Traffic flow state real-time evaluation method based on ETC portal frame data

Abstract

The invention discloses a traffic flow state real-time evaluation method based on ETC portal frame data. The method comprises the following steps: step 1, acquiring and processing historical data; step 2, data cleaning; counting the flow and calculating the average vehicle speed V; fitting to obtain a speed-flow relation function and calculating a confidence interval; step 24, constructing a historical reference model library; step 3, acquiring real-time data of a road section and cleaning the data; the real-time flow and the average vehicle speed are calculated, and the congestion degree is judged according to the average vehicle speed; 4, evaluating whether the real-time data point falls within the confidence interval, and if the real-time data point falls outside the confidence interval, determining that the traffic state is abnormal; and step 5, outputting a real-time traffic flow state evaluation result. In conclusion, the vehicle passing data of the adjacent ETC door frames of the expressway are collected in real time, the average vehicle speed of the road section is calculated in combination with the fixed time interval, and the average vehicle speed is dynamically compared with the historical reference data in the same period, so that accurate evaluation of the traffic state is achieved.

Description

Technical Field

[0001] This invention belongs to the field of intelligent transportation system technology, and relates to traffic condition assessment and big data analysis technology, specifically to a method for real-time assessment of traffic flow status based on ETC gantry data. Background Technology

[0002] With the increasing traffic flow on highways and urban roads year by year, traffic congestion has become a pressing problem. ETC (Electronic Toll Collection) system, a commonly used toll collection system on highways, also has the ability to collect real-time traffic data. Effectively utilizing ETC data to identify traffic conditions provides a basis for early warning of traffic anomalies and road management, making it a way to improve highway capacity and driving safety.

[0003] Currently, most traffic condition assessment studies utilize inductive loop detectors, video sensors, or GPS to collect traffic data. For example, Chinese invention patent CN102779410A, a parallel implementation of multi-source heterogeneous massive traffic data fusion, processes data collected from inductive loop detectors and GPS to obtain average vehicle speeds, then fuses these average speeds from the two different data sources for traffic condition assessment. However, its limitations include the limited coverage of inductive loop detectors, which can only collect localized data from their installation locations, and high deployment costs. GPS relies on device penetration, requiring a sufficient number of vehicles to install it, and data acquisition is difficult. Existing research includes some traffic condition assessment studies based on ETC gantry data, but these also have certain shortcomings. For example, the Chinese invention patent "A Real-time Traffic Flow Estimation Method for Highway Sections Based on Gantry Data" (CN114596700A) collects traffic data in real time through ETC gantries and combines it with historical static information. It dynamically determines the statistical period using the average travel time and estimates the real-time traffic flow of the road section by combining the number of vehicles at the upstream gantries and the net flow at entrances and exits. Its shortcomings are that it fails to consider the impact of the possible presence of ramps or service areas on traffic volume and travel time in the studied road section, and it does not consider the impact of different weather conditions on traffic flow.

[0004] Based on the aforementioned shortcomings of existing research, there is an urgent need for a dynamic traffic status assessment method based on ETC data. This method should fully utilize the wide coverage and real-time nature of ETC gantry data, consider the impact of vehicles entering and exiting the main road and entering service areas, and take into account the impact of different weather conditions on traffic flow, so as to achieve accurate assessment of traffic flow and provide decision support for traffic management. Summary of the Invention

[0005] The purpose of this invention is to provide a real-time traffic flow status assessment method based on ETC gantry data, in order to solve the problem of low accuracy in existing traffic flow status assessment methods.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A method for real-time assessment of traffic flow status based on ETC gantry data includes the following steps:

[0008] Step 1: Historical data acquisition and processing; including the following sub-steps:

[0009] Step 11: Select a highway section between two adjacent ETC gantries. This section includes one interchange and one service area. Collect traffic flow data, speed limit values, and weather data for a consecutive week. The two adjacent ETC gantries are denoted as upstream gantry A and downstream gantry B.

[0010] Step 12: Group and store the collected data according to the days of the week. The dataset from 0:00 to 24:00 on Monday is denoted as file x1, which contains data from a Mondays. This process continues until Sunday, resulting in a total of 7 files, denoted as x1 to x7.

