Study on control and optimization algorithm of dynamic green wave used in city traffic intelligent control

3.1 Optimization scheme of signal control

When a city traffic trunk traffic direction is another direction of more than 1.5 times in different periods or city tide phenomenon obviously exists, the heavy traffic flow direction one-way green wave coordinated control methods can alleviate the heavy traffic pressure in the direction of traffic flow.

(1) Signal cycle optimization. The duration of signal affects the efficiency of vehicles passing through the main lines. The best signal cycle is to ensure that vehicles cross the street safely at the shortest time and through pedestrians and pedestrians. a. the best cycle. C 0 i = L 1 - ∑ i = 1 n max { y i , y i ′ , … … }(1)

In the form: C0i is the best period of time, s; L is the length of the total signal loss, s; yi, y’i, … the flow ratio of each lane in the phase i phase. The best period of each intersection is calculated and the maximum period is taken as the public cycle. b. the half cycle. In coordinated control, when the traffic intersection part is small, does not require a longer period, and in order to avoid the long time non coordination of the direction of traffic waiting, the public half cycle, by adjusting the phase difference, the key from the intersection of exit of vehicles are able to pass through the adjacent intersections and continuous, coming from the adjacent the key to the intersection of intersection vehicles every half cycle coordination time, namely half cycle green wave coordinated control. C = C 0 i 2(2)

C is the signal cycle of the intersection with less i traffic, s.

(2) Optimization of phase difference. Before phase difference coordination optimization, we first optimize and adjust the phase release sequence of intersections. The characteristics of tidal traffic are mainly the uneven distribution of traffic flow, especially the waste of intersection timing due to uneven steering. In order to improve the utilization efficiency of green phase phase large flow capacity and intersection, the method for the single phase to release, release to lap the most basic, the release method as shown in Fig. 1. The phase difference is an important concept in coordination, an important parameter for green wave coordination, its main purpose is to reduce the vehicle waiting in the intersection before parking, ensure the vehicle in parking through the intersection. O = L + L j v ¯ - L t v ¯ - L t 2(3)

The O, the phase difference between the downstream intersection and the upstream intersection, s, L, the length of the road section, m; Lt, the queue length of the downstream intersection, m, Lj, the length of the upstream intersection, m; V, the average speed of the section, and m/s. In solving the coordinated phase difference, the average speed of the road section is generally estimated by the measured or observed. If the actual speed is greater than this value, it can reduce the phase difference value; if the actual speed is less than this value, you can appropriately increase the phase difference value, finally get the coordination phase and realize green wave coordination.

OPEN IN VIEWER

Fig.1

Release way diagram.

Change to traffic refers to the different directions of traffic time transform certain lanes of traffic or traffic, is a transformation in the direction of travel of traffic organization, is the main characteristics of a certain period of time the traffic capacity is not balanced, every day presents periodic space-time similar unbalanced traffic state changes, presented in morning and the evening peak congestion form. The lane changing direction is transformed into a variable direction lane, which includes the main line tidal lane and the steering lane. For unbalanced distribution problem, if we only control the signal, it is difficult to solve the problem of imbalance in space. Only balancing the traffic flow per lane can we get the heavy traffic flow.

(1) Unbalance calculation of tidal traffic flow distribution. The setting of tidal lanes should satisfy certain theoretical conditions, that is, the traffic capacity that does not affect the direction of light traffic flow, the number of two-way no more than 3 lanes or the number of vehicle lanes in the middle, and the most important point is the severity of traffic flow distribution imbalance. The directional distribution coefficient K of traffic flow reflects the imbalance of traffic distribution in 2 directions on a section. It is a symbol of whether the traffic flow has tidal character, and it is also an important index to decide whether or not to set tidal lanes. K = α q z q z + q q(4)

In the formula: qq is traffic flow in the direction of light traffic flow, pcu/h; qz is the traffic flow in the direction of i heavy traffic flow, pcu/h; α is the influence factor of traffic flow, taking into account the influence of other factors outside the environment and the traffic volume statistical error, and correcting the traffic volume in all directions. When K≥2/3, the traffic flow distribution in the main line is not evenly distributed, and the conditions of the tidal lane are met.

