A method for calculating the number of vehicles queuing across the entire road segment based on single data from checkpoints.

Abstract

This invention discloses a method for calculating the number of vehicles queuing across a road segment based on single data from checkpoints, relating to the field of traffic management and control. The method includes acquiring checkpoint vehicle data using checkpoint equipment and merging and calculating the vehicle data from upstream and downstream intersections for each vehicle, eliminating speeding vehicles; using the k-means algorithm to obtain non-queuing passage time; filtering out queuing vehicles by comparing vehicle passage time with non-queuing passage time and obtaining the queuing time interval for each queuing vehicle; statistically analyzing the peak number of all queuing vehicles within the research period; correcting the queuing vehicle count to obtain the maximum number of queuing vehicles per lane within the statistical period; and calculating the number of queuing vehicles by associating upstream and downstream intersection vehicle data. This method solves problems such as data loss and clock asynchrony caused by associating data from multiple system devices, improving the simplicity and effectiveness of detecting and calculating the number of queuing vehicles on a road segment.

Description

Technical fields:

[0001] This invention relates to the field of traffic management and control, and in particular to a method for calculating the number of vehicles queuing across the entire road segment based on single data of vehicles passing through checkpoints. Background technology:

[0002] The number of vehicles queuing can assess road traffic demand and intersection capacity, and is an important indicator for measuring the smoothness of intersections, providing a basis for decision-making in urban road traffic management and traffic congestion control.

[0003] Current methods for detecting the number of vehicles queuing on roads mainly rely on intersection radar and video detection equipment to detect vehicles entering from the intersection. However, due to the limited detection distance of these devices, it is difficult to calculate the number of vehicles queuing across the entire road segment, which to some extent leads to an underestimation of the number of vehicles queuing at intersections. Some current methods have taken into account segment-based queue detection, such as ZL201610652009X, which combines intersection checkpoint data and floating car data (floating cars generally refer to buses and taxis equipped with onboard GPS positioning devices and driving on urban main roads) to locate the floating cars' positions and speeds on the road in real time and further analyze the length of vehicle queues at intersections. This calculation method relies on a large amount of floating car data and is not applicable in cities or road segments with a small number of floating cars. In addition, vehicle queues often occur due to vehicles stopping at intersections. The number of vehicles queuing is caused by waiting at red lights, so it is related to the color of the traffic lights at the intersection. Currently, the calculation method for the number of vehicles queuing is basically to match the data from the checkpoint with the traffic light status data of the intersection signal controller. This involves accurately judging the changes in the color of the traffic lights for each direction at the intersection, recording the changes in the traffic lights in real time, and matching them with the data of vehicles arriving at the intersection. It is necessary to accurately obtain the number of vehicles queuing at the intersection at the moment when the green light turns on for each direction as the maximum number of vehicles queuing for that direction in the current cycle. However, since the current intersection signal controllers and checkpoint equipment are all clocked by their respective servers, and data is often delayed or lost during transmission, it is difficult for the traffic light change data and checkpoint data from different systems to be consistent in the time dimension. This makes it difficult to achieve accurate data matching, resulting in a large deviation in the number of vehicles queuing at the intersection.

[0004] Based on this, the present invention needs to propose a reasonable method for calculating the number of vehicles queuing across the entire road segment, so as to more accurately calculate the number of vehicles queuing at intersections. Summary of the Invention:

[0005] To overcome the shortcomings of existing technologies, the purpose of this invention is to provide a method for calculating the number of vehicles queuing across a road segment based on single data from checkpoints. This method calculates the number of vehicles queuing by correlating vehicle data from upstream and downstream checkpoints, solving the problem that a single detection device at an intersection can only detect the queuing status of vehicles within a certain range. Electronic police checkpoint devices are installed at all traffic light-controlled intersections in cities, and the data source is relatively common and easy to obtain. Furthermore, this method, based on single data from electronic police checkpoints, can effectively calculate the number of vehicles queuing across a road segment, offering the advantage of using a single data source. This avoids problems such as data delays, data loss, and clock asynchrony caused by correlating data from multiple system devices, and particularly addresses the shortcomings caused by the difficulty in matching data with traffic light status. Therefore, it effectively improves the simplicity and effectiveness of calculating the number of vehicles queuing across a road segment.