[0011] Step 2: Building the history repository; including the following sub-steps:

[0012] Step 21: Clean the data obtained in Step 1, specifically by removing vehicle records that enter / exit the main line of the road segment between two adjacent ETC gantries, stop at service areas, or have abnormal travel times.

[0013] Step 22: For the cleaned data obtained in Step 21, count the number of vehicles passing through downstream gantry B at fixed time intervals as the flow rate Q, and calculate the average vehicle speed between two adjacent ETC gantries.

[0014] Specifically, the calculation formula is as follows:

[0015] average speed Where d is the gantry spacing, in kilometers; Δt i Let n be the time (in hours) for vehicle i to pass through the section between gantry frames; n is the total number of vehicles in the statistical period.

[0016] Step 23: For specific days of the week, time periods, and weather conditions, collect all historical data for the same period, using traffic flow Q as the independent variable and average vehicle speed as the independent variable. Using Q as the dependent variable, the velocity-flow relationship function V = f(Q) is fitted, and the confidence interval of this function at a predetermined confidence level is calculated.

[0017] Step 24: Repeat steps 22 and 23 to store the speed-flow relationship functions and their parameters corresponding to all specific conditions, and build a historical benchmark model library;

[0018] Step 3: Real-time data acquisition and processing:

[0019] The license plate numbers, vehicle passage times, and weather conditions of two adjacent ETC gantries in the road segment are acquired in real time, and invalid data is removed according to the data cleaning method in step 21. Following the same method as in step 22, the number of vehicles passing through the downstream gantry B is calculated at fixed time intervals and recorded as the real-time flow Q. r Calculate the average vehicle speed between two adjacent ETC gantries. And according to the GA / T115—2020 Road Traffic Congestion Evaluation Method, the congestion level is determined by average vehicle speed;

[0020] Step 4: Dynamic State Assessment

[0021] Match the historical velocity-flow relationship function V=f(Q) under real-time conditions; evaluate the real-time data points obtained in step 3. Whether it falls within the confidence interval; if it falls outside the confidence interval, it is determined that the traffic condition is abnormal.

[0022] Calculate real-time average vehicle speed The relational function with respect to the real-time traffic Q r Predicted value at location The residual R is calculated using the following formula: The anomaly level is determined based on the magnitude of |R|;

[0023] Step 5: Output the real-time traffic flow status assessment results.

[0024] Compared with the prior art, the beneficial effects of the present invention are:

[0025] 1. By making full use of the real-time data of the ETC gantry system, the accuracy of traffic flow status assessment is improved. Especially in the collaborative monitoring of highways and multiple road segments, the assessment results can be more accurate and reliable. Moreover, it does not require additional manpower or additional equipment in the highway network, making it universal and easy to promote.

[0026] 2. By utilizing real historical traffic data of road sections, abnormal data caused by accidents or construction can be effectively filtered out, enabling a more scientific estimation of real-time traffic conditions. The assessment results are accurate and have strong anti-interference capabilities.

[0027] 3. This invention can determine real-time traffic conditions and abnormal situations, helping traffic management departments to respond quickly and take corresponding measures to optimize traffic flow.

[0028] 4. A speed-flow relationship model was introduced for state assessment, achieving a leap from single-parameter threshold comparison to multi-parameter relationship consistency diagnosis. It can keenly identify hidden anomalies where speed and flow appear normal individually, but their combined relationship deviates from historical patterns. Its sensitivity and accuracy in detecting events such as traffic accidents and temporary construction are far superior to traditional methods, providing a deeper level of decision-making basis for traffic management.

[0029] In summary, this invention achieves accurate assessment of traffic conditions by collecting real-time vehicle traffic data from adjacent ETC gantries on highways, calculating the average vehicle speed of the road segment at fixed time intervals, and dynamically comparing it with benchmark data under historical conditions. It fully considers the impact of vehicles entering and exiting the main line and entering service areas, as well as the impact of weather on traffic flow, thus improving the accuracy of the assessment results. Attached Figure Description

[0030] Figure 1 This is a flowchart of the traffic state dynamic estimation method based on ETC gantry data of the present invention;

[0031] Figure 2 This is a schematic diagram of the research section in this invention;

[0032] Figure 3 Data cleaning flowchart;

[0033] Figure 4 Flowchart for building a historical data benchmark.

