Road interruption monitoring and grading early warning system and method based on ETC data

Abstract

The invention discloses a road interruption monitoring and grading early warning system and method based on ETC data. The method comprises the following steps: acquiring transaction data, snapshot data and video data of an ETC door frame of a vehicle in real time to form ETC door frame data; the collected ETC door frame data are preprocessed, repeated data are cleaned and removed, and error data are corrected; on the basis of the ETC portal historical information, the shunting probability of the vehicle shunting from the ETC portal to the next portal is predicted; the downstream portal flow is predicted according to the section passing time and the shunting probability; calculating the flow change rate of the downstream ETC portal in combination with the historical traffic flow of the downstream ETC portal so as to judge whether the flow of the portal drops suddenly or not; and performing flow interruption grading alarm and emergency response according to the flow change rate. According to the invention, accurate detection, graded alarm and emergency response of traffic interruption of the expressway are realized, and the operation efficiency and safety of the expressway can be effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of expressways, and particularly to a road interruption monitoring and grading warning system and method based on ETC data. Background Art

[0002] Due to the characteristics of expressways such as being closed, having a long distance between exits, high vehicle speeds, and large traffic volumes, once an accident occurs, the consequences are often more serious. In recent years, the problem of road interruptions caused by traffic accidents and natural disasters has become increasingly severe. These accidents not only threaten the safety of public travel but also cause serious economic losses. Therefore, it is of great significance to detect expressway accidents in a timely manner and formulate effective emergency rescue strategies to ensure road traffic safety and reduce accident losses.

[0003] At present, there are problems of untimely, incomplete, and non-closed reporting of expressway road interruption detection information. On the one hand, ETC data has a lag and cannot reflect real-time road condition changes, making it difficult to meet the requirements of rapid response; on the other hand, there is a black box problem within the ETC section, and the road state is judged only based on the portal traffic monitoring data. When the traffic flow returns to zero, it is easily misjudged as a blocking event, and the actual road condition cannot be accurately reflected. These problems seriously affect the efficiency and accuracy of expressway emergency management, and it is urgent to optimize the monitoring and reporting process through technical means to improve the integrity and closed-loop nature of information processing. Summary of the Invention

[0004] The purpose of the present invention is to provide a road interruption monitoring and grading warning system and method based on ETC data. Flow drop detection based on the flow change rate has important practical significance for improving the driving safety of expressways.

[0005] The technical solution adopted by the present invention is as follows:

[0006] A road interruption monitoring and grading warning method based on ETC data, which includes the following steps:

[0007] Step 1, collect transaction data, capture data, and video data of ETC portals of vehicles in real time to form ETC portal data;

[0008] Step 2, preprocess the collected ETC portal data, clean and remove duplicate data, and correct incorrect data;

[0009] Step 3, predict the diversion probability of vehicles from an ETC portal to the next portal based on historical ETC portal information;

[0010] Step 4, predict the downstream portal traffic based on the section passing time and the diversion probability;

[0011] Step 5: Calculate the flow rate change rate of the downstream ETC gantry in combination with the historical traffic volume of the downstream ETC gantry, compare it with the set sudden flow rate drop threshold, determine whether the flow rate of the gantry drops suddenly, and issue a warning when the flow rate drops suddenly;

[0012] Step 6: Perform traffic interruption classification warning and emergency response according to the flow rate change rate.

[0013] Furthermore, in Step 1, the transaction data includes information such as the passing time, license plate number, and vehicle type of the vehicle. The captured data records the driving trajectory and license plate recognition information of the vehicle, and the video data is used to assist in verifying the driving state of the vehicle and traffic events.

[0014] Furthermore, in Step 1, when collecting ETC gantry data, data is collected in combination with the RSU, radar speedometer, millimeter-wave radar, and vision sensor fusion system to enrich the data dimension and provide more comprehensive support for subsequent analysis.

[0015] Furthermore, in Step 2, the radar speedometer and millimeter-wave radar are used to collect the speed and position data of the vehicle to fill in the vehicle information not captured by the ETC gantry; for the vehicle data in the area not covered by the RSU, interpolation methods are used for filling to ensure the continuity of the data; the time registration algorithm is used to synchronize and align the time of the captured data and ETC transaction data to solve the problem of inconsistent timestamps of multi-source data.

[0016] Furthermore, Step 3 specifically includes the following steps:

[0017] Step 3-1: Obtain the traffic volume data of the ETC gantry within the same time slice in the past one-year period, and at the same time obtain the passing time data of different vehicle types within each time slice;

[0018] Step 3-2: Classify and organize the traffic volume data and passing time data by date, time slice, and vehicle type to form a structured data set;

[0019] Step 3-3: Statistically analyze the diverted traffic volume of the ETC gantry at the same time slice on the corresponding date of the previous year;

[0020] Step 3-4: Screen all vehicle passing records at the same time slice on the corresponding date of the previous year, and calculate the diversion probability from the ETC gantry to the next gantry, and use it as the current diversion probability prediction value.

