Intelligent traffic control method and system for asymmetric intersection

Abstract

The invention relates to an asymmetric intersection signal intelligent control method and system, and belongs to the technical field of intelligent traffic control. According to the method, traffic flow data of a main road and branch roads are collected in real time, data processing and analysis technologies are utilized, timing of signal lamps is dynamically adjusted, and especially before a green light signal of the main road is finished, if no vehicle waits on the branch roads, the green light time of the main road is automatically prolonged, so that the passing efficiency of the intersection is improved. The system comprises a data detection unit, a data processing unit, a signal control unit and a data storage unit, can monitor traffic flow and vehicle queuing conditions in real time, and performs signal timing adjustment according to real-time data. According to the invention, the vehicle waiting time can be effectively reduced, the traffic flow is improved, the congestion is reduced, and the traffic safety and passing efficiency are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of traffic control, and particularly relates to an intelligent traffic control method and system for an asymmetric intersection, which is applicable to the intelligent control of urban traffic signal lights, especially for intersections where the traffic flows of the main road and the branch road are asymmetric. Background Art

[0002] With the acceleration of the urbanization process, the problem of traffic congestion has become increasingly serious, especially at intersections, where problems such as long waiting times for vehicles and low traffic efficiency are particularly prominent. Traditional traffic signal control methods usually adopt fixed signal timing schemes and cannot be dynamically adjusted according to real-time traffic flows. As a result, when the traffic flow changes greatly, the signal timing of the traffic lights cannot meet the actual needs, resulting in a waste of traffic resources and a reduction in traffic efficiency.

[0003] In urban traffic, an asymmetric intersection refers to an intersection with asymmetry in traffic flow, geometric layout, or signal control. The signal control of an asymmetric intersection (where the main road intersects with the branch road) is a complex and important issue. Traditional signal control methods usually adopt fixed-time allocation and cannot dynamically adjust the signal time according to real-time traffic flows. This fixed mode is prone to causing traffic congestion during peak hours, while it may waste green light time and reduce traffic efficiency during off-peak hours. In addition, although existing intelligent traffic systems can detect vehicle flows, in the application of asymmetric intersections, they often cannot effectively handle the dynamic relationship between the main road and the branch road. For example, when there are no vehicles on the branch road, the green light time of the main road ends and switches to the green light of the branch road, while the vehicles on the main road queue up and wait, extending the green light duration of the main road. And when there are vehicles on the branch road, the signal switch may not be timely enough, resulting in too long waiting times for the vehicles on the branch road.

[0004] Although existing intelligent traffic control systems can make certain degrees of dynamic adjustments according to traffic flows, in asymmetric intersections, the traffic flow differences between the main road and the branch road are relatively large, and existing systems often cannot effectively handle this asymmetric traffic condition, resulting in too long waiting times for the vehicles on the main road and low traffic efficiency for the vehicles on the branch road. Therefore, there is an urgent need for an intelligent control method that can dynamically adjust signal timing according to real-time traffic flows to improve the traffic efficiency of intersections and reduce vehicle waiting times. Developing an intelligent control method and system that can dynamically adjust signal time is of great significance for improving the traffic efficiency and safety of asymmetric intersections. Summary of the Invention

[0005] The invention provides an intelligent traffic control method and system for an asymmetric intersection. By collecting traffic flow data of the main road and the branch road in real time and using data processing and analysis techniques, the signal timing is dynamically adjusted. Especially before the end of the green signal of the main road, if there are no vehicles waiting on the branch road, the green time of the main road is automatically extended to improve the traffic efficiency of the intersection. The system includes a data detection unit, a data processing unit, a signal control unit, and a data storage unit, which can monitor traffic flow and vehicle queuing conditions in real time and adjust signal timing according to real-time data. The invention improves the traffic capacity and operation efficiency of the intersection by monitoring traffic flow and signal control, can effectively reduce vehicle waiting time, increase traffic flow, reduce congestion, and enhance traffic safety and traffic efficiency.

[0006] The present invention is realized through the following technologies: The intelligent control method is realized through traffic data collection, data processing and analysis, signal timing decision-making, signal control, monitoring and adjustment strategies. When the green signal of the main road at the intersection is about to end and there are no vehicles on the intersecting branch road, signal adjustment and timing control are automatically carried out to extend the green signal duration of the main road and improve the traffic efficiency of the intersection.

[0007] For the traffic data collection mentioned above, traffic flow detectors and vehicle detectors are used to collect traffic flow and vehicle queue length data of the main road and the branch road in real time as the basis for signal timing.

[0008] For the data processing and analysis, the collected traffic data is transmitted to the intelligent traffic control system, and the system processes and analyzes the data. By analyzing traffic flow and vehicle queuing conditions, the traffic conditions of the main road and the branch road are judged.

[0009] For the signal timing decision-making, according to the results of data analysis, the intelligent traffic control system makes signal timing decisions according to preset algorithms and strategies. The system judges that there are no vehicles on the intersecting branch road before the end of the green signal of the main road and extends the green time of the main road.

[0010] For the signal control, the intelligent traffic control system transmits the signal timing decision results to the signal control equipment to control the signal lights of the main road and the branch road. The system extends the green time of the main road so that the vehicles on the main road have enough time to pass, while the waiting time of the branch road signal lights will be extended accordingly.

[0011] For the monitoring and adjustment, the intelligent traffic control system continuously monitors traffic flow and vehicle queuing conditions and makes adjustments according to real-time data. If the traffic conditions of the main road change or there are vehicles waiting on the branch road, the system makes a new signal timing decision according to the new data to ensure traffic efficiency and safety.

