodiments, and all equivalent transformations done on the basis of the technical solutions of the present application all fall within the scope of protection of the present invention.

Example:

As shown in Figure 1 at the intersection, the east-west direction is set as the main road, the distance between the vehicle detector and the stop line is L _a =80m, the north-south direction is the secondary road, the distance between the vehicle detector and the stop line is _Lb =60m, and the distance between the vehicle detector and the stop line is Lb=60m. The maximum total green light duration G _main is 80s, the maximum green light duration G _minor of secondary roads is 45s, the minimum green light duration G _min is 15s, and the unit green light change duration g _lay is 3s.

Table 1 shows the timing status of signal lamps obtained from 10:00:00 in this embodiment.

hypothetical moment It is the starting moment of a red light on the main road, and at this moment, the vehicle detector data and signal light phase transformation data of the previous period are collected statistically, such as: the transition time point T _A of the straight road at 9:59:10, and the transition time T _B of the straight road 9:59 :24, T _C 9:59:30, the transition time of the straight lane, the start time of the straight green light Turn left when the green light starts The equivalent number of vehicles passing the vehicle detector within the time period from T _A to n _L =20, the equivalent number of vehicles passing through the vehicle detector within the time from T _B to T _C is n _j =2, there is no transition time T _A and T on the left turn _B , T _C. Calculate saturated traffic density: k _j =n _L /L _a =20/80=0.25, queue dissipation speed and $v_m=L_a/(T_{G,L}^0-T_C)=80/(44-30)=5.71m/\mathrm{the\; s},$ Since vm<v, there is a surplus of green light time, the queue length Q _S =L _a +n _j /k _j =80+2/0.25=88m, so the next cycle straight green light time length T _S =T _S -g _lay =30s-3s =27s, the green light time for left turn is T _L =G _min =15s.

while the moment It is the start time of the red light on the secondary road. At this moment, the vehicle detector data and signal light phase transformation data of the previous cycle are collected statistically, such as: the start time of the green light is 10:00:00, the duration of the green light T _G is 30s, and there is no transition point T _A , so the duration of the green light in the next cycle is T _G =G _min =15s;

time It is the starting time of another red light on the main road, and at this moment, the vehicle detector data and signal light phase transformation data of the previous period are collected statistically, such as: the starting time of the straight green light It is 10:00:31, there is no transition time T _A , and the time when the red light turns left 10:00:00, left turn transition time T _A 10:00:47, left turn transition time T _B 10:01:08, left turn transition time T _C 10:01:11, left turn green light start time The equivalent number of vehicles passing the vehicle detector within the time period from T _A is n _L =19, and the equivalent number of vehicles passing the vehicle detector within the time period from T _B to T _C is n _j =1. Calculate the saturated traffic density of the left-turning lane: k _j = n _L /L _a = 19/80 = 0.23, queue dissipation speed $v=L_a/(T_B-T_{G,L}^1)=80/10=8m/\mathrm{the\; s}$ and $v_m=L_a/(T_{R,L}^2-T_C)=80/2=40m/\mathrm{the\; s},$ Since v _m □v has no spare green light time, Q _L ＝L _a +n _j /k _j ＝80+1/0.23＝84m, so the green light duration for going straight in the next cycle is T _S ＝G _min ＝15s, and the green light duration for left turns is T _L =T _L +g _lay =15s+3s=18s;

time It is the start time of another red light on the sub-arterial road. At this moment, the vehicle detector data and signal light phase transformation data of the previous cycle are collected statistically, such as: red light start time is 10:00:31, the transition time T _A is 10:01:02, the time when the green light starts is 10:01:13, the transition time T _B is 10:01:21, and the transition time T _C is 10:01:25, The equivalent number of vehicles passing the vehicle detector within the time period from T _A is n _L =14, and the equivalent number of vehicles passing the vehicle detector within the time period from T _B to T _C is n _j =3. Calculate the saturated traffic density of the lane: k _j = n _L /L _b = 14/60 = 0.23, queue dissipation speed $v=L_b/(T_B-T_G^1)=60/8=7.5m/\mathrm{the\; s}$ and $v_m=L_b/(T_{R,C}^2-T_C)=60/4=15m/\mathrm{the\; s},$ Since v _m >v, there is no spare green light time, Q=L _b +n _j /k _j ＝60+3/0.23＝73m, so the green light time in the next cycle is T _G ＝T _G +g _lay ＝15s+3s＝18s ;

