技术领域technical field
本发明涉及道路交通控制领域，尤其是一种城市交叉口可变导向车道自适应控制方法。The invention relates to the field of road traffic control, in particular to an adaptive control method for variable guiding lanes at urban intersections.
背景技术Background technique
对于信号控制交叉口的管理与控制，通常在保持车道功能不变的情况下，利用信号相位调整的方式适应实际的交通需求。该管控方式具有一定的局限性，尤其当各转向交通流在不同时段的波动明显时，往往造成不必要的时空资源浪费，或者引起交通拥堵。近年来，部分城市交叉口设置了可变导向车道，然而，可变导向车道转向功能的变换往往由值班交警采用人工控制或者采取固定时段的车道变换方案的方式进行调控，主观性较强，导致可变导向车道利用率低下的问题。For the management and control of signalized intersections, usually the signal phase adjustment is used to adapt to the actual traffic demand while keeping the lane function unchanged. This control method has certain limitations, especially when the traffic flow of each turn fluctuates significantly at different time periods, it often causes unnecessary waste of time and space resources, or causes traffic congestion. In recent years, variable directional lanes have been set up at some urban intersections. However, the transformation of the steering function of the variable directional lanes is often regulated by the traffic police on duty using manual control or a fixed-time lane change scheme, which is highly subjective and leads to The problem of low utilization of variable guiding lanes.
已有研究主要针对在信号控制交叉口交通流具有潮汐特性的情况，而针对同一进口道上左转和直行交通流在不同时间段波动显著的情况研究较少，相关文献提出了手动定时控制、感应控制、双停车线控制和主、预信号控制等方法，但所提出的流程实际执行起来可操作性不强，参数的获取和采集较困难，难以实现实时的可变导向车道转向功能的变换，并且未采用合适的主预信号之间的配时协调方案。Existing research mainly focuses on the situation that the traffic flow at signal-controlled intersections has tidal characteristics, but there are few studies on the situation that the left-turn and straight-going traffic flow fluctuates significantly in different time periods on the same approach road. The relevant literature proposes manual timing control, sensor Control, double stop line control, main and pre-signal control, etc., but the proposed process is not very operable in actual execution, and the acquisition and collection of parameters are difficult, and it is difficult to realize the real-time transformation of the steering function of the variable guiding lane. Moreover, a suitable timing coordination scheme between main pre-signals is not adopted.
发明内容Contents of the invention
本发明所要解决的技术问题在于，提供一种城市交叉口可变导向车道自适应控制方法，能够实现可变导向车道的转向功能变换。The technical problem to be solved by the present invention is to provide an adaptive control method for a variable directional lane at an urban intersection, which can realize the transformation of the steering function of the variable directional lane.
为解决上述技术问题，本发明提供一种城市交叉口可变导向车道自适应控制方法，包括如下步骤：In order to solve the above-mentioned technical problems, the present invention provides a method for adaptive control of variable guiding lanes at urban intersections, which includes the following steps:
(1)历史和实时交通流量数据的观测和收集；(1) Observation and collection of historical and real-time traffic flow data;
(2)基于k近邻非参数回归对所述步骤(1)得到的历史和实时交通流量数据，进行交通量数据库的整合、状态向量的选取、相似机制的选取、近邻个数的确定以及回归函数的构建，得到短时交通量预测数据；(2) The historical and real-time traffic flow data that described step (1) obtains based on k-nearest neighbor non-parametric regression, carry out the integration of traffic volume database, the selection of state vector, the selection of similar mechanism, the determination of the number of neighbors and the regression function The construction of short-term traffic volume prediction data is obtained;
(3)根据基于步骤(1)得到的历史交通量数据，通过计算和比较各进口道左转、直行交通流的拥堵程度，得到历史可变导向车道属性建议值；(3) According to the historical traffic volume data obtained based on step (1), by calculating and comparing the congestion degree of the left-turn and straight-going traffic flow of each entrance road, the suggested value of the attribute of the historical variable guiding lane is obtained;
(4)基于所述步骤(1)得到的历史交通量数据和所述步骤(3)得到的历史可变导向车道属性建议值，基于判别分析法，得到可变导向车道左转和直行属性相应的判别 分析函数；(4) Based on the historical traffic volume data obtained in the step (1) and the suggested value of the historical variable directional lane attribute obtained in the step (3), based on the discriminant analysis method, the corresponding left-turn and straight-going attributes of the variable directional lane are obtained. The discriminant analysis function;
(5)使用所述步骤(2)得到的短时交通流量数据和所述步骤(4)得到的判别分析函数，得到实时的可变导向车道属性建议值；(5) using the short-term traffic flow data that described step (2) obtains and the discriminant analysis function that described step (4) obtains, obtain the real-time variable guiding lane property suggestion value;
(6)根据所述步骤(5)得到实时的可变导向车道属性建议值，由信号控制机实现可变导向车道主预信号之间的配时协调，包括可变导向车道转向功能的变换时刻的选择和清空时间的计算，并由信号控制机控制相应主预信号灯的显示。(6) According to the step (5), the real-time variable steering lane attribute suggestion value is obtained, and the timing coordination among the main pre-signals of the variable steering lane is realized by the signal control machine, including the changing moment of the steering function of the variable steering lane The selection and calculation of the emptying time, and the display of the corresponding main pre-signal lights is controlled by the signal control machine.
