Traffic control system for real-time data transmission

Abstract

The invention relates to the technical field of road vehicle traffic control, in particular to a traffic control system for real-time data transmission, and the system comprises a traffic data collection module which is used for monitoring vehicle arrival time sequence and queue length data through installing a camera at a traffic intersection, carrying out the real-time synchronization and storage of the collected data, and generating a real-time traffic data flow; and analyzing by using the real-time traffic data flow, identifying traffic peak and valley periods, and generating a traffic flow peak and valley identification result. According to the method, the cameras are arranged at the traffic intersection to monitor the arrival time sequence and queue length of the vehicles in real time, the vehicle data are collected and synchronously stored to form the real-time traffic data flow, and the peak and valley periods of the traffic flow are identified, so that the green light duration of the intersection is dynamically optimized to adapt to the actual change of the traffic flow; the matching of signal lamp control and actual traffic demands is realized, and the road congestion degree is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of road vehicle traffic control, and particularly to a traffic control system for real-time data transmission. Background Art

[0002] In the technical field of road vehicle traffic control, it mainly involves the monitoring, command, and management of the operation of various vehicles in the road traffic system. Through the real-time monitoring and analysis of the running state, traffic flow, speed, and congestion situation of road vehicles, combined with technical means such as traffic signal control, traffic guidance, vehicle guidance, and coordinated management, the road traffic efficiency is optimized, traffic congestion is alleviated, and the accident incidence rate is reduced.

[0003] In the actual operation process of the prior art, although the monitoring and control of basic information such as vehicle running state and traffic flow can be realized, there are still problems of insufficient data real-time performance and signal response accuracy. Traditional road traffic control usually adopts a signal timing plan with fixed time periods, lacking the flexible response ability to actual traffic demand changes, resulting in vehicle backlogs at sudden peak periods or accident scenes, causing regional congestion and reducing vehicle passing efficiency. Therefore, improvements are needed. Summary of the Invention

[0004] The purpose of the present invention is to solve the deficiencies existing in the prior art, and a traffic control system for real-time data transmission is proposed.

[0005] To achieve the above purpose, the present invention adopts the following technical solutions: A traffic control system for real-time data transmission includes:

[0006] A traffic data collection module, which monitors the arrival time sequence and queue length data of vehicles by installing cameras at traffic intersections, synchronizes and stores the collected data in real time to generate real-time traffic data streams; analyzes the real-time traffic data streams to identify traffic peak and trough periods and generates traffic flow peak and trough identification results;

[0007] A signal optimization decision module, based on the traffic flow peak and trough identification results, dynamically optimizes the calculation of the green light time for each intersection, calculates and generates an optimized green light time plan; combines the optimized green light time plan with environmental temperature and rainfall to adjust the red light extension strategy and generates an environment-adaptive signal adjustment result;

[0008] An emergency response coordination module, which receives the position and expected travel route information of emergency vehicles in real time, dynamically adjusts the signal lights based on the expected travel route information and the environment-adaptive signal adjustment result, gives priority to turning on the green light for emergency vehicles, and generates an optimized result for the passage of emergency vehicles; based on the position and expected travel route information of emergency vehicles, sends a notice of the approach of emergency vehicles to surrounding drivers to improve road surface safety and generates a driver prompt notice result;

[0009] The comprehensive application module of traffic management performs real-time information feedback based on the environmental adaptability signal adjustment results and the emergency vehicle passage optimization results, and generates real-time feedback results of traffic status.

[0010] Preferably, the steps of acquiring the real-time traffic data stream are:

[0011] Install cameras at traffic intersections to obtain vehicle arrival timing and queue length data. Based on the image data obtained by the camera, analyze the arrival time, stay time, vehicle spacing and queue status of each vehicle, match the motion trajectory between adjacent frames, and calculate the cumulative number of vehicles in different time periods to establish the original traffic monitoring data set;

[0012] Based on the original traffic monitoring data set, the arrival timestamps of each vehicle are parsed, the time deviations between different cameras are corrected, and a real-time traffic data stream is formed.

[0013] Preferably, the steps for obtaining the traffic flow peak and valley identification result are:

[0014] Based on the real-time traffic data stream, extract the hourly vehicle flow data to obtain hourly flow analysis results;

[0015] Based on the hourly flow analysis results, the flow change rate is calculated using the following formula:

[0016] ;

[0017] in, is the flow rate of the current hour, For the next hour's traffic, for The acceleration of time, is the average vehicle length in the current period, is the vehicle density in the current period, is the flow rate change rate;

[0018] Based on the flow rate change rate, the upper and lower limits of the threshold are set to define the peak and valley, determine the peak and valley periods of the traffic flow, and obtain the traffic flow peak and valley identification results.

[0019] Preferably, the steps for obtaining the optimized green light time scheme are:

[0020] Based on the traffic flow peak and valley identification results, the flow data of each intersection during peak hours and valley hours are extracted, and the average traffic speed, lane capacity, signal cycle, lane occupancy rate and number of pedestrian crossing requests of each intersection in different time periods are segmented and counted to form a traffic flow period characteristic data set;

[0021] Based on the traffic flow period feature dataset, calculate the green light time for each intersection, and the calculation formula is:

[0022] ;

[0023] Wherein, is the green light time, is the current number of lanes at this intersection, is the average flow velocity of the traffic flow, is the number of vehicles passing through per unit time, is the queue length at this intersection, is the average vehicle length at this intersection, is the number of pedestrian crossing requests, is the lane occupancy rate of the current intersection, is the average waiting duration at this intersection, is the average traffic flow within the current green light cycle;

[0024] Based on the green light time, adjust the green light time parameter of the signal lamp to form an optimized green light time plan.

[0025] Preferably, the steps for obtaining the environmentally adaptable signal adjustment result are:

[0026] Based on the optimized green light time plan, calculate the red light extension time, and the calculation formula is:

[0027] ;

[0028] Wherein, is the red light extension time, is the real-time rainfall, is the real-time road surface humidity, is the real-time visibility at the current intersection, is the number of pedestrians waiting during the red light period, is the detected average vehicle speed, is the average braking distance of the vehicles monitored at this intersection;

[0029] Based on the red light extension time, adjust the signal lamp cycle to generate an environmentally adaptable signal adjustment result.

[0030] Preferably, the steps for obtaining the optimized result for emergency vehicle passage are:

[0031] Receive the position and expected travel route information of the emergency vehicle, parse the current coordinates, moving speed and expected arrival time of the emergency vehicle, and combine with the real-time road conditions to generate emergency vehicle travel path prediction data;

[0032] Based on the predicted driving path data of the emergency vehicle, judge the expected signal status when the emergency vehicle passes through each intersection, and generate a dynamic signal light adjustment plan;

[0033] Based on the dynamic signal light adjustment plan, adjust the signal control parameters of each intersection along the way to generate an optimized result for the passage of the emergency vehicle.

[0034] Preferably, the step of obtaining the driver prompt notification result is as follows:

[0035] Based on the position and predicted travel route information of the emergency vehicle, calculate the time for sending the notification. The calculation formula is:

[0036] ;

[0037] Wherein, is the time for sending the notification, and are the current position coordinates of the emergency vehicle, and are the position coordinates of the surrounding driver's vehicle, is the speed of the emergency vehicle, is the current travel angle of the emergency vehicle, is the time increment from the current moment to the time when the emergency vehicle is expected to reach the relative position;

[0038] Based on the time for sending the notification, send a warning message to the surrounding drivers, including the expected arrival time and avoidance strategy of the emergency vehicle, to generate a driver prompt notification result.