[0021] Furthermore, the calculation formula for the diversion probability from the ETC gantry to the next gantry is as follows:

[0022]

[0023] where P div (t,M i ,Mj ) represents the diversion probability from the M i -th gantry to the M j -th gantry at time t, C div (t, M i , M j ) represents the traffic flow from the M i -th gantry to the M j -th gantry, and C total (t, M i ) represents the total traffic flow of the M i -th gantry at time t;

[0024] Furthermore, step 4 specifically includes the following steps:

[0025] Step 4-1: Screen out the vehicle passing records of different vehicle models within the same time slice on the same day in the past year, and predict the predicted value of the average passing time of each vehicle model on the same day. The calculation formula for the average passing time is:

[0026]

[0027] where T avg (t, i, v) represents the average passing time of vehicle model v in the i-th passing section at time t, and T n (t, i, v) represents the passing time of vehicle model v in the n-th transaction record, and N v is the total number of transaction records of vehicle model v;

[0028] Step 4-2: Use the differential algorithm to calculate the traffic flow per minute of the current gantry in the target section. The formula is:

[0029]

[0030] where C in (t, i, v) represents the cumulative traffic flow of vehicle model v at the current gantry in the i-th section when the time is t;

[0031] Step 4-3: Predict the vehicle passing time of the downstream gantry in the target section according to the transaction time of the current gantry and the average passing time of the section:

[0032] t exit (t, j, v) = t + T avg (i, v);

[0033] where t exit (t, i, v) represents the predicted time for the vehicle of vehicle model v entering the i-th section at time t to pass through the downstream gantry;

[0034] Step 4-4, diversion probability and current gantry flow per minute, predict the downstream gantry flow of the target section:

[0035] F out (t exit (t,i,v),M j ,v)=F in (t,M i ,v)×P div (t exit (t,i,v),M i ,M j ,v);

[0036] Step 4-5, based on the time slice of the downstream gantry transaction time, calculate the final downstream gantry predicted flow:

[0037] F pred (t,M j )=∑ i ∑ v F out ((t exit (t,i,v),M j ,v));

[0038] Among them, F pred (t,M j ) indicates that at time t, the downstream gantry M of the jth section j Predicted flow, i is all possible current gantry sections.

[0039] Furthermore, step 5 specifically includes the following steps:

[0040] Step 5-1: Count the historical traffic volume F of the ETC gantry at time slice t on the same day last year actual (t,M j );

[0041] Step 5-2: predict the flow data F based on the downstream gantry pred (t,M j ) and historical traffic flow F actual (t,M j ) Calculate the flow rate change rate, the formula is:

[0042]

[0043] Step 5-3, calculate the traffic flow change rate in two adjacent time slices (t, t-1) of the ETC gantry, the formula is:

[0044]

[0045] Step 5-4: Count the traffic change rate R of all time slices histand R adj The distribution of and calculate the cumulative distribution function; select a threshold k1 and k2 for sudden drops for two flow change rates respectively according to the distribution of the flow change rate; and then calculate the comprehensive weight coefficient k, and the formula for the comprehensive weight coefficient is:

[0046] k = α×k1+(1 - 0α)×k2;

[0047] where α is the weight coefficient of the flow change rate R hist ;

[0048] Step 5-4, count the distribution of the flow change rate R hist and R adj and calculate the cumulative distribution function; select a threshold k1 and k2 for sudden drops for two flow change rates respectively according to the distribution of the flow change rate; and then calculate the comprehensive weight coefficient k, and the formula for the comprehensive weight coefficient is:

[0049] k = α×k1+(1 - α)×k2;

[0050] where α is the weight coefficient of the flow change rate R hist ;

[0051] Step 5-5, conduct multi-parameter monitoring of sudden drops in traffic through traffic sudden drop threshold monitoring and ETC gantry heartbeat monitoring; monitor the traffic change rate in real time, and when the traffic change rate is lower than the comprehensive threshold k, trigger a traffic sudden drop alarm;

[0052] Furthermore, in Step 5-5, conduct heartbeat monitoring on each ETC gantry to ensure the normal operation of the gantry; when the gantry goes offline, immediately issue an alarm and mark the traffic data of the corresponding gantry as unavailable to avoid misjudging sudden drops in traffic.

