Method, device and equipment for generating traffic signal control scheme

Abstract

The invention relates to a traffic signal control scheme generation method, device and equipment. Comprising the steps of obtaining a green light starting moment of each phase in an intersection; dividing the information control scheme management time corresponding to the intersection to obtain a plurality of target time windows, and calculating the initial dispersion of each target time window, including mapping the green light starting time corresponding to the target time window, calculating the center of gravity corresponding to the mapping result, and calculating the initial dispersion of each target time window at least based on the period duration and the center of gravity. Calculating to obtain initial dispersion; obtaining an initial dispersion sequence based on the initial dispersion; determining a local maximum value in the initial dispersion sequence as a segmentation point, and obtaining a plurality of time periods through the segmentation point; and calculating the actual period duration and the actual phase duration corresponding to each time period, and matching to obtain a target signal control scheme corresponding to the intersection. According to the invention, the efficiency and accuracy of information acquisition in the credit control scheme are improved, and the credit control scheme can be obtained without relying on excessive data.

Description

Technical Field

[0001] The present application relates to the field of intelligent transportation technology, and in particular to a method, device and equipment for generating a traffic signal control scheme. Background Art

[0002] Traffic light signal control scheme information is of great significance for establishing and analyzing traffic indicators. It can evaluate the relationship between traffic supply and demand and guide traffic light control. However, due to the management and communication problems of signal machine manufacturers and different regions, in some practical scenarios, there are data barriers for various types of traffic data at intersections, and it is difficult to directly obtain various information about traffic lights. Therefore, establishing a unified intersection signal control scheme that estimates intersection signal control schemes by observing vehicle data has become a core issue. The general idea to solve this problem is to extract and analyze features from the GPS data of Internet floating cars, but existing methods generally rely on multiple days of data to complete feature extraction and analysis, and the robustness of the information estimation errors and abnormal data in the signal control scheme is poor.

[0003] At present, no effective solution has been proposed for the problem that the calculation of intersection signal control schemes in the existing technology needs to rely on data from multiple days, and the efficiency and accuracy of information acquisition in the signal control schemes are low. Summary of the invention

[0004] Based on this, it is necessary to provide a method, device and equipment for generating a traffic signal control solution to address the above technical problems.

[0005] In a first aspect, the present application provides a method for generating a traffic signal control scheme. The method comprises:

[0006] Obtaining the green light start time of the vehicle corresponding to each phase in the intersection, wherein the green light start time of the vehicle reflects the green light start time of the corresponding phase;

[0007] The signal control scheme management time corresponding to the intersection is divided to obtain a plurality of target time windows, and the initial discreteness of each target time window is calculated respectively, wherein the initial discreteness corresponding to any target time window includes mapping the green light start time corresponding to the target time window according to the cycle duration corresponding to the target time window, calculating the center of gravity corresponding to the mapping result, and calculating the initial discreteness based on at least the cycle duration and the center of gravity;

[0008] Arrange the initial discreteness corresponding to all target time windows in chronological order to obtain an initial discreteness sequence;

[0009] The local maximum value in the initial discreteness sequence is determined as a segmentation point, and the signal control scheme management time is divided by the segmentation point to obtain multiple time periods; wherein the cycle lengths of the signal lights in different time periods are different;

[0010] The actual cycle duration and the actual phase duration corresponding to each time period are calculated, and the target signal control scheme corresponding to the intersection is obtained according to the actual cycle duration and the actual phase duration.

[0011] In one embodiment, obtaining a cycle duration corresponding to any target time window includes:

[0012] Enumerate the values ​​of all initial cycle durations corresponding to the target time window;

[0013] The green light start time is mapped based on each initial cycle duration in turn, and the center of gravity and discreteness corresponding to each mapping result are calculated. The initial cycle duration with the smallest discreteness value is taken as the cycle duration corresponding to the target time window.

[0014] In one embodiment, obtaining the green light start time of the vehicle corresponding to each phase at the intersection includes:

[0015] Identify the stop line at the intersection as the origin position, and based on the preset upstream vehicle trajectory and the preset downstream vehicle trajectory corresponding to the origin position, jointly determine the preset upstream vehicle trajectory and the preset downstream vehicle trajectory as sub-trajectories of the corresponding phase;

[0016] A preset target speed sequence including a speed threshold is obtained, and the target speed sequence is matched with the vehicle driving speed sequence corresponding to the collected sub-trajectory. If a match is detected, the green light start time in the driving speed sequence is determined based on the speed threshold.

[0017] In one embodiment, the green light start time is mapped based on each initial cycle duration in turn, and the center of gravity and dispersion corresponding to each mapping result are calculated, including:

[0018] Calculate the value of the center of gravity corresponding to each initial cycle duration, and perform calculations based on the center of gravity value and the initial cycle duration to obtain a reference moment within the initial cycle duration, map the green light start moment of the same phase to the initial cycle duration based on the reference moment, and obtain multiple corresponding discrete points. Calculate the initial discreteness corresponding to the target time window based on the discrete points, the initial cycle duration, the preset penalty term and the center of gravity, where the center of gravity is the point with the minimum sum of distances to all other discrete points, and the penalty term is composed of the preset number of phases and the actual number of phases contained in the initial cycle duration.

[0019] In one embodiment, a plurality of target time windows are determined based on a preset signal control solution management time, including:

[0020] Determine multiple initial time windows in chronological order based on the signal control solution management time, take each initial time window as the center, and combine a preset first number of initial time windows before the initial time window and a preset second number of initial time windows after the initial time window to form a target time window;

[0021] The local maximum is the local maximum discreteness corresponding to the local maximum time window in the target time window, wherein the local maximum discreteness is not less than the initial discreteness corresponding to each of the preset third number of target time windows before and after.

[0022] In one embodiment, the preset signal control scheme management time is divided by the split point to obtain multiple time periods, including:

[0023] Traversing all initial time windows included in the target time window where the segmentation point is located, dividing the target time window once respectively by different division methods, dividing the target time window into a first time period and a second time period respectively based on each division method, and calculating the discreteness of the first time period and the second time period respectively, wherein the division method is determined based on the position of each initial time window in the target time window; identifying the division method in which the discreteness of the first time period and the second time period after division is less than a preset discreteness threshold as the target division method, and determining the position of the corresponding initial time window as the target segmentation point based on the target division method;

[0024] Repeat the above steps of determining the target segmentation points until all target time windows containing the local maximum discreteness are traversed, and the target segmentation points corresponding to each target time window are obtained. The signal control scheme management time is divided by all target segmentation points to obtain multiple time periods.

[0025] In one embodiment, calculating the actual cycle duration corresponding to each time period includes:

[0026] The enumeration method is used to obtain multiple period cycle lengths and period center of gravity corresponding to the period, and calculations are performed based on the period cycle length and period center of each period in turn to obtain multiple discrete degrees corresponding to the period. The discrete degree with the smallest value is taken as the target discrete degree, and the period length corresponding to the target discrete degree is determined as the actual period length.

