Variable Lane Recognition Method, Device, Equipment and Storage Medium

Abstract

The present application discloses a method, apparatus, device, and storage medium for identifying variable lanes. The method includes: obtaining traffic perception data of an intersection section within a preset time period; clustering the traffic perception data according to a preset clustering algorithm to obtain traffic period division result data; inputting the traffic perception data and the traffic period division result data into a preset variable lane identification model to identify variable lane approach lanes that simultaneously meet the steering imbalance condition and the lane imbalance condition, and the traffic periods for setting variable lanes. According to the variable lane identification method provided by the present application, all intersections where variable lanes can be set can be automatically identified, and the traffic periods for setting variable lanes can be recommended, providing a basis for variable lane setting for relevant management departments, thereby improving the urban traffic congestion situation.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent transportation, and particularly to a method, device, equipment and storage medium for identifying variable lanes. Background Art

[0002] With the increasing number of motor vehicles in China, the phenomenon of traffic congestion is becoming increasingly serious. In particular, intersections, as the nodes of urban traffic, are an important part of the urban road system. Therefore, solving the congestion problem at intersections plays a crucial role in alleviating urban traffic congestion. Many intersections will have uneven flows in various directions, which easily leads to low utilization efficiency of the approach lanes and congestion. In order to make more full use of road resources and improve road traffic capacity, variable lanes are now set at many intersections in many cities.

[0003] However, the currently common variable lane method in cities is fixed-time control. This method mainly forms solutions for different time periods through manual research or experience. This means not only time-consuming and laborious, but also overly subjective, and it cannot adapt to the different traffic characteristics of each city, resulting in poor applicability and low reusability of the final governance plan. Moreover, the current research focus on variable lanes is the generation of setting schemes, ignoring the feasibility discrimination of variable lanes at intersections in the early stage. Some intersections themselves do not have the environment for setting variable lanes, or reasonable fixed traffic organization can solve the traffic congestion problem. Blindly setting variable lanes based on experience may lead to the phenomenon of getting worse after governance. Furthermore, the existing variable lane control schemes mostly determine the set time periods based on the peak travel time periods in the city, lacking in-depth analysis of each intersection and ignoring the off-peak periods and some non-congested peak periods. In fact, some intersections can solve the traffic congestion problem through reasonable fixed traffic organization. Blindly setting variable lanes is difficult to ensure the maximum utilization of traffic time-space resources. Summary of the Invention

[0004] The embodiments of the present application provide a method, device, equipment and storage medium for identifying variable lanes. To provide a basic understanding of some aspects of the disclosed embodiments, a simple summary is given below. This summary part is not a general review, nor is it intended to identify key / important constituent elements or delineate the protection scope of these embodiments. Its sole purpose is to present some concepts in a simple form as a preamble to the subsequent detailed description.

[0005] In a first aspect, the embodiments of the present application provide a method for identifying variable lanes, including:

[0006] Obtaining traffic perception data of an intersection section within a preset time period;

[0007] Clustering the traffic perception data according to a preset clustering algorithm to obtain traffic time period division result data;

[0008] Input traffic perception data and traffic period division result data into a preset variable lane recognition model to identify variable lane approach roads that simultaneously meet the turning imbalance condition and the lane imbalance condition, and the traffic periods for setting variable lanes.

[0009] In an optional embodiment, obtain traffic perception data for an intersection section within a preset time period, including:

[0010] Obtain one or more types of traffic perception data among the intersection number, road direction, lane number, lane function, lane traffic flow, lane utilization rate, and push time for the intersection section within the preset time period.

[0011] In an optional embodiment, cluster the traffic perception data according to a preset clustering algorithm to obtain traffic period division result data, including:

[0012] Input the traffic perception data into an improved K-means clustering algorithm to obtain the initial traffic periods after clustering;

[0013] Sort the initial traffic periods in chronological order, and calculate the mean value of each initial traffic period according to the start and end times of each initial traffic period;

[0014] Merge the initial traffic periods with a mean value less than the preset period threshold to obtain the traffic periods after the first merge;

[0015] Merge the traffic periods after the first merge with continuous start and end times and a lane traffic flow mean ratio greater than the preset threshold for a second time to obtain the traffic periods after the second merge;

[0016] Perform period annotation based on the lane traffic flow corresponding to the traffic periods after the second merge to obtain traffic period division result data, where the annotated traffic periods include morning rush hour, evening rush hour, daytime flat peak, and night-time flat peak.

[0017] In an optional embodiment, input the traffic perception data and the traffic period division result data into a preset variable lane recognition model to identify the approach roads that simultaneously meet the turning imbalance condition and the lane imbalance condition, including:

[0018] Input the traffic perception data and the traffic period division result data into the turning imbalance detection unit of the preset variable lane recognition model to obtain the approach roads with turning imbalance;

[0019] Input the traffic perception data and the traffic period division result data corresponding to the approach roads with turning imbalance into the lane imbalance detection unit to obtain the approach roads with lane imbalance.

