A method and system for intelligent traffic control at asymmetric intersections

Abstract

This invention relates to an intelligent signal control method and system for asymmetric intersections, belonging to the field of intelligent traffic control technology. The method dynamically adjusts traffic light timings by collecting real-time traffic flow data from main roads and side roads, utilizing data processing and analysis techniques. Specifically, if there are no vehicles waiting on side roads before the main road's green light ends, the main road's green light time is automatically extended to improve intersection efficiency. The system includes a data detection unit, a data processing unit, a signal control unit, and a data storage unit, capable of real-time monitoring of traffic flow and vehicle queuing, and adjusting signal timings based on real-time data. This invention effectively reduces vehicle waiting time, increases traffic flow, reduces congestion, and improves traffic safety and efficiency.

Description

Technical Field

[0001] This invention belongs to the field of traffic control technology, specifically relating to an intelligent traffic control method and system for asymmetrical intersections, applicable to the intelligent control of urban traffic lights, particularly for intersections where the traffic flow of main roads and branch roads is asymmetrical. Background Technology

[0002] With the acceleration of urbanization, traffic congestion has become increasingly serious, especially at intersections, where problems such as long waiting times and low traffic efficiency are particularly prominent. Traditional traffic signal control methods typically use fixed signal timing schemes, which cannot be dynamically adjusted according to real-time traffic flow. As a result, when traffic flow changes significantly, the signal timing cannot adapt to the actual needs, leading to a waste of traffic resources and a reduction in traffic efficiency.

[0003] In urban traffic, asymmetrical intersections refer to intersections that exhibit asymmetry in traffic flow, geometric layout, or signal control. Signal control at asymmetrical intersections (where main roads and side roads intersect) is a complex and critical issue. Traditional signal control methods typically employ fixed time allocations, failing to dynamically adjust signal timings based on real-time traffic flow. This fixed pattern easily leads to traffic congestion during peak hours, while during off-peak hours it may waste green light time and reduce traffic efficiency. Furthermore, while existing intelligent transportation systems can detect vehicle flow, their application at asymmetrical intersections often fails to effectively handle the dynamic relationship between main roads and side roads. For example, when there are no vehicles on the side road, the green light for the main road may switch to the side road when the green light ends, causing vehicles on the main road to queue and prolong the main road's green light duration. Conversely, when there are vehicles on the side road, the signal switch may not be timely enough, resulting in excessively long waiting times for vehicles on the side road.

[0004] While existing intelligent traffic control systems can dynamically adjust to some extent based on traffic flow, they often fail to effectively handle the significant differences in traffic flow between main roads and side roads at asymmetrical intersections. This results in excessively long waiting times for vehicles on main roads and low traffic efficiency on side roads. Therefore, there is an urgent need for an intelligent control method that can dynamically adjust signal timing based on real-time traffic flow to improve intersection efficiency and reduce vehicle waiting times. Developing an intelligent control method and system capable of dynamically adjusting signal timing is of great significance for improving the efficiency and safety of asymmetrical intersections. Summary of the Invention

[0005] This invention provides an intelligent traffic control method and system for asymmetric intersections. By collecting real-time traffic flow data from main roads and side roads, and utilizing data processing and analysis techniques, it dynamically adjusts traffic light timings. Specifically, if there are no vehicles waiting on side roads before the main road's green light ends, the main road's green light time is automatically extended to improve intersection efficiency. The system includes a data detection unit, a data processing unit, a signal control unit, and a data storage unit. It can monitor traffic flow and vehicle queuing in real time and adjust signal timings based on real-time data. This invention improves intersection capacity and operational efficiency by monitoring traffic flow and controlling signals, effectively reducing vehicle waiting time, increasing traffic flow, reducing congestion, and enhancing traffic safety and efficiency.

[0006] This invention is achieved through the following technologies: 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 on the main road of an intersection ends and there are no vehicles on the intersecting side road, the signal timing adjustment is automatically performed to extend the green light signal duration on the main road and improve the traffic efficiency of the intersection.

[0007] The traffic data collection uses traffic flow detectors and vehicle detectors to detect vehicles entering the waiting area at each intersection, as well as pedestrians waiting to cross the road at intersections. It monitors the vehicle flow, time distribution, traffic patterns, and vehicle type distribution on main roads and side roads. It also collects real-time traffic flow and vehicle queue length data on main roads and side roads as a basis for signal timing.