[0053] Furthermore, Step 6 specifically includes the following steps:

[0054] Step 6-1, screen out the data with a flow change rate lower than the threshold based on the flow change rate of historical data and the flow change rate of adjacent time slices:

[0055]

[0056] where φ is the indicator function, taking 1 when the condition is satisfied and 0 otherwise;

[0057] Step 6-2, screen out the data with the ratio of actual traffic to predicted traffic less than k as b, and at the same time count the data c with the ratio of actual traffic to predicted traffic being 0, and the total number of involved sections is N:

[0058]

[0059] Step 6-3, when the number of filtered b data is greater than 0, a general alarm is output, including the involved section and the sudden drop in traffic data b;

[0060] Step 6-4, when the number of filtered c data is greater than 0, it indicates that there is a complete interruption of traffic. Traffic detection is performed on subsequent consecutive time slices; for each section i, check whether the actual traffic L in the next three consecutive time slices is all 0: If so, a severe alarm message is output, including the involved area and the time slice information of continuous zero traffic; otherwise, if the actual traffic in three consecutive time slices is not 0, that is, there is traffic generated in a time slice, continue to monitor the traffic in subsequent time slices and execute Step 6-5;

[0061] Among them, the judgment expression is: L = φ(F actual (t + 1, i) = 0 ∧ F factual (t + 2, i) = 0 ∧ F factual (t + 3, i) = 0);

[0062] Step 6-5, check whether the predicted traffic FL in the next three time slices is not all 0; if so, output a general alarm message; otherwise, when the traffic in the next three time slices is all 0, judge that a traffic interruption event occurs and output a severe alarm message;

[0063] Among them, the judgment expression is: FL = φ(F pred (t + 1, i) ≠ 0 ∧ F pred (t + 2, i) ≠ 0 ∧ F pred (t + 3, i) ≠ 0).

[0064] Furthermore, in Step 6, for general alarms, immediately verify the alarm information through the monitoring system, confirm the specific location and scope of the sudden drop in traffic, and contact on-site patrol personnel or traffic police to obtain the actual on-site situation; for severe alarms, immediately activate the severe alarm emergency plan, establish an emergency command team, and coordinate and handle it uniformly.

[0065] The road interruption monitoring and grading early warning system based on ETC data includes:

[0066] A data acquisition module, including an ETC gantry hardware unit for collecting vehicle data and an interface unit for accessing multi-source data;

[0067] A section traffic prediction module, including a unit for processing historical data, a unit for predicting section traffic, and a unit for outputting prediction results;

[0068] The described traffic calculation module, including a unit for calculating real-time traffic, a unit for detecting traffic interruption, and a unit for generating alarm information;

[0069] The described information transmission module includes a unit for internal system communication, a unit for communicating with external systems, and a unit for monitoring communication status.

[0070] With the above technical solutions, the present invention realizes precise detection, hierarchical warning, and emergency response of highway traffic flow interruption through the fusion of ETC gantry data and multi-source perception data, which can effectively improve the operation efficiency and safety of highways and provide scientific decision-making support for traffic management departments. The present invention can improve the real-time performance and accuracy of highway traffic flow interruption detection, optimize the information reporting process at the same time, realize rapid response and closed-loop management of traffic blockage events, thus effectively ensuring road traffic safety and reducing accident losses. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] The following further elaborates on the present invention in detail in conjunction with the drawings and specific embodiments;

[0072] Figure 1 It is a schematic flow chart of the method for road interruption monitoring and hierarchical warning based on ETC data of the present invention;

[0073] Figure 2 It is a schematic flow chart of the hierarchical warning for traffic flow interruption of the present invention;

[0074] Figure 3 It is a schematic flow chart of the monitoring of sudden drop in gantry traffic flow of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0075] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0076] As Figures 1 to 3 shown in one of them, the present invention discloses a road interruption monitoring and hierarchical warning system based on ETC data, which includes the following steps:

[0077] Step 1, data collection and multi-source data supplementation: Real-time collection of transaction data, capture data, and video data of the ETC gantries of vehicles to form ETC gantry data;

[0078] Step 2, data preprocessing: Preprocess the collected ETC gantry data, clean and remove duplicate data, and correct incorrect data;

[0079] Step 3, predicting the vehicle diversion probability: Predict the diversion probability of vehicles from an ETC gantry to the next gantry based on the historical information of the ETC gantry;

[0080] Step 4, predicting the traffic flow of the downstream gantry: Predict the traffic flow of the downstream gantry according to the section passing time and the diversion probability;

[0081] Step 5, Multi-parameter and Multi-time Dimension Traffic Sudden Drop Monitoring: Based on the predicted downstream gantry traffic and the historical traffic volume of the downstream ETC gantry, calculate the traffic change rates of the downstream ETC gantry for the current time slice and the adjacent time slices respectively, and compare them with the set traffic sudden drop threshold to determine whether the gantry has a traffic sudden drop;

[0082] Step 6, Traffic Interruption Classification Alarm and Emergency Response: Conduct traffic interruption classification alarm and emergency response according to the traffic change rate.