[0039] Preferably, the step of obtaining the real-time traffic status feedback result is as follows:

[0040] Based on the environmental adaptability signal adjustment result and the optimized result for the passage of the emergency vehicle, extract the signal adjustment duration, signal switching times and emergency vehicle passage time periods of each intersection, and count the average vehicle passing volume within each time period to establish traffic signal adjustment analysis data;

[0041] Based on the traffic signal adjustment analysis data, judge the traffic flow differences at each intersection, analyze the traffic efficiency of different sections, identify abnormal congestion situations, and establish a real-time traffic status feedback result.

[0042] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0043] In the present invention, by deploying cameras at traffic intersections to monitor the arrival time sequence and queue length of vehicles in real time, collecting and synchronously storing vehicle data to form a real-time traffic data stream, identifying peak and trough periods of traffic flow, and then dynamically optimizing the green light duration at intersections to adapt to the actual changes in traffic flow, the matching of signal control and actual traffic demand is achieved, and the degree of road congestion is reduced. Further combined with environmental factors, the red light extension strategy is dynamically adjusted according to temperature and rainfall to enhance the adaptability of signal control under variable weather conditions and reduce the negative impact of bad weather on traffic flow. In addition, by receiving the position information and expected route of emergency vehicles in real time, the signal lights are automatically adjusted to give priority to emergency vehicles the right of way with a green light, and at the same time, a prompt notice of the approach of emergency vehicles is actively sent to surrounding drivers, improving the passing efficiency of emergency vehicles and road surface safety, alleviating the road congestion condition, and enhancing the operation safety and reliability of the road network. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a system flowchart of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0045] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0046] Please refer to Figure 1 , the present invention provides a technical solution: A traffic control system for real-time data transmission includes:

[0047] A traffic data collection module, which monitors the arrival time sequence and queue length data of vehicles by installing cameras at traffic intersections, synchronizes and stores the collected data in real time to generate a real-time traffic data stream; analyzes the real-time traffic data stream to identify peak and trough periods of traffic flow and generates a traffic flow peak and trough identification result;

[0048] A signal optimization decision module, based on the traffic flow peak and trough identification result, dynamically optimizes and calculates the green light time at each intersection, calculates and generates an optimized green light time plan; combines the optimized green light time plan with environmental temperature and rainfall to adjust the red light extension strategy and generates an environment-adaptive signal adjustment result;

[0049] An emergency response coordination module, which receives the position and expected travel route information of emergency vehicles in real time, dynamically adjusts the signal lights based on the expected travel route information and the environment-adaptive signal adjustment result to give priority to emergency vehicles with a green light and generates an optimized emergency vehicle passing result; based on the position and expected travel route information of emergency vehicles, sends a notice of the approach of emergency vehicles to surrounding drivers to improve road surface safety and generates a driver prompt notice result;

[0050] The comprehensive application module of traffic management provides real-time information feedback based on the environmental adaptability signal adjustment results and emergency vehicle traffic optimization results, and generates real-time feedback results on traffic status.

[0051] The steps to obtain real-time traffic data stream are:

[0052] Install cameras at traffic intersections to obtain vehicle arrival timing and queue length data. Based on the image data obtained by the camera, analyze the arrival time, stay time, vehicle spacing and queue status of each vehicle, match the motion trajectory between adjacent frames, and calculate the cumulative number of vehicles in different time periods to establish the original traffic monitoring data set;

[0053] Based on the original traffic monitoring data set, the arrival timestamp of each vehicle is parsed, the time deviation between different cameras is corrected, and a real-time traffic data stream is formed.

[0054] Specifically, first, on-site measurements are carried out according to the number of lanes and the visible range at the intersection to determine the number and location of cameras. Usually, three to five cameras are set up around each intersection to cover the traffic flow images in different directions. This number is obtained by evaluating the road width in the range of 15 meters to 30 meters and comparing the coverage of the images collected historically. After the cameras are installed, three basic indicators, namely brightness, contrast, and clarity, are recorded in the images collected per second and compared with the preset collection standards. For example, the pixel value range for brightness comparison is 0 to 255, the visual effect quantization range for contrast comparison is 10 to 50, and the resolution range for clarity comparison is 300 lines to 600 lines. If any of the indicators exceeds the above range, the camera focus and external lighting conditions are checked again. Subsequently, vehicle detection and analysis are performed on the images that meet the collection standards in frame order. A convolutional neural network model with a structure of five convolutional layers and two fully connected layers is used to achieve vehicle detection and analysis. The input data of the model is the image pixel matrix arranged according to the frame number and timestamp, and each pixel retains three RGB channels. The output data of the model is the coordinates of the detected vehicle rectangular area and the mask annotation of the vehicle within the image. The training process of the model includes five main stages: data preparation, parameter initialization, forward propagation, backward propagation, and parameter iteration. In the data preparation stage, 10,000 images are first extracted from all the collected images, and the vehicle positions are manually marked to obtain the true labels of the training samples. These images cover the intersection scenes at different times and in different weather conditions. In the parameter initialization stage, the weights of the five convolutional kernels are randomly set to floating-point numbers between -0.05 and 0.05, and the initial weights of the fully connected layers are between -0.01 and 0.01. In the forward propagation stage, convolutional operations are performed on each image sample and passed through the ReLU activation function to ensure that the output is non-negative. Subsequently, the max-pooling layer is used to reduce the number of parameters. In the backward propagation stage, the difference between the prediction result and the true label is calculated according to the loss function, and the weights are updated layer by layer. In the parameter iteration stage, the learning rate is usually set to 0.001 for each iteration, and 50 consecutive iterations are performed. The mean square error is used as the loss function, and the accuracy and error changes of the training set and the validation set are statistically analyzed after each round of iteration. If the accuracy does not improve for five consecutive rounds of iteration, 0.0002 is reduced on the basis of the original learning rate until the training stop condition is met. The neural network obtained through such training can be used to automatically segment the vehicle body area, match the same vehicle positions in subsequent frames, and calculate the vehicle movement trajectory. After vehicle segmentation and tracking are completed, the vehicle information within the same period is numbered and summarized. The pixel positions of the vehicle in adjacent frames are compared. If the displacement of the same vehicle between any two frames is less than 3 pixel units and the difference in the motion vector direction does not exceed 5 degrees, it is determined as the same target. Finally, the number of vehicles is counted at different times, and the vehicle queuing situation is recorded. All vehicle detection results, queuing duration, and other parameters are summarized to form the original traffic monitoring dataset.

[0055] Based on the obtained original traffic monitoring dataset, read the vehicle identification, the timestamp of each vehicle in each frame, and the vehicle queuing information, and align the time tags corresponding to each vehicle. The alignment method is to select the time starting point of each camera under the same intersection as the reference and calculate the difference between each camera and this reference point. For the case where the difference is between -0.1 second and 0.1 second, it is directly recorded as synchronized and no further correction is required. For the difference exceeding this interval, a secondary correction is performed in combination with the order of vehicles entering the picture. Specifically, during the alignment process, read the appearance order of the vehicle in the pictures of different cameras and refer to the multi-frame comparison result of the queuing length. If the difference in the queuing length observed across cameras at the same moment exceeds two vehicles, it is determined that the timestamp deviation of this camera is relatively large and interpolation calculation is performed based on the historical image sequence, shifting the timestamp forward or backward by an amount ranging from 0.05 second to 0.2 second. Taking the difference in the queuing length after offset being less than two vehicles as the completion standard of the correction, this correction standard is obtained by combining the statistical result that the vehicle queuing peak value on the same road section within 30 days is between 10 and 20 vehicles and through segmented comparison. After completion of the alignment, store the accurate arrival timestamp of each vehicle corresponding to its queue position in a classified manner. Finally, uniformly combine all the normalized time data to form a real-time traffic data stream.