[0020] In an alternative embodiment, traffic perception data and traffic period division result data are input into a preset variable lane recognition model to identify variable lane approach roads that simultaneously meet the turning imbalance condition and the lane imbalance condition, and the traffic periods for setting variable lanes, including:

[0021] Input the traffic perception data and the traffic period division result data into a preset variable lane recognition model to identify approach roads that simultaneously meet the turning imbalance condition and the lane imbalance condition;

[0022] Count the number of days of approach roads that simultaneously meet the turning imbalance condition and the lane imbalance condition within a preset time period. If the number of days of the approach road that simultaneously meets the turning imbalance condition and the lane imbalance condition is greater than a preset number of days threshold, then the approach road meets the frequently occurring feature of imbalance, and the approach road is a variable lane approach road;

[0023] According to a preset cycle, calculate the union of the lane imbalance time periods corresponding to the variable lane approach roads to obtain the traffic periods for setting variable lanes.

[0024] In an alternative embodiment, input the traffic perception data and the traffic period division result data into the turning imbalance detection unit of a preset variable lane recognition model to obtain approach roads with turning imbalance, including:

[0025] Calculate the maximum left-turn flow ratio and the minimum left-turn flow ratio of each approach road during the morning peak period and the evening peak period according to a preset cycle;

[0026] Calculate the maximum left-turn flow ratio and the minimum left-turn flow ratio of each approach road during the daytime flat peak period;

[0027] Calculate the absolute value of the first difference between the maximum left-turn flow ratio during the morning peak period and the evening peak period and the minimum left-turn flow ratio during the daytime flat peak period, and the absolute value of the second difference between the minimum left-turn flow ratio during the morning peak period and the evening peak period and the maximum left-turn flow ratio during the daytime flat peak period. Obtain the maximum difference value according to the larger value of the absolute value of the first difference and the absolute value of the second difference;

[0028] Multiply the maximum difference value by the sum of the number of left-turn and straight-through lanes of the approach road to obtain the turning imbalance coefficient;

[0029] When the turning imbalance coefficient is greater than or equal to a preset first coefficient threshold, determine that the approach road has turning imbalance.

[0030] In an alternative embodiment, input the traffic perception data corresponding to the approach road with turning imbalance and the traffic period division result data into the lane imbalance detection unit to obtain approach roads with lane imbalance, including:

[0031] Calculate the average value of the maximum lane utilization rate and the average value of the minimum lane utilization rate for each import lane during the first period corresponding to the morning peak and the second period corresponding to the evening peak according to a preset period;

[0032] Obtain the lane imbalance coefficient of the import lane according to the difference between the average value of the maximum lane utilization rate and the average value of the minimum lane utilization rate;

[0033] When the lane imbalance coefficient of the import lane is greater than or equal to a preset second coefficient threshold and the average value of the minimum lane utilization rate is less than or equal to a preset third coefficient threshold, determine that the lane of the import lane is imbalanced.

[0034] In a second aspect, an embodiment of the present application provides a variable lane identification device, including:

[0035] An acquisition module for acquiring traffic perception data of an intersection section within a preset time period;

[0036] A clustering module for clustering the traffic perception data according to a preset clustering algorithm to obtain traffic period division result data;

[0037] An identification module for inputting the traffic perception data and the traffic period division result data into a preset variable lane identification model to identify a variable lane import lane that simultaneously satisfies the steering imbalance condition and the lane imbalance condition, and the traffic period for setting the variable lane.

[0038] In a third aspect, an embodiment of the present application provides an electronic device, including a processor and a memory storing program instructions, and the processor is configured to execute the variable lane identification method provided in the above embodiment when executing the program instructions.

[0039] In a fourth aspect, an embodiment of the present application provides a computer-readable medium, on which computer-readable instructions are stored, and the computer-readable instructions are executed by a processor to implement a variable lane identification method provided in the above embodiment.

[0040] The technical solutions provided by the embodiments of the present application may include the following beneficial effects:

[0041] The variable lane recognition method provided by the embodiments of this application mines the traffic flow period characteristics at intersections through clustering analysis of multi-source traffic perception data, inputs the traffic perception data and the traffic period division result data into a preset variable lane recognition model, and identifies, according to the standard model, the variable lane approach lanes that simultaneously meet the turning imbalance condition and the lane imbalance condition, as well as the traffic periods for setting variable lanes. This method can achieve a global road network scan, automatically identify all intersections where variable lanes can be set, and scientifically solve the defects in the traditional variable lane setting, such as time-consuming and laborious governance processes, strong subjectivity and low reusability of governance plans due to complete reliance on expert experience. It provides a basis for variable lane setting for relevant management departments, thereby improving the urban traffic congestion situation.

[0042] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The accompanying drawings herein are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with the present invention and used together with the specification to explain the principles of the present invention.

[0044] Figure 1 is a flowchart showing a method for recognizing a variable lane according to an exemplary embodiment;

[0045] Figure 2 is a schematic diagram showing a method for recognizing a variable lane according to an exemplary embodiment;

[0046] Figure 3 is a schematic diagram showing the output result of a K-means clustering algorithm according to an exemplary embodiment;

[0047] Figure 4 is a schematic diagram showing the output result of an improved K-means clustering algorithm according to an exemplary embodiment;

[0048] Figure 5 is a schematic diagram showing the structure of a variable lane recognition device according to an exemplary embodiment;

[0049] Figure 6 is a schematic diagram showing the structure of an electronic device according to an exemplary embodiment;

[0050] Figure 7 is a schematic diagram showing a computer storage medium according to an exemplary embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] The following description and the accompanying drawings fully illustrate the specific embodiments of the present invention, enabling those skilled in the art to practice them.