[0008] The data processing and analysis involves transmitting the collected traffic data to the data processing unit. The system processes and analyzes the data, performing preprocessing, feature extraction, and parameter optimization on the main road and branch road traffic flow data based on the traffic data from the data detection unit. Through Fourier transform, periodic features are extracted from the complex time-domain signal, analyzing the periodicity of traffic flow data, including the periodic changes during morning peak, evening peak, and off-peak periods, the differences in traffic flow between weekdays and weekends, and the changes in traffic flow during different seasons or holidays. The analysis results serve as the basis for intelligent signal adjustment. By analyzing traffic flow and vehicle queuing, the system determines the traffic conditions of main roads and branch roads and sets the basic green light time for main roads. T mb Basic green light time for side streets T sb The system uses historical traffic data to train a regression model through self-learning, and employs an embedded system control algorithm to calculate the function. f ( Q h ), calculate the main road green light extension time Δ TIt generates signal control commands and calculates intersection timing schemes and real-time signal adjustment parameters based on the data of vehicles and pedestrians waiting to cross the road in the waiting areas of each intersection.

[0009] The signal timing decision is made based on data analysis results. The intelligent traffic control system performs signal timing decisions according to preset algorithms and strategies. If the system determines that there are no vehicles on the intersecting side roads before the main road green light signal ends, it adjusts the main road green light extension time Δ. T .

[0010] The signal control system transmits the signal timing decision results to the signal control equipment to control the traffic lights on the main road and the side roads. When there are no vehicles on the side roads intersecting with the main road, the system extends the green light time on the main road to allow vehicles on the main road sufficient time to pass, while the waiting time of the traffic lights on the side roads will be extended accordingly.

[0011] The aforementioned monitoring and adjustment system continuously monitors traffic flow and vehicle queuing, and makes adjustments based on real-time data. If the traffic conditions on the main road change or vehicles are waiting on the side roads, the system re-determines signal timing based on the new data to ensure efficient and safe traffic.

[0012] Furthermore, the specific control method is as follows: when the main road has a basic green light time T mb Buffer time before end T bf Inside, there are still vehicles in the main road waiting area, and the system checks whether there are vehicles waiting in the side road waiting areas;

[0013] If there are vehicles waiting in the waiting area on the side road, the main road's green light time will be approximately [time missing]. T mb After the event, the main road light turns red, and the side road light turns green.

[0014] If there are no vehicles waiting in the side road waiting area, the system will automatically extend the green light time on the main road. ΔT ;

[0015] Extend the green light time on the main road ΔT Buffer time before end T bf Inside, there are still vehicles in the main road waiting area, and the system detects whether any vehicles from side roads have entered the waiting area;

[0016] If a vehicle enters the waiting area from a side road, the extended waiting time will continue on the main road. ΔT After the green light signal turns red, the basic green light time for the branch road begins. T sb Signal;

[0017] If the time is extended ΔT Buffer time before end Tbf Inside, no vehicles have entered the waiting area on the side road, but there are still vehicles in the waiting area on the main road. The waiting time on the main road will be extended by another period. ΔT ;

[0018] During the second extended time ΔT If a vehicle enters the waiting area from the side road, the main road light turns yellow as a buffer signal, then red, and the side road then enters its basic green light period. T sb Signal;

[0019] If the second extension time ΔT No vehicles entered the waiting area on the middle branch road, and the extended time for the main road was extended. ΔT After a 3-5 second buffer period (from green to yellow), the light turns red, marking the start of the basic green light time for the secondary road. T sb Signal;

[0020] If the main road has a second extended time ΔT When the system detects pedestrians waiting to cross the main road, the main road light turns yellow for a buffer period, then turns red, and the secondary road enters its basic green light period. T sb Signal;

[0021] If all vehicles in the main road waiting area have left within the first or second extended time ΔT, the main road light turns red, and the secondary road light becomes essentially green. T sb Signal;

[0022] Basic green light time for side roads T sb After the green light ends, the main road will return to its basic green light period. T mb Moving on to the next cycle.