[0056] The steps for obtaining the traffic flow peak and valley recognition results are as follows:

[0057] Based on the real-time traffic data stream, extract the vehicle flow data per hour to obtain the flow analysis result per hour;

[0058] Based on the flow analysis result per hour, calculate the flow change rate. The calculation formula is:

[0059] ;

[0060] where, is the flow of the current hour, is the flow of the next hour, is the acceleration at the moment of is the average vehicle length of the current period, is the vehicle density of the current period, is the flow change rate;

[0061] Based on the flow change rate, set the upper and lower threshold limits to define the peak and valley, determine the peak and valley periods of the traffic flow, and obtain the traffic flow peak and valley recognition results.

[0062] Specifically, based on the real-time traffic data stream obtained previously, read the time series records corresponding to each intersection and divide them hourly. When traversing each record in each hourly segment one by one, it is necessary to first collect the number of vehicles passing through during this time segment, the arrival time and departure time of the vehicles when passing through this intersection, and reconfirm duplicate or missing information. The confirmation methods include comparing the vehicle identification records in adjacent time periods or adjacent intersections again to exclude the situation of the same vehicle being counted multiple times. For the missing records caused by camera occlusion or blurred images, the comparison of several frames before and after can be further used to determine whether a certain vehicle actually passed through this intersection during this time period. Subsequently, summarize all accurate vehicle records and fix the total number of vehicles in this hour as the traffic flow value for this time segment. For example, if a total of 276 valid vehicle records are found in the statistical interval from 1 to 2 o'clock and overlapping statistics are excluded, then 276 is used as the traffic flow value for this time segment. Then, arrange the traffic flow values in different time segments in sequence, establish an hourly traffic flow sequence by continuously recording the vehicle flow, and save it with a precision of one or two decimal places. For example, if 314 valid records are counted from 2 to 3 o'clock, then mark 314 in the corresponding time segment. When checking the relevant records, it is necessary to follow the pre-established time overlap judgment criteria, such as if the same vehicle has appeared at 1:59, it will no longer be included in the traffic flow list from 2 to 3 o'clock, so as to ensure the accurate distinction of the traffic flow data for each hour. Then, conduct a simple search and check on the hourly traffic flow sequence to prevent traffic flow values exceeding a reasonable range. For example, during the night from 2 to 4 o'clock, the traffic flow is usually small. A range upper and lower limit can be preset according to the historical statistical rules of the previous continuous week, such as 0 to 500 vehicles. If it exceeds this range, mark the data as abnormal and compare it with the adjacent time period of the same road section for further positioning. In this way, the vehicle statistics for each intersection are completed hour by hour and numbered according to the intersection number, and finally the hourly traffic flow analysis results are generated.

[0063] The advantage of the formula is that it simultaneously considers multiple important factors such as the difference in vehicle traffic flow between adjacent hours, the average acceleration of vehicles, the average vehicle length, and vehicle density, and introduces the superposition term of vehicle length and density in the denominator, making the calculated traffic flow change rate more capable of reflecting the comprehensive influence of multiple conditions in different time periods, and thus more accurately measuring the change range and fluctuation of traffic flow between adjacent hours.

[0064] The acquisition steps are as follows: After determining the hourly statistical period, record the number of vehicles passing through the specified intersection during that period and summarize it into numerical data. The number of vehicle passages included in each period can be obtained by counting each vehicle based on the camera recognition results. To clarify the source of this value, set up cameras at a road section with a width of 30 meters and three lanes and collect 24-hour image data. Eliminate duplicate counts by matching the vehicle positions between frames and combining the timestamps of vehicle passages. For example, the statistical value obtained from 7 to 8 am is 600 vehicles, and record these 600 as .

[0065] The acquisition steps are as follows: Directly continue with the traffic volume value of the next hour after the previous hour's statistical period. The specific operation is the same as that for obtaining , and it is still based on the vehicle recognition results of the same camera for statistics. If the previous period is from 8 to 9 am, then this period is from 9 to 10 am. Assume that the number of vehicles recorded from 9 to 10 am is 715. Detect each vehicle that appears from 9 am to 9:59:59 am by checking the image frames and use the same camera position and annotation method as the previous period for statistics. If a vehicle is detected to be double-counted, further check its entry and exit boundaries within adjacent time periods to ensure that the result of 715 only corresponds to the total number of vehicles actually passing through within the range from exactly 9 am to exactly 10 am. At this time, assign 715 to .

[0066] The acquisition steps are as follows: During the process of tracking vehicles for the current period , by detecting the speed values of each vehicle at different detection positions and combining the timestamps of the vehicle entering and leaving this position before and after, use the uniformly accelerated motion equation to calculate the acceleration of the corresponding vehicle. Average the acceleration results sampled within the same period to obtain the average acceleration of this period. This process can obtain the acceleration value by dividing the speed difference between adjacent speed measurement points by the actual driving time difference, and obtain by taking the arithmetic mean of the accelerations within each hour. For example, the acceleration values of 100 vehicles are collected from 8 to 9 am, and the average value is 2.3 meters per second squared. When incorporating it into the subsequent formula, record it as .

[0067] The acquisition steps are as follows: Summarize the average vehicle length of the recognized vehicles within the current period . The length of each vehicle can be estimated by image detection. First, place a reference object at the intersection to calibrate the corresponding relationship between the pixel length and the actual length. After positioning the contour of each vehicle captured by the camera, obtain the actual length of each vehicle by multiplying the pixel size of the vehicle by the corresponding calibration conversion coefficient, and then average the length results of all vehicles in this period to obtain , for example, 640 vehicle outlines were captured from 8 to 9 o'clock. Among them, the average length of cars is about 4.3 meters, and the average length of large buses and trucks is 12.0 meters. The weighted average length of the mixed traffic flow measured on that day is 5.8 meters, and its calculation method can be defined as , so from 8 to 9 o'clock, = 5.8 meters.

[0068] The acquisition steps of are as follows: count the density of identified vehicles in the road cross-section during the current period. The density represents the number of vehicles distributed per unit road section length. Capture the traffic flow at a top-down or oblique angle of the images at the same intersection, locate the relative positions of all vehicles in the picture with the lane lines during this period and count the space occupancy ratio of vehicles at a certain moment. Limit the unit road section length to the common 100 meters to reduce deviation. For the traffic flow density calculation from 8 to 9 o'clock, obtain the average value at the sampling time points of multiple detections. For example, at 8:10, 8:30, and 8:50, record that there are 10 vehicles, 12 vehicles, and 11 vehicles within the 100-meter length of this road section respectively, then the average is 11 vehicles, and regard it as = 11. After completing the conversion of the ratio of the number of vehicles to the unit road section length, record it as 11 / 100 and obtain the value 0.11.

[0069] Calculation process:

[0070] The first step: Determine the specific data, let = 640, = 715, = 2.3, = 5.8, = 0.11;

[0071] The second step: Calculate the flow difference ;

[0072] The third step: Multiply the flow difference by the acceleration to get ;

[0073] The fourth step: Calculate the internal term of the denominator ;

[0074] The fifth step: Take the square root of the denominator ;

[0075] The sixth step: Divide the numerator by the denominator ;

[0076] The seventh step: Multiply by Converted to a percentage, it is ;

[0077] The results show that when the hourly traffic volume increases from 640 vehicles to 715 vehicles and the average acceleration, average vehicle length, and traffic flow density are 2.3, 5.8, and 0.11 respectively, the traffic flow change rate in this period is approximately 18%. If a higher or lower change rate occurs in subsequent periods, it means different degrees of traffic flow fluctuations.