优选的，步骤(1)中选取交通量指标作为分析左转、直行车流量比例的特征参数，包括历史和实时交通流量数据，其中实时交通量数据由感应线圈检测器现场检测得到，在每个信号周期，对进口道的左转、直行车道上分车型的车辆数进行统计，并将数据实时上传至控制中心；历史交通量数据是由实时交通量数据累积而成，为了便于数据的储存、计算以及保证车道属性转换的实时性，本发明每隔半个月更新一次历史交通量数据库，并且保持为最近一个月的交通量数据。Preferably, in the step (1), the traffic volume index is selected as the characteristic parameter for analyzing the ratio of left-turn and straight traffic, including historical and real-time traffic flow data, wherein the real-time traffic volume data is detected on-site by the induction coil detector, and in each The signal cycle counts the number of vehicles of different types on the left-turn and straight lanes of the entrance road, and uploads the data to the control center in real time; the historical traffic volume data is accumulated from real-time traffic volume data, in order to facilitate data storage, To calculate and ensure the real-time performance of lane attribute conversion, the present invention updates the historical traffic volume database every half a month, and keeps the latest month's traffic volume data.
优选的，步骤(2)中的k近邻非参数回归是在控制中心实现的，对交通量数据进行处理具体包括如下步骤：Preferably, the k-nearest neighbor non-parametric regression in step (2) is realized in the control center, and the processing of the traffic volume data specifically includes the following steps:
(21)对交通流量数据库的整合，将历史与实时交通流量数据库整合为一个复合数据库，并且复合数据库中的数据是动态更新的；(21) For the integration of the traffic flow database, the historical and real-time traffic flow databases are integrated into a composite database, and the data in the composite database is dynamically updated;
(22)对状态向量的选取，定义状态向量：(22) For the selection of the state vector, define the state vector:
X＝[vh(t),vh(t-1),vh(t-2),v(t),v(t-1),v(t-2)] (1)X=[vh (t), vh (t-1), vh (t-2), v(t), v(t-1), v(t-2)] (1)
其中，v(t)表示复合数据库中t时段的实时交通量，单位：pcu/h；Among them, v(t) represents the real-time traffic volume in the time period t in the composite database, unit: pcu/h;
vh(t)表示复合数据库中t时段的历史交通量，单位：pcu/h；vh (t) represents the historical traffic volume in the period t in the composite database, unit: pcu/h;
(23)相似机制的选取，采用欧氏距离定义相似性：(23) Selection of similarity mechanism, using Euclidean distance to define similarity:
其中，di为复合数据库中相邻三个时段的实时交通量和历史交通量之间的欧氏距离，单位：pcu/h；d为复合数据库中k个最近邻的欧氏距离，单位：pcu/h；Among them, di is the Euclidean distance between real-time traffic volume and historical traffic volume in three adjacent time periods in the composite database, unit: pcu/h; d is the Euclidean distance of the k nearest neighbors in the composite database, unit: pcu/h;
(24)近邻个数的确定，一般k设定在1到20之间，具体地满足：(24) To determine the number of neighbors, generally k is set between 1 and 20, specifically satisfying:
其中，Lmin(k)代表实际交通量和回归预测得到的交通量之间的残差平方和最小时对应的k，即最优近邻个数；Among them, Lmin (k) represents the k corresponding to the smallest residual square sum between the actual traffic volume and the traffic volume predicted by regression, that is, the number of optimal neighbors;
(25)回归函数的构建，下一时段的交通流量K(t+1)的计算公式：(25) The construction of the regression function, the calculation formula of the traffic flow K(t+1) in the next period:
其中，vhi(t)是t时段第i个近邻的交通量，i＝1，2，…，k，单位：pcu/h；Among them, vhi (t) is the traffic volume of the i-th neighbor in the t period, i=1, 2, ..., k, unit: pcu/h;
βi为第k个近邻的回归系数。βi is the regression coefficient of the kth nearest neighbor.
优选的，步骤(3)是在控制中心实现，方法具体包括如下步骤：Preferably, step (3) is realized in the control center, and the method specifically includes the following steps:
(31)计算每条车道的设计通行能力，(31) Calculate the design capacity of each lane,
其中，Nl为进口道设有专用左转车道而未设有专用右转车道时，专用左转车道的设计通行能力，单位：pcu/h；Among them, Nl is the design traffic capacity of the dedicated left-turn lane when the entrance road has a dedicated left-turn lane but no dedicated right-turn lane, unit: pcu/h;
Ns为直行车道的设计通行能力，单位：pcu/h；Ns is the design traffic capacity of the straight lane, unit: pcu/h;
Nsr为直行车道及直右车道设计通行能力之和，单位：pcu/h；Nsr is the sum of the design capacity of the straight lane and the straight right lane, unit: pcu/h;
βl为本面左转车辆比例；βl is the proportion of left-turning vehicles on this plane;
ψs为直行车道通行能力折减系数，可采用0.9；ψs is the capacity reduction coefficient of the through lane, which can be 0.9;
tg为信号周期内的绿灯时间，单位：s；tg is the green light time in the signal cycle, unit: s;
t1为变为绿灯后第一辆车启动并通过停止线的时间，单位：s,可采用2.3s；t1 is the time for the first vehicle to start and pass the stop line after the light turns green, unit: s, 2.3s can be used;
tis为直行或右行车辆通过停止线的平均间隔时间，单位：s/pcu；tis the average interval time for straight or right vehicles passing the stop line, unit: s/pcu;
tc为信号周期，单位：s；tc is the signal period, unit: s;
(32)计算左转、直行车道的饱和度Si，分别为：(32) Calculate the saturation Si of the left-turn and straight lanes, respectively:
Si＝Qi/Ni (6)Si =Qi /Ni (6)
其中,i＝l,s,分别代表左转和直行；Among them, i=l, s represent turning left and going straight respectively;
Qi为车道单个信号周期内的交通量，单位：pcu/h；Qi is the traffic volume in a single signal cycle of the lane, unit: pcu/h;
Ni为车道的设计通行能力，单位:pcu/h。Ni is the design capacity of the lane, unit: pcu/h.