[0027] In one embodiment, calculating the actual phase duration corresponding to each time period includes:

[0028] Select the phase with the most observation values ​​of the green light start time as the reference phase, and define the green light start time in the reference phase as the reference time;

[0029] The reference time closest to each green light start time is used as the target reference time corresponding to each green light start time, and the corresponding green light start time is mapped based on each target reference time and the actual cycle duration, and the mapping result corresponding to the actual cycle duration is obtained according to the mapping results of all green light start times;

[0030] Obtain a preset solution template, match the solution template with the mapping result, determine the target solution template from the solution template according to the matching result, and determine the phase sequence in the signal control solution management time according to the target solution template, and calculate the actual phase duration based on the phase sequence, wherein the solution template represents the initial phase sequence corresponding to each phase.

[0031] In a second aspect, the present application also provides a device for acquiring a traffic signal control scheme. The device comprises:

[0032] An acquisition module, used to acquire the green light start time of the vehicle corresponding to each phase in the intersection, wherein the green light start time of the vehicle reflects the green light start time of the corresponding phase;

[0033] A calculation module is used to divide the signal control scheme management time corresponding to the intersection to obtain multiple target time windows, and calculate the initial discreteness of each target time window respectively, wherein calculating the initial discreteness corresponding to any target time window includes mapping the green light start time corresponding to the target time window according to the cycle duration corresponding to the target time window, calculating the center of gravity corresponding to the mapping result, and calculating the initial discreteness based on at least the cycle duration and the center of gravity; arranging the initial discreteness corresponding to all the target time windows in chronological order to obtain an initial discreteness sequence;

[0034] A generation module is used to determine the local maximum value in the initial discreteness sequence as a segmentation point, and divide the signal control scheme management time through the segmentation point to obtain multiple time periods; wherein the cycle duration of the traffic lights in different time periods is different; calculate the actual cycle duration and the actual phase duration corresponding to each time period, and obtain the target signal control scheme corresponding to the intersection according to the actual cycle duration and the actual phase duration.

[0035] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0036] Obtaining the green light start time of the vehicle corresponding to each phase in the intersection, wherein the green light start time of the vehicle reflects the green light start time of the corresponding phase;

[0037] The signal control scheme management time corresponding to the intersection is divided to obtain a plurality of target time windows, and the initial discreteness of each target time window is calculated respectively, wherein the initial discreteness corresponding to any target time window includes mapping the green light start time corresponding to the target time window according to the cycle duration corresponding to the target time window, calculating the center of gravity corresponding to the mapping result, and calculating the initial discreteness based on at least the cycle duration and the center of gravity;

[0038] Arrange the initial discreteness corresponding to all target time windows in chronological order to obtain an initial discreteness sequence;

[0039] The local maximum value in the initial discreteness sequence is determined as a segmentation point, and the signal control scheme management time is divided by the segmentation point to obtain multiple time periods; wherein the cycle lengths of the signal lights in different time periods are different;

[0040] The actual cycle duration and the actual phase duration corresponding to each time period are calculated, and the target signal control scheme corresponding to the intersection is obtained according to the actual cycle duration and the actual phase duration.

[0041] The above-mentioned method, device and equipment for generating a traffic signal control scheme first obtains the green light start time of the vehicle corresponding to each phase in the intersection; then divides the signal control scheme management time corresponding to the intersection to obtain multiple target time windows, and calculates the initial discreteness of each target time window respectively, wherein the calculation of the initial discreteness includes mapping the green light start time corresponding to the target time window according to the cycle duration corresponding to the target time window, calculating the center of gravity corresponding to the mapping result, and obtaining the initial discreteness based on at least the cycle duration and the center of gravity; then arranging the initial discreteness corresponding to all target time windows in chronological order to obtain an initial discreteness sequence; finally, determining the local maximum value in the initial discreteness sequence as the segmentation point, dividing the signal control scheme management time through the segmentation point to obtain multiple time periods, calculating the actual cycle duration and the actual phase duration corresponding to each time period, and obtaining the target signal control scheme corresponding to the intersection according to the actual cycle duration and the actual phase duration. Through this application, the signal control scheme of the intersection can be inferred according to the vehicle driving conditions at the intersection. Among them, this application uses the calculation of the initial discreteness of the target time window to accurately locate the segmentation points of different time periods, and then generates the target signal control scheme corresponding to the intersection according to the actual cycle duration and actual phase duration corresponding to the different time periods. Compared with the existing technology, this application has a small estimation observation error for the signal control scheme, strong robustness to abnormal data, and does not need to rely on vehicle driving data at the intersection for multiple days, but only vehicle driving data in a shorter period of time can complete the calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 A schematic diagram of a flow chart of a method for generating a signal control scheme in an embodiment;

[0043] Figure 2 A schematic diagram of a sub-trajectory in an embodiment;

[0044] Figure 3 A minimum discreteness sequence line graph in one embodiment;

[0045] Figure 4 A schematic diagram of a target time window including multiple initial time windows in one embodiment;

[0046] Figure 5 A schematic diagram of a common solution template in an embodiment;

[0047] Figure 6 A schematic diagram of a flow chart of a method for generating a signal control scheme in a preferred embodiment;

[0048] Figure 7 is a structural block diagram of a signal control scheme generating device in an embodiment;

[0049] Figure 8 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0051] In one embodiment, Figure 1 As shown, a method for generating a traffic signal control scheme is provided, comprising the following steps:

[0052] Step S110, obtaining the green light start time of the vehicle corresponding to each phase in the intersection, wherein the green light start time of the vehicle reflects the green light start time of the corresponding phase.

[0053] The phase refers to a certain vehicle flow direction at the intersection, such as the north straight phase, the north left turn phase, etc. For example, at an intersection, only the left turn flow and the straight flow are calculated, and the intersection should have a total of 8 phases. The green light start time of the vehicle in each phase is obtained. The green light start time is the time when the vehicle changes from a stopped state to a driving state, that is, the green light start time reflects the start time of the green light corresponding to the phase.

[0054] Step S120, divide the signal control scheme management time corresponding to the intersection to obtain multiple target time windows, and calculate the initial discreteness of each target time window respectively, wherein calculating the initial discreteness corresponding to any target time window includes mapping the green light start time corresponding to the target time window according to the cycle duration corresponding to the target time window, calculating the center of gravity corresponding to the mapping result, and calculating the initial discreteness based on at least the cycle duration and the center of gravity.