[0034] The present invention will be further explained and described below with reference to the accompanying drawings and embodiments. Detailed Implementation

[0035] like Figure 1 As shown, the real-time traffic flow status assessment method based on ETC gantry data provided by this invention includes the following steps:

[0036] Step 1: Historical data acquisition and processing; including the following sub-steps:

[0037] Step 11: Select a highway section between two adjacent ETC gantries. This section includes one interchange and one service area. Collect traffic flow data (i.e., ETC passage records), speed limit values, and weather data for a consecutive week (excluding holidays). The two adjacent ETC gantries are denoted as upstream gantry A and downstream gantry B.

[0038] Step 12: Group and store the collected data according to the days of the week. The dataset from 0:00 to 24:00 on Monday is denoted as file x1, which contains data from a Mondays. This process continues until Sunday, resulting in a total of 7 files, denoted as x1 to x7.

[0039] Specifically, the data file includes the time point, license plate number, time the vehicle passes through gantry A, time the vehicle passes through gantry B, speed limit for the road segment, and weather conditions.

[0040] Step 2: Building the history repository; including the following sub-steps:

[0041] Step 21: Clean the data obtained in Step 1, specifically by removing vehicle records that enter / exit the main line of the road segment between two adjacent ETC gantries, stop at service areas, or have abnormal travel times.

[0042] Specifically, the data cleaning methods are as follows:

[0043] Because there is an interchange between upstream gantry A and downstream gantry B, upstream gantry A records the vehicle's passage time. If there is no record of the vehicle's passage on downstream gantry B, it means that the vehicle left the main line through the interchange after passing upstream gantry A. Therefore, the vehicle's passage data needs to be removed when calculating the average vehicle speed.

[0044] Similarly, if there is no passage record for upstream gantry A but a vehicle passage time is recorded for downstream gantry B, it means that the vehicle entered the main line from the interchange between the gantry, and the passage data is removed when calculating the average speed.

[0045] If there is a service area between upstream gantry A and downstream gantry B, and a vehicle passes through both upstream gantry A and downstream gantry B consecutively, but the passage time exceeds 1.5 times the average running time of similar vehicles, then it is determined that the vehicle is staying in the service area or has encountered other abnormalities, and the vehicle's passage data will be removed when calculating the average vehicle speed.

[0046] Step 22: For the cleaned data obtained in Step 21, count the number of vehicles passing through downstream gantry B at fixed time intervals as the flow rate Q, and calculate the average vehicle speed between two adjacent ETC gantries.

[0047] Specifically, the calculation formula is as follows:

[0048] average speed Where d (km) is the gantry spacing, Δt i (hours) represents the time it takes for vehicle i to pass through the section between gantry frames, and n represents the total number of vehicles in the statistical period.

[0049] Step 23: For specific days of the week, time periods, and weather conditions, collect all historical data for the same period, using traffic flow Q as the independent variable and average vehicle speed as the independent variable. Using Q as the dependent variable, the velocity-flow relationship function V = f(Q) is fitted, and the confidence interval of this function at a predetermined confidence level is calculated.

[0050] Specifically, the speed-flow relationship function is a linear function V = a - bQ, where a and b are parameters obtained by linear regression fitting;

[0051] Step 24: Repeat steps 22 and 23 to store the speed-flow relationship functions and their parameters corresponding to all specific conditions, and build a historical benchmark model library;

[0052] Step 3: Real-time data acquisition and processing:

[0053] The license plate numbers, vehicle passage times, and weather conditions of two adjacent ETC gantries in the road segment are acquired in real time, and invalid data is removed according to the data cleaning method in step 21. Following the same method as in step 22, the number of vehicles passing through the downstream gantry B is calculated at fixed time intervals and recorded as the real-time flow Q. r Calculate the average vehicle speed between two adjacent ETC gantries. And according to the GA / T115—2020 Road Traffic Congestion Evaluation Method, the congestion level is determined by average vehicle speed;

[0054] Step 4: Dynamic State Assessment

[0055] Match the historical speed-flow relationship function V=f(Q) under real-time conditions (day of the week, time of day, weather); evaluate the real-time data points obtained in step 3. Whether it falls within the confidence interval; if it falls outside the confidence interval, it is determined that the traffic condition is abnormal.