time It is the start time of another red light on the main road, and at this moment, the vehicle detector data and signal light phase transformation data of the previous period are collected statistically. For example: the transition time T _A of the straight road is 10:01:20, and the green light start time is 10:01:28, the transition time T _B is 10:01:38, and the transition time T _C is 10:01:42, The equivalent number of vehicles passing the vehicle detector within the time period from T _A is n _L =18, and the equivalent number of vehicles passing the vehicle detector within the time period from T _B to T _C is n _j =2. Calculate the saturated traffic density of the through lane: k _j =n _L /L _a =18/80=0.225, queue dissipation speed and Due to v _m □v, there is no spare green light time, so the straight queue length Q _S =L _a +n _j /k _j =80+2/0.225=88m; the transition time T _A of the left turn lane is 10:01:27, green light start time is 10:01:43, the transition time T _B is 10:01:55, there is no transition time T _C , The number of equivalent vehicles passing the vehicle detector within the time from T _B to n _L =20, from T _B to The equivalent number of vehicles passing the vehicle detector within a given time is n _j = 4, and the saturated traffic flow density of the left-turning lane is calculated: k _j = n _L /L _a = 20/80 = 0.25, the queue dissipation speed Therefore, the left-turn queue length Q _L >L _a +n _j /k _j =80+4/0.25=96m, so the duration of the straight green light in the next cycle T _S =T _S +g _lay =15s+3s=18s, and the duration of the left-turn green light T _L = G _main - T _S = 80s-18s = 62s;

time It is the starting time of another red light on the secondary road. At this moment, the vehicle detector data and signal light phase transformation data of the previous cycle are collected statistically. For example: the transition time T _A of the secondary road is 10:01:45, and the starting time of the green light is 10:02:01, the transition time T _B is 10:02:10, there is no transition time T _C , The number of equivalent vehicles passing the vehicle detector within the time from T _B to n _L =14, from T _B to The number of equivalent vehicles passing through the vehicle detector within a certain period of time is n _j = 11, calculate the saturated traffic flow density of the last cycle lane: k _j = n _L /L _b = 14/60 = 0.23, the queue length Q>L _b +n _j /k _j ＝60+11/0.23＝108m, when the effective green light ends, there is still a queue behind the vehicle detector, so the green light duration of the next cycle is T _G ＝G _minor ＝45s;

time It is the start time of another red light on the main road. At this moment, the vehicle detector data and signal light phase transformation data of the previous cycle are collected statistically _. is 10:02:19, the transition time T _B is 10:02:30, there is no transition time T _C , The number of equivalent vehicles passing the vehicle detector within the time from T _B to n _L =19, from T _B to The equivalent number of vehicles passing through the vehicle detector within a certain period of time is n _j = 14, and the saturated traffic flow density of the through lane in the last cycle is calculated: k _j = n _L /L _a = 19/80 = 0.2375, and the queue length Q _S >L _a +n _j /k _j ＝80+14/0.2375＝139m; the transition time T _A of the left turn road is 10:02:29, the time when the green light starts is 10:02:37, the transition time T _B is 10:02:55, there is no transition time T _C , The number of equivalent vehicles passing through the detector within T _B is n _L =20, and T _B to The number of equivalent vehicles passing through the vehicle detector n _j =27, calculate the saturated traffic flow density of the left-turn lane in the last period: k _j =n _L /L _a =20/80=0.25 long queue Q _L >L _a +n _j / k _j ＝80+27/0.25＝188m At the end of the effective green light, there are still queues behind the detectors of the two lanes, so the duration of the straight green light in the next cycle is