[0078] According to the traffic flow change rate calculated above, for each hourly period, compare and count item by item whether this value is within a pre-determined upper and lower limit range. First, it is necessary to make a preliminary detection by combining the traffic flow change rate data in the corresponding period within at least 7 days to 30 days in the past, and then select the periods with relatively close traffic flow levels as a reference to determine the upper and lower limit range. For example, from 8:00 to 9:00 every day, the traffic volume on this road section is concentrated, and the traffic flow change rate is usually between 5% and 25%. Therefore, 5% can be selected as the lower limit and 25% as the upper limit. If it is found during the statistics that the traffic flow change rate in a certain hourly period is lower than 5%, it is determined to be at a low level, and if it exceeds 25%, it is determined to be in the stage of rapid traffic flow increase. During the implementation of this step, the recent traffic flow change rate can be queried at a granularity of half an hour or 1 hour and compared with the preset upper limit of 25% and lower limit of 5%. For situations exceeding this range, the vehicle tracking records can be read again to verify the acceleration and density information, so as to confirm and exclude the numerical fluctuations caused by abnormal identification and incorporate the finally confirmed valid change rate into the full data list. Then, by comparing each period one by one, match and mark the change rate of all periods with the upper and lower limits. Any value between 5% and 25% is marked as a normal period, and a note is added to the periods with values outside this range. Based on the above marking results, identify the peak periods exceeding 25% and the low periods below 5% among all hourly periods. When the comparison and marking of all periods are completed, the traffic flow peak and valley identification results can be obtained.

[0079] The steps to obtain the optimized green light time plan are as follows:

[0080] Based on the traffic flow peak and valley identification results, extract the traffic volume data of the peak and valley periods at each intersection, and conduct segmented statistics on the average traffic flow speed, lane passing capacity, signal cycle, lane occupancy rate, and pedestrian crossing request times at each intersection in different time periods to form a traffic flow period feature data set;

[0081] Based on the traffic flow period feature data set, calculate the green light time for each intersection. The calculation formula is:

[0082] ;

[0083] Among them, is the green light time, is the current number of lanes at this intersection, is the average flow velocity of the traffic flow, is the number of vehicles passing through per unit time, is the queue length at the intersection, is the average vehicle length at the intersection, The number of pedestrian crossing requests, is the lane occupancy rate of the current intersection, is the average waiting time at the intersection, is the average traffic volume in the current green light cycle;

[0084] Based on the green light time, the green light time parameters of the traffic light are adjusted to form an optimized green light time plan.

[0085] Specifically, based on the results of traffic flow peak and valley identification, the data content of the peak and valley periods of each intersection that have been divided is read, and information such as traffic speed, lane capacity, signal cycle, lane occupancy rate and number of pedestrian crossing requests are extracted in these time periods. First, the specific start and end times are located through the previously obtained time period division records, and the traffic operation parameters of the intersection within the time range are retrieved. The traffic speed is then recorded in kilometers per hour or meters per second and the speed is measured by using the method of parsing the vehicle's driving trajectory frame by frame using the camera equipped with the speed measuring device mentioned above. The lane capacity is evaluated and recorded based on the actual vehicle traffic volume at the hourly or half-hourly level on the premise of determining the current number of lanes and lane widths at the intersection. The signal cycle reads the actual traffic light switching frequency in the traffic signal equipment at the intersection according to the working configuration and makes a time mark. The lane occupancy rate can be obtained by measuring the number of vehicles in the picture during the same time period. The proportion of occupied lane area is quantitatively counted and recorded as a real number between 0 and 1. For example, when collecting data on the daytime peak period of the intersection within 30 days, if the occupancy rate of a certain period is 0.75, the value is saved in the occupancy rate list of the corresponding period of the day. The number of pedestrian crossing requests can be read from the infrared detection equipment installed on one side of the zebra crossing and the number of triggers in the statistical period is calculated one by one. For each intersection, the same statistical method is used to record the above indicators one by one, and the recorded values are distinguished and marked according to the peak period and the trough period. Subsequently, these marked multi-period data are merged to form an overall segmented statistical result and arranged in order according to the intersection number and time period. In this way, on the basis of peak and trough, the average traffic speed, lane capacity, signal cycle, lane occupancy rate and number of pedestrian crossing requests at each intersection in different time periods are segmented and summarized, and finally a traffic flow period characteristic data set is formed.

[0086] The advantage of the formula is that multiple core parameters are incorporated into the same calculation expression. It includes elements such as the current number of lanes and traffic flow speed that reflect the scale and flow characteristics of the traffic flow, and also synthesizes dynamic traffic indicators such as the number of vehicles per unit time, queue length, and lane occupancy rate. By simultaneously considering the number of pedestrian crossing requests and the average traffic flow within the current green light cycle, a mutually restrictive relationship is formed between the numerator and the denominator, enabling a more systematic and accurate calculation reference for the green light time and taking into account the traffic needs of both vehicles and pedestrians during the actual application process.

[0087] Parameter Obtaining steps: represents the current number of lanes at this intersection. In actual monitoring, the specific number of lanes can be obtained through on-site investigation or by checking the road network planning documents. For example, if there are 4 lanes in the east-west direction at a certain intersection and all are open to social vehicles, then = 4.

[0088] Parameter Obtaining steps: is the average flow velocity of the traffic flow, indicating the average value of the actual driving speeds of vehicles within a certain period at the current intersection or section. To obtain this value, it is necessary to continuously monitor the speeds of several vehicles on each main lane at the intersection, and then perform a weighted average on all the detected speed values. For example, if 200 vehicles are continuously sampled in the west entrance direction during the peak period and the average speed is 28 kilometers per hour, and 180 vehicles are sampled in the east entrance direction with an average speed of 25 kilometers per hour, then the overall average flow velocity can be obtained by comprehensively weighting these two average speeds according to the traffic flow ratio of the two directions. If the traffic flow ratio is approximately 60% for the west entrance and 40% for the east entrance, then it can be calculated kilometers per hour, that is, .

[0089] Parameter Obtaining steps: is the number of vehicles passing through per unit time. For example, if 525 vehicles are recorded passing through in 15 minutes, then the number of vehicles passing through per unit time can be recorded as vehicles per minute.

[0090] Parameter Obtaining steps: represents the queue length at this intersection. It is necessary to record the total length occupied by the queuing vehicles. There are often more queuing vehicles during the peak traffic flow period. The length can be obtained by taking pictures with a camera installed downstream of the intersection and measuring the distance between the end position of the vehicle queue and the stop line. If on-site monitoring shows that the common queue length at this intersection during the morning peak is about 65 meters, then it can be recorded as meters.

[0091] Parameter Obtaining steps: is the average vehicle length at this intersection. It is necessary to measure the vehicle length of each vehicle passing through this intersection one by one within a specified time period and calculate the average value. The measurement can calculate the pixel difference between the head and tail of the vehicle by means of the calibrated distance superimposed by the camera, and then multiply by the corresponding conversion coefficient to obtain the true length of the vehicle. Subsequently, the sum of the lengths of all vehicles is added and divided by the total number of vehicles to obtain the average vehicle length. For example, it is monitored that 700 vehicles pass through during a certain peak period, with the average length of small passenger cars being 4.3 meters and the average length of large passenger cars or trucks being 12 meters. After statistics, the average vehicle length of the mixed traffic flow is about 5.6 meters. At this time, record and include it in the dataset for calculation.

[0092] Parameter Obtaining steps: represents the number of pedestrian crossing requests. It is necessary to record the cumulative number of times pedestrians press the crossing button at the crosswalk of the corresponding intersection. If 180 pedestrian trigger amounts are recorded during a certain period at peak hours, then .