[0012] The specific control method is: When the basic green time of the main road T mbBuffer time before the end T bf Within this time, if there are still vehicles in the main road waiting area and the system detects that there are vehicles waiting in the branch road waiting area, the basic green light time of the main road T mb After it ends, the main road turns red and the branch road turns green. If there are no vehicles waiting in the branch road waiting area, the system automatically controls the extended green light time of the main road ΔT ; During the extended green light time of the main road ΔT Buffer time before the end T bf Within this time, if there are still vehicles in the main road waiting area and the system detects whether there are vehicles driving into the branch road waiting area, if there are vehicles driving into the branch road waiting area, the main road continues to complete the extended time ΔT After the green light signal, it turns red and switches to the basic green light time of the branch road T sb signal. If the extended time ΔT Buffer time before the end T bf Within this time, if there are still no vehicles driving into the branch road waiting area and there are still vehicles in the main road waiting area, the main road extends for another extended time ΔT ; During the second extended time ΔT If there are vehicles driving into the branch road waiting area, the main road turns into a yellow light for buffer reminder and then turns red, switching to the basic green light time of the branch road T sb signal. If there are no vehicles driving into the branch road waiting area during the second extended time ΔT the main road ends the extended time ΔT The green light signal turns into a yellow light for buffer (3 - 5 seconds) reminder and then turns red, switching to the basic green light time of the branch road T sb signal; If during the second extended time of the main road ΔT the system detects that there are pedestrians waiting to cross the main road in the branch road, the main road turns into a yellow light for buffer reminder and then turns red, switching to the basic green light time of the branch road T sb signal; If during the first and second extended times ΔT all the vehicles in the main road waiting area have driven out, the main road turns red and switches to the basic green light time of the branch road T sb signal; After the basic green light time of the branch road T sb ends, it switches back to the basic green light time of the main road T mb and enters the next cycle.

[0013] The extended green light time Δ of the main road mentioned above T ​, by performing the Discrete Fourier Transform (DFT) on the real-time collected main road and branch road traffic data, the traffic flow data is transformed from the time domain to the frequency domain to detect the periodicity in the data, for analyzing the periodic characteristics of the traffic flow data, and thereby determining the extended time of traffic signal control; the specific steps are as follows: 1) Data normalization, scaling the flow data to the range [0, 1] to improve the convergence speed and prediction accuracy of the model. The formula is as follows: , where: Q ′ is the original traffic flow data, Q min is the minimum value in the data, Q max is the maximum value in the data; 2) Perform the discrete Fourier transform on the normalized flow data Q to obtain the frequency domain signal F ( u ). The Fourier transform formula is: , where: Q n is the traffic flow data at time point n , N is the total number of data points, u is the frequency index, ([[]] u = 0, 1, 2, …, N−1), F ( u ) is the transformation result in the frequency domain; 3) Calculate the amplitude ∣ F ( u )∣; find the frequencies u corresponding to the first few peaks with the largest amplitudes. When the peak appears at u = k , the corresponding period is: T = N / k , k is the peak position of the frequency domain index; select the periods corresponding to the first few peaks as candidate signal control periods; 4) Determine the period T of the traffic flow through the Fourier transform, and adjust the extended time of the signal control accordingly; According to the basic time T mb of the main road green light, calculate the extended time Δ T : , where: Q h ​The historical traffic flow data of the main road, including the vehicle flow, time distribution, traffic patterns, and vehicle type distribution of the main road and branch roads, Q t is the flow threshold, used to determine whether the current flow requires adjusting the signal time, α and β is the adjustment coefficient, used to control the range of Δ T and α control the increase and decrease amplitude of the extended time, β is the basic extended time.

[0014] The historical traffic flow data of the main road described above Q h is calculated using the historical average model, which uses the flow data within the historical time to predict the future flow. The calculation formula is: , where: t is the time, Q ( t ) is the actual observed flow at time t , γ is the smoothing coefficient, and its value range is [0, 1], used to adjust the weight of historical data.

[0015] The basic green light time of the main road described above T mb is calculated by the following formula: , where: Q k is the maximum vehicle flow to pass through, h is the saturated headway (unit: second), indicating the average time interval for vehicles to pass through the intersection in the saturated state, t s is the start-up loss.

[0016] The basic green light time of the branch road described above T sb is calculated by the following formula: , where, T min is the minimum green light time of the branch road, T d is the time difference from the time point when a vehicle on the branch road is detected to enter the waiting area to the current time, T bf is the buffer time.

[0017] The present invention provides a system for an intelligent traffic control method for an asymmetric intersection, characterized by including: A data detection unit, which is used to detect the vehicles driving into the waiting area at each intersection and the pedestrians waiting to cross the road at the intersection, collect the traffic data of the main road and the branch road in real time, and monitor the vehicle flow, time distribution, traffic pattern and vehicle type distribution of the main road and the branch road; A data processing unit, according to the traffic data of the data detection unit, preprocesses, extracts features and optimizes parameters for the traffic flow data of the main road and the branch road. Through Fourier transform, periodic features are extracted from complex time-domain signals, and the periodicity of traffic flow data, the periodic changes during the morning rush hour, evening rush hour and low peak hours, the flow differences between weekdays and weekends, and the flow changes in different seasons or holidays are analyzed. The analysis results are used as the basis for intelligent signal adjustment; The system uses historical traffic data to self-learn and train a regression model to calculate α 、 β and Q t the optimal values; Adopt the control algorithm of the embedded system to calculate the function f ( Q h ) to calculate the extended time Δ T of the main road, generate signal control instructions, and calculate the intersection timing plan and real-time signal adjustment parameters according to the data of the vehicles in the waiting area and the pedestrians waiting to cross the road detected at each intersection; A signal control unit, according to the timing plan and real-time signal adjustment parameters of the data processing unit, adjusts the extended time ΔT of the main road in real time, controls the traffic lights of the main road and the branch road, realizes the switching of traffic signals, and displays the remaining time of the traffic lights through a digital tube or a display screen; A data storage unit, which records traffic flow data and system operation logs, stores the collected traffic flow data and system status information locally or in the cloud, analyzes the stored data, generates a traffic flow report, and provides decision-making support for traffic management.