[0056] Calculate real-time average vehicle speed The relational function with respect to the real-time traffic Q r Predicted value at location The residual R is calculated using the following formula: The anomaly level is determined based on the magnitude of |R|;

[0057] Specifically, the deviation of the relationship is calculated based on the residual R, and a threshold range is set to classify different levels of traffic condition anomalies.

[0058] Step 5: Output the real-time traffic flow status assessment results, which include:

[0059] Real-time traffic flow, traffic status anomalies, and anomaly levels;

[0060] Average speed of real-time traffic flow

[0061] Congestion level C of real-time traffic flow;

[0062] The number N of vehicles exiting the mainline via the interchange between upstream gantry A and downstream gantry B in real-time traffic flow. A ;

[0063] The number N of vehicles entering the mainline via the interchange between upstream gantry A and downstream gantry B in real-time traffic flow. B ;

[0064] The increase or decrease N of the number of vehicles passing through upstream gantry A and downstream gantry B in real-time traffic flow compared to the historical number of vehicles passing through (obtained in step 23). C .

[0065] Example

[0066] To verify the feasibility and effectiveness of the present invention, the following section of the Chengdu-Nanchong Expressway from gantry G004251004001610 (gantry A) to gantry G004251004001710 (gantry B), with a length of 3.6km, a speed limit of 120km / h, and including one interchange, is used to implement the dynamic traffic condition assessment method of the present invention.

[0067] 1. Historical data acquisition and processing

[0068] Historical traffic flow data for four consecutive weeks was collected from the ETC toll records of the aforementioned road sections. Taking all Mondays in July 2023 (July 3, July 10, July 17, and July 24) as an example, the data from 0:00 to 24:00 on each Monday was recorded as file x1. This file contains historical data for four Mondays. The toll information contained in the files is shown in Table 1.

[0069] Table 1 contains data files containing travel information.

[0070] license plate number Time through gantry A Time through gantry B Travel time (s) Sichuan A34N9H 2023-07-03 10:26:45 2023-07-03 10:29:04 139 Sichuan A4S4R8 2023-07-10 14:49:49 2023-07-10 14:52:05 136 Sichuan AH87W0 2023-07-17 20:23:46 2023-07-17 20:25:56 130 Sichuan A3GS24 2023-07-17 23:01:43 2023-07-17 23:04:08 145 Anhui LC3398 2023-07-24 12:14:55 2023-07-24 12:17:36 161 …

[0071] 2. Historical benchmark construction

[0072] Abnormal and invalid passage data, such as those entering / leaving midway or those with excessively long passage times, are removed. The information contained in the abnormal data records is shown in Table 2.

[0073] Table 2 Information on Abnormal Access Records

[0074]

[0075] After removing invalid data, starting from 0:00, the number of vehicles passing through adjacent gantries every 5 minutes (flow rate Q) is counted, and the average travel time, average speed, and standard deviation of speed between adjacent gantries are calculated. Taking 07:55-08:10 as an example, the above information is collected for three consecutive 5-minute time intervals, as shown in Table 3:

[0076] Table 3. Calculation results for Mondays in July, 07:55-08:10

[0077]

[0078] After obtaining the above basic data, the key step of this invention is to, for specific conditions (such as "Monday, sunny"), extract the traffic flow Q and average vehicle speed from all time intervals in the historical data. Using the data points, a speed-flow relationship function is fitted. Using the data in this example, the function V = 100.21 - 0.063Q is obtained for the conditions "Monday, sunny, 07:55-08:00". The 95% confidence interval for its slope b is (-0.140, 0.014). This function represents the historical relationship model of this road segment under specific conditions.

[0079] Meanwhile, congestion levels are determined according to the GA / T 115-2020 standard (the smooth flow threshold for road sections with a speed limit of 120 km / h is 70 km / h). The GA / T 115-2020 standard is shown in Table 4. Based on this, a historical benchmark database of flow-speed relationship functions and congestion levels is constructed, as shown in Table 5.