[0006] The technical solution of the present invention is as follows:

[0007] (1) Establish the relationship between checkpoint equipment and road sections, and build a system for calculating the number of vehicles queuing on the entire road section by using vehicle data collected by checkpoint equipment that is widely available and easy to obtain.

[0008] (2) Obtain vehicle data at upstream and downstream intersection checkpoints, clean the data, and complete the fusion calculation of vehicle data at upstream and downstream intersections for each vehicle. Generate a vehicle travel dataset including vehicle license plate, upstream and downstream intersection number, turning direction and passage time. Calculate the passage time of each vehicle through upstream and downstream intersections. Based on the distance between checkpoint equipment at upstream and downstream intersections and the maximum speed limit of the road section, obtain the minimum passage time of a vehicle without speeding through upstream and downstream intersections. Remove all speeding vehicles whose passage time is less than the minimum passage time.

[0009] (3) The k-means algorithm was used to perform cluster analysis on the remaining vehicle passage time. The vehicle passage time within the analysis period was clustered into k classes, and the mean of the cluster centers of the m data clusters with smaller passage time values ​​was selected as the non-queuing passage time t of the road segment under study during that period. z (To ensure the calculation of non-queuing passage time t) z The selected sample data should be representative, and the number of data in the m-class data clusters should not be less than 10% of the total number of data.

[0010] (4) Calculate the number of vehicles in the queue.

[0011] (4.1) Calculate the vehicle queuing time interval: Based on the time t when vehicle i passes the upstream. i1 With downstream time t i2 Calculate the vehicle passage time and compare it with the non-queue passage time t calculated in the previous step. z For comparison, if the time t for passing the vehicle is... i2 -t i1 Greater than t zIf vehicles are queuing, then the passing time t of vehicle i at the upstream intersection of the road segment is further calculated. i1 Non-queueing passage time t z The sum of these values ​​represents the start time of the vehicle queue, and the time t for passing through the downstream intersection of the road segment. i2 As the end time of queuing, the queuing time interval (t) of the vehicle in that road segment at the current time period is obtained. i1 +t z , t i2 );

[0012] (4.2) Count the number of vehicles queuing on the road segment per unit time: Use t as the unit of statistical time to calculate the real-time number of vehicles queuing on the road segment (to ensure the accuracy of the calculation results of the number of vehicles queuing, it is recommended to take 1 second for t);

[0013] (4.3) Queue Peak Calculation (Number of vehicles queuing at the moment the green light turns on): Under normal circumstances, the number of vehicles queuing on a road segment increases continuously during the red light due to stopping and waiting, and decreases continuously during the green light period as the queue clears, reaching its peak at the moment the green light turns on. Based on the trend of the number of vehicles queuing at various times on the road segment (number of vehicles queuing per second), the peak value of the vehicle queue is obtained so that the number of vehicles queuing is based on the trend of the number of vehicles queuing, thus obtaining the peak value of all vehicles queuing during the study period for that road segment;

[0014] (5) Correction calculation:

[0015] (5.1) Equivalent car correction: Since different vehicle types occupy different road space (large vehicles occupy more road space and small vehicles occupy less), it is necessary to convert the vehicle type into the equivalent car number by multiplying it by the equivalent car correction coefficient.

[0016] (5.2) Queue interval correction: Since the above calculations all use the vehicle passage time t at the upstream intersection. i1 Non-queueing passage time t z The sum of these values ​​is used as the starting time for the vehicle queue. However, usually there are already some vehicles initially queuing at the intersection, so the actual starting time for most vehicles to queue is earlier than t. i1 +t z To ensure the accuracy of the calculation results, the number of vehicles in the queue obtained above needs to be multiplied by an amplification factor in a manner that is proportional to the current number of vehicles in the queue.