[0018] The data detection unit sets vehicle detection sensors at the entrance and exit of the waiting area of each lane at each intersection. The vehicle detection sensors include geomagnetic sensors for detecting the presence and passing of vehicles, radar sensors for measuring vehicle speed and queue length, and high-definition cameras for license plate recognition and vehicle type classification; Infrared thermal imaging sensors and image collectors are set in the pedestrian waiting area for crossing the road. The infrared thermal imaging sensors sense the presence of pedestrians in the waiting area for crossing the road. The image collectors use YOLO for image segmentation, feature extraction, bounding box prediction, and non-maximum suppression. Through computer vision and deep learning algorithms, the image collectors use the YOLO algorithm to detect pedestrians in real time and identify the targets waiting to cross the road.

[0019] The described data processing unit is an embedded processing system, including a core control module, a user interaction module, a communication module, and a fault detection and alarm module; the core control module is the center of the entire traffic signal control system, dynamically adjusts the duration of the signal lights according to traffic flow data and preset control algorithms, receives traffic flow data from sensors, analyzes and processes it, monitors the system operation status, and takes measures when abnormalities are found to ensure the stable operation of the system; the user interaction module conducts system settings and adjustments through buttons, knobs or touch screens, uses an LCD or OLED display screen to display the system status and signal light cycle information in real time, provides an operation interface for maintenance personnel and administrators, and is used for system configuration and status monitoring; the communication module is responsible for realizing data transmission and instruction interaction between systems, realizing synchronization and coordination between signal light controllers, conducting data transmission with the traffic management center, and supporting remote monitoring and management; the fault detection and alarm module is responsible for monitoring the operation status of the system, promptly detecting and handling abnormalities, detecting key parameters of the system, and when abnormalities occur, giving an audible and visual alarm or remotely notifying maintenance personnel.

[0020] The beneficial effects of the present invention are as follows: By real-time monitoring of traffic flow and dynamically adjusting the signal timing, especially before the end of the green signal on the main road, if there are no vehicles waiting on the branch road, the green time of the main road is automatically extended, reducing the vehicle waiting time and improving the traffic efficiency at intersections. Through intelligent signal control, it can effectively reduce the vehicle queue length and the probability of traffic congestion, especially during peak hours, and can significantly improve the traffic capacity of traffic flow. By real-time monitoring and adjusting the signal timing, it can effectively avoid traffic accidents caused by unreasonable signal timing and improve traffic safety. The system can self-learn and adjust according to real-time traffic flow data, adapt to traffic flow changes in different time periods and different seasons, and has strong adaptability. Brief Description of the Drawings

[0021] Figure 1 Block diagram of the intelligent traffic control method of the present invention; Figure 2 Plan view of the asymmetric intersection of the present invention; Figure 3 Flow chart of the intelligent traffic control of the present invention; Figure 4 Traffic signal timing diagram of Embodiment 1; Figure 5 Traffic signal timing diagram of Embodiment 2; Figure 6 Traffic signal timing diagram of Embodiment 3.

[0022] In the figure: 1 - geomagnetic sensor, 2 - radar sensor, 3 - high-definition camera, 4 - infrared thermal imaging sensor, 5 - image collector. Detailed Implementation Manner

[0023] For better understanding of the present invention by those skilled in the art, in combination with Figures 1 - 6 the present application is further described. In the description of this specification, the orientation or positional relationship indicated by terms such as "upper", "lower", "left", "right", etc. is based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship commonly understood by those skilled in the art. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the components or parts referred to must have a specific orientation, be constructed and operated in a specific orientation. The content mentioned in the implementation manner does not limit the present invention.

[0024] The intelligent traffic control method for an asymmetric intersection is realized through traffic data collection at the intersection, data processing and analysis, signal timing decision-making, signal control, monitoring and adjustment strategies, as shown in Figure 1 the block diagram of the intelligent traffic control method. The specific solution is as follows: By collecting traffic data in all directions of the intersection, when the green light signal of the main road at the intersection is about to end and there is no vehicle on the intersecting branch road, the system automatically adjusts the signal timing control, extends the green light signal duration of the main road, and improves the traffic efficiency of the intersection.

[0025] Traffic data collection: Use traffic flow detectors and vehicle detector devices to collect the traffic flow and vehicle queue length data of the main road and the branch road in real time as the basis for signal timing.

[0026] Data processing and analysis: Transmit the collected traffic data to the intelligent traffic control system. The system processes and analyzes the data, and judges the traffic conditions of the main road and the branch road by analyzing the traffic flow and vehicle queue situation.

[0027] Signal timing decision-making: According to the results of data analysis, the intelligent traffic control system makes signal timing decisions according to preset algorithms and strategies. When the system judges that there is no vehicle on the intersecting branch road before the green light signal of the main road ends, it extends the green light time of the main road.