选择饱和度0.8作为车道属性转换的阈值，则有：Select saturation 0.8 as the threshold of lane attribute conversion, then:
(33)进行判别：若两种车道的饱和度均小于0.8，那么此时两种车道均未达到拥堵状态，保持原有的可变导向车道的属性；若左转车道的饱和度大于0.8而直行车道的饱和度小于0.8，那么对应左转车道达到拥堵状态但直行车道尚未达到拥堵状态，将历史可变导向车道的建议属性设为左转；若直行车道的饱和度大于0.8而左转车道的饱和度小于0.8，那么对应直行车道达到拥堵状态但左转车道尚未达到拥堵状态，将历史可变导向车道的建议属性设为直行；若两种车道的饱和度均大于0.8，那么此时两种车道达到拥堵状态，为了避免此时车道属性转换可能引起的附加交叉口延误，仍保持原有的可变导向车道的属性。(33) Discriminate: if the saturation of the two lanes is less than 0.8, then the two lanes have not reached the congested state at this time, and keep the original attribute of the variable steering lane; if the saturation of the left-turn lane is greater than 0.8 and If the saturation of the straight lane is less than 0.8, then the corresponding left-turn lane has reached the congestion state but the straight lane has not yet reached the congestion state, and the suggested attribute of the historical variable steering lane is set to turn left; if the saturation of the straight lane is greater than 0.8 and the left-turn lane If the saturation of the two lanes is less than 0.8, then the corresponding straight lane has reached the congestion state but the left-turn lane has not yet reached the congestion state, and the suggested attribute of the historical variable steering lane is set to go straight; if the saturation of the two lanes is greater than 0.8, then the two In order to avoid additional intersection delays that may be caused by the conversion of lane attributes at this time, the attributes of the original variable guidance lanes are still maintained.
优选的，步骤(4)是在控制中心实现的，方法具体包括如下步骤：Preferably, step (4) is realized in the control center, and the method specifically includes the following steps:
(41)明确有两个判别类型，即可变导向车道的属性为左转或直行，分别设为类和类，对应有两个观察指标，即左转交通量和直行交通量，其中类有s组数据，类中有t组数据；(41) It is clear that there are two types of discrimination, that is, the attribute of the variable guiding lane is turning left or going straight, and they are respectively set to class and class, corresponding to two observation indicators, that is, left-turn traffic volume and straight-going traffic volume, where class has s sets of data, There are t sets of data in the class;
(42)将这些数据写成矩阵形式，有：(42) Write these data into matrix form, have:
其中，aij为类中第i组数据的第j个变量的值，i＝1,2，…s，j＝1,2，单位：pcu/h；Among them, aij is The value of the jth variable of the i-th group of data in the class, i=1,2,...s, j=1,2, unit: pcu/h;
bij为类中第i组数据的第j个变量的值，i＝1,2，…t，j＝1,2，单位：pcu/h；bij is The value of the jth variable of the i-th group of data in the class, i=1,2,...t, j=1,2, unit: pcu/h;
计算出各类数据的平均值：Calculate the average of various types of data:
其中，分别代表类和类数据的第j各变量的平均值，单位：pcu/h；作出两类数据矩阵与相应平均值差值的矩阵A、B：in, Representing class and The average value of the jth variables of the class data, unit: pcu/h; make the matrix A and B of the difference between the two types of data matrix and the corresponding average value:
由此可通过以下公式求得两类数据的离差矩阵S：Therefore, the dispersion matrix S of the two types of data can be obtained by the following formula:
求解二元方程组：即Solve a system of equations in two variables: which is
由此可得出：From this it can be concluded that:
其中，第一个公式为待判别对象对应的判别函数，单位：pcu/h；Among them, the first formula is the discriminant function corresponding to the object to be discriminated, unit: pcu/h;
第二、三个公式用于计算类和类数据的判别平均值，单位：pcu/h；The second and third formulas are used to calculate class and The discriminant mean value of class data, unit: pcu/h;
第四个公式用于计算判别临界值，单位：pcu/h；The fourth formula is used to calculate the discriminant critical value, unit: pcu/h;
(43)对判别对象进行判别，假设有一待判别对象，其数据为X(x1,x2)，则其判别值为y＝c1x1+c2x2，单位：pcu/h，如果满足yA>yB，则建议判别准则：若y>y0，则可以判定反之则可以判定如果满足yA<yB，则建议判别准则：若y>y0，则可以判定反之则可以判定即可以简化为如下的判别准则：(43) Discriminate the object to be discriminated. Suppose there is an object to be discriminated, whose data is X(x1 , x2 ), then its discriminant value is y=c1 x1 +c2 x2 , unit: pcu/h, If it satisfies yA >yB , then the criterion of discrimination is suggested: if y>y0 , then it can be judged Otherwise, it can be determined If it satisfies yA <yB , then it is recommended to judge the criterion: if y>y0 , then it can be judged On the contrary, it can be determined That is, it can be simplified to the following criterion:
由于判别分析函数是基于历史交通量数据和历史可变导向车道属性确定的，而历史交通量数据库每隔半个月更新一次，并且保持为最近一个月的交通量数据，因此得到的判别分析函数亦是每隔半个月更新一次参数。Since the discriminant analysis function is determined based on the historical traffic volume data and the attributes of the historical variable directional lanes, and the historical traffic volume database is updated every half a month, and is kept as the traffic volume data of the latest month, the obtained discriminant analysis function The parameters are also updated every half a month.