[0052] It should be clear that the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0053] When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. On the contrary, they are merely examples of systems and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0054] This application provides a method for identifying variable-lane intersections. Through comprehensive research and analysis of traffic big data, combined with national standards and machine learning algorithms, the characteristics of traffic flow periods at intersections are mined, and indicators that can quantitatively describe the spatio-temporal imbalance characteristics of intersections are proposed, thereby identifying intersections where variable lanes can be set. The present invention can achieve a full-range road network scan, automatically identify all intersections where variable lanes can be set, and recommend the periods for switching existing lanes, providing a basis for variable lane setting for relevant management departments, thereby improving the urban traffic congestion situation.

[0055] The following will introduce in detail the method for identifying variable lanes provided in the embodiments of this application with reference to the accompanying drawings. See Figure 1 , and the method specifically includes the following steps.

[0056] S101 Obtain traffic perception data of the intersection section within a preset time period.

[0057] In one implementation, the traffic operation status of intersections is extremely susceptible to the influence of the intersections themselves and the surrounding environment. To ensure the timeliness of the variable lane setting scheme, it is necessary to extract recent data for corresponding analysis. At the same time, since the traffic flows on weekdays and holidays have different traffic characteristics, it is recommended to select the traffic perception data of the intersection within the past 30 days and divide the data set into two parts for weekdays and holidays for analysis.

[0058] For example, the required traffic perception data can be generated based on the vehicle passing data of the electronic police, or the corresponding data output by the radar can be directly used for calculation. As shown in the following table, the obtained traffic perception data includes one or more of the intersection number, road direction, lane number, lane function, lane traffic flow, lane utilization rate, and push time of the intersection section.

[0059] Field Name Field Meaning Data Type junction_id Intersection Number string id Direction string lane_id Lane Number string Lane_turning Lane Function string flow Flow string saturation Lane Utilization Rate string date_time Push Time (Time Interval: 5 min) datetime

[0060] According to this step, traffic perception data of the intersection section within a preset time period can be obtained.

[0061] S102 clusters the traffic perception data according to a preset clustering algorithm to obtain traffic period division result data.

[0062] When setting variable lanes, experts first roughly divide urban traffic into peak, off-peak, and flat-peak according to experience, and then conduct research on problem intersections during peak hours. However, in fact, for large cities, different intersections have different traffic characteristics, which may not necessarily coincide with the morning and evening peak hours of the city. At the same time, the traffic flow at intersections will change to a certain extent in different seasons. To more accurately automatically analyze the traffic flow characteristics of each intersection, this application proposes a traffic flow multi-period division algorithm.

[0063] In a possible implementation, the traffic flow sequence is ordered and cannot be disrupted during division. Therefore, an ordered clustering algorithm can be used for division. However, the ordered clustering algorithm has the defects of extremely large computational complexity and slow running speed, resulting in poor implementation.

[0064] In an alternative embodiment, this application performs period division based on an improved K-means clustering algorithm, inputs the traffic perception data into the improved K-means clustering algorithm, and obtains traffic period division result data.

[0065] Specifically, first input the traffic perception data into the improved K-means clustering algorithm to obtain the initial traffic periods after clustering.

[0066] The algorithm inputs the number of k-clusters. Usually, the elbow method can be used to identify the optimal number of classifications k. However, since the traffic period division itself requires a certain amount of expert experience intervention, in this embodiment, k is directly set to 5, and this value can be adjusted according to the urban traffic characteristics. This application does not make specific limitations.

[0067] Furthermore, input the D-dataset. To avoid differences in analysis results caused by daily traffic changes, this application selects the average traffic volume every 5 minutes in the 30-day dataset (divided into weekdays / holidays) for analysis, and the total sample size is 288. Among them, the specific dataset selection can be set according to the actual situation.

[0068] After inputting the data, execute the algorithm process, including the following steps:

[0069] ① Initialization: Randomly select 5 objects from D as the initial cluster centers;

[0070] ② Calculate the distance from each clustering object to the cluster center, and assign each object to the most similar cluster;

[0071] ③ Recalculate the mean of the objects in each cluster as the new center point of each cluster;

[0072] ④ Repeat steps ② and ③ until the various cluster centers no longer change, and then complete the clustering.

[0073] Obtain the initial traffic periods after clustering. The improved K-means clustering algorithm provided in the embodiments of the present application adjusts the output results.