[0028] Signal control: The intelligent traffic control system transmits the signal timing decision result to the signal control device to control the signal lights of the main road and the branch road. The system extends the green light time of the main road so that the vehicles on the main road have enough time to pass, while the signal lights of the branch road will correspondingly extend the waiting time.

[0029] Monitoring and adjustment: The intelligent traffic control system continuously monitors the traffic flow and vehicle queue situation and makes adjustments according to the real-time data. If the traffic conditions of the main road change or there are vehicles waiting on the branch road, the system makes a new signal timing decision according to the new data to ensure traffic efficiency and safety.

[0030] The plan view of the asymmetric intersection is shown in Figure 2 ,Figure 2 is only a typical example of an asymmetric intersection, Figure 2 where intersection 1 and intersection 2 are the main roads, and intersection 3 and intersection 4 are the branch roads. The specific control method of the intelligent transportation control system is as follows: Set the basic green light time for the main road T mb and the basic green light time for the branch road T sb , the basic green light time for the main road T mb is calculated by the following formula: , where: Q k is the maximum traffic volume to pass through, h is the saturated headway (unit: second), indicating the average time interval for vehicles to pass through the intersection in the saturated state, t s is the start-up loss. The basic green light time for the main road T mb includes the straight and left-turn times of intersection 1 and intersection 2 in the middle. The basic green light time for the branch road T sb is calculated by the following formula: , where, T min is the minimum green light time for the branch road, T d is the time difference from the time point when a vehicle on the branch road is detected to enter the waiting area to the current time, T bf is the buffer time. The basic green light time for the main road T mb and the basic green light time for the branch road T sb can be fixed during a day, or can be calculated as different times according to the traffic volume during the morning and evening rush hours and normal periods.

[0031] Detect the real-time traffic conditions at the intersection through the data detection unit set at the intersection. When there are still vehicles in the main road waiting area within the buffer time T mb before the end of the basic green light time for the main road T bf (3 - 5 seconds), detect whether there are vehicles waiting in the branch road waiting area. If there are vehicles waiting in the branch road waiting area, after the basic green light time for the main road T mb ends, the main road turns red and the branch road turns green. If there are no vehicles waiting in the branch road waiting area, the system automatically controls the extension time of the green light for the main road ΔT . During the extension time of the green light for the main roadΔ T Buffer time before ending T bf Within (3 - 5 seconds), if there are still vehicles in the main road waiting area, the system detects whether there are vehicles entering the waiting area from the branch road. If there are vehicles entering the waiting area from the branch road, the main road continues to complete the extended time ΔT The green light signal turns into a yellow light buffer (3 - 5 seconds) and then turns into a red light, switching to the basic green light time of the branch road T sb Signal; if the extended time ΔT Buffer time before ending T bf Within (3 - 5 seconds), if there are no vehicles entering the waiting area from the branch road and there are still vehicles in the main road waiting area, the main road extends for another extended time Δ T ; During the second extended time ΔT If there are vehicles entering the waiting area from the branch road, the main road turns into a yellow light buffer (3 - 5 seconds) and then turns into a red light, switching to the basic green light time of the branch road T sb Signal; if there are still no vehicles entering the waiting area from the branch road during the second extended cycle, the main road ends the extended time ΔT After the green light signal turns into a red light, it switches to the basic green light time of the branch road T sb Signal; if during the second extended time of the main road ΔT The system detects that there are pedestrians waiting to cross the main road from the branch road. The main road turns into a yellow light buffer and then turns into a red light, switching to the basic green light time of the branch road T sb Signal; if during the first and second extended times ΔT All the vehicles in the main road waiting area have driven out. The main road turns into a red light and switches to the basic green light time of the branch road T sb Signal; After the basic green light time of the branch road T sb ends, it switches back to the basic green light time of the main road again T mb Enter the next cycle. The block diagram of the intelligent traffic control process of the present invention is shown in Figure 3 .

[0032] The extended green light time Δ of the main road T , is calculated and determined according to the number of vehicles detected in real time in the main road waiting area. It is a dynamic time, and the extended green light time Δ of the green light phase of the main road in each signal cycle T is different. The extended green light time Δ of the main road T, by performing Fourier transform (DFT) on the real-time collected main road and branch road traffic data, the traffic flow data is transformed from the time domain to the frequency domain, thereby detecting the periodicity in the data for analyzing the periodic characteristics of the traffic flow data, and determining the extended green time Δ of the main road based on this. T , and the specific steps are as follows: 1) First, normalize the real-time collected main road and branch road traffic data. Remove outliers and fill in missing values before normalization. Scale the flow data to the range [0, 1] to improve the convergence speed and prediction accuracy of the model. Use min-max normalization (Min-Max Scaling), and the formula is as follows: , where: Q ′ is the original traffic flow data, Q min is the minimum value in the data, Q max is the maximum value in the data; 2) Perform discrete Fourier transform on the normalized flow data Q to obtain the frequency domain signal F ( u ). The Fourier transform formula is: , where: Q n is the traffic flow data at time point n , N is the total number of data points, u is the frequency index, ([[]] u = 0, 1, 2, …, N−1), F ( u ) is the transformation result in the frequency domain; 3) Calculate the amplitude ∣ F ( u )∣ of the frequency domain signal. The frequency domain signal F ( u ) is a complex number, and its amplitude AMP( u ) can be obtained by calculating the modulus of the complex number. For the complex number F ( u ) = a + b , its amplitude is: , where: a is F ( u )'s real part, b is F ( u ​) The imaginary part. Find the frequencies corresponding to the first few peaks with the largest amplitudes u When the peak appears at u = k , the corresponding period is: T = N / k , k is the peak position of the frequency domain index; Select the periods corresponding to the first few peaks as the candidate signal control periods; 4) Determine the period of traffic flow through Fourier transform T , and adjust the extended time of signal control accordingly. According to the basic time of the main road green light T mb , calculate the extended time Δ T : , Where: Q h is the historical traffic flow data of the main road, including the vehicle flow, time distribution, traffic pattern, and vehicle type distribution of the main road and the branch road, Q t is the flow threshold, used to determine whether the current flow requires signal time adjustment, α and β are adjustment coefficients, used to control the range of Δ T , α control the increase and decrease amplitude of the extended time, β is the basic extended time, and the period T is used for dynamic adjustment of α and β .