[0023] The basic green light time on the main road T mb Basic green light time for side roads T sb Extended green light time on main road Δ T The system dynamically adjusts its signals based on real-time traffic data from main roads and secondary roads, combined with historical traffic flow data. Historical traffic flow data is obtained by performing a Fourier Transform (DFT) on the real-time collected main road and secondary road traffic data, transforming the traffic flow data from the time domain to the frequency domain. This detects periodicity in the data and allows for analysis of its periodic characteristics, including periodic variations during morning peak, evening peak, and off-peak hours, weekday and weekend traffic differences, and traffic changes across different seasons and holidays. The analysis results serve as the basis for intelligent signal adjustment. The specific steps for calculating the traffic fluctuation cycle using Fourier Transform are as follows:

[0024] 1) Data normalization scales the traffic data to the range [0, 1] to improve the model's convergence speed and prediction accuracy. The formula is as follows:

[0025] ,

[0026] in: Q ′ represents the raw traffic flow data. Q min It is the minimum value in the data. Q max It is the maximum value in the data;

[0027] 2) Normalized flow data Q Perform a discrete Fourier transform to obtain the frequency domain signal. F ( u The Fourier transform formula is:

[0028] ,

[0029] in: Q [ n [This refers to a specific point in time.] n Traffic flow data, N It is the total number of data points. u It is a frequency index, ( u =0,1,2,…,N-1). F ( u ) is the result of the frequency domain transformation;

[0030] 3) Calculate the amplitude of the frequency domain signal | F ( u Find the frequencies corresponding to the first few peak values ​​with the largest amplitude. u When the peak appears u=k When the time is specified, the corresponding period is: T=N / k , k The peak position is indexed in the frequency domain; the periods corresponding to the first few peaks are selected as the control periods for the candidate signals.

[0031] Furthermore, different traffic fluctuation cycles were determined using Fourier transform. T For different traffic fluctuation cycles T The detected traffic flow is used as historical traffic flow data. Q h This data is saved and used to extract matching data during traffic signal control, enabling dynamic adjustment of the basic green light time on the main road. T mb Basic green light time for side roads T sb Extended green light time on main road Δ T .

[0032] The historical traffic flow data Q h The historical average model is used to predict future traffic flow using historical traffic data. The calculation formula is as follows:

[0033] Q h ( t )= c · Q h ( t -1)+(1- c )· Q ( t ),

[0034] in: t For time, Q ( t ) is in time t The actual observed flow rate c It is a smoothing coefficient with a value range of [0,1], used to adjust the weight of historical data.

[0035] The basic green light time on the main road T mb Calculated using the following formula:

[0036] T mb = Q k · h+t s ,

[0037] in: Q k The maximum number of vehicles that need to pass through a green light is the number of vehicles queuing up to pass through the main road during that signal cycle. h Saturated headway is the average time interval between vehicles passing through an intersection under saturated conditions, i.e., the average time taken per vehicle, in seconds per vehicle. t s The starting loss (in seconds) is the total time lost by several cars (usually about 3) from starting to passing the stop line when the green light starts.

[0038] The basic green light time of the branch road T sb Calculated using the following formula:

[0039] T sb =max( T min , T d +T bf ),

[0040] in, T min The minimum green light time (in seconds) for a branch road. T d The time difference (in seconds) from the moment a vehicle on the side road was detected entering the waiting area to the present. T bf Buffer time (seconds).

[0041] The main road green light extension time Δ T Calculated using the following formula:

[0042] Δ T = α ·( Q h - Q t )+ β ,

[0043] in: Q h Historical traffic flow data for the main road (vehicles / unit of time). Q t It is a traffic flow threshold (vehicles / unit of time), used to determine whether the current traffic flow requires adjustment of the signal timing. α and β It is an adjustment coefficient used to control Δ T Scope α Control the increase or decrease of the extension time (seconds / (vehicle / unit time)). β It is the base extension time (seconds).

[0044] This invention provides a system for intelligent traffic control at asymmetric intersections, comprising: a data detection unit, a data processing unit, a signal control unit, and a data storage unit.

[0045] The data detection unit is used to detect vehicles entering the waiting area at each intersection, as well as pedestrians waiting to cross the road at the intersection. It collects traffic data of the main road and the side roads in real time and monitors the vehicle flow, time distribution, traffic patterns and vehicle type distribution of the main road and the side roads. The traffic data includes traffic flow, vehicle queue length, vehicle presence information and pedestrian waiting information.