[0074] Specifically, first obtain the time attributes of the segmented sample points. First, sort the initial traffic periods after segmentation in chronological order, and calculate the mean value x of each initial traffic period according to the start and end times of each initial traffic period. i 。

[0075] Further, merge the initial traffic periods with mean values less than the preset period threshold to obtain the traffic periods after the first merge. To avoid the problems of efficiency and safety hazards caused by frequent changes of the variable lane plan at intersections, the setting duration of the variable lane usually cannot be less than 1h. Therefore, set the preset period threshold to 1h, and it can also be set to 1.5h, 2h. The embodiments of the present application do not limit the specific value of the preset period threshold. Extract the initial traffic period i less than 1h, calculate the period with a smaller mean difference between this initial traffic period and the mean values of the two adjacent periods before and after. If |x i -x i-1 |>|x i -x i+1 |, then merge this period into the i + 1 period. If |x i -x i-1 |<|x i -x i+1 |, then merge this period into the i - 1 period. If they are equal, then merge it into the period with a shorter time.

[0076] Further, perform a second merge on the traffic periods after the first merge with continuous start and end times and a lane flow mean ratio greater than the preset threshold to obtain the traffic periods after the second merge. Obtain the mean values of the lane flows corresponding to two consecutive periods. For example, the mean traffic flow from 9:00 to 10:00 is 400, and the mean traffic flow from 10:00 to 11:00 is 410. Calculate the mean ratio of the two consecutive periods, 400 / 410 = 0.97. If the mean ratio is greater than the preset threshold, then merge these two periods. Among them, the preset threshold is 0.8. Those skilled in the art can also set the value of the preset threshold according to the actual situation.

[0077] Further, perform period annotation according to the lane flows corresponding to the traffic periods after the second merge to obtain the traffic period division result data. Among them, the annotated traffic periods include the morning peak, evening peak, daytime flat peak, and night-time flat peak.

[0078] Calculate the average lane flow corresponding to the traffic periods after the secondary merger. Take the top two traffic periods with the largest average lane flow as the peak periods, and represent the morning peak and the evening peak respectively according to the time attribute. The remaining periods can be marked as the night off-peak and the day off-peak respectively according to the time attribute.

[0079] In an exemplary scenario, taking the data of the intersection of Yuetan North Street - Fuchengmen South Street in Beijing on July 28, 2022 (weekday) as an example, a comparison of the outputs before and after the improvement is made. As Figure 3 shown, it is a schematic diagram of the output result of a K-means clustering algorithm. Figure 3 In it, the horizontal axis represents time, the vertical axis represents flow, and multiple vertical lines parallel to the vertical axis represent the divided periods. It can be seen that the period duration before the improvement is short and difficult to meet the requirements. Figure 4 It is a schematic diagram of the output result of an improved K-means clustering algorithm. As Figure 4 shown, the periods divided by the multiple vertical lines after the improvement are more reasonable. The present application improves the output result of the k-means clustering algorithm and can generate a traffic period division result that meets the variable lane setting.

[0080] S103 Input the traffic perception data and the traffic period division result data into a preset variable lane recognition model to identify the variable lane approach that simultaneously meets the steering imbalance condition and the lane imbalance condition, as well as the traffic periods for setting variable lanes.

[0081] The traffic organization and governance measures at intersections can be divided into fixed lane function optimization and dynamic lane function optimization. Fixed lane function optimization is more acceptable to drivers. Usually, if a series of problems can be solved through fixed traffic organization methods at an intersection, variable lanes are not considered. In existing research, the traffic conditions (idle running, imbalance, etc.) during the off-peak periods of intersections are usually ignored, and only the lane function optimization is carried out during the congestion periods or peak periods. In this way, some intersections that do not need to set variable lanes are "forced" to set variable lanes, and the generated governance plan is not the optimal governance plan for this intersection. Therefore, from the perspectives of time and space, the present application proposes two imbalance coefficients to identify the approach and periods that need to set variable lanes.

[0082] Specifically, first input the traffic perception data and the traffic period division result data into the steering imbalance detection unit of the preset variable lane recognition model, and obtain the approach with steering imbalance according to the steering imbalance detection unit.

[0083] During different time periods, there are certain differences in the proportion of traffic flows in each turning direction at the intersection approach. When this difference exceeds a certain threshold and the traffic demand is large, it will be difficult for the fixed lane function to meet the turning demands in different time periods simultaneously, and traffic congestion may occur in some turning directions during certain time periods. This application proposes a turning imbalance coefficient to identify the approach with a time-periodic imbalance in the proportion of each turning demand.

[0084] Identifying the approach with turning imbalance includes: calculating the maximum left-turn flow proportion and the minimum left-turn flow proportion of each approach during the morning peak period and the evening peak period. It should be noted that both the morning peak period and the evening peak period refer to the time periods in the traffic period division result obtained by the improved K-means clustering algorithm mentioned above. The embodiments of this application mainly consider the left-turn and straight-ahead traffic demands, and with a 5-minute data statistics granularity, the proportion of the left-turn flow in each cycle of approach i, denoted as Q′ i , where Q i，1 is the left-turn flow, and Q i，2 is the straight-ahead flow.