[0033] Historical traffic flow data of the main road Q h is calculated using the historical average model, and the flow data within the historical time is used to predict the future flow. The calculation formula is: , Where: t is the time, Q ( t ) is the actual observed flow at time t , γ is the smoothing coefficient, and its value range is [0, 1], used to adjust the weight of historical data. The value of the smoothing coefficient γ is dynamically adjusted according to the traffic period. During the morning peak period (7:00~9:00) and the evening peak period (17:00~19:00), γ = 0.8 to enhance the weight of historical data; During the flat peak period (9:00~17:00), γ = 0.6; During the night period (22:00~6:00), γ = 0.3 to enhance the sensitivity of real-time data.

[0034] The present invention adjusts signal timing according to real-time data. According to the above method, within each basic green time of the main road T mb , the extended green time Δ of the main road is calculated based on the collected traffic data of the main road and the branch road T . Within a time period, the basic green time of the main road T mb is fixed, and the extended green time Δ T is dynamically variable. By means of traffic flow and signal control, the passing capacity and operation efficiency of the intersection are improved, the waiting time of vehicles can be effectively reduced, the traffic flow can be increased, congestion can be reduced, and traffic safety and passing efficiency can be enhanced.

[0035] The present invention provides a system for an intelligent traffic control method for an asymmetric intersection, which is characterized by comprising: a data detection unit, a data processing unit, a signal control unit, and a data storage unit.

[0036] Data detection unit, which is used to detect vehicles driving into the waiting area at each intersection and pedestrians waiting to cross the road at the intersection, collect traffic data on the main road and branch road in real time, and monitor the vehicle flow, time distribution, traffic patterns and vehicle type distribution on the main road and branch road. The data detection unit sets vehicle detection sensors at the entrance and exit of the waiting area of each lane at each intersection. The vehicle detection sensors include a geomagnetic sensor 1, which is used to detect the presence of vehicles. By counting the vehicles entering and leaving the waiting area, the system calculates the number of vehicles in the waiting area for intelligent control of lane signals. A radar sensor 2, which is used to measure vehicle speed and queue length, and can calculate the time required to leave the waiting area. A high-definition camera 3, which is used for license plate recognition and vehicle type classification. If a vehicle changes lanes on a solid line, it can take pictures and transmit them to the management department as a basis for correcting violations. The high-definition camera 3 can also be used as a sensor to detect vehicles entering and leaving the waiting area. An infrared thermal imaging sensor 4 and an image collector 5 are set in the pedestrian waiting area for crossing the main road. The infrared thermal imaging sensor 4 senses the presence of pedestrians in the pedestrian waiting area for crossing the road, and sets a pedestrian stagnation time threshold. The stagnation time threshold is preferably 5 to 10 seconds. The infrared thermal imaging sensor 4 combines the continuous frame analysis of the image collector 5. If a pedestrian stays longer than the threshold and is in the waiting area, it is determined as waiting to cross the road. If it is less than the stagnation time threshold, it is judged as a passing pedestrian and is not recorded as a pedestrian waiting to cross the road. The infrared thermal imaging sensor 4 and the image collector 5 are set on the sidewalk at the main road intersection and should be able to cover the pedestrian waiting area for crossing the main road. The image collector 5 uses YOLO for image segmentation, feature extraction, bounding box prediction, and non-maximum suppression. Through computer vision and deep learning algorithms, the image collector 5 uses the YOLO algorithm to detect pedestrians in real time and identify the targets waiting to cross the road. YOLO is a real-time object detection algorithm based on deep learning. Its core function is to quickly and accurately identify and locate target objects in images or videos. YOLO transforms the object detection task into a single regression problem and directly predicts the bounding boxes and class probabilities of targets from the input image through a convolutional neural network (CNN). YOLO can quickly identify and locate various types of objects in real-time camera videos. It can not only detect the presence of targets, but also classify them, such as distinguishing cars, people, animals, etc., and can give the exact position of each target in the image, usually represented in the form of a bounding box. The infrared thermal imaging sensor 4 cooperates with the image collector 5 to identify people waiting to cross the main road and eliminate passing pedestrians.