(2) Turn to unbalanced calculation. Intersection variable lane orientation is changed to a lane, can according to the different traffic in the different periods, to guide rational intersection import way to apply to adopt an inherent management of the road. Is mainly due to the import way to disequilibrium commuter travel, arrive at a certain place in place, all vehicles converge to the same place, to turn left or right turn vehicles increases a certain period of time, a so-called steering imbalance. This apparent difference, can make more than one direction lane cycle is always in a saturated state, and the other with imported other direction lane saturation, smaller road resources wasted, therefore, to introduce to unbalanced coefficient to measure the balance of import way each lane. In the observation time, the number of left turns is Ql, the number of vehicles in the direct line is Qs, the number of the right transfer vehicles is Qr, and the ratio of the left turns to the total vehicle is b = Q 1 Q 1 + Q s + Q r . At peak time, the maximum number of left turning vehicles is bgl, max, minimum value bgl, min; Pingfeng left when the maximum proportion of vehicles for bpl, max, a minimum value of bpl, min, by putting the proportion of peak and off-peak vehicle maximum difference evaluation standards, and max{ |b_gl,max - b_pl,min|, |b_gl,min - b_pl,mx| } turned to the unbalanced coefficient is: d 1 = n max { | b gl , max - b pl , min | , | b gl , min - b pl , mx | }(5)

In the formula, n is the number of direct traffic lane and the number of left lane. Turn right turn right by the same token, the unbalanced coefficient but vehicles would not have left the intertexture of vehicles in the intersection and go straight, need to introduce correction factor β, according to the actual right turn vehicles in the traffic impact analysis of operation condition, β values in 0.6 to 0.8, that is: d r = n β max { | b gl , max - b pl , min | , | b gl , min - b pl , mx | }(6)

Through the analysis of the actual empirical investigation and the ideal state, when d≥1, the inlet channel has obvious unbalanced characteristics; When 1 > d≥0.5, it has weak unbalanced characteristics; when d < 0.5, the imbalance does not exist. Based on this judgment, when the unbalanced coefficient is greater than 0. 5 but when less than 1, the direction of the steering is less balanced, and it can be considered to set the variable direction lane, and it can be manually adjusted in time. When the imbalance coefficient is greater than 1, it is necessary to induce the traffic flow in the direction of traffic volume.

In order to coordinate multiple route intersection control as the research object, the average vehicle delay as the control target, the objective function was established. The city multi intersection coordination control design of timing, to maintain coordination of traffic signals between adjacent intersections, in the same coordinate control system within the scope of each cross exit using the same signal cycle length. Therefore, optimization of coordinated control is in the cycle condition is relatively fixed, phase sequence for each intersection signal timing. According to the traffic engineering theory, the average delay time of vehicles in a single signal control intersection is: d = c ( 1 - λ ) 2 2 ( 1 - λ x ) + x 2 2 q ( 1 - x ) .

In the formula: d is the average delay of each car; c is the duration of the signal cycle; λ is the green letter ratio; q is traffic flow; x is the saturation. In the typical four-phase intersection, the total delay of the single intersections can be obtained: D = ∑ i = 1 4 ∑ j = 1 2 { q ij [ c ( 1 - λ i ) 2 2 ( 1 - λ i x ij ) + x ij 2 2 q ij ( 1 - x ) ] }(7)

In the formula: qij is the traffic flow of the first phase j inlet; Xij is the flow saturation of the ith phase and j inlet; λi for the phase i of the green letter ratio, λi = ti/c. The sequence of the phase is optimization process for each phase of the signal control intersection phase sequence in real time and effective green light time optimization, so the four phase time to satisfy the constraint conditions for:

t 1 + t 2 + t 3 + t 4 = c - l(8)

Considering the safety of pedestrians crossing the street at signalized intersection, the minimum green time of each phase can not be less than a threshold K, and the minimum green time is set to 10 seconds, so the signal timing of each phase at intersection needs to be satisfied.

K ≤ t i ≤ c - l - 3 × 10(9)

l in the formula is the total loss of time, s. Considering the biggest intersection saturation binding, design reasonable signal timing scheme and reasonable given the sequence of the phase signal control intersection, should be in the right intersection signal timing situation. Ensuring that everyone signal phase saturation is large. However, avoid causing coordination control each intersection congestion, so ever one assumes that the signal control intersection phase saturation is not more than 0.95, namely:

x = q N = q s g c c = cq sg c ≤ 0.95(10)

In the type: q is the actual flow; N is the capacity; s is saturated flow; ge is the effective green time, that is:

g c ≥ cq 0.95 s = cy 0.95(11)

In the formula, y is the traffic ratio. For each signal control intersection, the maximum flow ratio of each phase is substituted into the minimum green time requirement of this phase, namely:

t i = g ei cy i , max 0.95(12)

The average delay of vehicles at the intersection of the coordinated control system is:

E = ∑ k = 1 n D k Q(13)

In the formula, n is the number of coordinated control intersection. Q is the total vehicle flow, and Q = ∑ k = 1 n q k .

Coordinated control system of phase is the main task of the phase sequence problems will phase signal sequence seqi combinatorial optimization, and assigned to each intersection on the unit, to adjust the benefits of coordinated control system, reduce the vehicle in the intersection delay. Each intersection of coordinated control system for the relationship between the phase difference should satisfy the constraints 0 ≤ Δ i ≤ c(14)

Δi in the formula for the intersection of absolute phase difference. In a typical three phase and four phase signal timing as the research object, the hypothesis does not consider the phase lap. intersection signal phase combination of phase sequence encoding is shown in Fig. 2.