[0093] Parameter Obtaining steps: represents the lane occupancy rate of the current intersection. Continuously photograph the number of covered pixels of the traffic flow within the lane range and calculate the ratio with the total number of lane pixels. For example, photograph the west import lane during a certain period. The total lane pixels in the statistical image are 200,000, and the total vehicle pixels are 100,000. Then the occupancy rate can be recorded as .

[0094] Parameter Obtaining steps: is the average waiting time at this intersection, which represents the average time between when a vehicle starts waiting at the intersection until it gets the permission to pass. To obtain this value, it is necessary to track the time interval when the actual vehicle waits in front of the stop line, identify the arrival and start times of the same vehicle by reading the license plate or vehicle body identification and subtract them. After continuously recording the waiting times of several vehicles at different time periods, take the average value. For example, in the case of counting 100 vehicles, it is found that the total waiting time of these vehicles is 2,700 seconds. Then calculate seconds.

[0095] Parameter Obtaining steps: represents the average traffic flow within the current green light cycle. It is necessary to count the number of vehicles passing through from the start to the end of a complete green light duration, and then divide this value by the green light duration to obtain the traffic flow per second or per minute. Then, observe multiple consecutive green light cycles according to the above recording method and calculate the average value. For example, if 40 vehicles actually pass through during a 30 - second green light cycle, then the counted traffic flow is .

[0096] Calculation process:

[0097] Step 1: For the corresponding intersection (four normally passable lanes), km / h (converted to approximately 7.44 m / s for unit consistency), vehicles per minute (converted to 0.58 vehicles per second), m, m, , , s, When substituting vehicles per second into the formula, unit consistency processing must be performed first.

[0098] Step 2: Unify the speed to m / s, then , with seconds as the time unit, then vehicles per second, with seconds as the calculation period, then s. After substituting into the calculation, we get ; This result indicates that for this combination of parameters, a green light duration of about 28 seconds may be required at the intersection. If the actual road conditions are close to the above collection conditions, this result can be used for signal control. If this result is greater than 30 seconds or less than 25 seconds, it means that the corresponding input parameters have changed significantly, and the corresponding items need to be re-measured or a larger-scale collection needs to be carried out again.

[0099] Based on the green light time, read the calculated green light duration in seconds and compare it with the corresponding configured signal cycle at each intersection before. Then match and register the green light duration in the traffic flow in each direction. Assuming that the current intersection is two-way traffic and additionally includes a dedicated left-turn phase, for the straight-through lanes, first apply the calculated green light time to the corresponding control hardware, and then record the timing difference between the original red light switching point and this green light duration. If this difference exceeds the existing preset threshold, check the specific source of the preset threshold. Such thresholds are usually obtained by the management department through historical traffic flow data statistics within one year and have an acceptable error range of 3 seconds to 5 seconds. Selecting the threshold for the corresponding intersection as 4 seconds means that if it exceeds 4 seconds, the green light cycle needs to be corrected again. For the left-turn lanes in the same direction, the traffic flow during the peak period needs to be separately counted and converted into a second-level demand, and then the green light time of the dedicated phase is added or subtracted. By repeatedly recording the difference values under each channel and comparing them with the threshold, finally, the green light time parameters of the signal lights are adjusted as a whole to form an optimized green light time plan.

[0100] The steps to obtain the environmental adaptability signal adjustment result are as follows:

[0101] Based on the optimized green light time plan, calculate the red light extension time. The calculation formula is:

[0102] ;

[0103] Among them, is the red light extension time, is the real-time rainfall, is the real-time road surface humidity, is the real-time visibility at the current intersection, is the number of pedestrians waiting during the red light period, is the detected average vehicle speed, is the average braking distance of vehicles monitored at this intersection;

[0104] Based on the red light extension time, adjust the signal light cycle to generate the environmental adaptability signal adjustment result.

[0105] Specifically, the advantage of the formula is that it integrates factors such as real-time rainfall, road surface humidity, visibility, number of pedestrians waiting, average vehicle speed, and vehicle braking distance into a unified expression, and presents the comprehensive impact on the signal light period adjustment demand through the comparison of the numerator and denominator. Thus, in situations such as heavy rainfall, high road surface humidity, or large pedestrian density, it can relatively accurately calculate the required red light extension duration, and combine the vehicle speed and braking distance to achieve safer intersection management.

[0106] Parameter acquisition steps: represents the real-time rainfall. If the rainfall is 10 millimeters per hour, record at this time.

[0107] Parameter acquisition steps: represents the real-time road surface humidity, which can be detected by a road surface humidity sensor. This sensor usually consists of two electrodes, and converts the corresponding humidity value by measuring the change in the minute conductivity of the road surface. Read the electrode resistance value every five minutes, and obtain the arithmetic mean of multiple measurement results within each hour. The humidity percentage formula can be defined as . Assuming that the reference conductivity of the dry road surface is 0.001 Siemens and the conductivity currently measured by the sensor is 0.002 Siemens, then the humidity is , which is converted to a numerical value such as . Record .

[0108] Parameter acquisition steps: Indicates the real-time visibility of the current intersection. The acquisition method is to set a laser ranging type or transmissive visibility meter at the intersection. This device continuously emits light into the front environment and quantifies the concentration of rain and fog particles in the air by measuring the degree of light attenuation, thereby obtaining the visibility distance. For example, if the visibility of this intersection is measured to be around 300 meters and shows little fluctuation during a certain period, and the arithmetic mean is 310 meters after multiple measurements, it is recorded as 。

[0109] Parameter Acquisition steps: Indicates the number of pedestrians waiting during the red light period. It is necessary to count the number of pedestrians at the stop line or the edge of the zebra crossing during the duration of the red light, and obtain the number information through video image analysis. For example, six rounds of red lights were recorded between 8 am and 9 am, and the pedestrian distributions were 30 people, 28 people, 35 people, 27 people, 29 people, and 31 people respectively. Calculate the average value ,At this time, 。

[0110] Parameter Acquisition steps: Is the detected average vehicle speed. It is obtained by sequentially setting vehicle speed detection devices at the entrance and exit of the intersection, counting the driving speeds of vehicles before and after passing through the intersection, and calculating the average. If 100 vehicles are detected during a certain observation period, the average value is obtained by dividing the total speed of all vehicles by 100. For example, if the detection shows that the total is 2650 kilometers per hour, then Kilometers per hour, and for easier subsequent calculation, it can be converted to meters per second ,Record as 。

[0111] Parameter Acquisition steps: Indicates the average braking distance of the vehicles monitored at this intersection. It is necessary to make on-site records when multiple vehicles are approaching the stop line. Specifically, a laser ranging device on the side of the lane measures the spatial distance that the vehicle travels from the start of braking to a complete stop, and obtains the average value from the data of several vehicles. The calculation formula for the vehicle braking distance can be defined as the braking distance is ,Where Is the initial vehicle speed, Is the friction coefficient between the road surface and the tire, Is the gravitational acceleration of 9.8 meters per second squared. The average braking distance is obtained by comparing the deviation between the actual measurement value and the theoretical value through multiple experiments. If the braking processes of 30 vehicles are tested during a peak period and the average value is measured to be 20 meters, this result is recorded as 。

[0112] Calculation process:

[0113] Step 1: Unify the units and take the observed values at that time. For example, millimeters per hour, , meters, persons, meters per second, meters.

[0114] Step 2: For the numerator part First calculate as .

[0115] Step 3: After summing, get .

[0116] Step 4: For the denominator part .

[0117] Step 5: Finally seconds.

[0118] This result indicates that at an intersection with a rainfall of 10 millimeters per hour, a road surface humidity of 2.0, a visibility of 310 meters, 30 people waiting at the red light, an average vehicle speed of 7.36 meters per second, and a braking distance of 20 meters, it is necessary to extend the existing red light by approximately 27.7 seconds. After registering this result in the daily red light delay record, the signal cycle adjustment can be directly executed. If it is detected later that the rainfall decreases or the humidity drops, the new extended time value should be re-monitored and calculated.