[0055] Among them, the above-mentioned signal control scheme management time is the total time length for collecting intersection vehicle driving information in this application, which can be preferably set to 24 hours. The signal control scheme management time is divided to obtain multiple target time windows. In this embodiment, the target time window is obtained by sliding a window of preset length for the signal control scheme management time. The window length can be set by relevant technical personnel, such as 1h, 30min, etc. There may be partial overlapping areas between adjacent target time windows. For example, the first target time window can be set to 00:00 to 1:00, and the second target time window can be set to 00:30 to 1:30, where the overlapping area is 00:30 to 1:00, and so on. In actual application, the specific division method of the signal control scheme management time can be set by relevant technical personnel according to actual needs. After obtaining multiple target time windows, the initial discreteness of each target time window is calculated respectively, including first mapping the green light start time corresponding to the target time window according to the cycle duration corresponding to the target time window, wherein the cycle duration refers to the time required for the traffic light to change one cycle in a preset phase sequence. For example, a cycle including 8 phases is generally divided into four stages. The first stage starts with the simultaneous green light for north straight driving and south straight driving, and the second stage starts with the simultaneous green light for south left turn and north left turn (it can be understood that some two phases can obtain green lights at the same time, but some two phases are prohibited from being green lights at the same time), etc. The sum of the time for the green light, yellow light and full red light to be on in sequence from the first stage to the fourth stage is a cycle. In practical applications, the duration of a cycle is commonly 125s. The above-mentioned mapping of the green light start time in the target time window according to the cycle duration includes:

[0056] x=mod(x 0 -x r , C)

[0057] Among them, x 0 is the green light start time to be mapped, x r is the reference time, which can be any time in the cycle duration, C is the cycle duration x is x 0 The mapping result.

[0058] After the mapping is completed, that is, all the green light start times of a phase within a cycle duration have been mapped, then the center of gravity is calculated based on the mapping result. The center of gravity refers to the point with the smallest sum of distances to all other points. Finally, the initial discreteness is calculated based on at least the center of gravity and the cycle duration. The initial discreteness reflects the dispersion of the green light start time after mapping. Furthermore, the cycle duration acquisition method in this embodiment includes direct input by relevant technical personnel; or enumerating the value of each possible cycle duration in the target time window, and calculating the above discreteness based on different cycle durations, and selecting the cycle duration corresponding to the calculation result with the smallest discreteness value as the actual cycle duration corresponding to the target time window. Furthermore, the cycle durations corresponding to different target time windows may be different from each other.

[0059] Step S130: Arrange the initial discretenesses corresponding to all target time windows in chronological order to obtain an initial discreteness sequence.

[0060] Among them, the initial discreteness of each target time window is calculated respectively according to the above method, and each initial discreteness is arranged in chronological order to obtain the above initial discreteness sequence.

[0061] Step S140, determining the local maximum value in the initial discreteness sequence as a segmentation point, dividing the signal control scheme management time by the segmentation point to obtain a plurality of time periods; wherein the cycle durations of the signal lights in different time periods are different.

[0062] Among them, the local maximum value in the initial discrete sequence is determined. There are many methods for selecting the local maximum value, including but not limited to determining the discreteness greater than a preset discreteness threshold as the local maximum value; or determining the discreteness corresponding to several adjacent time windows with the largest value as the local maximum value, etc. After determining the local maximum value, the local maximum value is used as a segmentation point, and the management time of the signal control scheme is divided based on the segmentation point to obtain multiple time periods, and the cycle lengths of the signal lights in different time periods are different. It can be understood that if the duration selected in the above text is closer to the actual cycle duration, the distribution of all mapped green light start times of a phase on [0, C) will be relatively concentrated, that is, distributed near the green light start time corresponding to the phase. On the contrary, if the selected cycle duration is significantly different from the actual cycle duration, the mapped x value will be diffuse on the time axis, that is, the data has a large degree of discreteness. Therefore, the discreteness can be used as an accuracy indicator of the cycle duration. In this embodiment, if the selected target time window spans two different time periods, the periodic law of the green light start time will be destroyed (the cycle duration of different time periods is different). Therefore, the discreteness value calculated at this time is relatively large. In summary, the division point of the time period can be queried according to the value of the discreteness.

[0063] Step S150, calculating the actual cycle duration and the actual phase duration corresponding to each time period, and obtaining a target signal control solution corresponding to the intersection according to the actual cycle duration and the actual phase duration.

[0064] Among them, after completing the division of the time period, the actual cycle duration corresponding to each time period can be determined according to the possible values ​​of all cycle durations in each time period, wherein the calculation of the actual cycle duration can follow the above idea, that is, enumerate each possible value of the cycle duration, calculate each corresponding discreteness according to all different values, and take the cycle duration corresponding to the minimum discreteness as the actual cycle duration of the time period. The above-mentioned actual phase duration is the difference between the green light start time of the phase and the green light start time of the subsequent phase. The calculation method of the above-mentioned actual phase duration includes but is not limited to, for each time period, after determining the actual cycle duration, the green light start time of each phase is mapped to the time axis of the actual cycle duration, so that the position of the green light start time of each phase in the cycle and the relative position relationship between each phase can be obtained, that is, the above-mentioned actual phase duration. This embodiment can match the above-mentioned actual cycle duration and the actual phase duration with the preset signal control template to screen out the target signal control scheme that meets the actual cycle duration and the actual phase duration.

[0065] Through steps S110 to S150, the present application uses the vehicle trajectory data detected at the intersection during the signal control scheme management time to obtain the green light start time of each phase, and through its periodic law, obtains the time period division of the intersection and the period within each period, and automatically matches the phase sequence according to the actual cycle length and the actual phase length, thereby estimating the complete signal control scheme. And according to field tests, taking the signal control scheme management time of the intersection as 24h as an example, when the detector for detecting the vehicle driving condition can capture at least 20% of the green light start time, the method of the present application can complete a relatively accurate estimation of the signal control scheme without spending a lot of time, wherein the detector for detecting the vehicle driving condition includes but is not limited to a radar video all-in-one machine.

[0066] In practical applications, the signal control scheme management time is commonly set to 24 hours, which is also the preferred duration of this application. For existing traditional control scheme generation methods, they all require multiple days of vehicle driving data, and assume that the daily schemes of these days are the same in order to realize the estimation of the control scheme, and considering that the existing schemes all need to collect a large amount of data, it is impossible to estimate the scheme in some low-traffic periods (such as 0-3 am) or the estimation error of the scheme is large. In this application, combining the dual advantages of the proposed scheme and the adopted data, it is possible to accurately estimate the scheme for all periods of the day by relying only on 24 hours of data.

[0067] In some embodiments, the method further comprises:

[0068] Enumerate the values ​​of all initial cycle durations corresponding to the target time window;

[0069] The green light start time is mapped based on each initial cycle duration in turn, and the center of gravity and discreteness corresponding to each mapping result are calculated. The initial cycle duration with the smallest discreteness value is taken as the cycle duration corresponding to the target time window.