[0037] The data processing unit preprocesses, extracts features, and optimizes parameters for the traffic flow data of the main road and the branch road according to the traffic data of the data detection unit. Through Fourier transform, it extracts periodic features from complex time-domain signals, analyzes the periodicity of traffic flow data, the periodic changes during the morning rush hour, evening rush hour, and low-traffic periods, the traffic flow differences between weekdays and weekends, and the traffic flow changes in different seasons or holidays. The analysis results are used as the basis for intelligent signal adjustment. The system uses historical traffic data to self-learn and train a regression model to calculate α , β and Q t optimal values. It uses the control algorithm of the embedded system to calculate the function f ( Q h ), calculates the extended green time Δ T of the main road, generates signal control instructions, and calculates the intersection timing plan and real-time signal adjustment parameters according to the data of the vehicles waiting in the stop area and the pedestrians waiting to cross the road at each intersection.

[0038] The data processing unit is an embedded processing system, including a core control module, a user interaction module, a communication module, and a fault detection and alarm module; the core control module is the center of the entire traffic signal control system. According to the traffic flow data and the preset control algorithm, it dynamically adjusts the duration of the signal lights, receives the traffic flow data from the sensors, analyzes and processes it, monitors the operation status of the system, and takes measures when abnormalities are found to ensure the stable operation of the system. The user interaction module performs system settings and adjustments through buttons, knobs, or touchscreens, uses an LCD or OLED display to display the system status and signal light cycle information in real time, and provides an operation interface for maintenance personnel and administrators for system configuration and status monitoring. The communication module is responsible for realizing data transmission and instruction interaction between systems, realizing synchronization and coordination between signal light controllers, transmitting data to the traffic management center, and supporting remote monitoring and management. The fault detection and alarm module is responsible for monitoring the operation status of the system, promptly detecting and handling abnormalities, detecting the key parameters of the system, and when abnormalities occur, notifying maintenance personnel through sound and light alarms or remotely.

[0039] The signal control unit, according to the timing plan and real-time signal adjustment parameters of the data processing unit, adjusts the extended green time Δ T of the main road in real time, controls the signal lights of the main road and the branch road, realizes the switching of traffic signals, and displays the remaining time of the signal lights through a digital tube or a display screen.

[0040] The data storage unit records the traffic flow data and the system operation log, stores the collected traffic flow data and system status information locally or in the cloud, analyzes the stored data, generates a traffic flow report, and provides decision-making support for traffic management.

[0041] Example 1: The main road and the branch road intersect at a crossroads. The intersection plan view is shown in Figure 1 . Intersection 1 and Intersection 2 are the main roads, and Intersection 3 and Intersection 4 are the branch roads. In the first phase, vehicles turning left at the main road intersections 1 and 2 have a green light signal duration of 25 seconds, and a 5-second yellow light signal indicates that the left-turn green light signal changes to a red light signal; in the second phase, there are more straight-going vehicles on the main road, and the green light signal duration for straight + right-turn is 45 seconds. Before the end of the 45-second green light signal, within the buffer time T bf (taking 5 seconds), the system detects that there are vehicles waiting in the branch road's waiting area. After the 5-second yellow light signal indicates the change to a red light signal after the end of the main road's straight + right-turn green light signal; in the third phase, there are relatively few straight-going, left-turning, and right-turning vehicles on the branch road. The method of straight + left + right-turn is adopted simultaneously in both directions, with left-turning vehicles giving way to straight-going vehicles. The green light signal duration is 25 seconds, and a 5-second yellow light indicates the end of the previous signal cycle and enters the next cycle. The traffic signal timing diagram of Example 1 is shown in Figure 4 .

[0042] Example 2: The intersection form is the same as that of Example 1. In the first phase, vehicles turning left at the main road intersections 1 and 2 have a green light signal duration of 25 seconds, and a 4-second yellow light signal indicates that the left-turn green light signal changes to a red light signal; in the second phase, there are more straight-going vehicles on the main road, and the green light signal duration for straight + right-turn is 45 seconds. Before the end of the 45-second green light signal, within the buffer time T bf (taking 5 seconds), there are still vehicles in the main road's waiting area, and the system detects that there are no vehicles waiting in the branch road's waiting area; in the third phase, the system automatically controls the green light signal duration for the main road's straight + right-turn to be extended by 30 seconds ΔT , during the 30-second extension of the main road's green light ΔT Before the end of the extension, within the buffer time T bf (taking 5 seconds), the system detects that vehicles have entered the branch road's waiting area. After the main road completes the 30-second extension of the main road's green light ΔT , it turns red; in the fourth phase, there are relatively few straight-going, left-turning, and right-turning vehicles on the branch road. The method of straight + left + right-turn is adopted simultaneously in both directions, with left-turning vehicles giving way to straight-going vehicles. The green light signal duration is 25 seconds, and a 4-second yellow light indicates the end of the previous signal cycle and enters the next cycle. The traffic signal timing diagram of Example 2 is shown in Figure 5 .

[0043] Example 3: The intersection form is the same as that of Example 1. The first phase and the second phase are the same as those of Example 2. Before the end of the 45-second green light signal in the second phase, within the buffer time T bfWithin (taking 5 seconds), the system detects that there are many vehicles in the straight and left-turn waiting areas at intersection 1, while there are no vehicles or fewer vehicles in the straight and left-turn waiting areas at the opposite intersection 2; in the 3rd phase, the system automatically controls the green light straight + right-turn green light signal on the intersection 1 side of the main road to be extended by 30 seconds (extension time). ΔT , the straight and left turns at intersection 2 turn into red lights, and the green light at intersection 1 is extended by 30 seconds (extension time). ΔT Buffer time before the end T bf Within (taking 5 seconds), the system detects that a vehicle has entered the waiting area on the branch road, and intersection 1 completes the 30-second extension time. ΔT After the green light signal of... (the specific number or description is missing here) for 30 seconds, the yellow light is on for 4 seconds and then prompts to turn into a red light; in the 4th phase, since there are relatively few vehicles going straight, turning left, and turning right on the branch road, a mode of simultaneous straight + left + right turns in both directions is adopted, with the left turn giving way to the straight vehicles. The green light signal duration is 25 seconds, and after the yellow light is on for 4 seconds, a signal cycle ends and enters the next cycle. The traffic signal timing diagram of Embodiment 3 is shown in Figure 6 .