[0119] Based on the red light extension time, read the second-level extension duration calculated previously and add it to the existing red light timing sequence of this intersection to complete the cycle redefinition. First, locate the signal timing structure in each direction currently and record the original red light duration as 45 seconds. Then, mark the monitored extended time in digital form in the system. If the extended time is greater than a preset threshold, such as 15 seconds, compare it with the original timing strategy of the system and check whether there are factors such as increased traffic flow density or abnormal visibility in the current road conditions. This threshold is set by the road management department between 10 seconds and 15 seconds based on the statistical data of the average extended duration in the past 30 days to distinguish a relatively large range of delays. When the calculated value is between 10 seconds and 15 seconds, maintain the conventional warning method. When the calculated value reaches more than 15 seconds, conduct a secondary verification according to the local management regulations and continue to check whether the current rainfall or humidity value is in the suddenly increasing range. If it is indeed detected that the rainfall and humidity rapidly climb within 5 minutes, add the extended time in full to the total cycle. If it is less than 10 seconds, make a confirmation at the lowest threshold of more than 7 seconds to determine whether it is actually necessary to fine-tune the red light. Subsequently, perform the cycle refresh operation on the signal controller in each direction in sequence, unify the records of all links in the daily environmental adaptability scheduling list and organize them according to the intersection numbers, and finally generate the environmental adaptability signal adjustment result.

[0120] The steps for obtaining the optimized results of emergency vehicle passage are as follows:

[0121] Receive the position and expected travel route information of the emergency vehicle, parse the current coordinates, moving speed and expected arrival time of the emergency vehicle, and generate the predicted data of the emergency vehicle's driving route in combination with the real-time road conditions;

[0122] Based on the predicted data of the emergency vehicle's driving route, judge the expected signal status when the emergency vehicle passes through each intersection, and generate a dynamic signal light adjustment plan;

[0123] Based on the dynamic signal light adjustment plan, adjust the signal control parameters of each intersection along the way to generate the optimized results of emergency vehicle passage.

[0124] Specifically, to receive the position and expected travel route information of the emergency vehicle, first read the current coordinates from the in-vehicle positioning device of the emergency vehicle and obtain the specific position of the vehicle in the road network in the form of latitude and longitude data. Subsequently, read the moving speed information uploaded by the emergency vehicle and compare this speed with the established road speed limit parameter library. For example, compare the speed with the range of 20 kilometers per hour to 80 kilometers per hour. If it is found that the speed exceeds 80 kilometers per hour, then recheck the speed data, analyze the corresponding relationship between the vehicle displacement and time between the previous and next sampling moments, exclude occasional outliers, then compare the expected arrival time of the emergency vehicle with the established average traffic flow statistical data, incorporate the road congestion conditions within multiple time segments for reference to correct the driving duration, and disassemble the possible driving routes of the vehicle according to kilometer-level or finer-grained network sections. Compare the current road conditions information of each section, such as the number of road accidents, the current meteorological conditions, and the signal cycle of adjacent intersections, and record item by item parameters such as the planned driving direction of the vehicle and the road turning radius. If there are multiple consecutive narrow curves in a section and the road width is less than 6 meters, then mark the potential traffic restriction situation in this section. Connect these marked sections in series to form a candidate set of lines arranged by priority, and then select a main line with a higher accessibility rate and a lower congestion possibility based on the actual departure coordinates and end coordinates of the vehicle. Summarize the expected driving time of each section of the emergency vehicle on this line in a segmented calculation manner. After the summary, obtain the overall estimated driving time of the emergency vehicle and make dynamic corrections in combination with the coordinates, moving speed, and road conditions obtained previously. Integrate all the matched time series results and mark the passing moments of each key intersection and main turning point, and finally generate the predicted data of the emergency vehicle's driving route.

[0125] Based on the predicted driving path data of emergency vehicles, read the passing moments of emergency vehicles at each road section and key intersections previously and compare them one by one with the current traffic signal status. First, look up the intersection number and its signal control cycle to which the road section belongs in the road section database. For example, if there is a cycle of 40 seconds of green light and 30 seconds of red light at a certain intersection, extract a complete green and red light sequence within the predicted passing period of the emergency vehicle. Then, refine the comparison between the time stamp when the emergency vehicle is expected to reach the intersection and this sequence. When it is found that there may be an overlap with the red light, calculate the specific number of seconds of the difference. If this difference value is greater than a certain predetermined standard, such as 15 seconds, it is marked as a period when signal light adjustment can be carried out. This predetermined standard is set between 10 seconds and 15 seconds by the management department within the past week based on multiple factors such as the needs of emergency vehicles and the signal timing frequency at intersections, and is verified by recording multiple batches of data. When the overlap value between the vehicle travel time and the signal light conversion period is within 10 seconds to 15 seconds, it meets the adjustable threshold. Immediately, make targeted revisions to the red and green light time sequences at the current intersection. If it is found that the traffic flow information is still in the medium or low load range, the green light can be directly extended by a certain number of seconds to match the time window when the emergency vehicle arrives. If it is in the high load range, first reduce the time period for adjacent directions in a segmented manner and verify the green light utilization rate of relevant lanes. Through this cycle of comparison and time rearrangement in different directions, accumulate the possible signal light adjustment operations at each intersection one by one, and form a dynamic signal light adjustment plan after summarization.

[0126] Based on the dynamic signal light adjustment plan, select each intersection where the emergency vehicle passes through, and compare the adjusted value of the green light duration or the reduced value of the red light listed in the plan with the actual signal cycle of this intersection. Divide the signal cycle into multiple segments and check whether each segment meets the adjusted number of seconds specified in the plan. For example, if the original cycle duration of a certain intersection is set to 45 seconds of red light and 35 seconds of green light, when the plan proposes to expand the green light to 40 seconds and reduce the red light to 40 seconds, change the original 45 - second red light to 40 - second red light and record it within the threshold range set in the plan. This threshold range is set to 3 seconds to 5 seconds based on the shortest adjustment interval of green and red lights statistically calculated by the traffic department from the data in the past 90 days. When the number of seconds to be adjusted exceeds 5 seconds, view the background traffic flow statistics of this road section and evaluate whether it will affect the surrounding road sections. If there is no conflict within the limited inspection range, make a substantial update to the signal cycle, and then send it to the controllers of the corresponding intersections one by one. Record the complete list of signal light control parameters and mark them corresponding to the driving time of the emergency vehicle. Finally, obtain the optimized result for the passage of emergency vehicles.

[0127] The steps to obtain the driver prompt notification result are as follows:

[0128] Based on the position and predicted travel route information of the emergency vehicle, calculate the time for sending the notification. The calculation formula is:

[0129] ;

[0130] Among them, is the time to send the notice, and are the current position coordinates of the emergency vehicle, and are the position coordinates of the surrounding driver vehicles, is the speed of the emergency vehicle, is the current traveling angle of the emergency vehicle, is the time increment from the current moment to the expected arrival of the emergency vehicle at the relative position;

[0131] Based on the time of sending the notice, send a warning message to the surrounding drivers, including the expected arrival time of the emergency vehicle and the avoidance strategy, and generate the driver prompt notice result.

[0132] Specifically, the benefit of the formula is that, in addition to the spatio-temporal parameters composed of distance and speed, it also takes into account the traveling angle and time increment of the emergency vehicle, and reflects the traffic differences of the emergency vehicle in different directions and routes through the coupling of the angle value and time increment, so as to more precisely obtain the best time period for sending the warning notice.