$T_S=G_{\mathrm{main}}\&times;\frac{Q_S}{Q_S+Q_L}=80\mathrm{the\; s}\&times;\frac{139m}{139m+180m}=35\mathrm{the\; s},$ The duration of the left turn green light is

${\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; main</span>}}<span\; class="notranslate">\; \&times;</span>\frac{{\mathrm{<span\; class="notranslate">\; Q</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}{{\mathrm{<span\; class="notranslate">\; Q</span>}}_{\mathrm{<span\; class="notranslate">\; S</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; Q</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 80</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; \&times;</span>\frac{\mathrm{<span\; class="notranslate">\; 188</span>}\mathrm{<span\; class="notranslate">\; m</span>}}{\mathrm{<span\; class="notranslate">\; 139</span>}\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 188</span>}\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 45</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; ;</span>$

time It is the start time of another red light on the sub-arterial road. At this _moment , the vehicle detector data and signal light phase transformation data of the previous cycle are collected statistically. is 10:03:39, the transition time T _B is 10:03:50, and the transition time T _C is 10:03:58, The equivalent number of vehicles passing the vehicle detector within the time period from T _B to T B is n _L = 14, the equivalent number of vehicles passing the vehicle detector within the time period from T _B to T _C is n _j = 8, and the saturated traffic flow density of the last cycle lane is calculated as k _j =n _L /L _b =14/60=0.23, queue length Q=L _b +n _j /k _j =60+8/0.23=94m, queue dissipation speed $v=L_b/(T_B-T_G^3)=60/11=5.45m/\mathrm{the\; s},$ The queue dissipates completely, and v _m < v, there is a surplus of green light time, then the green light duration of the next cycle T _G = T _G -g _lay = 45s-3s = 42s;

time It is the start time of a red light on the main road. At _this moment, the vehicle detector data and signal light phase transformation data of the previous period are collected statistically. is 10:04:59, the transition time T _B is 10:04:35, and the transition time T _C is 10:04:40, The equivalent number of vehicles passing the vehicle detector within the time period from T _B to T B is n _L =20, and the equivalent number of vehicles passing the vehicle detector from T _B to T _C is n _j =2. Calculate the saturated traffic flow density k _j = n _L /L _a = 20/80 = 0.25, queue dissipation speed The queue dissipates completely, and $v_m=L_a/(T_{G,L}^4-T_C)=80/19=4.\mathrm{twenty\; one}m/\mathrm{the\; s},$ v _m < v, there is a surplus of green light time, the queue length Q _S =L _a +n _j /k _j =80+2/0.25=88m; the transition time T _A of the left turn lane is 10:04:38, the time when the green light starts is 10:04:59, the transition time T _B is 10:05:20, there is no transition time T _C , The number of equivalent vehicles passing the vehicle detector within the time from T _B to T B is n _L =21, T _B to The equivalent number of vehicles passing through the vehicle detector within a certain period of time is n _j = 20, calculate the saturated traffic flow density of the left turn road in the previous cycle k _j = n _L /L _a =21/80 = 0.26, and the queue length Q _L >L _a +n _j /k _j ＝80+20/0.26＝157m, when the effective green light ends, there is still a queue behind the detector, so the duration of the straight green light in the next cycle T _S ＝T _S -g _lay ＝36s-3s＝33s, turn left at the green light The duration is T _L =G _main -T _S =80s-33s=47s.