[0083] Furthermore, in Step 1, the transaction data includes information such as the passing time, license plate number, and vehicle type of the vehicle. The captured data records the driving trajectory and license plate recognition information of the vehicle, and the video data is used to assist in verifying the vehicle driving status and traffic events.

[0084] Furthermore, in Step 1, when collecting ETC gantry data, combine the RSU, radar speedometer, millimeter wave radar, and vision sensor fusion system to collect data, enrich the data dimension, and provide more comprehensive support for subsequent analysis.

[0085] Specifically, in Step 1, at key sections of the highway, the ETC gantry system is used to collect transaction data such as the passing time, license plate number, and vehicle type of the vehicle in real time, as well as the captured data of the vehicle driving trajectory and license plate recognition and the video data for assisting in verifying traffic events. At the same time, in areas with large traffic volume and complex road conditions, combine the RSU, radar speedometer, millimeter wave radar, and vision sensor fusion system to collect data such as vehicle speed and position, enrich the data dimension, and provide more comprehensive support for traffic monitoring.

[0086] Furthermore, in Step 2, use the radar speedometer and millimeter wave radar to collect the speed and position data of the vehicle to fill in the vehicle information not captured by the ETC gantry; for the vehicle data in areas not covered by the RSU, fill it in by interpolation method to ensure the continuity of the data; adopt the time registration algorithm to synchronize and align the time of the captured data and the ETC transaction data to solve the problem of inconsistent timestamps of multi-source data.

[0087] Specifically, in Step 2, clean the captured data of the ETC gantry, remove duplicate data, and correct incorrect data; use the speed and position data of the vehicle collected by the radar speedometer and millimeter wave radar to fill in the vehicle information not captured by the ETC gantry; for the vehicle data in areas not covered by the RSU, fill it in by interpolation method to ensure the continuity of the data; adopt the time registration algorithm to synchronize the time of the captured data and the ETC transaction data to solve the problem of inconsistent timestamps of multi-source data.

[0088] Furthermore, Step 3 specifically includes the following steps:

[0089] Step 3-1: Obtain the traffic flow data of the ETC gantry within the same time slice in the past one-year period, and at the same time obtain the passing time data of different vehicle types within each time slice;

[0090] Step 3-2: Classify and organize the traffic flow data and passing time data by date, time slice, and vehicle type to form a structured data set;

[0091] Step 3-3: Statistically calculate the diversion flow of the ETC gantry within the same time slice on the corresponding date of the previous year;

[0092] Step 3-4: Screen all vehicle passing records within the same time slice on the corresponding date of the previous year, and calculate the diversion probability from the ETC gantry to the next gantry, which is used as the current diversion probability prediction value.

[0093] Furthermore, the calculation formula for the diversion probability from the ETC gantry to the next gantry in Step 3-4 is as follows:

[0094]

[0095] Among them, P div (t,M i ,M j ) represents the diversion probability from the M i th gantry to the M j th gantry at time t, C div (t,M i ,M j ) represents the traffic flow from the M i th gantry to the M j th gantry, and C total (t,M i ) represents the total traffic flow of the M i th gantry at time t.

[0096] Specifically, Step 3 collects the traffic flow data of the ETC gantry within the same time slice in the past one year, and at the same time collects the passing time data of different vehicle types within each time slice; classifies and organizes the data by date, time slice, and vehicle type to form a structured data set; for the ETC gantry, statistically calculates the diversion flow within the same time slice on the corresponding date of the previous year; screens all vehicle passing records within the same time slice on the corresponding date of last year, and calculates the traffic flow from this gantry to the next gantry to obtain the diversion probability; uses the diversion probability within the same time slice on the corresponding date of the previous year as the diversion probability prediction value for the current day.

[0097] Furthermore, Step 4 specifically includes the following steps:

[0098] Step 4-1: Screen out the vehicle passing records of different vehicle models within the same time slice at the same time on the same day in the past year, and predict the predicted value of the average passing time of each vehicle model on that day. The calculation formula for the average passing time is as follows:

[0099]

[0100] Among them, T avg (t, i, v) represents the average passing time of vehicle model v in the i-th passing section at time t, and T n (t, i, v) represents the passing time of vehicle model v in the n-th transaction record, and N v is the total number of transaction records of vehicle model v;

[0101] Step 4-2: Use the differential algorithm to calculate the traffic flow per minute of the current gantry in the target section. The formula is:

[0102]

[0103] Among them, C in (t, i, v) represents the cumulative traffic volume of vehicle model v of the current gantry in the i-th section when the time is t;

[0104] Step 4-3: Predict the vehicle passing time of the downstream gantry in the target section according to the transaction time of the current gantry and the average passing time of the section:

[0105] t exit (t, j, v) = t + T avg (i, v);

[0106] Among them, t exit (t, i, v) represents the estimated time for the vehicle of vehicle model v entering the i-th section at time t to pass through the downstream gantry;

[0107] Step 4-4: Predict the traffic flow of the downstream gantry based on the shunt probability and the traffic flow per minute of the current gantry:

[0108] F out (t exit (t, i, v), M j , v) = F in (t, M i , v) × P div (t exit (t, i, v), M i , M j , v);

[0109] Step 4-5: Calculate the final predicted traffic flow of the downstream gantry based on the time slice where the transaction time of the downstream gantry is located:

[0110] F pred(t,M j )=∑ i ∑ v F out ((t exit (t,i,v),M j ,v));

[0111] Among them, F pred (t,M j ) indicates that at time t, the downstream gantry M of the jth section j Predicted flow, i is all possible current gantry sections.

[0112] Specifically, in step 4, the vehicle traffic records of different models in the same time slice of the same day last year are screened out, and the average travel time of each model is calculated; the average travel time in the same time slice of the same day last year is used as the average travel time prediction value of the day; the per-minute flow rate of the current gantry in the target section is calculated using a differential algorithm; the estimated passing time of the downstream gantry vehicles in the target section is calculated based on the current gantry transaction time and the average travel time of the section; the downstream gantry flow rate of the target section is predicted based on the diversion probability and the current gantry flow rate described in step 3; the final downstream gantry predicted flow rate is calculated based on the time slice where the downstream gantry transaction time is located; the downstream gantry predicted flow rates of all sections are accumulated to obtain the total predicted flow rate of the section.

[0113] Furthermore, step 5 specifically includes the following steps:

[0114] Step 5-1: Count the historical traffic volume F of the ETC gantry at time slice t on the same day last year actual (t,M j );

[0115] Step 5-2: predict the flow data F based on the downstream gantry pred (t,M j ) and historical traffic flow F actual (t,M j ) Calculate the flow rate change rate, the formula is:

[0116]

[0117] Step 5-3, calculate the traffic flow change rate in two adjacent time slices (t, t-1) of the ETC gantry, the formula is:

[0118]

[0119] Step 5-4: Count the traffic change rate R of all time slices hist and R adjThe distribution is obtained, and the cumulative distribution function is calculated. Based on the distribution of the traffic change rate, a k1 and a k2 are selected as the thresholds for measuring whether the above two traffic change rates show a sharp drop; according to the historical data and the importance of adjacent time slices, a weight coefficient α is selected, and combined with k1 and k2, the comprehensive weight coefficient k is calculated. The formula is:

[0120] k = α × k1 + (1 - α) × k2;

[0121] Step 5-5, multi-parameter monitoring of traffic drops is performed through traffic drop threshold monitoring and ETC gantry heartbeat monitoring; the traffic change rate is monitored in real time. When the traffic change rate is lower than the comprehensive threshold k, a traffic drop warning is triggered;

[0122] Furthermore, in step 5-5, the heartbeat of each ETC gantry is monitored to ensure the normal operation of the gantry; when the gantry goes offline, an alarm is immediately issued, and the traffic data of the corresponding gantry is marked as unavailable to avoid misjudging traffic drops.

[0123] Specifically, in step 5, the historical traffic volume of the ETC gantry at the same time slice on the same day last year is counted; the traffic change rate is calculated based on the predicted traffic volume data of the downstream gantry and the historical traffic volume in step 4; for the ETC gantry, the traffic change rate within two adjacent time slices is calculated; the distribution of the traffic change rates of all time slices is counted, and the cumulative distribution function is calculated. Thresholds are selected based on the distribution of the traffic change rate; according to the historical data and the importance of adjacent time slices, a weight coefficient is selected, and combined with the threshold, the comprehensive weight coefficient is calculated; multi-parameter monitoring of traffic drops is performed through traffic drop threshold monitoring and ETC gantry heartbeat monitoring; the traffic change rate is monitored in real time. If it is lower than the comprehensive threshold, a traffic drop warning is triggered; the heartbeat of each ETC gantry is monitored to ensure the normal operation of the gantry. If the gantry goes offline, an alarm is immediately issued, and the traffic data of the gantry is marked as unavailable to avoid misjudging traffic drops.