优选的，步骤(6)中，得到的实时可变导向车道属性由信号控制机实现车道属性转换，在控制中心进行存储，并且为了防止车道属性频繁切换，有且仅有连续三个周期的建议属性与当前情况不同时，可变导向车道属性才会改变，具体地：Preferably, in step (6), the obtained real-time variable guidance lane attributes are converted by the signal control machine and stored in the control center, and in order to prevent frequent switching of lane attributes, there are and only suggestions for three consecutive cycles The attributes of the variable guidance lane will only change when the attributes are different from the current situation, specifically:
当可变导向车道属性由直行变为左转时，为了清空主信号左转绿灯启亮之前的主预信号停车线之间道路上的直行车辆，预信号的直行红灯需要提前截止，并且为了弥补主预信号左转绿灯同时启亮时的绿灯时间损失，预信号的左转绿灯需要提前启亮：When the attribute of the variable guiding lane changes from going straight to turning left, in order to clear the straight-going vehicles on the road between the main pre-signal stop line before the main signal left-turn green light is turned on, the pre-signal straight red light needs to be cut off in advance, and in order To make up for the green light time loss when the main pre-signal turns left and the green light is turned on at the same time, the left-turned green light of the pre-signal needs to be turned on in advance:
其中，t1为预信号的直行红灯需要提前截止时间，单位：s；Among them, t1 is the advance cut-off time for the straight-going red light of the pre-signal, unit: s;
t2为预信号的左转绿灯需要提前启亮时间，单位：s；t2 is the pre-signal left-turn green light that needs to be turned on in advance, unit: s;
v为车辆的平均速度，单位：m/s；v is the average speed of the vehicle, unit: m/s;
l1为主信号停车线与预信号停车线之间的距离,单位：m；l1 The distance between the main signal stop line and the pre-signal stop line, unit: m;
a为车辆的平均启动加速度，单位：m/s2；a is the average starting acceleration of the vehicle, unit: m/s2 ;
当可变导向车道属性由左转变为直行时，为了清空主信号直行绿灯启亮之前的主预信号停车线之间道路上的左转车辆，预信号的左转红灯需要提前截止，并且为了弥补主预信号直行绿灯同时启亮时的绿灯时间损失，预信号的直行绿灯需要提前启亮：When the attribute of the variable steering lane changes from left to straight, in order to clear the left-turning vehicles on the road between the main pre-signal stop line before the main signal straight green light turns on, the left-turn red light of the pre-signal needs to be cut off in advance, and in order To make up for the green light time loss when the main pre-signal straight-going green light is turned on at the same time, the straight-going green light of the pre-signal needs to be turned on in advance:
其中，t3为预信号的左转红灯需要提前截止时间，单位：s；Among them,t3 is the pre-signal left turn red light needs to advance the cut-off time, unit: s;
t4为预信号的直行绿灯需要提前启亮时间，单位：s；t4 is the time for the straight green light of the pre-signal to be turned on in advance, unit: s;
gL为主信号左转绿灯时间，单位：s；gL is the main signal turn left green light time, unit: s;
qL为左转车的到达率，单位：pcu/s；qL is the arrival rate of left-turn vehicles, unit: pcu/s;
C为信号周期，单位：s；C is the signal period, unit: s;
SL为左转车的饱和流率，单位：pcu/s。SL is the saturation flow rate of a left-turning vehicle, unit: pcu/s.
本发明的有益效果为：(1)可以实现实时的可变导向车道转向功能的变换，在采用k近邻非参数回归和判别分析法时，历史交通量数据和历史可变导向车道属性建议值并不是固定不变的，而是由实时的交通量数据不断更新的；由历史数据得到的判别分析函数中的参数也是动态变化的，根据实时的交通量数据和实时的可变导向车道属性值不断滚动调整，由此可以提高可变导向车道的利用率，降低了车辆在交叉口的延误；(2)可以实现主预信号之间的配时协调，为消散主预信号停车线之间的车辆，提出了在可变导向车道转向功能转变时清空时间的计算，以及为防止相应的绿灯时间损失，设置了预信号早启时间，明确了可变导向车道上车辆的路权，使得驾驶员及时调整行车方向，减少了不必要的等待时间。The beneficial effect of the present invention is: (1) can realize the conversion of the steering function of variable steering lane in real time, when adopting k-nearest neighbor nonparametric regression and discriminant analysis method, historical traffic volume data and history variable steering lane attribute suggestion value and It is not fixed, but is constantly updated by real-time traffic volume data; the parameters in the discriminant analysis function obtained from historical data are also dynamically changing, according to real-time traffic volume data and real-time variable guiding lane attribute values. Rolling adjustment, which can improve the utilization rate of variable guiding lanes and reduce the delay of vehicles at intersections; (2) can realize the timing coordination between the main pre-signals, in order to dissipate the vehicles between the main pre-signal stop lines , put forward the calculation of the clearing time when the turning function of the variable steering lane changes, and in order to prevent the loss of the corresponding green light time, set the early start time of the pre-signal, clarify the right of way of the vehicle on the variable steering lane, so that the driver can timely Adjust the driving direction to reduce unnecessary waiting time.
附图说明Description of drawings
图1为本发明的方法流程示意图。Fig. 1 is a schematic flow chart of the method of the present invention.
图2为本发明的进口道的车道分布、主预信号停车线的位置相关关系示意图。Fig. 2 is a schematic diagram of the relationship between the lane distribution of the entrance road and the location of the main pre-signal stop line of the present invention.
图3为本发明的k近邻非参数回归方法的流程示意图。Fig. 3 is a schematic flow chart of the k-nearest neighbor non-parametric regression method of the present invention.
图4为本发明的优化k近邻非参数回归方法中k的取值不同时，左转和直行的残差平方和随k变化的曲线示意图。Fig. 4 is a schematic diagram of curves showing the sum of squares of the residual error of turning left and going straight when k varies with k in the optimized k-nearest neighbor non-parametric regression method of the present invention.