[0080] Table 4. GA / T 115-2020 Method for Evaluating Road Traffic Congestion

[0081]

[0082] Table 5: Calculation results from historical databases for Mondays in July.

[0083]

[0084] 3. Real-time data acquisition and processing

[0085] Traffic flow data collected by two adjacent ETC gantries on the aforementioned road section on August 7, 2023, was used as the data for real-time assessment.

[0086] Taking 07:55-08:10 as an example again, after removing invalid data, the real-time parameters are calculated as shown in Table 6:

[0087] Table 6. Real-time Data Related Parameters

[0088]

[0089] 4. Dynamic Status Assessment

[0090] This invention achieves dynamic state assessment by comparing real-time data with historical relationship models. Based on the weekday, weather type, and time period of the real-time data, it matches the corresponding historical speed-flow relationship model and its confidence interval. For example, using real-time data (Monday, sunny, time period: 07:55-08:00, flow rate Q...),... r =102, average speed Let's take an example for evaluation:

[0091] (1) Model matching and prediction: retrieve the historical relation function V = 100.21 - 0.063Q and its confidence interval established by the condition "Monday, sunny" for this road segment;

[0092] (2) Residual calculation: Q r Substituting 102 into the model, we obtain the predicted vehicle speed. Calculate the residual between real-time vehicle speed and predicted vehicle speed

[0093] (3) Confidence interval determination: The real-time point (102, 93.7) is within the historical interval; 0 <T1<T2<T3

[0094] (4) State determination: Let T1 = 5 km / h, T2 = 10 km / h, T3 = 20 km / h; |R| = 0.08 km / h <T1;

[0095] If the combined residual is extremely small and falls within the historical confidence interval, the current traffic flow state is determined to be "normal". This method can effectively identify latent anomalies at the relational level that are difficult to detect with traditional single-parameter thresholding.

[0096] 5. Traffic flow state assessment parameter report output

[0097] A report was generated based on the traffic flow status assessment results above, as shown in Table 8:

[0098] Table 8 Real-time Traffic Flow Assessment Results

[0099]

[0100] The real-time traffic flow status assessment method based on ETC gantry data provided by this invention can achieve real-time and accurate assessment of highway traffic flow status. The data foundation of this invention relies on the widely deployed ETC gantry system on highways, offering advantages such as no additional equipment investment, ease of implementation, and ease of promotion. It fully considers the impact of factors such as weather conditions, interchanges, and service areas on traffic flow, establishing a historical database. Dynamic comparison with historical benchmark data under the same conditions as real-time data makes the traffic status assessment results more accurate and reliable. The output assessment parameters, such as average vehicle speed, number of vehicles leaving the mainline, speed deviation, and standard deviation ratio, can provide a scientific basis for traffic management department decision-making, improving the accuracy of traffic early warning and risk prevention.