[0124] Furthermore, step 6 specifically includes the following steps:

[0125] Step 6-1, data with a traffic change rate lower than the threshold is screened out based on the traffic change rate of historical data and the traffic change rate of adjacent time slices:

[0126]

[0127] where φ is an indicator function, taking 1 when the condition is satisfied and 0 otherwise;

[0128] Step 6-2, data with a ratio of actual traffic to predicted traffic less than k is screened as b, and at the same time, the number of data c with a ratio of actual traffic to predicted traffic of 0 is counted. The total number of involved sections is N:

[0129]

[0130] Step 6-3, when the number of filtered b data is greater than 0, output a general alarm, including the involved section and the sudden drop in traffic data b;

[0131] Step 6-4, when the number of filtered c data is greater than 0, it indicates that there is a complete interruption of traffic. Perform traffic detection on subsequent consecutive time slices; for each section i, check whether the actual traffic L in the next three consecutive time slices is 0: If so, output a severe alarm message, including the involved area and the time slice information of consecutive zero traffic; otherwise, if it does not meet the condition that the actual traffic in three consecutive time slices is 0, that is, there is traffic generated in the time slice, continue to monitor the traffic in subsequent time slices and execute Step 6-5;

[0132] Among them, the judgment expression is: L = φ(F actual (t + 1, i) = 0 ∧ F factual (t + 2, i) = 0 ∧ F factual (t + 3, i) = 0);

[0133] Step 6-5, check whether the predicted traffic FL in the next three time slices is not 0; If so, output a general alarm message; otherwise, when the traffic in the next three time slices is all 0, it is determined that a traffic interruption event occurs, and a severe alarm message is output;

[0134] Among them, the judgment expression is: FL = φ(F pred (t + 1, i) ≠ 0 ∧ F pred (t + 2, i) ≠ 0 ∧ F pred (t + 3, i) ≠ 0).

[0135] Furthermore, in Step 6, for general alarms, immediately verify the alarm information through the monitoring system, confirm the specific location and scope of the sudden drop in traffic, and contact on-site patrol personnel or traffic police to obtain the actual on-site situation; for severe alarms, immediately activate the severe alarm emergency plan, establish an emergency command team, and coordinate and handle it uniformly.

[0136] Specifically, in step 6, data with a traffic flow change rate lower than the threshold is screened out based on the traffic flow change rate of historical data and the traffic flow change rate of adjacent time slices; data with a ratio of actual to predicted traffic flow less than the threshold is screened out, and at the same time, the total number of sections where the ratio of actual to predicted traffic flow is 0 is counted; the screened data is analyzed. If the traffic flow change rate is lower than the threshold, a general alarm is output, including the sections involved and the traffic sudden drop data; if the traffic is completely interrupted, traffic detection is performed on subsequent consecutive time slices; for each section, it is checked whether the actual traffic flow in the next three consecutive time slices is 0, and whether the predicted traffic flow in the subsequent three time slices is not 0; if the conditions are met, a severe alarm message is output, including the area involved and the time slice information of consecutive zero traffic; for general alarms, the alarm information is verified through the monitoring system to confirm the specific location and scope of the traffic sudden drop, and the on-site patrol personnel or traffic police are contacted to obtain the actual on-site situation; for severe alarms, the severe alarm emergency plan is immediately activated, an emergency command group is established, and unified coordination and disposal are carried out.

[0137] The road interruption monitoring and grading warning system based on ETC data includes:

[0138] The data acquisition module includes an ETC gantry hardware unit for collecting vehicle data and an interface unit for accessing multi-source data;

[0139] The section traffic prediction module includes a unit for processing historical data, a unit for predicting section traffic, and a unit for outputting prediction results;

[0140] The described traffic calculation module includes a unit for calculating real-time traffic, a unit for detecting traffic interruption, and a unit for generating alarm information;

[0141] The information transmission module includes a unit for internal system communication, a unit for communicating with external systems, and a unit for monitoring communication status.

[0142] Specifically, the ETC gantry hardware unit real-time collects vehicle transaction data, capture data, and video data. At the same time, radar speedometer data is accessed through the multi-source data interface unit to supplement vehicle speed information and enrich the data dimension.

[0143] The historical data processing unit cleans and statistically analyzes the historical traffic flow data, calculates the average passing time of each section. Based on historical data and the probability of vehicle flow diversion, the downstream gantry traffic flow of this section within the next t minutes is predicted, and the prediction result output unit formats and stores the prediction result for subsequent module calls.

[0144] The real-time traffic calculation unit calculates the traffic change rate at the current moment. The traffic interruption detection unit determines whether the traffic change rate is lower than the threshold k. If it is lower than the threshold, a general alarm is triggered. The alarm information generation unit generates general alarm information, including the section number involved and the traffic sudden drop data.

[0145] The system continues to monitor the traffic data of this section. If the actual traffic is 0 in the next three consecutive time slices and the predicted traffic in the subsequent three time slices is not 0, a serious alarm is triggered. The alarm information generation unit generates serious alarm information, including the section number involved, the time slice information of consecutive zero traffic, and the predicted traffic data.