图5为本发明的可变导向车道属性为左转和直行时验证交通量数据准确性的示意图。Fig. 5 is a schematic diagram of verifying the accuracy of traffic volume data when the attribute of the variable guiding lane of the present invention is turning left and going straight.
图6(a)为本发明的可变导向车道属性由直行变为左转时与信号周期图的相关关系示意图。Fig. 6(a) is a schematic diagram of the correlation relationship between the attribute of the variable guiding lane of the present invention and the signal cycle diagram when the attribute changes from going straight to turning left.
图6(b)为本发明的本发明的可变导向车道转向功能变换时刻与信号周期图的相关关系示意图。Fig. 6(b) is a schematic diagram of the correlation relationship between the steering function change time of the variable steering lane and the signal cycle diagram of the present invention.
图6(c)为本发明的可变导向车道属性由左转变为直行时与信号周期图的相关关系示意图。Fig. 6(c) is a schematic diagram of the correlation relationship between the attribute of the variable guiding lane of the present invention and the signal cycle diagram when the attribute changes from left to straight.
图6(d)为本发明的可变导向车道转向功能变换时刻与信号周期图的相关关系示意图。Fig. 6(d) is a schematic diagram of the relationship between the steering function change time of the variable steering lane and the signal cycle diagram in the present invention.
具体实施方式detailed description
本发明提出的一种城市交叉口可变导向车道自适应控制算法，其流程图见附图1，主要包括如下步骤：A kind of adaptive control algorithm of variable guiding lanes at urban intersections proposed by the present invention, its flow chart is shown in accompanying drawing 1, mainly comprises the following steps:
(1)历史和实时交通量数据的观测和收集。(1) Observation and collection of historical and real-time traffic volume data.
本发明城市交叉口可变导向车道自适应控制算法的训练和测试都采用杭州市上塘路和文晖路南进口道的交通量数据。该交叉口南进口道共四条车道，从内向外第一条为左转车道，第二条为左转/直行可变车道，第三条为直行车道，第四条为直右车道。本发明实施例的车道分布、主预信号停车线的位置等关系如附图2。该路口目前由交警人工控制可变车道的变化，工作日上午七点到九点的高峰期可变车道为左转车道，其余时间均为直行车道。历史交通量数据采用直接向相关部门获取的方式，实时的交通量数据的调查采用人工调查与感应线圈检测器检测相结合的方式，人工调查交叉口的几何尺寸及信号配时，以每个信号周期时长为单位时间间隔输出该进口道左转和直行的交通流量数据，单位为pcu/h。本发明实施例采用268组做训练，12组数据做验证，采用的各转向交通量的比例描述见表1：The training and testing of the self-adaptive control algorithm of the variable guiding lane at the urban intersection of the present invention all use the traffic volume data of Shangtang Road and Wenhui Road South Entrance Road in Hangzhou. There are four lanes in the south entrance of the intersection. From inside to outside, the first lane is a left-turn lane, the second lane is a left-turn/straight variable lane, the third lane is a straight lane, and the fourth lane is a straight right lane. The relationship between the lane distribution and the position of the main pre-signal stop line in the embodiment of the present invention is shown in Figure 2. The intersection is currently controlled by the traffic police to change the variable lanes. The variable lanes are left-turn lanes during peak hours from 7:00 a.m. to 9:00 a.m. on weekdays, and the rest of the time are straight lanes. The historical traffic volume data is obtained directly from the relevant departments. The real-time traffic volume data investigation adopts the combination of manual investigation and induction coil detector detection. The geometric dimensions and signal timing of the intersection are manually investigated, and each signal The cycle duration is the unit time interval to output the traffic flow data of the entrance road turning left and going straight, and the unit is pcu/h. The embodiment of the present invention adopts 268 groups to do training, 12 groups of data to do verification, and the proportion description of each turning traffic volume that adopts is shown in Table 1:
表1本发明实施例所用的各转向交通量的比例The ratio of each turn traffic volume that table 1 embodiment of the present invention is used
(2)基于k近邻非参数回归对历史和实时交通量数据，进行交通量数据库的整合、状态向量的选取、相似机制的选取、近邻个数的确定以及回归函数的构建，得到短时交通量数据，本发明实施例的k近邻非参数回归方法的流程详见附图3。根据训练数据，分别计算了当近邻个数k在1到20之间时对应的左转、直行的残差平方和Lmin左转(k)、Lmin直行(k)，见表2，并且绘制了附图4。结合表2和附图4，发现对于本发明实例的左转、直行的短时交通量预测的最优近邻个数均为2个。(2) Based on k-nearest neighbor non-parametric regression for historical and real-time traffic volume data, integrate the traffic volume database, select the state vector, select the similar mechanism, determine the number of neighbors, and construct the regression function to obtain the short-term traffic volume For data, see Figure 3 for the flow of the k-nearest neighbor non-parametric regression method in the embodiment of the present invention. According to the training data, when the number of neighbors k is between 1 and 20, the corresponding left-turn and straight-going residual square sums Lmin turn left (k) and Lmin go straight (k) are calculated, see Table 2, and Figure 4 is drawn. In combination with Table 2 and accompanying drawing 4, it is found that the number of optimal neighbors for the short-term traffic volume prediction of the left-turn and straight-going examples of the present invention is 2.
表2本发明实施例残差平方和与近邻个数的关系Table 2 The relationship between the residual sum of squares and the number of neighbors in the embodiment of the present invention
(3)根据历史交通量数据，计算和比较各进口道左转、直行交通流的饱和度，得到历史可变导向车道属性建议值。(3) According to the historical traffic volume data, calculate and compare the saturation of the left-turn and straight-going traffic flow of each entrance road, and obtain the suggested value of the attribute of the historical variable guiding lane.