[0085] In a possible implementation, with a 30-minute minimum cycle, starting from the start time of the morning peak period and the evening peak period respectively, with a 5-minute time moving window, find the maximum left-turn flow proportion and the minimum left-turn flow proportion of approach i during the morning peak period and the evening peak period, denoted as Q′ i，Amax and Q′ i，Amin respectively. For example, the specific search method is to take the start time of the morning peak period as the time starting point, and gradually move backward with a 5-minute window. The traffic flow in the 9:00 - 9:30 period is statistically counted for the first time, the traffic flow in the 9:05 - 9:35 period is statistically counted for the second time, and so on, until the traffic flow in each cycle during the morning peak period is calculated.

[0086] Calculate the maximum left-turn flow proportion and the minimum left-turn flow proportion of each approach during the daytime off-peak period. Similarly, with a 30-minute minimum cycle, starting from the start time of multiple daytime off-peak periods respectively, with a 5-minute moving window, find the maximum and minimum Q′ i of approach i during the daytime off-peak period, denoted as Q′ i，Bma x and Q′ i，Bmin respectively.

[0087] Calculate the absolute value of the first difference between the maximum left-turn flow proportion during the morning peak period and the evening peak period and the minimum left-turn flow proportion during the daytime off-peak period, and the absolute value of the second difference between the minimum left-turn flow proportion during the morning peak period and the evening peak period and the maximum left-turn flow proportion during the daytime off-peak period. Obtain the maximum difference value based on the larger value between the absolute value of the first difference and the absolute value of the second difference. Calculate the maximum difference value Q′ of Q′i i，AB. Among them, Q' i，AB = max{|Q' i，Amax - Q' i，Bmin |, |Q' i，Bmax - Q' i，Amin |}

[0088] The steering imbalance coefficient is obtained according to the product of the maximum difference value and the sum of the number of left-turn and straight-through lanes in the approach lane.

[0089] In a possible implementation, the steering imbalance coefficient Q of approach lane i i，AB = Q' i，AB * N i，ls , where N i，ls represents the total number of left-turn and straight-through lanes in approach lane i.

[0090] When the steering imbalance coefficient is greater than or equal to a preset first coefficient threshold, it is determined that the approach lane is steered unevenly. In an exemplary scenario, assuming that in an ideal situation, when the turning traffic flows are in a balanced state, Q i，1 : Q i，2 = n i，1 : n i，2 , and at this time Q i，AB = 0; when the traffic flow changes with time, Q i，1 : Q i，2 = (n i，1 + 0.5): (n i，2 - 0.5), whether the left-turn lane needs to be increased or remain unchanged depends on the situation, and at this time Q i，AB = 0.5. When the traffic flow changes with time and one left-turn lane needs to be added and one straight-through lane needs to be reduced, at this time Q i，AB = 1.

[0091] In summary, there are three extreme critical values. When Q i，AB ≥ 1, the steering imbalance phenomenon is serious and the variable characteristics are prominent. For better practical application, the Q i，AB threshold can be appropriately adjusted downwards and set to 0.8. Those skilled in the art can also set it according to the actual situation.

[0092] Furthermore, the traffic perception data corresponding to the approach lane with steering imbalance and the traffic period division result data are input into the lane imbalance detection unit, and the approach lane with lane imbalance is obtained according to the lane imbalance detection unit.

[0093] Due to the imbalance in traffic demand, when the spatio-temporal resources at the intersection are divided inaccurately, some lanes in the approach lane will be congested, while the traffic capacity of some other lanes will be surplus. This application proposes a lane imbalance coefficient to identify the approach lanes with spatial imbalance in the utilization rate of each lane.

[0094] Based on the calculation results of the turning imbalance coefficient, relevant data of approach i with turning imbalance characteristics are extracted, and are respectively denoted as period 1 (start time: 10 minutes before the start time of the morning peak period, end time: 10 minutes after the end time of the morning peak period) and period 2 (start time: 10 minutes before the start time of the evening peak period, end time: 10 minutes after the end time of the evening peak period).

[0095] For the relevant data of approach i with turning imbalance characteristics extracted, calculate the average value of the maximum lane utilization rate and the average value of the minimum lane utilization rate of each approach during the first period corresponding to the morning peak and the second period corresponding to the evening peak. Specifically, with 15 minutes as the minimum cycle, starting from the start time of period 1 and period 2 respectively, with a 5-minute time moving window, calculate the average value M of the lane utilization rate of each lane of approach i for 15 minutes i，l . Among them, q i，l , g i，l , h i，l respectively represent the traffic volume, saturated headway (taking 2.5 s), and effective green time of lane l of approach i, and c represents the signal cycle length. According to this calculation method, the average value M of the maximum lane utilization rate and the average value M of the minimum lane utilization rate of each approach during the first period corresponding to the morning peak and the second period corresponding to the evening peak are calculated i，lmax and the average value M of the minimum lane utilization rate i，lmin .

[0096] According to the difference between the average value of the maximum lane utilization rate and the average value of the minimum lane utilization rate, the lane imbalance coefficient M of the approach is obtained i . Among them, M i = M i，lmax - M i，lmin .