Table 1 The timing status of signal lights obtained from 10:00:00

period of time signal light status duration(s) 10:00:00-10:00:30 Go straight on the main road and turn left are red lights, secondary roads are green 30 10:00:31-10:00:57 Main road straight ahead green light, left turn red light, secondary arterial road red light 27 10:00:58-10:01:12 Go straight on the main road with a red light, turn left with a green light, and a secondary road with a red light 15 10:01:13-10:01:27 Go straight on the main road and turn left are red lights, secondary roads are green 15 10:01:28-10:01:42 Main road straight ahead green light, left turn red light, secondary arterial road red light 15 10:01:43-10:02:00 Go straight on the main road with a red light, turn left with a green light, and a secondary road with a red light 18 10:02:01-10:02:18 Go straight on the main road and turn left are red lights, secondary roads are green 18 10:02:19-10:02:36 Main road straight ahead green light, left turn red light, secondary arterial road red light 18 10:02:37:10:03:38 Go straight on the main road with a red light, turn left with a green light, and a secondary road with a red light 62 10:03:39-10:04:23 Go straight on the main road and turn left are red lights, secondary roads are green 45 10:04:24-10:04:58 Main road straight ahead green light, left turn red light, secondary arterial road red light 35 10:04:59-10:05:44 Go straight on the main road with a red light, turn left with a green light, and a secondary road with a red light 45 10:05:45-10:06:26 Go straight on the main road and turn left are red lights, secondary roads are green 42 10:06:27-10:06:59 Main road straight ahead green light, left turn red light, secondary arterial road red light 33 10:07:00-10:07:47 Go straight on the main road with a red light, turn left with a green light, and a secondary road with a red light 47

Claims (1)

1. an adaptive intersection signal control method, is characterized in that, the method is carried out according to following steps:

Step one, 1 wagon detector is installed in each track, crossing, and the distance of wagon detector and stop line is designated as L, and each wagon detector opens the bright moment for the red light gathering each phase place n-th cycle , the n-th cycle green light open the bright moment , the (n+1)th cycle red light open the bright moment , each vehicle by wagon detector time holding time t _o, wagon detector free time interval t _fand pass through the vehicle number n of detecting device;

Described phase place comprises the craspedodrome phase place at craspedodrome queue place and the phase place of turning left at queue place of turning left;

One-period extinguishes the moment for opening the bright moment from red light to green light, note G _minfor minimum green time, note g _layfor unit green light extends or the time of shortening, note G _mainfor maximum total green time of major trunk roads;

Step 2, at the end of one-period effective green time, detection data are transferred to back-end data disposal system by wagon detector, analyze the change of traffic behavior in this cycle, find out the transition moment T of wagon detector eigenstate _a, T _band T _c;

T _athat red light opens bright rear vehicle v ₁start the transition between states moment taking wagon detector for a long time, T _aafter moment, wagon detector is by vehicle v ₁take for a long time, moment T _acan be used to determine whether this phase place track exists longer vehicle platoon;

T _bthat green light opens bright rear vehicle v ₁sail out of the transition moment of wagon detector, green light open bright after, T _bbefore moment, the vehicle before stop line sails out of crossing with saturation volume rate successively, but is 0 by the vehicle number of detecting device, at T _bafterwards, vehicle is with maximum current flow rate by wagon detector, and the holding time of wagon detector and interval time reduce;

T _clast car v in vehicle queue's queue ₂by the transition between states moment of wagon detector, determine last car of queue v ₂according to being: in queue queue, during continuous vehicle passing detection device, interval time is less, when a car and its Following Car pass through detecting device interval greater than threshold value A, illustrate in Following Car not Queue here queue;

The value obtaining threshold value A according to current traffic survey data statistics is 3s;

Step 3, adopt from moment is to T _aby the vehicle number n of detecting device in time period in moment _lcalculate the vehicle density k under saturated conditions _j, then k _j=n _l/ L; Adopt from moment is to T _bdistance L between time period in moment and stop line to wagon detector calculates the dissipation speed v of vehicle queue's queue, then and then pass through T _band T _cby the vehicle number n of detecting device in time period _jcalculate the maximum queue length Q in this cycle;

(1) if there is not T _amoment, then Q < L;

(2) if there is T _amoment, then determine T further _cmoment;