[0070] Specifically, when calculating the cycle duration corresponding to the target time window, all possible initial cycle duration values ​​corresponding to the target time window can be enumerated one by one. Each green light start time is mapped based on each initial cycle duration in turn, and the center of gravity and discreteness corresponding to each mapping result are calculated, where the center of gravity is the point with the smallest sum of distances from all other points. In this embodiment, on the time axis [0, C), the two mapped green light start times x i , x j The periodic distance between them is:

[0071]

[0072] Furthermore, the center of gravity is calculated as:

[0073] Therefore, the distance between two points can be calculated one by one through the above formula, so as to screen out the center of gravity corresponding to the initial cycle duration.

[0074] After obtaining the initial cycle duration and the center of gravity corresponding to each initial cycle duration, the dispersion is calculated according to the following formula:

[0075]

[0076] In the formula, the numerator describes the discreteness of the green light start time distribution with the sum of the squares of the distances between the center of gravity and other points, and the denominator plays a normalization role to reduce the influence of the cycle duration on the discreteness calculation. By summing the discreteness of each phase, the total discreteness of the green light start time at the intersection can be obtained. When the selected cycle duration C is equal to the actual cycle duration, the total discreteness calculated at this time should take a minimum value. In summary, the cycle duration corresponding to the target time window can be calculated by enumeration. The accurate cycle duration can be calculated through this embodiment, and the present embodiment does not require a large amount of vehicle driving characteristic data to complete the calculation of the cycle duration.

[0077] In some embodiments, the method further comprises:

[0078] Identify the stop line at the intersection as the origin position, and based on the preset upstream vehicle trajectory and the preset downstream vehicle trajectory corresponding to the origin position, jointly determine the preset upstream vehicle trajectory and the preset downstream vehicle trajectory as sub-trajectories of the corresponding phase;

[0079] A preset target speed sequence including a speed threshold is obtained, and the target speed sequence is matched with the vehicle driving speed sequence corresponding to the collected sub-trajectory. If a match is detected, the green light start time in the driving speed sequence is determined based on the speed threshold.

[0080] Specifically, Figure 2 This is a schematic diagram of a sub-trajectory in an embodiment, where the stop line at the intersection is taken as the origin, an area upstream of the stop line is divided into an upstream vehicle trajectory, and an area downstream of the stop line is divided into a downstream vehicle trajectory. Preferably, the length of the upstream vehicle trajectory is generally 10m, and the length of the downstream vehicle trajectory is generally 5m. Furthermore, if the lane has a waiting area, the above-mentioned origin position is set as the stop line corresponding to the waiting area.

[0081] Further, the speed sequence of the vehicles traveling in the sub-trajectory is obtained, that is, the speed of the vehicles in each unit time in the sub-trajectory is obtained. Preferably, the speed of each vehicle in the sub-trajectory per second can be collected to obtain the speed sequence of each vehicle in the sub-trajectory. A preset target speed sequence is obtained, and the target speed sequence includes a speed threshold. In some preferred embodiments, the target speed sequence is defined as a speed V at three consecutive moments. t-2 、V t-1 、V t , the preset speed threshold is V s =2m / s, and V t-2 <V s , and V t-1 <V s , and V t ≥V s . In summary, after obtaining the target speed sequence, the target speed sequence is matched with the driving speed sequence of the vehicle corresponding to the collected sub-trajectory. If the match is successful, that is, there is a sequence in the vehicle driving speed sequence of the sub-trajectory that meets the standard of the target speed sequence, then the moment corresponding to the speed threshold in the successfully matched sequence is determined as the green light start time. Through the above method, the above analysis can be performed on each lane of each phase, so as to obtain all the green light start times of each phase in the signal control scheme management time. The method in this embodiment can be used to infer the green light start time of the traffic light based on the driving conditions of multiple vehicles, and because multiple vehicles are integrated, the difference between the multiple green light start times obtained should be small, that is, the error is small, and it is closer to the real green light start time of the traffic light.

[0082] In one embodiment, the method further comprises:

[0083] Calculate the centroid value corresponding to each initial cycle duration, and perform calculations based on the centroid value and the initial cycle duration to obtain a reference moment within the initial cycle duration. Based on the reference moment, map the green light start moment of the same phase to the initial cycle duration to obtain multiple corresponding discrete points. Calculate the initial discreteness corresponding to the target time window based on the discrete points, the initial cycle duration, the preset penalty term and the centroid, where the centroid is the point with the minimum sum of distances to all other discrete points, and the penalty term is composed of the preset number of phases and the actual number of phases contained in the initial cycle duration.

[0084] Specifically, since the cycle duration is generally an integer within a certain range, the enumeration method can be used to list all possible cycle duration values ​​that may correspond to the target time window. For the results based on each initial cycle duration mapping, there is a point with the minimum sum of distances between all other points, so that all possible values ​​of the initial cycle duration corresponding to the target time window and the corresponding centroid value can be obtained. Further, a reference moment within the initial cycle duration is obtained. The reference moment can be any point within the randomly obtained initial cycle duration, or a point specified by relevant personnel. Based on the reference moment, the green light start time of the same phase is mapped to the initial cycle duration to obtain multiple corresponding discrete points. And further based on the discrete points, the initial cycle duration, the preset penalty term and the corresponding centroid, the initial discreteness corresponding to the target time window is calculated. When calculating the discreteness, if the sum of several continuous phases in the scheme is exactly equal to half of the cycle duration, due to the influence of random trajectory observation errors and abnormal data (such as the temporary scheme of manual control changes the fixed timing cycle duration in a short time), ∑ p ψ p (0.5C 0 ) is less than ∑ p ψ p (C 0 ), which will cause the estimated value of the cycle duration to be erroneously taken as 0.5C0 (that is, in this application, the cycle duration corresponding to the smallest discreteness is taken as the actual cycle duration). To solve this problem, a penalty term is introduced in this embodiment, and the calculation of the discreteness is corrected as follows:

[0085]

[0086] Among them, n 0 is the number of phases of green light start vehicles expected to be observed in each cycle. For a crossroad, by default, each entrance lane only focuses on the straight and left turn directions. Therefore, let n 0 =8, that is, the preset number of phases should be 8, n p(C) is the average number of different phases in each cycle after the green light start time of all phases on the time axis corresponds to the selected cycle length C, that is, the actual number of phases. When C = 0.5C 0 On average, only about 0.5n can be obtained per cycle 0 different phase numbers, that is, the actual number of phases is about 4, so n p (0.5C 0 ) will be much smaller than n p (C 0 ), w is the weight of the penalty term. Considering the relative size of the discreteness and the penalty term, the preferred value is 0.5, where C 0 The actual cycle duration of the estimated time period is a fixed value. This embodiment can effectively avoid the situation where a half cycle is mistakenly selected as the estimated value of the entire cycle duration. Compared with the prior art that only calculates the discreteness through the cycle duration and the center of gravity, the method in this embodiment improves the accuracy of the cycle duration estimation.