本发明实例中，直行车道通行能力折减系数ψs采用0.9，信号周期内的绿灯时间tg为70s,信号灯变为绿灯后第一辆车启动并通过停止线的时间t1采用2.3s，直行或右行车辆通过停止线的平均间隔时间tis为2.4s/pcu,信号周期tc为180s,本面左转车辆比例βl为0.383，因此直行车道、直行车道及直右车道、专用左转车道的设计通行能力分别如下：In the example of the present invention, the capacity reduction factor ψs of the through lane adopts 0.9, the green light time tg in the signal cycle is 70s, and the time t1 for the first vehicle to start and pass the stop line after the signal light becomes green light adopts 2.3s, The average interval time tis for straight-going or right-going vehicles passing the stop line is 2.4s/pcu, the signal periodtc is 180s, and the ratio βl of vehicles turning left on this plane is 0.383. The design capacities of left-turn lanes are as follows:
根据历史的交通量数据以及各车道的设计通行能力，对左转和直行车道的饱和度分别进行计算，并判断历史可变导向车道属性的建议值，表3是部分数据判断结果。According to the historical traffic volume data and the design capacity of each lane, the saturation of the left-turn and through lanes are calculated respectively, and the suggested value of the attribute of the historical variable guiding lane is judged. Table 3 is the judgment result of some data.
表3判定历史可变导向车道属性Table 3 Determining attributes of historical variable guiding lanes
(4)基于历史交通量数据和历史可变导向车道属性建议值，基于判别分析法，得到进口道可变导向车道左转和直行属性分别的判别分析函数。(4) Based on the historical traffic volume data and the suggested values of historical variable-guided lane attributes, and based on the discriminant analysis method, the discriminant analysis functions for the left-turn and straight-going attributes of the variable-guided lane at the entrance are obtained.
式中，x1、x2分别为待判别的进口道上左转和直行交通量，单位:pcu/h；In the formula, x1 and x2 are respectively the volume of left-turn and straight-going traffic on the entrance road to be judged, unit: pcu/h;
L0、L1分别代表可变导向车道的属性为左转和直行。L0 and L1 respectively represent that the attributes of the variable guiding lane are left turn and straight.
(5)使用短时交通量数据和判别分析函数，得到实时的可变导向车道属性建议值。(5) Using the short-term traffic volume data and the discriminant analysis function to obtain the real-time suggested values of the variable guiding lane attributes.
将历史可变导向车道属性值和实时的可变导向车道属性建议值进行比较，得到本发明实施例的验证数据集的准确性，如附图5。The accuracy of the verification data set of the embodiment of the present invention is obtained by comparing the attribute value of the historical variable guiding lane with the real-time suggested value of the variable guiding lane attribute, as shown in FIG. 5 .
(6)根据可变导向车道属性建议值，对照当前时刻可变导向车道的属性，对可变导向车道主预信号之间的配时协调进行设置，如附图6。(6) According to the suggested value of the attribute of the variable directional lane, compare the attributes of the variable directional lane at the current moment, and set the timing coordination between the main pre-signals of the variable directional lane, as shown in Figure 6.
本发明实施例中，主信号停车线与预信号停车线之间的距离l1＝90m，车辆的平均速度v＝12m/s2，车辆的平均启动加速度v＝12m/s2，因此预信号的直行红灯需要提前截止时间预信号的左转绿灯需要提前启亮时间主信号左转绿灯时间gL＝74s，左转车的到达率qL＝0.11pcu/s，左转车的饱和流率SL＝0.53pcu/s，信号周期C＝180s，因此预信号的左转红灯需要提前截止时间预信号的直行绿灯需要提前启亮时间t4＝t2＝5s。In the embodiment of the present invention, the distance l1 between the main signal stop line and the pre-signal stop line is 90m, the average speed of the vehicle v=12m/s2 , and the average starting acceleration of the vehicle v=12m/s2 , so the pre-signal The straight-going red light needs to be cut off in advance The left-turn green light of the pre-signal needs to be turned on in advance The green light time of the main signal turning left gL =74s, the arrival rate qL of the left-turning vehicle =0.11pcu/s, the saturated flow rate SL of the left-turning vehicle =0.53pcu/s, and the signal period C=180s, so the pre-signal A left turn on a red light requires an early deadline The straight-going green light of the pre-signal needs to be turned on t4 =t2 =5s in advance.
本发明提出一种城市交叉口可变导向车道自适应控制算法。基于k近邻非参数回归预测短时交通量进行；利用判别分析法，得到交通量数据和可变导向车道的属性值之间的判别关系，并且得到实时的可变导向车道属性建议值；设置了主预信号之间的提前截止时间和提前启动时间。模型的准确性较高，可实现实时的可变导向车道转向功能的变换，保证主预信号的配时协调，提高可变导向车道的利用率，缩短交叉口的平均延误。The invention proposes an adaptive control algorithm for a variable guiding lane at an urban intersection. Predict short-term traffic volume based on k-nearest neighbor non-parametric regression; use discriminant analysis method to obtain the discriminant relationship between traffic volume data and attribute values of variable directional lanes, and obtain real-time variable directional lane attribute suggestion values; set Early cut-off time and early start time between main pre-signals. The accuracy of the model is high, and it can realize the real-time transformation of the steering function of the variable-guided lane, ensure the timing coordination of the main pre-signal, improve the utilization rate of the variable-guided lane, and shorten the average delay at the intersection.