If there is T in (a) _cmoment, then Q=L+n _j/ k _j;

If there is not T in (b) _cmoment, then red light of lower cycle is opened the bright moment as transition moment T _c, statistics T _bwith by the vehicle number n of detecting device in time period _j, then Q > L+n _j/ k _j;

Step 4, the signal control strategy of adjustment major trunk roads, detailed process is as described below:

Craspedodrome queue length Q is calculated respectively according to the method for step 3 _swith the queue length Q that turns left _l;

(1) if the length of two queues is all less than L, then turn left phase time and craspedodrome phase time all distribute minimum green time, i.e. T _g1=T _g2=G _min;

(2) if only have a certain queue length to be less than L, then the minimum green time T of the phase assignments at this queue place _g1=G _min, the queue that another length is more than or equal to L then needs to determine whether this queue exists transition moment T _c:

If there is transition moment T in (a) _c, then according to T _cwith to calculate in this time interval vehicle by minimum speed v needed for L _m, then have determine whether there is green light Slack Time:

If a) v _m, then there is green light Slack Time in < v, former long green light time is cut unit green light and change time g _lay, i.e. T _g2=T _g2-g _lay;

If b) v _m, then there is not green light Slack Time, namely indicate residual queue in>=v, if T _g2< G _main-T _g1, then g is added at former long green light time _lay, i.e. T _g2=T _g2+ g _layif, T _g2=G _main-T _g1, then T _g2=T _g2;

If there is not transition moment T in (b) _c, then represent at the end of green time, detecting device upstream also has queue queue, then the long green light time of this phase place is T _g2=G _main-T _g1;

(3) if the queue length of two queues is all more than or equal to L, then determine whether individual queue exists transition moment T _c, then have:

If both there is moment T in (a) _c, then determine whether each phase place has green time more than needed, perform green light according to the strategy (a) of step (2) and extend or reduce to obtain long green light time T _g1and T _g2, wherein: T _g1+ T _g2≤ G _main;

If there is moment T in (b) queue _c, there is not moment T in another _c, then there is moment T _cphase place according to the strategy (a) of step (2) perform green light extend or reduce to obtain long green light time T _g1, there is not moment T _cthe green time of phase place be T _g2=G _main-T _g1;

If neither there is transition moment T in (c) _c, then according to Q _sand Q _llength ratio distribute the green time of phase place and craspedodrome phase place of turning left, namely $\left\{\begin{array}{c}{T}_L=G_{\mathrm{main}}\&times;\frac{Q_L}{Q_L+Q_S}\\ T_S=G_{\mathrm{main}}\&times;\frac{Q_S}{Q_L+Q_S}\end{array}\right.$

Wherein: T _lthe long green light time of phase place of turning left, T _sthe long green light time of craspedodrome phase place, Q _scraspedodrome phase alignment length, Q _lphase alignment length of turning left;

Step 5, the signal control strategy of adjustment subsidiary road, detailed process is as described below:

Subsidiary road is kept straight on and is turned left and to be controlled by a phase place, remembers G _minorfor the maximum green time of subsidiary road, G _minfor minimum green time, g _layfor unit green light extends or the time of shortening, first calculate the queue length Q of subsidiary road according to the method for step 3, T _gfor subsidiary road green time;

(1) if queue length Q < is L, then Minimum Green Time T is distributed _g=G _min;

(2) if queue length Q>=L, then transition moment T is determined whether there is _c, then have:

If there is transition moment T in (a) _c, then according to T _cwith to calculate in this time interval vehicle by minimum speed v needed for L _m, then have determine whether there is green light Slack Time:

If a) v _m, then there is green light Slack Time, then T in < v _g=T _g-g _lay;

If b) v _m, then there is not green light Slack Time, namely indicate residual queue in>=v, if T _g< G _minor, then T _g=T _g+ g _layif, T _g=G _minor, then T _g=T _g;

If there is not transition moment T in (b) _c, then T _g=G _minor.