[0087] In some embodiments, the method further comprises:

[0088] Determine multiple initial time windows in chronological order based on the signal control solution management time, take each initial time window as the center, and combine a preset first number of initial time windows before the initial time window and a preset second number of initial time windows after the initial time window to form a target time window;

[0089] The local maximum is the local maximum discreteness corresponding to the local maximum time window in the target time window, wherein the local maximum discreteness is not less than the initial discreteness corresponding to each of the preset third number of target time windows before and after.

[0090] Specifically, for the signal control scheme management time, it is divided into multiple initial time windows in chronological order. For example, in some preferred embodiments, the above-mentioned signal control scheme management time may be 24 hours, and then the signal control scheme management time is divided into units of 15 minutes to form 96 initial time windows (96*15min=24 hours). Taking each initial time window as the center, and combining several initial time windows before the initial time window, and several initial time windows after the initial time window, together form a target time window. For example, after obtaining 96 initial time windows, take the fifth initial time window as the center, combine the first number x initial time windows before the initial time window, and the second number y initial time windows after the initial time window, together form a target time window. The length of the target time window is (x+y+1)*15min. It can be understood that the division of the initial time window in the present application allows proximity to the boundary of the signal control scheme management time. Taking a target time window composed of seven initial time windows (i.e., x=y=3) as an example, when the second window is selected as the center, then the five initial time windows 1, 2, 3, 4, and 5 are used. The window constitutes a target time window, that is, when the center is close to the boundary, the required number of initial time windows is selected as much as possible; similarly, when the center is not too close to the boundary, normal selection is maintained. For example, when the fifth window is selected as the center, the second, third, fourth, fifth, sixth, seventh, and eighth initial time windows constitute the corresponding target time window. On this basis, the combination of target time windows is (same as above, taking seven initial time windows as an example to constitute a target time window). For the first, second, third, fourth, fifth, sixth, and seventh initial time windows, they constitute the target time window together, and the fourth initial time window is the center. Similarly, for the second, third, fourth, fifth, sixth, seventh, and eighth initial time windows, they constitute the target time window together, and the fifth initial time window is the center. Similarly, the above method is repeated to obtain all target time windows.

[0091] Furthermore, the local maximum is the calculation result of the discreteness of each target time window, and the discrete value of a certain target time window is not less than the discreteness value corresponding to the third number of target time windows before and after, that is:

[0092]

[0093] Then the discrete value corresponding to the target time window is taken as the local maximum value, wherein the third number is preferably 3, and the third number can be equal to the first number and the second number, or can be different, and can be set by relevant technical personnel according to actual needs. Through this embodiment, the time window corresponding to the local maximum discrete value can be calculated. According to the above explanation, it can be understood that the place with a larger discreteness is generally close to the division point of the time period (because the cycle lengths of different time periods are different, when a target time window spans two different time periods, the periodic law of the green light start time is destroyed, and the calculated discreteness is larger at this time), so the approximate position of the time period division point can be quickly estimated by the above method.

[0094] In some embodiments, the method further comprises:

[0095] Traversing all initial time windows included in the target time window where the segmentation point is located, dividing the target time window once respectively by different division methods, dividing the target time window into a first time period and a second time period respectively based on each division method, and calculating the discreteness of the first time period and the second time period respectively, wherein the division method is determined based on the position of each initial time window in the target time window; identifying the division method in which the discreteness of the first time period and the second time period after division is less than a preset discreteness threshold as the target division method, and determining the position of the corresponding initial time window as the target segmentation point based on the target division method;

[0096] Repeat the above steps of determining the target segmentation points until all target time windows containing the maximum discrete value are traversed to obtain the target segmentation points corresponding to each target time window, and divide the signal control scheme management time by all target segmentation points to obtain multiple time periods.

[0097] Specifically, considering that the switching between different time periods is often accompanied by transition schemes and random errors in data, the time corresponding to the local maximum may not be strictly equal to the division point of the time period, but in its vicinity, such as Figure 3 As shown, it is a line graph of the minimum discreteness sequence in an embodiment. In the figure, the blue line is the connection line between the minimum discreteness of each time window, and the red dotted line is the actual segmentation point of the time period. It can be seen from the figure that the local maximum value is often not equal to the position of the actual segmentation point. The actual segmentation point is generally around the local maximum value. Therefore, it is necessary to make a more accurate judgment on the position of the segmentation point, that is, it is necessary to search for the exact position of the segmentation point in the target time window corresponding to the local maximum value.

[0098] Figure 4 In one embodiment, the target time window TW includes multiple initial time windows. i Schematic diagram, with a target time window TW i For example, seven initial time windows are included in thei-3 To T i+3 ) position determines six division methods, namely, the first division method is, in the first initial time window T i-3 and the second initial time window T i-2 The second division method is to divide the time window T in the second initial time window T i-2 and the third initial time window T i-1 The target time window is divided into the first time period and the second time period based on each division method, and the discreteness of the first time period and the second time period is calculated respectively, that is, Figure 3 The target time window TW i According to the corresponding division method, it is divided into two periods, namely, the first period TS b and the second period TS f , with lengths of n and 7-n respectively, where n is the length of an initial time window. b and TS f The dividing point of the two parts is consistent with the real dividing point of the time period, then the discrete calculation results of the two parts are b and ψ f will be smaller, otherwise, ψ b and ψ f There will be a larger value in . Therefore, in this embodiment, the division method in which the discreteness of the first time period and the second time period after division is less than the preset discreteness threshold is identified as the target division method. Similarly, the position of the division at this time is used as the actual segmentation point, so that the real segmentation point corresponding to the target time window can be found. In summary, the target time window TW i The problem of internal precise search period division split point is converted into the following optimization problem:

[0099] n s =arg n min(max(ψ b , ψ f ))

[0100] tS b ={T i-3 ,...,T i+n-4}

[0101] TS f ={T i+n-3 , ..., T i+3},n∈[1,6]

[0102] Similarly, the above steps of determining the target segmentation point are repeated until all target time windows containing the maximum discrete value are traversed to obtain the target segmentation point corresponding to each target time window, and the signal control scheme management time is divided by all target segmentation points to obtain multiple time periods. Through this embodiment, after a segmentation, an accurate search is performed again in the target time window, so that a more accurate location of the time period segmentation point can be obtained.

[0103] In some embodiments, the method further comprises:

[0104] The enumeration method is used to obtain multiple period cycle lengths and period center of gravity corresponding to the period, and calculations are performed based on the period cycle length and period center of each period in turn to obtain multiple discrete degrees corresponding to the period. The discrete degree with the smallest value is taken as the target discrete degree, and the period length corresponding to the target discrete degree is determined as the actual period length.