[0133] Parameter Obtaining steps: represents the current position abscissa of the emergency vehicle and needs to be detected by a positioning device. In a test, when the device detects that the lateral coordinate of the emergency vehicle is 483210.57 meters, this value can be recorded as .

[0134] Parameter Obtaining steps: represents the current position ordinate of the emergency vehicle, which is the same as . If the actual detection shows that the longitudinal coordinate of the emergency vehicle is 2730541.33 meters, then record .

[0135] Parameter Obtaining steps: is the abscissa of the surrounding driver vehicle. The acquisition method is similar to that of the emergency vehicle. After matching the license plate or shape of each vehicle recognized by the camera, map the pixel position of the vehicle in the image to the geographical coordinate. If the lateral coordinate of the driver vehicle is detected to be 480100.82 meters at the scene, then take .

[0136] Parameter Obtaining steps: is the ordinate of the surrounding driver vehicle and needs to be the same as Cooperate to represent the geographical plane position where the vehicle is located. If its longitudinal coordinate is measured to be 2,729,402.47 meters by the same principle, record .

[0137] Parameter Acquisition steps: is the speed of the emergency vehicle, which needs to be continuously detected by an on-vehicle speed measurement device. For example, during a patrol of an emergency vehicle, its speed is recorded as 54 kilometers per hour, then the corresponding meters per second, record .

[0138] Parameter Acquisition steps: represents the current traveling angle of the emergency vehicle, which is usually defined with reference to the interval from 0 degrees to 360 degrees in the due north direction. To obtain the angle, the heading information can be directly read from the vehicle's GPS / IMU integrated device, or detected by an on-vehicle compass after calibration. If the IMU solution is used, the data of the accelerometer and gyroscope need to be fused to obtain the heading angle. For example, detection shows that the heading angle of the emergency vehicle at a certain moment is 90 degrees, that is, traveling eastward, then .

[0139] Parameter Acquisition steps: is the time increment from the current moment to the time when the emergency vehicle is expected to reach a certain relative position. The monitoring personnel first locate the shortest passable path between the relative position and the vehicle's current position, and calculate it in segments by combining the vehicle speed and road conditions. The total time is obtained by adding the time consumed for each segment, that is . For example, if the total distance is 1800 meters, the speed is 15 meters per second and there is no blockage on the road section, then seconds. When it is found in the previous actual measurement that the road congestion causes the actual passing time of the vehicle to be prolonged, the speed value of this congested section is adjusted to a lower value, and finally the average value is taken to obtain the accurate .

[0140] Calculation process:

[0141] The first step: Confirm the parameter values measured currently, for example , , , , meters per second, degrees, seconds;

[0142] The second step: Calculate the difference in the plane distance between the two vehicles, ;

[0143] The third step: Take the square root of the result and divide by , , and then seconds;

[0144] Step 4: Calculate the angle increment, seconds;

[0145] Step 5: Add the step-by-step results to obtain seconds;

[0146] The result shows that when the coordinate gap and speed parameters between the emergency vehicle and the surrounding driver vehicles meet the above values, the warning system will send an avoidance notice to the driver about 251 seconds after the current moment. If it is found in further detection that the speed or traveling angle of the emergency vehicle changes, all parameters should be read again and this formula should be calculated again to obtain the updated notice time.

[0147] Based on the time of sending the notice, read the previously calculated ones to determine the specific notice timing. First, check various receiving methods registered in the current command center, such as in-vehicle displays, mobile terminal prompt sounds or flashing lights, etc. Locate the communication method of this vehicle item by item in the stored driver information and select the available reminder method according to the safety restrictions during vehicle driving. For example, a high-brightness flashing indicator light will be used as an intuitive warning measure, and then combined with the traffic conditions, mark the driver's vehicle with a countdown in seconds. If it is close to 0 seconds, immediately send a warning signal through sound and light. At the same time, compare and verify the lane-changing or deceleration strategies that the driver may take within a previously set range. This range is jointly formulated by vehicle manufacturers and traffic management departments and recorded in the road safety strategy table, including a feasible speed range from 5 km / h to 80 km / h and a safe lane-changing range from 0 m to 100 m and continuously monitored within the same time period. When it is between 10 seconds and 30 seconds, mark it as entering the "about to remind" state. If it is less than 10 seconds, convert it to the "urgent reminder" state and update the reminder light frequency again. Identify the position of each vehicle in the vehicle fleet item by item according to the above steps and conduct a comprehensive comparison in combination with information such as the road design speed limit, intersection signal time periods, and traffic restriction measures. Mark the records that do not exceed the reasonable range or do not trigger the danger warning threshold as the normal state. If the system detection result exceeds the specified speed limit or the vehicle cannot avoid, summarize and mark the concerned records again. Finally, summarize the notice sending time, the relative position of the vehicle, and the number of one or more prompts to form the driver reminder notice result.

[0148] The steps to obtain the real-time feedback result of the traffic status are as follows:

[0149] Based on the results of environmental adaptability signal adjustment and the optimization results of emergency vehicle passage, extract the signal adjustment duration, signal switching times, and emergency vehicle passage periods at each intersection, and count the average vehicle passing volume within each period to establish traffic signal adjustment analysis data;

[0150] Based on the traffic signal adjustment analysis data, judge the traffic flow differences at each intersection, analyze the traffic efficiency of different sections, identify abnormal congestion situations, and establish real-time traffic status feedback results.

[0151] Specifically, based on the previously obtained results of environmental adaptability signal adjustment and the optimization results of emergency vehicle passage, first read the signal adjustment duration within the corresponding period at each intersection from the records and mark the actual effective moments of these durations. Then, take the signal switching times as another statistical indicator and summarize them one by one. For example, number all the traffic light conversion events generated at each intersection within 24 hours and compare them with the previously set reference value. This reference value is the average daily switching times obtained by summing up all the conversions within 30 consecutive days and then dividing by 30, which is used to judge whether the current intersection has frequent or infrequent switching. Mark all those with more than 5 times higher than this reference value as high-switching intersections and pay attention to them, and mark those with 5 times lower than this reference value as low-switching intersections. In this way, gradually form a list of signal switching times. Subsequently, retrieve the information of the emergency vehicle passage period within the same time period, correspond the start passing time and the end passing time of the emergency vehicle passing through each intersection and record the seconds or minutes occupied by the passage. Combine this set of emergency vehicle periods with the previously summarized signal adjustment duration and switching times for interval overlap analysis. For the overlapping periods, continue to compare with the lane traffic conditions recorded by the vehicle monitoring equipment, mark the eligible passing periods in segments, and count the average vehicle passing volume within each segment. Obtain the total number of passing vehicles per minute or per hour by counting the vehicle data detected by the camera. For example, for the statistics from 23:00 at night to 1:00 in the early morning, the hourly distinction can be used to observe the total traffic volume, and for 7:00 to 9:00 during the day, the 30-minute distinction can be used to grasp the commuting peak situation. Compare the passing values obtained from these segment statistics with the previously set traffic flow range. For example, define this range as 100 vehicles per hour to 800 vehicles per hour. Mark all the periods exceeding 800 vehicles as saturated periods in the table and make additional explanations in combination with the emergency vehicle passing information. Finally, uniformly arrange the signal adjustment duration, signal switching times, and emergency vehicle passage periods at all intersections, as well as the corresponding average vehicle passing volume, and sort them with the intersection number and time as the index to form a multi-dimensional data list and mark keyword fields such as time, passing volume, and switching frequency, and finally obtain the traffic signal adjustment analysis data.