[0146] The present invention adopts the above technical solutions. Through the fusion of ETC gantry data and multi-source perception data, it realizes the accurate detection, hierarchical alarm, and emergency response of highway traffic interruption, can effectively improve the operation efficiency and safety of highways, and provides scientific decision-making support for traffic management departments. The present invention can improve the real-time performance and accuracy of highway traffic interruption detection, optimize the information reporting process at the same time, realize the rapid response and closed-loop management of traffic blockage events, thus effectively ensuring road traffic safety and reducing accident losses.

[0147] Obviously, the described embodiments are part of the embodiments of the present application, rather than all embodiments. Without conflict, the embodiments in the present application and the features in the embodiments can be combined with each other. Usually, the components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the detailed description of the embodiments of the present application is not intended to limit the scope of the present application claimed, but merely represents the selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

Claims

1. A method for monitoring road interruptions and hierarchical early warning based on ETC data, characterized in that: It includes the following steps: Step 1, collect the transaction data, capture data, and video data of the ETC gantry of the vehicle in real time to form ETC gantry data; Step 2, preprocess the collected ETC gantry data, clean and remove duplicate data, and correct error data; Step 3, predict the diversion probability of the vehicle from the ETC gantry to the next gantry based on the historical information of the ETC gantry; Step 4, predict the downstream gantry traffic based on the section passing time and diversion probability; Step 5, calculate the traffic change rate of the downstream ETC gantry in combination with the historical traffic volume of the downstream ETC gantry, compare it with the set traffic sudden drop threshold, judge whether the gantry has a sudden drop in traffic, and issue a warning when the traffic suddenly drops; Step 6, perform traffic interruption classification warning and emergency response according to the traffic change rate.

2. The method for monitoring road interruption and grading early warning based on ETC data according to claim 1, characterized in that: In Step 1, the transaction data includes the passing time, license plate number, and vehicle type information of the vehicle. The capture data records the driving trajectory and license plate recognition information of the vehicle, and the video data is used to assist in verifying the driving state of the vehicle and traffic events.

3. The method for monitoring road interruptions and hierarchical early warning based on ETC data according to claim 1, characterized in that: When collecting ETC gantry data, combine the RSU, radar speedometer, millimeter wave radar, and vision sensor fusion system to collect data; use the radar speedometer and millimeter wave radar to collect the speed and position data of the vehicle to fill in the vehicle information not captured by the ETC gantry; for the vehicle data in the area not covered by the RSU, use the interpolation method to fill it to ensure the continuity of the data; adopt the time registration algorithm to synchronize and align the time of the capture data and ETC transaction data.

4. The method for monitoring road interruption and grading early warning based on ETC data according to claim 1, characterized in that: Step 3 specifically includes the following steps: Step 3-1, obtain the traffic volume data of the ETC gantry within the same time slice in the past one-year period, and at the same time obtain the passing time data of different vehicle types within each time slice; Step 3-2, classify and organize the traffic volume data and passing time data by date, time slice, and vehicle type to form a structured data set; Step 3-3, count the diversion traffic volume of the ETC gantry within the same time slice on the corresponding date of the previous year; Step 3-4, screen all vehicle passing records within the same time slice on the corresponding date of the previous year, and calculate the diversion probability from the ETC gantry to the next gantry, and use it as the current diversion probability prediction value; the calculation formula of the diversion probability is as follows: Among them, P div (t, M i , M j ) represents the diversion probability from the M i -th gantry to the M j -th gantry at time t. C div (t, M i , M j ) represents the traffic flow from the M i -th gantry to the M j -th gantry. C total (t, M i ) represents the total traffic flow of the M i -th gantry at time t.