[0097] When the lane imbalance coefficient of the approach is greater than or equal to the preset second coefficient threshold and the average value of the minimum lane utilization rate is less than or equal to the preset third coefficient threshold, it is determined that the lane of the approach is unbalanced. In an exemplary scenario, the second coefficient threshold is 0.4. To avoid unreasonable selection of the second coefficient threshold and the situation where the lane utilization rates of all lanes are very large, a minimum expected lane utilization rate limit is added, and the third coefficient threshold is set to 0.6. Among them, the coefficient thresholds can be appropriately adjusted according to the actual situation, and the embodiments of the present application do not make specific limitations

[0098] In an alternative embodiment, if the duration of intersection period 1 / period 2 is not a multiple of 15 minutes, the remaining 5 minutes or 10 minutes can be directly calculated, and at this time, the influence of the time length on the traffic flow can be ignored. Since the variable lane should not be switched frequently, the best period for setting the variable lane is tentatively set to 1 h (adjustable), that is, if there are at least 4 consecutive cycles, the lane imbalance coefficient Mi If it is greater than or equal to the second coefficient threshold, then this time period meets the prerequisite for setting a variable lane.

[0099] Further, input the traffic perception data and the traffic time period division result data into a preset variable lane recognition model to identify the approach that simultaneously meets the turning imbalance condition and the lane imbalance condition.

[0100] Count the number of days of the approach that simultaneously meets the turning imbalance condition and the lane imbalance condition within a preset time period. If the number of days that this approach simultaneously meets the turning imbalance condition and the lane imbalance condition is greater than the preset number of days threshold, then this approach meets the characteristic of frequent occurrence of imbalance, and this approach is a variable lane approach.

[0101] In an exemplary scenario, after calculating the above two indicators, it is also necessary to ensure that the imbalance characteristic has frequent occurrence. In the 30-day dataset (divided into weekdays / holidays) selected in the embodiments of the present application, if 70% or more days meet the turning imbalance constraint, then this approach meets the turning imbalance characteristic. After meeting this characteristic, if 70% or more days meet the lane imbalance constraint, then this approach meets the lane imbalance characteristic. Determine the approach of the variable lane by simultaneously meeting the turning imbalance characteristic and the lane imbalance characteristic.

[0102] Further, obtain the union of the turning imbalance and lane imbalance time periods that meet the constraint days. Calculate the union of the lane imbalance time periods corresponding to the variable lane approach according to a preset cycle to obtain the traffic time period for setting the variable lane.

[0103] Finally, the variable lane detection model outputs the approach of the intersection where the variable lane can be set and the time period for recommending the lane change function.

[0104] To facilitate understanding of the variable lane recognition method provided by the embodiments of the present application, the following is combined with the attached

[0105] Figure 2 for illustration. As Figure 2 shown, this method includes the following steps.

[0106] First, obtain the intersection basic data, the traffic police vehicle passing data or the radar data, and the intersection signal timing plan data, and process the obtained multi-source data, including intersection matching, deleting empty license records, and cleaning short-term repeated records of the same vehicle. Obtain the traffic perception data of various indicators such as the intersection number, road direction, lane number, lane function, lane traffic flow, lane utilization rate, and push time of the intersection section according to the processed data.

[0107] Further, input the traffic perception data of various indicators into an improved K-means clustering algorithm to obtain the traffic time period division result data.

[0108] Furthermore, based on the traffic period division result data, time imbalance features are identified (by calculating the turning imbalance coefficient), and spatial imbalance features are identified (by calculating the lane imbalance coefficient), to obtain the approach lanes where variable lanes can be set and the time periods for recommending lane-changing functions.

[0109] Variable lanes are a simple and effective traffic congestion mitigation measure, but not all intersections are suitable for setting variable lanes. Blind setting may lead to worse conditions after governance. The variable lane identification method proposed in this application generates a multi-period division result of traffic flow that meets the requirements for setting variable lanes by improving the output of the k-means clustering algorithm. At the same time, it proposes indicators that can quantitatively describe the spatio-temporal imbalance features of intersections to achieve the identification of intersections where variable lanes can be set throughout the region.

[0110] Compared with the existing identification of variable lanes, most are "expert experience-based", which makes it difficult to reuse in each city or even at each intersection. At the same time, blindly setting variable lanes based on experience may lead to worse conditions after governance. This application quantifies the process standard for variable lane identification through a simple and effective algorithm + rules. The model has a fast calculation speed, is easy to operate, and has strong practicality, and can be reused in different cities. It scientifically solves the defects in the traditional setting of variable lanes, such as time-consuming and laborious governance processes, strong subjectivity of governance plans, and low reusability due to complete reliance on expert experience.

[0111] The embodiment of this application also provides a variable lane identification device, which is used to execute the variable lane identification method of the above embodiment, as Figure 5 shown. The device includes:

[0112] An acquisition module 501, configured to acquire traffic perception data of an intersection section within a preset time period;

[0113] A clustering module 502, configured to cluster the traffic perception data according to a preset clustering algorithm to obtain traffic period division result data;

[0114] An identification module 503, configured to input the traffic perception data and the traffic period division result data into a preset variable lane identification model to identify the variable lane approach lanes that simultaneously meet the turning imbalance condition and the lane imbalance condition, and the traffic periods for setting variable lanes.