[0152] Based on the traffic signal adjustment analysis data, first conduct a horizontal comparison of the average vehicle passing volume at each intersection during the same time period and record the flow difference value. For example, set a reference volume range of 200 vehicles per hour to 600 vehicles per hour to identify whether the traffic flow is above or below the normal level. Mark the intersections with a traffic flow above 600 vehicles per hour as having a high flow rate and compare the data of the adjacent sections to determine whether it is a local concentration or a large-scale traffic jam. Mark the records with a traffic flow below 200 vehicles per hour as having a low flow rate and refer to the traffic conditions of adjacent intersections to confirm whether there is a regional vacancy or other reasons leading to an extremely low traffic flow. Subsequently, conduct a comprehensive summary based on the signal switching situation and the passing situation of emergency vehicles at each intersection. When it is found that the traffic flow at this intersection continuously exceeds 600 vehicles per hour during certain periods and the vehicle queue observation exceeds the preset threshold, it is regarded as having an obvious traffic jam. Set the threshold at the time when the end of the queue reaches 30 meters downstream of this intersection every day and lasts for more than 15 minutes to establish a mark. If the data shows similar characteristics after comparing with the adjacent time periods of the same section or other intersections in the same area again, it is determined that the traffic efficiency here has decreased significantly and there is an abnormal congestion, and it is added to the key viewing list. After completing the traffic flow statistics for each intersection and different sections, all the results are matched with the signal adjustment analysis data for the corresponding time periods, so as to lock the time period and location of the congestion. The remaining records within the reference range are uniformly regarded as the normal distribution and marked. Finally, summarize all the high flow difference, abnormal congestion, and normal flow time periods and list them separately in combination with the intersection numbers, organize the overall traffic flow at the same time for this section and extract the most obvious difference value section to form the real-time traffic status feedback result.

Claims

1. A traffic control system for real-time data transmission, characterized in that, The system comprises: The traffic data collection module installs cameras at traffic intersections to monitor vehicle arrival timing and queue length data, synchronizes and stores the collected data in real time, and generates a real-time traffic data stream; uses the real-time traffic data stream for analysis to identify traffic peak and valley periods, and generates traffic flow peak and valley identification results; The signal optimization decision module dynamically optimizes and calculates the green light time of each intersection based on the traffic flow peak and valley identification results, and calculates and generates an optimized green light time plan; combines the optimized green light time plan with the ambient temperature and rainfall, adjusts the red light extension strategy, and generates an environmentally adaptable signal adjustment result; The emergency response coordination module receives the location and expected route information of the emergency vehicle in real time, dynamically adjusts the traffic light based on the expected route information and the environmental adaptability signal adjustment result, turns on the green light for the emergency vehicle first, and generates an emergency vehicle traffic optimization result; based on the location and expected route information of the emergency vehicle, sends an emergency vehicle approach notification to surrounding drivers to improve road safety and generate a driver prompt notification result; The comprehensive application module of traffic management performs real-time information feedback based on the environmental adaptability signal adjustment results and the emergency vehicle passage optimization results, and generates real-time feedback results of traffic status.

2. The traffic control system for real-time data transmission according to claim 1, wherein The steps of acquiring the real-time traffic data stream are: Install cameras at traffic intersections to obtain vehicle arrival timing and queue length data. Based on the image data obtained by the camera, analyze the arrival time, stay time, vehicle spacing and queue status of each vehicle, match the motion trajectory between adjacent frames, and calculate the cumulative number of vehicles in different time periods to establish the original traffic monitoring data set; Based on the original traffic monitoring data set, the arrival timestamps of each vehicle are parsed, the time deviations between different cameras are corrected, and a real-time traffic data stream is formed.

3. The traffic control system for real-time data transmission according to claim 1, characterized in that The steps for obtaining the traffic flow peak and valley recognition results are as follows: Based on the real-time traffic data stream, extract the hourly vehicle flow data to obtain hourly flow analysis results; Based on the hourly flow analysis results, the flow change rate is calculated using the following formula: Among them, F t is the flow rate in the current hour, F t+1 is the flow rate in the next hour, a t is the acceleration at time t, L t is the average vehicle length in the current period, P t is the vehicle density in the current period, and ΔF is the flow rate change rate; Based on the flow rate change rate, the upper and lower limits of the threshold are set to define the peak and valley, determine the peak and valley periods of the traffic flow, and obtain the traffic flow peak and valley identification results.

4. The traffic control system for real-time data transmission according to claim 1, characterized in that, The steps for obtaining the optimized green light time scheme are: Based on the traffic flow peak and valley identification results, the flow data of each intersection during peak hours and valley hours are extracted, and the average traffic speed, lane capacity, signal cycle, lane occupancy rate and number of pedestrian crossing requests of each intersection in different time periods are segmented and counted to form a traffic flow period characteristic data set; Based on the traffic flow period characteristic data set, the green light time of each intersection is calculated using the following formula: Among them, T g is the green light time, S c is the number of current lanes at this intersection, V m is the average flow velocity of the traffic flow, D t is the number of vehicles passing through per unit time, Q p is the queue length at this intersection, L a is the average vehicle length at this intersection, P r is the number of pedestrian crossing requests, O c is the lane occupancy rate of the current intersection, W s is the average waiting time at this intersection, G v is the average traffic flow within the current green light cycle; Based on the green light time, the green light time parameters of the signal light are adjusted to form an optimized green light time solution.

5. The traffic control system for real-time data transmission according to claim 1, wherein The steps of obtaining the environmental adaptability signal adjustment result are: Based on the optimized green light time scheme, the red light extension time is calculated using the following formula: Among them, T r is the red light extension time, R d is the real-time rainfall, H m is the real-time road surface humidity, L v is the real-time visibility at the current intersection, D p is the number of pedestrians waiting during the red light period, V c is the detected average vehicle speed, B r is the average braking distance of vehicles monitored at this intersection; Based on the red light extension time, the signal light cycle is adjusted to generate an environmental adaptability signal adjustment result.

6. The traffic control system for real-time data transmission according to claim 1, wherein The steps for obtaining the optimized results of emergency vehicle passage are as follows: Receive the location and expected travel route information of the emergency vehicle, parse the current coordinates, moving speed, and expected arrival time of the emergency vehicle, and generate emergency vehicle travel path prediction data in combination with the real-time road conditions; Based on the emergency vehicle travel path prediction data, judge the expected signal status when the emergency vehicle passes through each intersection, and generate a dynamic signal light adjustment plan; Based on the dynamic signal light adjustment plan, adjust the signal control parameters of each intersection along the way to generate the optimized results of emergency vehicle passage.

7. The traffic control system for real-time data transmission according to claim 1, characterized in that, The steps for obtaining the driver prompt notification results are as follows: Based on the location and expected travel route information of the emergency vehicle, calculate the time for sending the notification, and the calculation formula is: Among them, T notify is the time to send the notice, x ev and y ev are the current position coordinates of the emergency vehicle, x drv and y drv are the position coordinates of the surrounding driver vehicles, v ev is the speed of the emergency vehicle, θ ev is the current traveling angle of the emergency vehicle, and Δt is the time increment from the current moment to the time when the emergency vehicle is expected to reach the relative position; Based on the time for sending the notification, send a warning message to surrounding drivers, including the expected arrival time of the emergency vehicle and the avoidance strategy, to generate the driver prompt notification results.

8. The traffic control system for real-time data transmission according to claim 1, wherein The steps for obtaining the real-time traffic status feedback results are as follows: Based on the environmental adaptability signal adjustment results and the optimized results of emergency vehicle passage, extract the signal adjustment duration, signal switching times, and emergency vehicle passage time periods of each intersection, and count the average vehicle passing volume within each time period to establish traffic signal adjustment analysis data; Based on the traffic signal adjustment analysis data, judge the traffic flow differences at each intersection, analyze the traffic efficiency of different sections, identify abnormal congestion situations, and establish the real-time traffic status feedback results.

Images