[0017] (5.3) Right turn correction: Since most intersections have right turn channelization, vehicles do not turn right through the intersection. The intersection checkpoint equipment is difficult to capture right-turning vehicles. To ensure the accuracy of the algorithm calculation, it is necessary to perform correction calculations for right-turning vehicles. First, calculate the number of vehicles leaving the downstream intersection straight ahead. Then, calculate the number of vehicles arriving at the upstream intersection straight ahead and turning left, and leaving the downstream intersection straight ahead. The quotient of the two is the straight-ahead queue correction coefficient. Similarly, calculate the number of vehicles leaving the downstream intersection by turning left and the number of vehicles arriving at the upstream intersection by turning left and turning left, and leaving the downstream intersection by turning left. The quotient of the two is the left-turn queue correction coefficient. The straight-ahead queue correction coefficient and the left-turn queue correction coefficient are used to correct the queued vehicles for straight ahead and left turn respectively.

[0018] (6) Calculation of the average number of vehicles queuing in a single lane: The peak number of vehicles queuing after correction of various dimensions during the study period of this road section is calculated by combining the peak number of vehicles queuing in each direction with the number of lanes corresponding to the direction of flow to obtain the maximum number of vehicles queuing in a single lane during the statistical period.

[0019] Compared with the prior art, the present invention has the following technical advantages:

[0020] 1. This invention uses widely available and easily accessible checkpoint data, and calculates the number of vehicles queuing on a road segment by linking the vehicle data of upstream and downstream intersections, thus solving the problem that a single detection device can only detect the queuing status of vehicles within a certain range of an intersection.

[0021] 2. This invention uses widely available and easily accessible checkpoint data, and calculates the number of vehicles queuing on a road segment by associating it with vehicle data from upstream and downstream intersections. This avoids problems such as data loss and clock asynchrony caused by associating data from multiple system devices, and especially the shortcomings caused by matching with traffic lights, thus improving the simplicity and effectiveness of detecting and calculating the number of vehicles queuing on a road segment. Attached image description:

[0022] Figure 1 This is a flowchart of the method of the present invention.

[0023] Figure 2 This is a schematic diagram of the checkpoint device of the present invention at the intersection.

[0024] Figure 3 This is a schematic diagram of the vehicle passage data structure of the checkpoint in this invention.

[0025] Figure 4 This is a schematic diagram of the gate vehicle matching data structure of the present invention.

[0026] Figure 5 This is a clustering diagram of non-queue passage time according to the present invention.

[0027] Figure 6 This is a schematic diagram of the logic for determining the number of vehicles in a queue according to the present invention.

[0028] Figure 7 This is a schematic diagram of an invention example. Detailed implementation method:

[0029] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0030] The following is combined with Figures 1 to 6 The present invention will be described in further detail below.

[0031] Step 1: Using widely available and easily accessible checkpoint vehicle data, build a system for calculating the number of vehicles queuing across the entire road segment. The checkpoint equipment is located at intersections such as... Figure 2 As shown, these checkpoints are typically positioned near the stop line on each direction of the intersection's approach lanes. They serve two purposes: firstly, to monitor and photograph vehicles running red lights, and secondly, to record the real-time data of every vehicle passing through the intersection. The structure of the vehicle passage data obtained using these checkpoint devices is as follows: Figure 3 As shown, the information includes vehicle ID, intersection number, direction of approach lane, flow direction, lane number, license plate number, vehicle type, and passage time.

[0032] Step 2: Since the flow rates for straight-ahead and left-turn traffic at intersections differ, the number of vehicles queuing also differs. Therefore, the number of vehicles queuing at intersections is calculated separately for straight-ahead and left-turn traffic. Traffic data for straight-ahead and left-turn traffic at upstream and downstream intersections is obtained. The data is cleaned, duplicate data is removed, and upstream and downstream traffic matching data for downstream straight-ahead and left-turn traffic are generated separately. The structure of the checkpoint traffic matching data is as follows: Figure 4 As shown, this includes vehicle ID, license plate number, vehicle type, transit time at upstream and downstream intersections, intersection numbers, direction of entrance lanes at upstream and downstream intersections, lane numbers at upstream and downstream intersections, and flow direction at upstream and downstream intersections. The transit time for each vehicle at the upstream and downstream intersections is calculated. Based on the distance between the upstream and downstream intersections and the maximum speed limit for this road segment, the minimum transit time that a vehicle can travel without exceeding the speed limit is obtained. Vehicles exceeding the speed limit with a transit time shorter than the minimum transit time are then eliminated.