OPEN IN VIEWER

Fig.2

Intersection signal phase phase sequence combination coding.

Multiple population genetic algorithm in the optimization process of seqi, optimization of multiple individual species at the same time, through connect the population immigration operator, in order to realize the optimal individual in every population exchange, so as to achieve the purpose of the collaborative optimization. The hypothesis does not consider the intersection spacing and vehicle average speed limit, select trunk road coordination control the direction of the mainstream to optimize the sequence of the phase adjusting. Multiple intersection control system of the final output is the optimal phase phase sequence of the individual, the individual optimizes the results of multiple populations. Solve the standard genetic algorithm into the local optimal solution to the problem, should be a reasonable choice seqi collection scope. The use of multiple population genetic algorithm of the roulette wheel selection operator, single point crossover operator, and basic bit mutation operator to determine relevant parameters.

4.1 Example verification analysis

(1) Analyses of tidal traffic characteristics on the demonstration road. Jiang Shan south road is an important main road in the west coast economic zone of Qingdao city, with two-way traffic of 6 lanes and a width of 3.5 m per lane. There are many residential areas and work units on both sides of the road, and the traffic volume of the morning and evening peak are large, which is in severe congestion. The traffic volume of Ping Feng is small and it has obvious tidal phenomena. In this paper, the section of three intersections of the intersection of Jiang Shan south road, Wu Tai mountain road, Jiang Shan south road – bonded port area and Jiang Shan south road – Zhuo Ting plaza intersection are analyzed.

By Jiang Shanna road, Wu Tai mountain road, Jiang Shanna road, Beijing road, Jiang Shanna road – three Zhuo Ting square intersection traffic flow investigating. Data analysis found that, during the morning rush high-density traffic into the workspace, from south to north, traffic congestion is serious. Queuing time is long, especially influenced by Zhuo Ting square for office, turn left Zhuo Ting square vehicles more. In the evening rush hour, the vehicle pulls away from the working area, forming a high-density, evening peak traffic flow, which is affected by the bonded port area as a residential area. The traffic imbalance coefficient is between 0. 6 and 0.7. Zhuo Ting square left turn vehicles are more, and the original signal control scheme phase, phase sequence and timing to meet the needs of morning and evening peak left vehicles, easy cause of intersection congestion, reduce the straight road traffic efficiency.

(2) Control strategy design. Jiang Shanna (WuTai mountain road – ZhuoTing square section) in the form of “4 + 3”, barriers are removed, the original 6 to 7 lanes, the middle lane lane, delimit tide during 0:00–12:00, tidal lane is in the south to north, exit only turn left at the ZhuoTing square vehicles are allowed. At the same time in tidal lane end setup guide variable lane, lane changes the transformation of the phase green attributes. When left to get through, the change to the lane for a left turn lane, straight for passage, turn lanes for straight lane. During the period of 12:00–24:00, the direction of the tidal lane is north to south, and the exit of the bonded port area is only allowed to enter the vehicle at the left of the bonded port area, as shown in Fig. 3.

OPEN IN VIEWER

Fig.3

Jiangshan south tidal lane design.

The signal cycle of each intersection is calculated on the basis of formula (5), and the maximum period length is taken as the public cycle. Due to the small traffic flow at the intersection of the Jiang Shan South Road Zhuo Ting square, the half cycle of the intersection is coordinated, and the cycle length is obtained according to the formula (6). For each intersection, the traffic flow is also unbalance. In order to improve the capacity of large flow phase and the utilization efficiency of interphase green light, the phase lap processing method can be adopted. The timing strategy for each intersection is shown in Table 1. In accordance with the intersection scheme, shown in Fig. 4 the formation of Jiang Shan South Road Green Wave coordinated effect to use TDP tools. Vehicles in the basic driving speed, can not stop by every intersection.

OPEN IN VIEWER

Fig.4

Jiang shan road coordination effect map.

Figure 5 (a) and (b) are the curve of the flow change of the cars in and out of the city at the early and evening peaks of each intersection. From Fig. 5, it can be seen that after the implementation of the tidal lane, the flow rate of the intersections of the import and export cities increased slightly, and the increase was about 10% to 20%. It can be seen that the implementation of the tidal Lane effectively dredges the traffic flow of the South Road. According to the plan implemented in this paper, the traffic delays and parking times of vehicles on the south section of Jiang Shan road are almost 0, and the running speed is increased by 10% on average. The traffic operation condition has been greatly improved. The problems of traffic congestion and left turn queuing are effectively solved by the implementation of the tidal lanes and variable direction lanes on Jiang Shan South Road. At the same time, green wave coordination of Jiang Shan Road, improve the pass rate and promote the smooth running of the surrounding traffic.

OPEN IN VIEWER

Fig.5

Morning and evening peak traffic comparison chart.