[0115] It should be noted that when the variable lane recognition device provided in the above embodiments executes the variable lane recognition method, only the division of the above functional modules is used for illustration. In practical applications, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the variable lane recognition device provided in the above embodiments and the variable lane recognition method embodiments belong to the same concept. The implementation process is detailed in the method embodiments and will not be repeated here.

[0116] An embodiment of the present application also provides an electronic device corresponding to the variable lane recognition method provided in the foregoing embodiments to execute the variable lane recognition method described above.

[0117] Please refer to Figure 6 , which shows a schematic diagram of an electronic device provided in some embodiments of the present application. As Figure 6 shown, the electronic device includes: a processor 600, a memory 601, a bus 602, and a communication interface 603. The processor 600, the communication interface 603, and the memory 601 are connected through the bus 602; a computer program that can run on the processor 600 is stored in the memory 601, and when the processor 600 runs the computer program, it executes the variable lane recognition method provided in any of the foregoing embodiments of the present application.

[0118] Among them, the memory 601 may include a high-speed random access memory (RAM: Random Access Memory), and may also include a non-volatile memory, such as at least one disk memory. Through at least one communication interface 603 (which can be wired or wireless), a communication connection is established between the system network element and at least one other network element, and the Internet, wide area network, local area network, metropolitan area network, etc. can be used.

[0119] The bus 602 may be an ISA bus, a PCI bus, or an EISA bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. Among them, the memory 601 is used to store a program. After receiving an execution instruction, the processor 600 executes the program. The variable lane recognition method disclosed in any of the foregoing embodiments of the present application can be applied to the processor 600 or implemented by the processor 600.

[0120] The processor 600 may be an integrated circuit chip with the ability to process signals. In the implementation process, each step of the above method may be completed by the integrated logic circuit of the hardware in the processor 600 or the instructions in the form of software. The above-mentioned processor 600 may be a general-purpose processor, including a central processing unit (CPU for short), a network processor (NP for short), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the embodiments of the present application may be directly embodied as being executed and completed by the hardware decoding processor, or executed and completed by a combination of the hardware and software modules in the decoding processor. The software module may be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. This storage medium is located in the memory 601, and the processor 600 reads the information in the memory 601 and combines its hardware to complete the steps of the above method.

[0121] The electronic device provided in the embodiments of the present application and the variable lane recognition method provided in the embodiments of the present application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run, or implemented by them.

[0122] The embodiments of the present application also provide a computer-readable storage medium corresponding to the variable lane recognition method provided in the foregoing embodiments. Please refer to Figure 7 , which shows that the computer-readable storage medium is an optical disc 700, on which a computer program (i.e., a program product) is stored. When the computer program is run by a processor, it will execute the variable lane recognition method provided in any of the foregoing embodiments.

[0123] It should be noted that examples of computer-readable storage media may also include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory, or other optical and magnetic storage media, which will not be elaborated here one by one.

[0124] The computer-readable storage medium provided by the above embodiments of the present application and the method for identifying variable lanes provided by the embodiments of the present application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run, or implemented by the application programs stored therein.

[0125] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, 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, it should be considered as the scope described in this specification.

[0126] The above embodiments only represent several implementation manners of the present invention, and the description thereof is relatively specific and detailed, but it should not be understood as a limitation on the scope of the patent of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the inventive concept of the present invention, several deformations and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention should be subject to the appended claims.

Claims

1. A method for identifying a variable lane, characterized in that, Including: Obtain traffic perception data of the intersection section within a preset time period; Cluster the traffic perception data according to a preset clustering algorithm to obtain traffic period division result data; among them, the marked traffic periods include morning rush hour, evening rush hour, daytime flat peak, and night-time flat peak; Input the traffic perception data and the traffic period division result data into a preset variable lane recognition model to identify variable lane approach lanes that simultaneously meet the turning imbalance condition and the lane imbalance condition, and the traffic periods for setting variable lanes; including: input the traffic perception data and the traffic period division result data into the turning imbalance detection unit of the preset variable lane recognition model, calculate the maximum left-turn flow ratio and the minimum left-turn flow ratio of each approach lane during the morning rush hour and the evening rush hour according to a preset cycle; calculate the maximum left-turn flow ratio and the minimum left-turn flow ratio of each approach lane during the daytime flat peak period; calculate the absolute value of the first difference between the maximum left-turn flow ratio during the morning rush hour and the evening rush hour and the minimum left-turn flow ratio during the daytime flat peak period, and the absolute value of the second difference between the minimum left-turn flow ratio during the morning rush hour and the evening rush hour and the maximum left-turn flow ratio during the daytime flat peak period, and obtain the maximum difference value according to the larger value of the absolute value of the first difference and the absolute value of the second difference; multiply the maximum difference value by the sum of the number of left-turn and straight-through lanes of the approach lane to obtain the turning imbalance coefficient; when the turning imbalance coefficient is greater than or equal to a preset first coefficient threshold, determine that the approach lane is turning-imbalanced; Including: input the traffic perception data and the traffic period division result data corresponding to the turning-imbalanced approach lane into the lane imbalance detection unit, calculate the maximum lane utilization rate mean value and the minimum lane utilization rate mean value of each turning-imbalanced approach lane during the first period corresponding to the morning rush hour and the second period corresponding to the evening rush hour according to a preset cycle; obtain the lane imbalance coefficient of the approach lane according to the difference between the maximum lane utilization rate mean value and the minimum lane utilization rate mean value; when the lane imbalance coefficient of the approach lane is greater than or equal to a preset second coefficient threshold and the minimum lane utilization rate mean value is less than or equal to a preset third coefficient threshold, determine that the approach lane is lane-imbalanced.