[0044] For the right turn in the above embodiments, signal control is adopted. If there is a dedicated right-turn lane at the intersection, it can be set so that vehicles turning right on a red light give way to vehicles and pedestrians can pass. In Embodiment 2 and Embodiment 3, a 1-second all-red signal is set after the yellow light, during which all directions are red lights, used to clear the vehicles within the intersection. This is just one form of control, not representing a mandatory setting, nor a limitation of this application.

[0045] The description and drawings of this application are only a specific implementation manner, not restrictive. Those skilled in the art, under the inspiration of this application and without departing from the scope of protection of the purpose of this application, can also make many forms, all of which are within the scope of protection of this application.

Claims

1. An intelligent traffic control method for an asymmetric intersection, characterized by: The intelligent control method is implemented through traffic data collection, data processing and analysis, signal timing decision-making, signal control, monitoring and adjustment strategies. When the green light signal of the main road at the intersection ends and there is no car on the branch road intersecting with it, the signal timing control is automatically adjusted to extend the duration of the green light signal of the main road, thereby improving the traffic efficiency of the intersection; The traffic data collection uses traffic flow detectors and vehicle detectors to collect real-time traffic flow and vehicle queue length data on main roads and branch roads as a basis for signal timing; The data processing and analysis described above transmits the collected traffic data to the intelligent traffic control system, which processes and analyzes the data, and determines the traffic conditions of the main roads and branch roads by analyzing the traffic flow and vehicle queues; The signal timing decision is made based on the results of data analysis. The intelligent traffic control system makes a signal timing decision based on preset algorithms and strategies. The system determines that there are no vehicles on the branch road that intersects with the main road before the green light signal of the main road ends, and extends the green light time of the main road. The signal control described above, the intelligent traffic control system transmits the signal timing decision result to the signal control device, controls the traffic lights of the main road and the branch road, and the system prolongs the green light time of the main road so that the vehicles on the main road have enough time to pass, while the waiting time of the traffic lights on the branch road will be extended accordingly; The monitoring and adjustment mentioned above, the intelligent traffic control system continuously monitors the traffic flow and vehicle queues, and makes adjustments based on real-time data. If the traffic conditions on the main road change or there are vehicles waiting on the branch road, the system will re-make signal timing decisions based on the new data to ensure efficient and safe traffic; The specific control method is: when the main road basic green light time T mb Buffer time before end T bf If there are still vehicles waiting in the waiting area of ​​the main road, the system will detect whether there are vehicles waiting in the waiting area of ​​the branch road. If there are vehicles waiting in the waiting area of ​​the branch road, the green light time of the main road will be T mb After the end, the main road turns red and the branch road turns green. If there is no vehicle waiting in the branch road waiting area, the system automatically controls the main road green light to extend the time. ΔT ; Extend the green light time on main roads ΔT Buffer time before end T bf If there are still vehicles in the waiting area of ​​the main road, the system will detect whether there are vehicles entering the waiting area on the branch road. If there are vehicles entering the waiting area on the branch road, the main road will continue to complete the extension time. ΔT The green light signal turns red after the branch road turns green. T sb Signal, if extended time ΔT Buffer time before end T bf If there are still no vehicles entering the waiting area on the branch road and there are still vehicles in the waiting area on the main road, the main road will be extended for another time. ΔT ; In the second extended time ΔT If a vehicle enters the waiting area on the branch road, the main road turns yellow and then turns red, and then turns to the basic green light time of the branch road. T sb signal, if the second extended time ΔT No vehicles enter the waiting area on the branch road, and the main road ends and the time is extended ΔT The green light signal changes to yellow light buffer (3~5 seconds) and then turns to red light, which is the basic green light time for the branch road. T sb signal; if the main road second extension time ΔT In the process, the system detects that there are pedestrians waiting to cross the main road on the branch road. The main road turns yellow for a buffer prompt and then turns red, turning to the basic green light time of the branch road. T sb signal; if the first and second extended time ΔT The main road turns red and the branch road turns green after all vehicles in the waiting area have left. T sb Signal; basic green light time for branch lines T sb After the end, the basic green light time for the main road to return T mb Enter the next cycle.

2. The asymmetric intersection intelligent traffic control method according to claim 1 is characterized by: The main road green light extension time Δ T , by performing Fourier transform (DFT) on the main road and branch road traffic data collected in real time, the traffic flow data is converted from the time domain to the frequency domain, so as to detect the periodicity in the data, which is used to analyze the periodic characteristics of the traffic flow data and determine the extension time of the traffic signal control; the specific steps are: 1) Data normalization: scaling the traffic data to the range of [0, 1] to improve the convergence speed and prediction accuracy of the model. The formula is as follows: ， in: Q ′ is the original traffic flow data, Q min is the minimum value in the data, Q max is the maximum value in the data; 2) Normalized traffic data Q Perform discrete Fourier transform to obtain the frequency domain signal F ( u ), the Fourier transform formula is: ， in: Q [ n ] is at the time point n Traffic flow data, N is the total number of data points, u is the frequency index, ( u =0,1,2,…,N−1), F ( u ) is the transformation result in the frequency domain; 3) Calculate the amplitude of the frequency domain signal | F ( u )|; Find the frequencies corresponding to the first few peaks with the largest amplitude u , when the peak value appears at u=k When , the corresponding period is: T=N / k , k is the peak position of the frequency domain index; the periods corresponding to the first few peaks are selected as candidate signal control periods; 4) The traffic flow cycle is determined by Fourier transform T , so as to adjust the extension time of signal control; According to the basic time of green light on main road T mb , calculate the extension time Δ T : ， in: Q h The historical traffic flow data of the main road includes the vehicle flow, time distribution, traffic mode and vehicle type distribution of the main road and branch roads. Q t It is the flow threshold, which is used to determine whether the current flow needs to adjust the signal time. α and β is the adjustment coefficient used to control Δ T range, α Control the increase or decrease of the extension time, β It is the basic extension time.