[0101] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0102] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for real-time assessment of traffic flow status based on ETC gantry data, characterized in that, It includes the following steps: Step 1: Historical data acquisition and processing; including the following sub-steps: Step 11, Select a highway section between two adjacent ETC gantries. This section contains 1 interchange and 1 service area. Collect traffic flow data, road section speed limit values, and weather data for consecutive a weeks. The two adjacent ETC gantries are denoted as the upstream gantry A and the downstream gantry B; Step 12, Group and store the collected data by day within a week. The data set from 0:00 to 24:00 on Monday of each week is denoted as file x1. File x1 contains data for a Mondays.类推至周日共7组文件,记为x1~x7; Step 2: Historical database construction; including the following sub-steps: Step 21, Clean the data obtained in Step 1. Specifically,剔除两相邻ETC门架间中途驶入 / 驶离路段主线、停留服务区及其他通行时间异常的车辆记录; Step 22: For the cleaned data obtained in Step 21, count the number of vehicles passing through downstream gantry B at fixed time intervals as the flow rate Q, and calculate the average vehicle speed between two adjacent ETC gantries. Specifically, the calculation formula is as follows: average speed Where d is the gantry spacing, in kilometers; Δt i Let n be the time (in hours) for vehicle i to pass through the section between gantry frames; n is the total number of vehicles in the statistical period. Step 23: For specific days of the week, time periods, and weather conditions, collect all historical data for the same period, using traffic flow Q as the independent variable and average vehicle speed as the independent variable. Using Q as the dependent variable, the velocity-flow relationship function V = f(Q) is fitted, and the confidence interval of this function at a predetermined confidence level is calculated. Step 24, Repeat Step 22 and Step 23, Store the speed-flow relationship functions and their parameters corresponding to all specific conditions, and construct a historical benchmark model library; Step 3: Real-time data acquisition and processing: The license plate numbers, vehicle passage times, and weather conditions of two adjacent ETC gantries in the road segment are acquired in real time, and invalid data is removed according to the data cleaning method in step 21. Following the same method as in step 22, the number of vehicles passing through the downstream gantry B is calculated at fixed time intervals and recorded as the real-time traffic flow Q. r Calculate the average vehicle speed between two adjacent ETC gantries. And according to the GA / T 115—2020 Road Traffic Congestion Evaluation Method, the congestion level is determined by the average vehicle speed; Step 4: Dynamic state assessment: Match the historical velocity-flow relationship function V=f(Q) under real-time conditions; evaluate the real-time data points obtained in step 3. Whether it falls within the confidence interval; if it falls outside the confidence interval, it is determined that the traffic condition is abnormal. Calculate real-time average vehicle speed The relational function with respect to the real-time traffic Q r Predicted value at location The residual R is calculated using the following formula: The anomaly level is determined based on the magnitude of |R|; Step 5: Output the real-time traffic flow state assessment result.

2. The method according to claim 1, characterized in that, In Step 1, the data file includes the time point, license plate number, the time when the vehicle passes gantry A, the time when the vehicle passes gantry B, the road section speed limit value, and the weather condition.

3. The method according to claim 1, characterized in that, In Step 2, the method of data cleaning is: Since there is an interchange between the upstream gantry A and the downstream gantry B, the upstream gantry A records the vehicle passing time. If there is no passing record of this vehicle at the downstream gantry B, it means that the vehicle has left the main line through the interchange after passing the upstream gantry A. Then, when calculating the average vehicle speed, the passing data of this vehicle needs to be剔除; Similarly, if there is no passing record at the upstream gantry A and the downstream gantry B records the vehicle passing time, it means that the vehicle enters the main line from the interchange between the gantries. When calculating the average vehicle speed,剔除该通行数据; If there is a service area between the upstream gantry A and the downstream gantry B, and the vehicle continuously passes the upstream gantry A and the downstream gantry B, but the passing time exceeds 1.5 times the average running time of similar vehicles, it is determined that the vehicle stays in the service area or has other abnormalities. When calculating the average vehicle speed,剔除该车辆通行数据.

4. The method according to claim 1, characterized in that, In Step 23, the speed-flow relationship function is a linear function V = a - bQ, where a and b are parameters obtained by linear regression fitting.

5. The method according to claim 1, characterized in that, In Step 4, the specific method for determining the abnormal level according to the magnitude of |R| is: When |R| < R1, it is considered that the deviation degree of the relationship is acceptable, and the state is determined to be normal; when R1 ≤ |R| < R2, it is determined as level I abnormality, that is, mild abnormality; when R2 ≤ |R| < T3, it is determined as level II abnormality, that is, moderate abnormality; when |R| ≥ T3, it is determined as level III abnormality, that is, severe abnormality; where, T1, T2, T3 are positive thresholds preset according to the historical data distribution characteristics and business requirements, satisfying 0 < T1 < T2 < T3.

6. The method according to claim 1, characterized in that, In Step 5, the content of the real-time traffic flow state assessment result includes: Abnormal conditions and abnormal levels of the real-time traffic flow; Average speed of real-time traffic flow The congestion degree C of the real-time traffic flow; The number of vehicles N passing from the main line to the interchange ramp between the upstream gantry A and the downstream gantry B in real-time traffic flow A ; Number of vehicles N in real-time traffic flow passing from the interchange ramp to the main line between the upstream gantry A and the downstream gantry B B ; The real-time traffic flow shows the increase or decrease N of the number of vehicles passing through upstream gantry A and downstream gantry B compared to the historical number of vehicles passing through. C .

Images