[0033] Step 3: As Figure 3As shown, the k-means algorithm is used to perform cluster analysis on vehicle transit time. The specific process is as follows: 1. The optimal number of clusters k is generally between 2 and 7, and k = 5 is selected; 2. Randomly select 5 data points as initial cluster centers; 3. Using the distance formula Δx = |h - h0|, where h is the time taken for each vehicle and h0 is the time taken for each initial cluster center, calculate the distance between all data points and each cluster center, and assign each data point to the nearest cluster center; 4. Calculate the average coordinates of all points in each cluster as the updated cluster center; 5. Repeat steps 3 and 4 above, iteratively calculating the distance from each data point to the updated cluster center, and continuously updating the cluster centers until the cluster centers of each data cluster no longer change; 6. Sort the data clusters from smallest to largest according to the value of the cluster centers, calculate the proportion of the number of data points in the first data cluster to the total number of data points. If it is less than 10%, then take the first two data clusters for calculation, until the number of data points in the first m data clusters is greater than 10% of the total number of data points, and take the average value of the cluster centers of the m data clusters as the non-queue passage time t. z ;

[0034] Step 4: Calculate the number of vehicles in the queue.

[0035] (1) Calculate the vehicle queuing time interval: such as Figure 4 As shown, based on the time t when vehicle i passes through the upstream intersection... i1 The time t of passing through the downstream intersection i2 Calculate the vehicle passage time and compare it with the non-queue passage time t calculated in the previous step. z For comparison, if the vehicle passage time (t) i2 -t i1 ) greater than t z Then vehicle i is in a queue, and further, t i1 With t z The sum of these values ​​represents the time at which the vehicle queue begins, t. i2 As the end time of queuing, the queuing time interval (t) of the vehicle in that road segment at the current time period is obtained. i1+ t z , t i2 );

[0036] (2) Count the number of vehicles queuing on a road segment per unit time: Determine whether a certain time T within the period is within the vehicle queuing interval (t). i1 +t z , t i2 If the vehicle is in queue at time T, increment the queue count by 1 until all vehicles in the period are traversed, finally obtaining the queue count at time T. Repeat the above steps, using t as the unit of time (t is recommended to be 1 second), to calculate the real-time queue count per second of the road segment within the statistical period.

[0037] (3) Queue peak calculation (number of vehicles queuing when the green light turns on): Under normal circumstances, the number of vehicles queuing on a road segment increases continuously during the red light due to stopping and waiting, and decreases continuously during the green light period as the queue clears, reaching its peak when the green light turns on. Based on the trend of the number of vehicles queuing at various times on the road segment (number of vehicles queuing per second), the peak number of all vehicles queuing during the study period of the road segment is obtained;

[0038] Step 5: Correcting the number of vehicles in the queue

[0039] (1) Equivalent car correction: Since different vehicle types occupy different road spaces (large vehicles occupy more road space, small vehicles occupy less), it is necessary to convert the vehicle type into the equivalent car quantity by multiplying it by the equivalent car correction coefficient. The equivalent car conversion coefficients for different vehicle types are shown in Table 1 below:

[0040] Table 1. Conversion Factors for Equivalent Cars

[0041]

[0042] (2) Queue interval correction: Since the above calculations all use the vehicle passage time t at the upstream intersection, i1 Non-queueing passage time t z The sum of these values ​​is used as the starting time for the vehicle queue. However, usually there are already some vehicles initially queuing at the intersection, so the actual starting time for most vehicles to queue is earlier than t. i1+ t z To ensure the accuracy of the calculation results, the number of vehicles in the queue obtained above needs to be multiplied by an amplification factor in a manner proportional to the current number of vehicles in the queue. The specific correction factors are shown in Table 2 below:

[0043] Table 2. Queue Interval Correction Coefficient Table

[0044]

[0045] (3) Right turn correction: Since most intersections have right turn channelization, vehicles do not turn right through the intersection, making it difficult for intersection checkpoint equipment to capture right-turning vehicles. Figure 2 As shown, the checkpoint equipment only detects vehicles going straight and turning left. To ensure the accuracy of the algorithm calculation, a correction calculation is needed for right-turning vehicles. First, calculate the number of vehicles leaving the downstream intersection going straight, then calculate the number of vehicles arriving at the upstream intersection going straight and turning left, and leaving the downstream intersection going straight. The quotient of these two numbers is the straight-going queue correction coefficient. Similarly, calculate the number of vehicles leaving the downstream intersection turning left, and the number of vehicles arriving at the upstream intersection going straight and turning left, and leaving the downstream intersection turning left. The quotient of these two numbers is the left-turn queue correction coefficient.

[0046] Step 6: Calculate the average number of vehicles queuing in a single lane: Combine the peak number of vehicles queuing in each dimension after correction during the study period of this road segment with the number of lanes corresponding to the flow direction to obtain the maximum number of vehicles queuing in a single lane within the statistical period.

[0047] Calculation example:

[0048] like Figure 7 Taking the straight-ahead direction at the west entrance of intersection B during the period from 12:00:00 to 12:30:00 as an example, calculate the average number of vehicles queuing.

[0049] Given that the distance between the center points of intersections A and B is L = 800m, the maximum speed limit of the road section is V = 60km / h, and the number of straight and left-turn lanes at the intersection entrances is 1 each, the electronic police equipment required in this example is located as shown in the figure. Vehicles enter from the west entrance of intersection A, turn left from the north entrance, and turn right from the south entrance, and leave from intersection B.

[0050] Step 1: Utilize readily available and widely accessible vehicle data from checkpoints to build a system for calculating the number of vehicles queuing along the entire road segment. The number of checkpoint devices required for this example and their specific locations are as follows: Figure 7 As shown, this includes checkpoint equipment at the west entrance of intersection A, the north entrance of intersection A, and the west entrance of intersection B. The checkpoint equipment is used to obtain vehicle ID, intersection number, entrance lane direction, flow direction, lane number, license plate number, vehicle type, and passage time.

[0051] Step 2: Obtain vehicle passage data from intersections A and B respectively. Clean the data, remove duplicates, and generate upstream and downstream vehicle passage matching data, including vehicle ID, license plate number, vehicle type, passage time at upstream and downstream intersections, intersection numbers, entrance direction at upstream and downstream intersections, lane numbers at upstream and downstream intersections, and flow direction at upstream and downstream intersections. Collect data on vehicles turning left from intersection A and going straight into intersection B. Calculate the travel time for these vehicles to travel straight through intersection B. Based on the upstream and downstream intersection distance L = 800m and the maximum speed limit V = 60km / h for this road segment, calculate the minimum travel time t for vehicles without exceeding the speed limit. min =L / V=48s, remove all speeding vehicles whose travel time is less than 48s;

[0052] Step 3: Use the k-means algorithm to perform cluster analysis on the remaining vehicle transit times, taking k=5. The cluster centers and number of data points for the 5 data clusters are shown in the table below:

[0053]

[0054] Based on the cluster center values, sort the data clusters from smallest to largest, and calculate the proportion of data in the first cluster to the total number of data points:

[0055] C1=15 / (15+22+86+45+12)×100%=8.33%,

[0056] 8.33% < 10%, therefore the calculation is based on the first two data clusters:

[0057] C 1+2 = (15+22) / (15+22+86+45+12)×100% = 20.55%

[0058] 20.55% > 10%, non-queueing time t z = (52 + 56) / 2 = 54s

[0059] Step 4: Calculate the number of vehicles in the queue.