3. The intelligent traffic control method for asymmetric intersections according to claim 2 is characterized in that: The main road historical traffic flow data Q h The historical average model is used to calculate and use the traffic data in the historical time to predict the future traffic. The calculation formula is: ， in: t For time, Q ( t ) is at time t The actual observed flow rate, γ is the smoothing coefficient, which ranges from [0,1] and is used to adjust the weight of historical data.

4. The intelligent traffic control method for asymmetric intersections according to claim 1 is characterized in that: The basic green light time of the main road T mb Calculated by the following formula: ， in: Q k is the maximum traffic volume that needs to pass. h is the saturated headway time (unit: seconds), which indicates the average time interval between vehicles passing through the intersection under saturation conditions. t s For startup loss.

5. The intelligent traffic control method for asymmetric intersections according to claim 1 is characterized in that: The basic green light time of the branch T sb Calculated by the following formula: ， in, T min is the minimum green light time for the branch road, T d is the time difference from the time when the branch road vehicle is detected to enter the waiting area to the current time, T bf For buffer time.

6. A system for the intelligent traffic control method for an asymmetric intersection according to claim 1, characterized in that: include: The data detection unit is used to detect vehicles entering the waiting area at each intersection and pedestrians waiting to cross the road at the intersection, collect traffic data of the main road and branch roads in real time, and monitor the vehicle flow, time distribution, traffic mode and vehicle type distribution of the main road and branch roads; The data processing unit pre-processes, extracts features and optimizes parameters of traffic flow data on main roads and branches based on the traffic data from the data detection unit. It extracts periodic features from complex time domain signals through Fourier transform, analyzes the periodicity of traffic flow data, the periodic changes in the morning peak, evening peak and valley periods, the difference in traffic flow between weekdays and weekends, and the changes in traffic flow in different seasons or holidays. The analysis results are used as the basis for intelligent signal adjustment. The system uses historical traffic data to self-learn and train a regression model to calculate α , β and Q t The optimal value of the embedded system control algorithm is used to calculate the function f ( Q h ), calculate the main circuit extension time Δ T , generate signal control instructions, calculate intersection timing plans and real-time signal adjustment parameters based on the data of vehicles and pedestrians waiting to cross the road in the waiting area of ​​each intersection; The signal control unit adjusts the main line extension time in real time according to the timing plan of the data processing unit and the real-time signal adjustment parameters ΔT , control the traffic lights of the main road and branch roads, realize the switching of traffic signals, and display the remaining time of the traffic lights through digital tubes or display screens; The data storage unit records traffic flow data and system operation logs, stores the collected traffic flow data and system status information locally or in the cloud, analyzes the stored data, generates traffic flow reports, and provides decision support for traffic management.

7. The system of the asymmetric intersection intelligent traffic control method according to claim 6 is characterized by: The data detection unit is equipped with vehicle detection sensors at the entrance and exit of the waiting area of ​​each lane at each intersection. The vehicle detection sensors include geomagnetic sensors for detecting the presence and passage of vehicles, radar sensors for measuring vehicle speed and queue length, and high-definition cameras for license plate recognition and vehicle type classification. Infrared thermal imaging sensors and image collectors are arranged in the pedestrian waiting area. The infrared thermal imaging sensors sense the presence of pedestrians in the waiting area. The image collector uses YOLO to perform image segmentation, feature extraction, bounding box prediction, and non-maximum suppression. Through computer vision and deep learning algorithms, the image collector uses the YOLO algorithm to detect pedestrians in real time and identify targets waiting to cross.

8. The system of the asymmetric intersection intelligent traffic control method according to claim 6 is characterized by: The data processing unit is an embedded processing system, including a core control module, a user interaction module, a communication module, and a fault detection and alarm module; the core control module is the hub of the entire traffic signal control system, dynamically adjusts the duration of the signal light according to traffic flow data and a preset control algorithm, receives flow data from sensors, analyzes and processes it, monitors the system operation status, and takes measures when abnormalities are found to ensure stable operation of the system; the user interaction module performs system settings and adjustments through buttons, knobs or touch screens, uses LCD or OLED display screens to display system status and signal light cycle information in real time, and provides an operation interface for maintenance personnel and administrators for system configuration and status monitoring; the communication module is responsible for realizing data transmission and command interaction between systems, realizing synchronization and coordination between signal light controllers, and transmitting data with the traffic management center to support remote monitoring and management; the fault detection and alarm module is responsible for monitoring the system operation status, timely discovering and handling abnormalities, detecting key parameters of the system, and notifying maintenance personnel through sound and light alarms or remote notifications when abnormalities occur.

Images