[0105] Specifically, in this embodiment, the method for calculating the cycle duration described above is reused, and the possible cycle durations and the corresponding centroid values ​​in the time period are listed by enumeration, and the calculation is performed based on the cycle duration and the centroid of each time period in turn, and the cycle duration with the smallest discreteness is taken as the actual cycle duration in each time period, that is:

[0106]

[0107] Further, if the difference between the estimated values ​​of the cycle durations of two adjacent time periods is small, the two time periods are considered to belong to the same time period. Preferably, if the difference between the estimated values ​​of the cycle durations of two adjacent time periods is less than 5s, it is considered that the difference between the estimated values ​​of the cycle durations is small. At this time, the two time periods can be merged, and the estimated value of the cycle duration of the merged time period can be recalculated again. Through this embodiment, the calculation of the cycle duration of the time period can be completed, and the two time periods can be merged according to the results of the estimated cycle durations of each time period, so as to obtain a more accurate time period division result.

[0108] In some embodiments, the method further comprises:

[0109] Select the phase with the most observation values ​​of the green light start time as the reference phase, and define the green light start time in the reference phase as the reference time;

[0110] The reference time closest to each green light start time is used as the target reference time corresponding to each green light start time, and the corresponding green light start time is mapped based on each target reference time and the actual cycle duration, and the mapping result corresponding to the actual cycle duration is obtained according to the mapping results of all green light start times;

[0111] Obtain a preset solution template, match the solution template with the mapping result, determine the target solution template from the solution template according to the matching result, and determine the phase sequence in the signal control solution management time according to the target solution template, and calculate the actual phase duration based on the phase sequence, wherein the solution template represents the initial phase sequence corresponding to each phase.

[0112] Specifically, after the time period division is completed and the time period cycle length corresponding to each time period is calculated, in this embodiment, for each time period, according to the time period cycle length corresponding to each time period, the green light start time of each phase is mapped to the time axis of the time period cycle length according to the following formula, so as to obtain the position of the green light start time of each phase in the cycle:

[0113] x=mod(x 0 -x rp , C) where x 0 is the green light start time for projection, x rp For x 0 The corresponding reference time, C is the period length corresponding to the period. In the above formula, the phase with the most observations of the green light start time in the period is selected as the reference phase, and all the green light start times in the reference phase are defined as the reference time. When mapping each green light start time, the distance from each green light start time x 0 The most recent reference time x rp , as the target reference time, let the target reference time be 0, and after mapping according to the above formula, the mapping results of each green light start time can be obtained, that is, the position of the green light start time of all phases on the long time axis of the period cycle.

[0114] Through this embodiment, X rp By mapping the green light start time of each phase as a reference time, the relative relationship between each phase and the green light start time of the reference phase can be obtained. Compared with randomly selecting a time as a reference time for mapping, in this embodiment, the nearest reference time X is used as the time to be mapped. rp As the target reference moment, the time distance between the two is only within the range of several cycles, which can effectively reduce the impact of the clock offset accumulation error and the cycle duration estimation error on the phase duration estimation, and the relative position relationship between the obtained phases is more robust.

[0115] Finally, the above preset solution template is a standard intersection signal control solution template. Generally, there are 13 common solution templates, such as Figure 5As shown, the matching scheme is performed for the east-west and north-south entrance road directions respectively. Taking the east-west direction as an example, according to the release order, overlap, single release and rotation release of the straight and left turn phases, the above 13 common templates are obtained. It can be seen from the figure that from top to bottom and from left to right, they are: 1. East straight and west straight are carried out at the same time, and after the end, the east straight and east left turn are carried out at the same time, and after the end, the east left turn and west left turn are carried out at the same time; 2. East straight and west straight are carried out at the same time, and after the end, the east left turn and west left turn are carried out at the same time; 3. East straight and west straight are carried out at the same time, and after the end, the west left turn and west straight are carried out at the same time, and after the end, the east left turn and west left turn are carried out at the same time; 4. East left turn and west left turn are carried out at the same time, and after the end, the east straight and east left turn are carried out at the same time, and after the end, the east straight and west straight are carried out at the same time... and so on. In this embodiment, the scheme template is matched with the mapping result, and the mapping result includes the distribution of each phase at each moment on the time axis of the period cycle. According to the mapping result, the scheme template is matched. If all phases and the time relationship between phases can match the above template, the target scheme template can be determined from multiple scheme templates, and the phase sequence in the entire signal control scheme cycle can also be determined, so as to calculate the actual phase duration based on the phase sequence, where the phase duration can be defined as the difference between the green light start time of the phase and the green light start time of the next phase.

[0116] The present application also provides a preferred embodiment of a method for generating a traffic signal control scheme. Figure 6 The figure is a flow chart of a method for generating a traffic signal control scheme in a preferred embodiment.

[0117] Step S610, first obtain the green light start time of each lane corresponding to the phase of the intersection within the signal control scheme management time. Among them, the signal control scheme management time is preferably 24 hours, that is, the green light start time of each lane corresponding to the phase of an intersection within 24 hours is collected. The stop line of each lane of the intersection can be used as the origin position, and for each trajectory, a section [b up , b down ] as the sub-trajectory of study, where the upstream vehicle trajectory b up Preferably 10m, downstream vehicle trajectory b down Preferably, it is 5m. If the lane has a waiting area, the origin position is set to the stop line corresponding to the waiting area. Set a preset target speed sequence, collect the driving speed sequence of the vehicle in each lane from the sub-trajectory, and match the driving speed sequence with the target speed sequence. Preferably, the target speed sequence can be a series of speed value standards, such as, the target speed sequence includes a speed threshold Vs=2m / s, and the target speed sequence includes:

[0118] Vt-2 , V t-1 <V s And V t >V s

[0119] If there is a driving speed sequence of a vehicle that meets the above target speed sequence in the sub-trajectory, then the above t is used as the green light start time for this phase based on the vehicle. Based on the above method, the above analysis is performed on each lane of each phase, and all green light start times of each phase in a day are recorded.

[0120] Step S620, calculate the time segmentation points in the signal control scheme management time. The signal control scheme management time is divided into 96 units of 15 minutes. With each unit i as the center, the three units before and after are combined to construct an initial time window TW with a length of 105 minutes. i , and calculate the minimum discreteness of each initial time window respectively. Since the cycle duration must be an integer within a certain range, the minimum discreteness of each initial time window can be solved by enumeration. Among them, the method for calculating the discreteness of each initial time window is to first map the green light start time of vehicles in the same phase to a time axis with a length of the cycle duration (list all the values ​​of the cycle duration by enumeration, and calculate the corresponding discreteness in turn), and calculate the corresponding center of gravity. After obtaining the values ​​of each cycle duration and its corresponding center of gravity, the discreteness corresponding to each time window can be calculated:

[0121]

[0122] Among them, n 0 is the number of preset phases in each cycle, that is, the number of phases of vehicles starting with green lights that are expected to be observed. Taking the intersection as an example, by default, each entrance lane only focuses on the two flow directions of straight and left turn, so let n 0 =8,n p (C) is the average number of different phases in each cycle after the green light start time of all phases on the time axis corresponds to the selected cycle length C, that is, the actual number of phases. When C = 0.5C, only about 0.5n can be obtained per cycle on average. 0 different phase numbers, so n p (0.5C 0 ) will be much smaller than n p (C 0 ), w is the weight of the penalty term. Considering the relative size of the discreteness and the penalty term, the preferred value is 0.5. The above method can effectively avoid the situation where, when the cycle length of the initial time window is determined according to the minimum discreteness, half a cycle is mistakenly regarded as the cycle length corresponding to the initial time window due to the influence of random trajectory observation errors and abnormal data.