尽管本发明就优选实施方式进行了示意和描述，但本领域的技术人员应当理解，只要不超出本发明的权利要求所限定的范围，可以对本发明进行各种变化和修改。Although the present invention has been illustrated and described in terms of preferred embodiments, those skilled in the art should understand that various changes and modifications can be made to the present invention without departing from the scope defined by the claims of the present invention.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710168676.5ACN107067764B (en) | 2017-03-21 | 2017-03-21 | Self-adaptive control method for variable guide lane of urban intersection |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710168676.5ACN107067764B (en) | 2017-03-21 | 2017-03-21 | Self-adaptive control method for variable guide lane of urban intersection |
| Publication Number | Publication Date |
|---|---|
| CN107067764Atrue CN107067764A (en) | 2017-08-18 |
| CN107067764B CN107067764B (en) | 2020-01-03 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710168676.5AActiveCN107067764B (en) | 2017-03-21 | 2017-03-21 | Self-adaptive control method for variable guide lane of urban intersection |
| Country | Link |
|---|---|
| CN (1) | CN107067764B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108198441A (en)* | 2018-01-26 | 2018-06-22 | 杨立群 | A kind of quick, intelligent traffic system and method |
| CN108281017A (en)* | 2018-03-16 | 2018-07-13 | 武汉理工大学 | A kind of intersection traffic modulating signal control method based on bus or train route cooperative system |
| CN108281013A (en)* | 2018-03-22 | 2018-07-13 | 安徽八六物联科技有限公司 | A kind of road traffic monitoring system |
| CN108510735A (en)* | 2018-04-09 | 2018-09-07 | 宁波工程学院 | A kind of urban road intersection morning evening peak divides the prediction technique of steering flow |
| CN108983771A (en)* | 2018-07-03 | 2018-12-11 | 天津英创汇智汽车技术有限公司 | Vehicle lane-changing decision-making technique and device |
| CN109448371A (en)* | 2018-11-05 | 2019-03-08 | 王晨 | A kind of real-time variable lane control method and control system |
| CN109559513A (en)* | 2018-12-12 | 2019-04-02 | 武汉理工大学 | A kind of Adaptive Signal Control method based on the prediction of adjacent periods flow difference |
| CN109920244A (en)* | 2017-12-12 | 2019-06-21 | 上海宝康电子控制工程有限公司 | Changeable driveway real-time control system and method |
| CN110097752A (en)* | 2019-03-27 | 2019-08-06 | 杭州远眺科技有限公司 | A kind of intelligent and variable guided vehicle road calculation method |
| CN111145564A (en)* | 2020-01-03 | 2020-05-12 | 山东大学 | Self-adaptive variable lane control method and system for signal control intersection |
| CN111915894A (en)* | 2020-08-06 | 2020-11-10 | 北京航空航天大学 | Variable lane and traffic signal cooperative control method based on deep reinforcement learning |
| CN112017434A (en)* | 2020-08-19 | 2020-12-01 | 公安部交通管理科学研究所 | A method and system for variable lane control based on space-time coordination |
| CN115171402A (en)* | 2022-06-24 | 2022-10-11 | 东南大学 | Method for setting reverse variable guide lanes between adjacent T-shaped intersections |
| CN119296343A (en)* | 2024-10-30 | 2025-01-10 | 河北省交通规划设计研究院有限公司 | A smart city planning management system and method based on big data |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR2668631A1 (en)* | 1990-10-29 | 1992-04-30 | Silec Liaisons Elec | Method for control of the signal lights of a crossroads |
| CN102034350A (en)* | 2009-09-30 | 2011-04-27 | 北京四通智能交通系统集成有限公司 | Short-time prediction method and system of traffic flow data |
| JP2011237921A (en)* | 2010-05-07 | 2011-11-24 | Sumitomo Electric Ind Ltd | Signal control device and computer program |
| CN102938204A (en)* | 2012-08-03 | 2013-02-20 | 东南大学 | Variable guiding lane steering function conversion control method of city intersections |
| CN103700273A (en)* | 2014-01-06 | 2014-04-02 | 东南大学 | Signal timing optimization method based on variable guide lane |
| CN104464320A (en)* | 2014-12-15 | 2015-03-25 | 东南大学 | Shortest path induction method based on real road network features and dynamic travel time |
| CN105336163A (en)* | 2015-10-26 | 2016-02-17 | 山东易构软件技术股份有限公司 | Short-term traffic flow forecasting method based on three-layer K nearest neighbor |
| CN106297326A (en)* | 2016-10-27 | 2017-01-04 | 深圳榕亨实业集团有限公司 | Based on holographic road network tide flow stream Lane use control method |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR2668631A1 (en)* | 1990-10-29 | 1992-04-30 | Silec Liaisons Elec | Method for control of the signal lights of a crossroads |
| CN102034350A (en)* | 2009-09-30 | 2011-04-27 | 北京四通智能交通系统集成有限公司 | Short-time prediction method and system of traffic flow data |
| JP2011237921A (en)* | 2010-05-07 | 2011-11-24 | Sumitomo Electric Ind Ltd | Signal control device and computer program |
| CN102938204A (en)* | 2012-08-03 | 2013-02-20 | 东南大学 | Variable guiding lane steering function conversion control method of city intersections |
| CN103700273A (en)* | 2014-01-06 | 2014-04-02 | 东南大学 | Signal timing optimization method based on variable guide lane |
| CN104464320A (en)* | 2014-12-15 | 2015-03-25 | 东南大学 | Shortest path induction method based on real road network features and dynamic travel time |
| CN105336163A (en)* | 2015-10-26 | 2016-02-17 | 山东易构软件技术股份有限公司 | Short-term traffic flow forecasting method based on three-layer K nearest neighbor |
| CN106297326A (en)* | 2016-10-27 | 2017-01-04 | 深圳榕亨实业集团有限公司 | Based on holographic road network tide flow stream Lane use control method |
| Title |
|---|
| SUN XIN ET.AL: ""A multi-objective model for cordon-based congestion pricing schemes with nonlinear distance tolls"", 《J.CENT.SOUTH UNIV》* |