[0060] (1) Calculate the vehicle queuing time interval:

[0061] Taking three vehicles traveling straight through intersection B within a statistical period as an example, the time t for the three vehicles to pass through the upstream and downstream intersections is... i1 t i2 As shown in the table below:

[0062]

[0063] Calculate the time (t) for each vehicle to pass through. i2 -t i1 ) and non-queue passage time t z Compare:

[0064] Car No. 1: t 12 -t 11 =50s<54s; Car No. 2: t 22 -t 21 =57s>54s; Car No. 3: t 32 -t 31 =65s>54s;

[0065] Therefore, we know that car number 1 is not in the queue, while cars 2 and 3 are in the queue. The queue interval is (t). i1 +54s, t i2 ): Car No. 2 (12:04:59, 12:05:02), Car No. 3 (12:05:04, 12:05:15).

[0066] Similarly, determine whether all vehicles are in a queue within the cycle and calculate the queue interval corresponding to the queued vehicles.

[0067] (2) Count the number of vehicles queuing on a road segment per unit time: Taking 12:05:05 as an example, determine whether this time is within the vehicle queuing interval (t). i1 +t z, t i2 Within. The time t for car number 2 to finish queuing. 22 =12:05:02<12:05:05, not in a queue state, car number 3 starts queuing at time t. 31 +t z =12:05:04<12:05:05, queuing end time t 32 =12:05:15>12:05:05, therefore car number 3 is in a queue at that time.

[0068] Similarly, iterate through all queuing intervals of queuing vehicles and count the total number of queuing vehicles at that moment.

[0069] Repeat the above steps, using t as the unit of time, with t recommended to be 1 second, to calculate the number of vehicles queuing per second on the road segment during the period from 12:00:00 to 13:00:00.

[0070] (3) Queue peak calculation: Under normal circumstances, the number of vehicles queuing on a road segment increases continuously due to stopping and waiting during red lights, and decreases continuously as the queue clears during green lights, reaching its peak when the green light turns on. Based on the trend of the number of vehicles queuing per second during the 12:00:00-12:30:00 cycle of the road segment, the peak number of all vehicles queuing on the road segment can be obtained as shown in the table below;

[0071]

[0072] Step 5: Correcting the number of vehicles in the queue

[0073] (1) Equivalent car conversion: Based on the vehicle type information of each vehicle, the conversion factor table of equivalent car for different types of vehicles in Table 1 is used for correction:

[0074] Table 1. Conversion Factors for Equivalent Cars

[0075]

[0076] Peak queue size (equivalent to small cars) = Number of large passenger and freight vehicles × 2 + Number of small cars. The calculation results are shown in the table below:

[0077]

[0078] (2) Queue interval correction: The number of vehicles in the queue calculated above is corrected according to the correction coefficients in Table 2:

[0079] Table 2. Queue Interval Correction Coefficient Table

[0080]

[0081] 0 < Queue peak (equivalent car correction) < 16

[0082] Queue peak (queue interval correction) = Queue peak (equivalent car correction) × 1.1;

[0083] 15 < Queue peak (equivalent car correction) < 31

[0084] Queue peak (queue interval correction) = Queue peak (equivalent car correction) × 1.2;

[0085] The calculation results are shown in the table below:

[0086]

[0087] (3) Correction for right turn: such as Figure 7 As shown, vehicles leaving intersection B going straight include those turning right from the south entrance of intersection A, but intersection A does not have a checkpoint to detect them. To ensure the accuracy of the algorithm, a correction calculation is performed for right-turning vehicles. The number a of vehicles leaving the downstream intersection going straight is counted. s =245, the number of vehicles (b) that arrive at the upstream intersection by going straight and turning left, and leave by going straight at the downstream intersection. s =180, straight-line queuing correction coefficient q s =a s ÷b s =245÷180=1.36;

[0088] Peak queue size (right turn correction) = Peak queue size (queue interval correction) × 1.36

[0089] The calculation results are shown in the table below:

[0090]

[0091] Step 6: Calculate the average number of vehicles queuing in a single lane: Average the peak number of vehicles queuing after correction for each dimension.