2. The method according to claim 1, characterized in that, Obtain traffic perception data of the intersection section within a preset time period, including: Obtain one or more pieces of traffic perception data among the intersection number, road direction, lane number, lane function, lane flow, lane utilization rate, and push time of the intersection section within a preset time period.

3. The method according to claim 1, wherein Cluster the traffic perception data according to a preset clustering algorithm to obtain traffic period division result data, including: Input the traffic perception data into an improved K-means clustering algorithm to obtain the initial traffic periods after clustering; Sort the initial traffic periods in chronological order, and calculate the mean value of each initial traffic period according to the start and end times of each initial traffic period; Merge the initial traffic periods with mean values less than the preset period threshold to obtain the traffic periods after the first merge; Perform secondary merging on the traffic periods after the first merging where the start and end times are continuous and the mean lane flow ratio is greater than a preset threshold to obtain the traffic periods after secondary merging; Perform period annotation based on the lane flows corresponding to the traffic periods after secondary merging to obtain traffic period division result data.

4. The method according to claim 1, wherein Input the traffic perception data and the traffic period division result data into a preset variable lane recognition model to identify the variable lane approach roads that simultaneously meet the turning imbalance condition and the lane imbalance condition, and the traffic periods for setting variable lanes, including: Input the traffic perception data and the traffic period division result data into a preset variable lane recognition model to identify the approach roads that simultaneously meet the turning imbalance condition and the lane imbalance condition; Count the number of days of the approach roads that simultaneously meet the turning imbalance condition and the lane imbalance condition within a preset time period. If the number of days of the approach road that simultaneously meets the turning imbalance condition and the lane imbalance condition is greater than a preset number-of-days threshold, then the approach road meets the frequent-occurrence feature of imbalance, and the approach road is a variable lane approach road; Calculate the union of the lane imbalance periods corresponding to the variable lane approach roads at a preset cycle to obtain the traffic periods for setting variable lanes.

5. An identification device for variable lanes, characterized in that, Including: An acquisition module for acquiring traffic perception data of the intersection section within a preset time period; A clustering module for clustering the traffic perception data according to a preset clustering algorithm to obtain traffic period division result data; among them, the labeled traffic periods include morning peak, evening peak, daytime flat peak, and night-time flat peak; An identification module for inputting the traffic perception data and the traffic period division result data into a preset variable lane recognition model to identify the variable lane approach roads that simultaneously meet the turning imbalance condition and the lane imbalance condition, and the traffic periods for setting variable lanes; including: inputting the traffic perception data and the traffic period division result data into the turning imbalance detection unit of the preset variable lane recognition model, calculating the maximum left-turn flow ratio and the minimum left-turn flow ratio of each approach road during the morning peak period and the evening peak period according to a preset cycle; calculating the maximum left-turn flow ratio and the minimum left-turn flow ratio of each approach road during the daytime flat peak period; calculating the absolute value of the first difference between the maximum left-turn flow ratio during the morning peak period and the evening peak period and the minimum left-turn flow ratio during the daytime flat peak period, and the absolute value of the second difference between the minimum left-turn flow ratio during the morning peak period and the evening peak period and the maximum left-turn flow ratio during the daytime flat peak period, and obtaining the maximum difference value according to the larger value of the absolute value of the first difference and the absolute value of the second difference; multiplying the maximum difference value by the sum of the left-turn and straight-through lane numbers of the approach road to obtain the turning imbalance coefficient; when the turning imbalance coefficient is greater than or equal to a preset first coefficient threshold, determine that the approach road has turning imbalance; Including: inputting traffic perception data corresponding to the turning-unbalanced approach and traffic period division result data into a lane unbalance detection unit, calculating the average value of the maximum lane utilization rate and the average value of the minimum lane utilization rate of each turning-unbalanced approach during the first period corresponding to the morning peak and the second period corresponding to the evening peak according to a preset period; obtaining a lane unbalance coefficient of the approach according to the difference between the average value of the maximum lane utilization rate and the average value of the minimum lane utilization rate; when the lane unbalance coefficient of the approach is greater than or equal to a preset second coefficient threshold and the average value of the minimum lane utilization rate is less than or equal to a preset third coefficient threshold, determining that the approach is lane-unbalanced.

6. An electronic device, characterized in that, Including a processor and a memory storing program instructions, the processor being configured to execute the variable lane recognition method according to any one of claims 1 to 4 when executing the program instructions.

7. A computer-readable medium, characterized in that, Stored thereon are computer-readable instructions, the computer-readable instructions being executed by a processor to implement a variable lane recognition method according to any one of claims 1 to 4.

Images