| 周鹏 等: ""智能可变车道的导向判决算法的研究与实现"", 《武汉理工大学学报》* |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109920244A (en)* | 2017-12-12 | 2019-06-21 | 上海宝康电子控制工程有限公司 | Changeable driveway real-time control system and method |
| CN108198441A (en)* | 2018-01-26 | 2018-06-22 | 杨立群 | A kind of quick, intelligent traffic system and method |
| CN108198441B (en)* | 2018-01-26 | 2021-06-29 | 杨立群 | Rapid intelligent traffic system and method |
| CN108281017A (en)* | 2018-03-16 | 2018-07-13 | 武汉理工大学 | A kind of intersection traffic modulating signal control method based on bus or train route cooperative system |
| CN108281013A (en)* | 2018-03-22 | 2018-07-13 | 安徽八六物联科技有限公司 | A kind of road traffic monitoring system |
| CN108510735A (en)* | 2018-04-09 | 2018-09-07 | 宁波工程学院 | A kind of urban road intersection morning evening peak divides the prediction technique of steering flow |
| CN108983771A (en)* | 2018-07-03 | 2018-12-11 | 天津英创汇智汽车技术有限公司 | Vehicle lane-changing decision-making technique and device |
| CN109448371A (en)* | 2018-11-05 | 2019-03-08 | 王晨 | A kind of real-time variable lane control method and control system |
| CN109559513A (en)* | 2018-12-12 | 2019-04-02 | 武汉理工大学 | A kind of Adaptive Signal Control method based on the prediction of adjacent periods flow difference |
| CN109559513B (en)* | 2018-12-12 | 2021-07-20 | 武汉理工大学 | An Adaptive Signal Control Method Based on Adjacent Period Flow Difference Prediction |
| CN110097752A (en)* | 2019-03-27 | 2019-08-06 | 杭州远眺科技有限公司 | A kind of intelligent and variable guided vehicle road calculation method |
| CN111145564A (en)* | 2020-01-03 | 2020-05-12 | 山东大学 | Self-adaptive variable lane control method and system for signal control intersection |
| CN111915894A (en)* | 2020-08-06 | 2020-11-10 | 北京航空航天大学 | Variable lane and traffic signal cooperative control method based on deep reinforcement learning |
| CN111915894B (en)* | 2020-08-06 | 2021-07-27 | 北京航空航天大学 | Variable lane and traffic signal cooperative control method based on deep reinforcement learning |
| CN112017434A (en)* | 2020-08-19 | 2020-12-01 | 公安部交通管理科学研究所 | A method and system for variable lane control based on space-time coordination |
| CN115171402A (en)* | 2022-06-24 | 2022-10-11 | 东南大学 | Method for setting reverse variable guide lanes between adjacent T-shaped intersections |
| CN115171402B (en)* | 2022-06-24 | 2023-08-29 | 东南大学 | A method for setting reverse variable guiding lanes between adjacent T-shaped intersections |
| CN119296343A (en)* | 2024-10-30 | 2025-01-10 | 河北省交通规划设计研究院有限公司 | A smart city planning management system and method based on big data |
| Publication number | Publication date |
|---|---|
| CN107067764B (en) | 2020-01-03 |
| Publication | Publication Date | Title |
|---|---|---|
| CN107067764B (en) | Self-adaptive control method for variable guide lane of urban intersection | |
| CN111145564B (en) | Self-adaptive variable lane control method and system for signal control intersection | |
| CN108335496B (en) | City-level traffic signal optimization method and system | |
| CN111791887A (en) | An energy-saving driving method for vehicles based on hierarchical speed planning | |
| CN105118308B (en) | Urban road intersection traffic signal optimization method based on cluster intensified learning | |
| CN105279982A (en) | Single intersection dynamic traffic signal control method based on data driving | |
| CN113012433B (en) | Vehicle-mounted networking energy-saving auxiliary driving control method and system | |
| CN109872544A (en) | A kind of control method and device of traffic signals | |
| CN111932909B (en) | Real-time variable lane dynamic allocation method under intelligent vehicle-road cooperative environment | |
| CN107038864B (en) | A method for judging the rationality of setting guide lanes at intersection entrances | |
| CN108932856B (en) | A method for setting the right of way at an intersection under automatic driving | |
| CN107730886A (en) | Dynamic optimization method for traffic signals at urban intersections in Internet of vehicles environment | |
| CN105632198A (en) | City area road traffic coordination control method and city area road traffic coordination system based on fuzzy control | |
| CN115083149B (en) | Reinforced learning variable duration signal lamp control method for real-time monitoring | |
| CN111754771B (en) | Individual travel time prediction method based on traffic signals and density delay | |
| CN117351734A (en) | Intelligent regulation and control method and system for vehicle delay | |
| CN110415522A (en) | A kind of control method and device of the changeable driveway based on multiple target radar | |
| CN105046990A (en) | Pavement signal lamp control method between adjacent intersections based on particle swarm algorithm | |
| CN106683441A (en) | Intersection signal timing plan evaluating method | |
| WO2023035666A1 (en) | Urban road network traffic light control method based on expected reward estimation | |
| CN117935532A (en) | Vehicle green wave passing planning method and device, electronic equipment and storage medium | |
| CN109859475B (en) | Intersection signal control method, device and system based on DBSCAN density clustering | |
| CN108961752A (en) | A kind of traffic signal control control effect on-line evaluation method and system | |
| CN109064760B (en) | Data-driven intelligent robust vehicle speed real-time planning method and system | |
| CN110097757B (en) | A key path recognition method for intersection groups based on depth-first search |
| 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 |