[0092] H 平均 = (27+31+23+23+30+26+26+26+30+30+26+26+23+23+26)

[0093] / 15=26

[0094] Combined with the number of lanes, given that the number of straight lanes at the west entrance of intersection B is y = 1, the final calculation yields the number of vehicles queuing for straight traffic in a single lane at the west entrance of intersection B, H:

[0095] H = H 平均 / y=26 / 1=26

[0096] Compared with existing technologies, this invention addresses a series of problems arising from existing methods of measuring queue length using intersection detection equipment, such as limited detection range, the need for matching most measurement methods with traffic light status data from intersection signal controllers, and limitations imposed by the number of data acquisition devices when using floating cars or other detection equipment. It utilizes widely available and easily accessible checkpoint data, and calculates the number of vehicles queuing on a road segment by correlating this data with vehicle data from upstream and downstream intersections. This solves the problem that a single detection device can only detect the queue status of vehicles within a certain range at an intersection, avoids data loss and clock asynchrony issues caused by correlating data from multiple system devices, and particularly addresses the shortcomings caused by traffic light matching issues. This improves the simplicity and effectiveness of calculating the number of vehicles queuing on a road segment.

[0097] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for calculating the number of vehicles queuing across an entire road segment based on single data from checkpoints, characterized in that, Includes the following steps: (1) Determine the analysis section for the research period and use checkpoint equipment to obtain checkpoint vehicle data near the stop line of each direction of the intersection. The checkpoint vehicle data structure includes vehicle ID, intersection number, direction of the entrance lane, flow direction, lane number, license plate number, vehicle type, and vehicle passage time. (2) Remove duplicate data from the checkpoint vehicle data in step (1), complete the fusion calculation of vehicle data at upstream and downstream intersections for each vehicle, and generate a vehicle travel dataset including vehicle ID, license plate number, vehicle type, vehicle time at upstream and downstream intersections, intersection number, direction of entrance lane at upstream and downstream intersections, lane number at upstream and downstream intersections, and flow direction at upstream and downstream intersections. Calculate the vehicle travel time for each vehicle at upstream and downstream intersections based on the vehicle travel dataset, and remove speeding vehicles whose vehicle travel time is less than the minimum vehicle travel time. (3) The k-means algorithm is used to perform cluster analysis on the remaining vehicle passage time. The vehicle passage time in the study period is clustered into k classes, and the mean of the cluster center of the m data cluster with the smaller passage time value is selected as the non-queuing passage time of the analyzed road segment in the study period. (4) Based on the vehicle travel dataset in step (2), obtain the vehicle passage time of each vehicle, compare the vehicle passage time with the non-queuing passage time, filter out vehicles with queues, obtain the queuing time interval of the queuing vehicles in the analysis section during the study period, count the number of queuing vehicles in the analysis section per unit time, and obtain the peak number of all queuing vehicles in the analysis section during the study period based on the changing trend of the number of queuing vehicles in the analysis section at each time. (5) To address the issues of different vehicle types occupying different roads, delays in vehicle queuing time intervals, and the difficulty of intersection checkpoints in capturing right-turning vehicles, resulting in inaccurate statistics of the number of vehicles in queues, the peak number of vehicles in queues is calculated and corrected. (6) The peak number of queuing vehicles after correction of various dimensions during the research period of the analyzed road section is combined with the number of corresponding flow lanes to obtain the maximum number of queuing vehicles per lane in the statistical period. The specific process for obtaining the peak number of all queued vehicles during the study period of the analyzed road segment is as follows: (4.1) Calculate the vehicle queuing time interval: based on the time t when vehicle i passes the upstream. i1 With downstream time t i2 Calculate the vehicle passage time and compare it with the non-queue passage time t calculated in the previous step. z For comparison, if the time t for passing the vehicle is... i2 -t i1 Greater than t z If vehicles are queuing, further analysis will be conducted on the passing time t of vehicle i at the upstream intersection of the road segment. i1 Non-queueing passage time t z The sum is taken as the start time of the vehicle queue, and the passing time t at the downstream intersection of the road segment is analyzed. i2 As the end time of queuing, the queuing time interval (t) of the vehicle in the analyzed road segment during the research period is obtained. i1 +t z , t i2 ); (4.2) Analyze the number of vehicles queuing on a road segment per unit time: Determine whether a certain time T within the period is within the vehicle queuing interval (t i1 +t z , t i2 If the vehicle is in the queue at time T, the number of vehicles in the queue at that time is incremented by 1 until all vehicles in the cycle are traversed, and the number of vehicles in the queue at time T is finally obtained. Repeat the above steps, using t as the unit of statistical time, to calculate the number of vehicles queuing per second on the road segment within the statistical period. (4.3) Queue peak calculation: Under normal circumstances, the number of vehicles queuing on the analyzed road segment increases continuously due to stopping and waiting at red lights, and decreases continuously as the queue is cleared during green lights, reaching its peak at the moment the green light turns on; Based on the trend of the number of vehicles queuing at various times in the analyzed road segment, the peak value of the vehicle queuing in the analyzed road segment is obtained, so that the number of vehicles queuing can be obtained according to the trend of the number of vehicles queuing, and the peak value of all vehicles queuing in the analyzed road segment during the research period is obtained.