[0123] It can be understood that there are multiple time periods in the signal control scheme management time, and the cycle durations followed in different time periods are different. If a time window spans two different time periods, the periodic law of the green light start time is destroyed. Therefore, the minimum discreteness of the time window is large. According to the above idea, the segmentation point of the time period division can be found through the discreteness of each time window, that is, the minimum discreteness of each initial time window is taken as the discreteness of the initial time window, and each minimum discreteness is arranged in chronological order to obtain a discreteness sequence, and the local maximum value in the discreteness sequence is determined as the segmentation point of the time period, wherein the local maximum value is preferably defined as the point where the minimum discreteness of the time window is not less than the minimum discreteness corresponding to the three previous and next time windows, as shown in the following formula:

[0124]

[0125] Thus, each local maximum value can be determined, and the local maximum value in the discreteness sequence is determined as the segmentation point.

[0126] Step S630, correct each segmentation point to obtain the target segmentation point. Considering that the time period switching is often accompanied by transition schemes and random errors in data, the time at which the segmentation point calculated above is located may not be strictly equal to the segmentation point of the time period division, but may be near it. Therefore, it is necessary to make a more accurate judgment on the target time window of the segmentation point position, that is, to search for the exact target segmentation point within the target time window. The target time window is divided into two parts, namely, the first time period TS b and the second period TS f , the lengths are n and 7-n respectively. If the dividing point between the first time period and the second time period is consistent with the segmentation point of the time period division, then the corresponding minimum discreteness calculated by the two parts will be small. On the contrary, there is at least one larger value in the first time period and the second time period. Therefore, the above method can be used to accurately search for the segmentation point of the time period division in the target time window. Among them, the above division method is determined based on the position of each initial time window in the target time window. Therefore, according to the above steps, the exact position of the segmentation point is searched in the target time window.

[0127] Step S640, calculate the period cycle duration corresponding to each period, wherein, if the difference between the estimated values ​​of the period durations of adjacent periods is calculated to be less than 5s, it is considered that the two periods belong to the same period, and the two periods need to be merged, and the estimated value of the period duration of the merged period is recalculated.

[0128] Step S650, calculate the duration of each phase and match it with the preset scheme template. After determining the cycle duration of each time period, map the green light start time of each phase in the time period to the time axis of the cycle duration, so as to obtain the position of the green light start time of each phase in the cycle. When mapping, select the phase with the most observations of the green light start time in the time period as the reference phase, and select the observation value of the reference phase that is closest to the time to be mapped as the reference time, and complete the position of the green light start time of all phases in the time period on the time axis according to the above method.

[0129] Furthermore, assuming that the reaction time of vehicles after the green light of each phase of the intersection is equal, the green light start time of each phase can be regarded as the time when the green light starts, and the green light start time of each flow direction can be reversed by the green light start time of each phase. The mapping results of the green light start time of all phases are traversed. On the long axis of the cycle time, if the difference between the green light start time of two or more phases is less than 10s, the green light start time of these two (or more) phases is considered to be the same, and the average value represents the green light start time of the phase. In particular, if these phases include a reference phase, its green light start time is set to 0. After determining the relationship between the green light start time of each phase, the scheme structure can be matched with the standard cross intersection signal control scheme template to obtain the target signal control scheme. If all phases can match the above template, the phase sequence in the entire signal control scheme cycle can be determined. For each phase, its phase duration can be determined by the green light start time of the phase and the green light start time of the next phase. In summary, we can obtain an estimate of the completed day plan including time period division, phase sequence, cycle duration and phase duration, which is the above-mentioned traffic signal control plan.

[0130] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0131] Based on the same inventive concept, the embodiment of the present application also provides a traffic signal control scheme generation device for implementing the above-mentioned traffic signal control scheme generation method. The implementation solution provided by the device to solve the problem is similar to the implementation solution recorded in the above-mentioned method, so the specific limitations in the embodiments of one or more traffic signal control scheme generation devices provided below can refer to the limitations of the traffic signal control scheme generation method above, and will not be repeated here.

[0132] In one embodiment, Figure 7 As shown, a device for generating a traffic signal control scheme is provided, comprising: an acquisition module 71, a calculation module 72 and a generation module 73, wherein:

[0133] An acquisition module 71 is used to acquire the green light start time of the vehicle corresponding to each phase in the intersection, wherein the green light start time of the vehicle reflects the green light start time of the corresponding phase;

[0134] The calculation module 72 is used to divide the signal control scheme management time corresponding to the intersection to obtain multiple target time windows, and calculate the initial discreteness of each target time window respectively, wherein calculating the initial discreteness corresponding to any target time window includes mapping the green light start time corresponding to the target time window according to the cycle duration corresponding to the target time window, calculating the center of gravity corresponding to the mapping result, and calculating the initial discreteness based on at least the cycle duration and the center of gravity; arranging the initial discreteness corresponding to all the target time windows in chronological order to obtain an initial discreteness sequence;

[0135] A generation module 73 is used to determine the local maximum value in the initial discreteness sequence as a segmentation point, and divide the preset signal control scheme management time by the segmentation point to obtain multiple time periods; wherein the cycle duration of the traffic lights in different time periods is different; calculate the actual cycle duration and the actual phase duration corresponding to each time period, and obtain the target signal control scheme corresponding to the intersection according to the actual cycle duration and the actual phase duration.

[0136] Specifically, the acquisition module 71 acquires the green light start time of the vehicle corresponding to each phase in the intersection, and sends the above green light start time to the calculation module 72, the calculation module 72 divides the signal control scheme management time corresponding to the intersection to obtain multiple target time windows, and calculates the initial discreteness of each target time window respectively, wherein calculating the initial discreteness corresponding to any target time window includes mapping the green light start time corresponding to the target time window according to the cycle duration corresponding to the target time window, calculating the center of gravity corresponding to the mapping result, and calculating the initial discreteness based on at least the cycle duration and the center of gravity, arranging the initial discreteness corresponding to all target time windows in chronological order to obtain an initial discreteness sequence; the generation module 73 determines the local maximum value in the initial discreteness sequence as the segmentation point, divides the preset signal control scheme management time by the segmentation point to obtain multiple time periods, calculates the actual cycle duration and the actual phase duration corresponding to each time period, and obtains the target signal control scheme corresponding to the intersection according to the actual cycle duration and the actual phase duration.

[0137] Each module in the above-mentioned traffic signal control scheme generation device can be implemented in whole or in part by software, hardware and a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0138] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 8 As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store relevant data generated by the traffic signal control scheme. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for generating a traffic signal control scheme is implemented.