5. The method for road interruption monitoring and grading early warning based on ETC data according to claim 1, wherein: Step 4 specifically includes the following steps: Step 4-1, screen the vehicle passing records of different vehicle types within the same time slice on the same day in the past year, predict the average passing time prediction value of each vehicle type on the same day, and the calculation formula of the average passing time is: Among them, T avg (t, i, v) represents the average passing time of vehicle type v in the i-th passing section at time t, T n (t, i, v) represents the passing time of vehicle type v in the n-th transaction record, N v is the total number of transaction records of vehicle type v; Step 4-2, use the differential algorithm to calculate the traffic volume per minute of the current gantry in the target section, and the formula is: Among them, C in (t, i, v) represents the cumulative traffic flow of vehicle type v of the current gantry of the i-th section at time t; Step 4-3, predict the passing time of the vehicle at the downstream gantry of the target section according to the transaction time of the current gantry and the average passing time of the section; t exit (t, j, v) = t + T avg (i, v); where t exit (t, i, v) represents the estimated time for a vehicle of model v entering the i-th section at time t to pass through the downstream gantry; Step 4-4, predict the traffic volume of the downstream gantry of the target section based on the diversion probability and the traffic volume per minute of the current gantry; F out (t exit (t, i, v), M j , v) = F in (t, M i , v) × P div (t exit (t, i, v), M i , M j , v); Step 4-5, calculate the final predicted traffic volume of the downstream gantry based on the time slice where the transaction time of the downstream gantry is located; F pred (t, M j ) = ∑ i ∑ v F out ((t exit (t, i, v), M j , v)); Among them, F pred (t, M j ) represents the predicted flow rate of the downstream gantry M of the j-th section at time t, where i is all possible current gantry sections. j ​ 6. The method for monitoring road interruption and grading early warning based on ETC data according to claim 1, wherein: Step 5 specifically includes the following steps: Step 5-1, count the historical traffic volume F of the ETC gantry at time slice t on the same day last year actual (t, M j ); Step 5-2: Calculate the flow rate change rate based on the predicted flow rate data F of the downstream gantry pred (t, M j ) and the historical traffic flow F actual (t, M j ), with the formula as follows: Step 5-3, calculate the traffic change rate within two adjacent time slices (t, t-1) of the ETC gantry, and the formula is: Step 5-4, count the flow rate change rate R of all time slices hist and R adj for their distribution, and calculate the cumulative distribution function; select a threshold k1 and k2 that cause sudden drops for two flow rate change rates respectively according to the distribution of the flow rate change rate; furthermore, calculate the comprehensive weight coefficient k, and the formula for the comprehensive weight coefficient is: k = α×k1+(1-α)×k2; where α is the weight coefficient of the flow rate change rate R hist ; Step 5-5: Monitor the sudden traffic drop through the monitoring of the sudden traffic drop threshold and the heartbeat monitoring of the ETC gantry; monitor the real-time traffic change rate. When the traffic change rate is lower than the comprehensive threshold k, trigger the sudden traffic drop alarm.

7. The method for monitoring road interruption and grading early warning based on ETC data according to claim 6, wherein: In Step 5-5, perform heartbeat monitoring on each ETC gantry to ensure the normal operation of the gantry; when the gantry goes offline, immediately issue an alarm and mark the traffic data of the corresponding gantry as unavailable to avoid misjudging the sudden traffic drop.

8. The method for monitoring road interruption and grading early warning based on ETC data according to claim 1, wherein: Step 6 specifically includes the following steps: Step 6-1: Screen out the data with a traffic change rate lower than the threshold based on the traffic change rate of historical data and the traffic change rate of adjacent time slices: Among them, φ is an indicator function, which takes 1 when the condition is satisfied and 0 otherwise; Step 6-2: Screen out the data where the ratio of the actual traffic to the predicted traffic is less than k as b, and at the same time count the data c where the ratio of the actual traffic to the predicted traffic is 0. The total number of involved sections is N: Step 6-3: When the number of screened b data is greater than 0, output a general alarm, including the involved sections and the sudden traffic drop data b; Step 6-4: When the number of screened c data is greater than 0, it indicates that there is a situation of complete traffic interruption, and perform traffic detection on subsequent consecutive time slices; for each section i, check whether the actual traffic L in the next three consecutive time slices is all 0: If so, output a severe alarm message, including the involved area and the time slice information of continuous zero traffic; otherwise, if it does not meet the condition that the actual traffic in three consecutive time slices is 0, that is, there is traffic generated in the time slice, then continue to monitor the traffic in the subsequent time slices and execute Step 6-5; Step 6-5: Check whether the predicted traffic FL in the next three time slices is not all 0; if so, output a general alarm message; otherwise, when the traffic in the next three time slices is all 0, judge that a traffic interruption event occurs and output a severe alarm message.

9. The method for monitoring road interruption and grading early warning based on ETC data according to claim 8, wherein: In Step 6, for the general alarm, immediately verify the alarm information through the monitoring system, confirm the specific location and scope of the sudden traffic drop, and contact the on-site patrol personnel or traffic police to obtain the actual on-site situation; for the severe alarm, immediately start the severe alarm emergency plan, establish an emergency command group, and coordinate and handle it uniformly.

10. The road interruption monitoring and grading warning system based on ETC data applies the road interruption monitoring and grading warning method based on ETC data according to any one of claims 1 to 9, and is characterized in that: The system includes: A data acquisition module, including an ETC gantry hardware unit for collecting vehicle data and an interface unit for accessing multi-source data; A section traffic prediction module, including a unit for processing historical data, a unit for predicting section traffic, and a unit for outputting prediction results; The described traffic calculation module, including a unit for calculating real-time traffic, a unit for detecting traffic interruption, and a unit for generating alarm information; The information transmission module, including a unit for internal system communication, a unit for communicating with external systems, and a unit for monitoring communication status.

Images