2. The method for calculating the number of vehicles queuing across the entire road segment based on single data from checkpoint passage, as described in claim 1, is characterized in that: In step (2), speeding vehicles with a passing time shorter than the minimum passing time are removed from the dataset, as follows: (2.1) Calculate the transit time of each vehicle through the upstream and downstream intersections based on the vehicle travel dataset; (2.2) Based on the spacing between checkpoints at upstream and downstream intersections and the maximum speed limit of the analyzed road section, the minimum time for a vehicle to pass through upstream and downstream intersections without exceeding the speed limit is obtained; (2.3) Remove all speeding vehicles whose passing time is less than the minimum passing time.

3. The method for calculating the number of vehicles queuing across the entire road segment based on single data from checkpoint passage, as described in claim 1, is characterized in that: The specific steps for correcting the peak number of queuing vehicles based on the different road occupancy situations for different vehicle types in step (5) are as follows: Since different vehicle types occupy different sections of the road, the number of equivalent cars is uniformly converted by multiplying the vehicle type by the equivalent car correction coefficient.

4. The method for calculating the number of vehicles queuing across the entire road segment based on single data from checkpoints, as described in claim 1, is characterized in that: The specific steps for correcting the peak number of vehicles in the queue for delays in the vehicle queuing time interval in step (5) are as follows: Since the calculation in step (4) uses t i1+ t z This is considered the start time for the vehicle queue, but usually there are already some vehicles initially queuing at the intersection, so the actual start time for most vehicles to queue is earlier than t. i1+ t z To ensure the accuracy of the calculation results, the peak number of queued vehicles obtained from the above calculation needs to be multiplied by an amplification factor in a manner that is proportional to the current peak number of queued vehicles.

5. The method for calculating the number of vehicles queuing across the entire road segment based on single data from checkpoint passage, as described in claim 1, is characterized in that: To address the issue of inaccurate vehicle counts in traffic queues caused by the inability of intersection checkpoint equipment to capture right-turning vehicles, a correction calculation for peak queue vehicle counts is performed, as detailed below: First, calculate the number of vehicles leaving the downstream intersection going straight. Then, calculate the number of vehicles arriving at the upstream intersection going straight or turning left, and leaving the downstream intersection going straight. The quotient of these two numbers is the straight-line queue correction coefficient. Similarly, calculate the number of vehicles leaving the downstream intersection by turning left, and the number of vehicles arriving at the upstream intersection by going straight or turning left, and leaving the downstream intersection by turning left. The quotient of these two numbers is the left-turn queue correction coefficient. The peak numbers of vehicles queuing for straight-line and left-turn queues are corrected using the straight-line queue correction coefficient and the left-turn queue correction coefficient, respectively.

Images