[0139] Those skilled in the art will understand that Figure 8 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0140] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0141] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0142] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0143] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for generating a traffic signal control scheme, characterized in that: The method comprises: Obtaining the green light start time of the vehicle corresponding to each phase in the intersection, wherein the green light start time of the vehicle reflects the green light start time of the corresponding phase; The signal control scheme management time corresponding to the intersection is divided to obtain a plurality of target time windows, and the initial discreteness of each target time window is calculated respectively, wherein the calculation of the initial discreteness corresponding to any target time window includes mapping the green light start time corresponding to the target time window according to the cycle duration corresponding to the target time window, calculating the center of gravity corresponding to the mapping result, and calculating the initial discreteness based on at least the cycle duration and the center of gravity; Arranging the initial discretenesses corresponding to all the target time windows in chronological order to obtain an initial discreteness sequence; Determine the local maximum value in the initial discreteness sequence as a segmentation point, and divide the signal control scheme management time by the segmentation point to obtain a plurality of time periods; wherein the cycle lengths of the signal lights in different time periods are different; The actual cycle duration and the actual phase duration corresponding to each of the time periods are calculated, and a target signal control scheme corresponding to the intersection is obtained according to the actual cycle duration and the actual phase duration.

2. The method according to claim 1, characterized in that Obtaining the cycle duration corresponding to any of the target time windows includes: Enumerate all initial cycle duration values ​​corresponding to the target time window; The green light start time is mapped based on each of the initial cycle durations in turn, and the center of gravity and the discreteness corresponding to each mapping result are calculated, and the initial cycle duration with the smallest discreteness value is taken as the cycle duration corresponding to the target time window.

3. The method according to claim 1, characterized in that The step of obtaining the green light start time of the vehicle corresponding to each phase at the intersection includes: Identifying a stop line at the intersection as an origin position, and based on a preset upstream vehicle trajectory and a preset downstream vehicle trajectory corresponding to the origin position, jointly determining the preset upstream vehicle trajectory and the preset downstream vehicle trajectory as a sub-trajectory corresponding to the phase; A preset target speed sequence including a speed threshold is obtained, and the target speed sequence is matched with a driving speed sequence of the vehicle corresponding to the collected sub-trajectory. If a successful match is detected, the green light start time in the driving speed sequence is determined based on the speed threshold.

4. The method according to claim 2, characterized in that: The mapping of the green light start time based on each of the initial cycle durations in turn, and calculating the center of gravity and the dispersion corresponding to each mapping result, includes: Calculate the value of the center of gravity corresponding to each of the above initial cycle durations; A reference moment within the initial cycle duration is obtained, and based on the reference moment, the green light start moment of the same phase is mapped to the initial cycle duration to obtain a corresponding plurality of discrete points, and the initial discreteness corresponding to the target time window is calculated based on the discrete points, the initial cycle duration, a preset penalty term, and the centroid, wherein the centroid is the point with the minimum sum of distances to all other discrete points, and the penalty term is composed of the preset number of phases and the actual number of phases contained in the initial cycle duration.

5. The method according to claim 1, characterized in that The determining of multiple target time windows based on the preset signal control solution management time includes: Determine a plurality of initial time windows in chronological order based on the signal control scheme management time, take each of the initial time windows as the center, and combine a preset first number of the initial time windows before the initial time window and a preset second number of the initial time windows after the initial time window to form the target time window; The local maximum is a local maximum discreteness corresponding to a local maximum time window in the target time window, wherein the local maximum discreteness is not less than the initial discreteness corresponding to each of a preset third number of target time windows before and after.

6. The method according to claim 5, characterized in that The preset signal control scheme management time is divided by the segmentation point to obtain multiple time periods, including: Traversing all the initial time windows included in the target time window where the segmentation point is located, dividing the target time window once respectively by different division methods, dividing the target time window into a first time period and a second time period based on each of the division methods, and calculating the discreteness of the first time period and the second time period respectively, wherein the division method is determined based on the position of each of the initial time windows in the target time window; identifying the division method in which the discreteness of the first time period and the second time period after division is less than a preset discreteness threshold as the target division method, and determining the position of the corresponding initial time window as the target segmentation point based on the target division method; Repeat the above steps of determining the target segmentation points until all the target time windows containing the local maximum discreteness are traversed, and the target segmentation points corresponding to each target time window are obtained. The signal control scheme management time is divided by all the target segmentation points to obtain multiple time periods.

7. The method according to claim 1, characterized in that The calculating the actual cycle duration corresponding to each time period includes: The multiple period cycle lengths and period centers corresponding to the period are obtained by enumeration method, and calculations are performed based on each period cycle length and the period center in turn to obtain multiple discreteness corresponding to the period, and the discreteness with the smallest value is taken as the target discreteness, and the period length corresponding to the target discreteness is determined as the actual period length.

8. The method according to claim 7, characterized in that The calculating the actual phase duration corresponding to each time period includes: Selecting the phase with the most observation values ​​of the green light start time as the reference phase, and defining the green light start time in the reference phase as the reference time; The reference time closest to each of the green light start times is respectively used as the target reference time corresponding to each of the green light start times, and the corresponding green light start times are mapped based on each of the target reference times and the actual cycle duration, and a mapping result corresponding to the actual cycle duration is obtained according to the mapping results of all the green light start times; Obtain a preset scheme template, match the scheme template with the mapping result, determine the target scheme template from the scheme template according to the matching result, and determine the phase sequence in the signal control scheme management time according to the target scheme template, and calculate the actual phase duration based on the phase sequence, wherein the scheme template represents the initial phase sequence corresponding to each phase.

9. A device for generating a traffic signal control scheme, characterized in that: The device comprises: An acquisition module, used to acquire the green light start time of the vehicle corresponding to each phase in the intersection, wherein the green light start time of the vehicle reflects the green light start time of the corresponding phase; A calculation module, used for dividing the signal control scheme management time corresponding to the intersection to obtain a plurality of target time windows, and respectively calculating the initial discreteness of each target time window, wherein calculating the initial discreteness corresponding to any target time window comprises mapping the green light start time corresponding to the target time window according to the cycle duration corresponding to the target time window, calculating the center of gravity corresponding to the mapping result, and calculating the initial discreteness based on at least the cycle duration and the center of gravity; arranging the initial discreteness corresponding to all the target time windows in chronological order to obtain an initial discreteness sequence; A generation module is used to determine the local maximum value in the initial discreteness sequence as a segmentation point, and divide the management time of the signal control scheme by the segmentation point to obtain multiple time periods; wherein the cycle durations of the traffic lights in different time periods are different; calculate the actual cycle duration and the actual phase duration corresponding to each of the time periods, and obtain the target signal control scheme corresponding to the intersection according to the actual cycle duration and the actual phase duration.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

Images