[0046] The data processing unit, connected to the data detection unit, is used for traffic data processing and analysis, signal timing decision-making, and generating control commands. The signal timing decision-making includes, based on the results of data analysis, the intelligent traffic control system makes signal timing decisions according to preset algorithms and strategies. If the system determines that there are no vehicles on the intersecting side roads before the main road green light signal ends, the main road green light time is extended.

[0047] The signal control unit, connected to the data processing unit, is used to control the switching of main road and branch road traffic lights according to the timing scheme and real-time signal adjustment parameters of the data processing unit, and to adjust the main road green light extension time Δ in real time. T It controls the traffic lights on main roads and side roads, realizing intelligent control of traffic signals, and displays the remaining time of the traffic lights through digital tubes or displays.

[0048] The data storage unit is connected to the data processing unit and is used to store traffic flow data, system operation logs, and reports that support data analysis and decision-making. It 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.

[0049] The data detection unit includes a geomagnetic sensor 1, a radar sensor 2, a high-definition camera 3, an infrared thermal imaging sensor 4, and an image acquisition device 5. The data detection unit is used to collect real-time traffic data from main roads and branch roads. The traffic data includes traffic flow, vehicle queue length, vehicle presence information, and pedestrian waiting information. The geomagnetic sensor 1, radar sensor 2, and high-definition camera 3 are installed at the entrance and exit of the waiting area for each lane at each intersection. The geomagnetic sensor 1 detects the presence of vehicles; by detecting vehicles entering and leaving the waiting area, the system calculates the number of vehicles in the waiting area for intelligent lane signal control. The radar sensor 2 measures vehicle speed and queue length, calculating the time required to exit the waiting area. The high-definition camera 3 is used for license plate recognition and vehicle type classification, and also serves as a sensor for detecting vehicles entering and leaving the waiting area.

[0050] The infrared thermal imaging sensor 4 and image acquisition device 5 are installed on the sidewalk at the main road intersection, covering the waiting area for pedestrians waiting to cross the main road. The infrared thermal imaging sensor 4 detects the presence of pedestrians in the main road waiting area, sets a pedestrian dwell time threshold, and combines the infrared thermal imaging sensor 4 with continuous frame analysis from the image acquisition device 5. If a pedestrian stays longer than the threshold and is located in the main road waiting area, they are determined to be a pedestrian waiting to cross the road. If the dwell time is less than the threshold, they are determined to be a pedestrian walking and are not recorded as a pedestrian waiting to cross the road. The image acquisition device 5 uses YOLO for image segmentation, feature extraction, bounding box prediction, and non-maximum suppression. Through computer vision and deep learning algorithms, it detects pedestrians in real time and identifies pedestrian targets waiting to cross the main road.

[0051] 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.

[0052] The core control module is the central hub of the entire traffic signal control system. It receives traffic flow data from the data detection unit, analyzes and processes it, dynamically adjusts the duration of traffic lights based on the traffic flow data and preset control algorithms, monitors the system's operating status, and takes measures when abnormalities are detected to ensure stable system operation.

[0053] The user interaction module allows for system settings and adjustments via buttons, knobs, or a touchscreen. It uses an LCD or OLED display to show system status and indicator light cycle information in real time, providing an operating interface for maintenance personnel and administrators for system configuration and status monitoring.

[0054] The communication module is responsible for data transmission and command interaction between systems, enabling synchronization and coordination between traffic light controllers, transmitting data with the traffic management center, and supporting remote monitoring and management.

[0055] The fault detection and alarm module is responsible for monitoring the system's operating status, promptly detecting and handling anomalies, detecting key system parameters, and notifying maintenance personnel via audible and visual alarms or remote notification when an anomaly occurs.

[0056] The beneficial effects of this invention are: by monitoring traffic flow in real time and dynamically adjusting traffic light timings, especially before the main road green light ends, if there are no vehicles waiting on the side roads, the main road green light time is automatically extended, reducing vehicle waiting time and improving intersection efficiency. Through intelligent signal control, vehicle queue lengths can be effectively reduced, lowering the probability of traffic congestion, especially during peak hours, significantly improving traffic flow capacity. By monitoring and adjusting traffic light timings in real time, traffic accidents caused by unreasonable traffic light timings can be effectively avoided, improving traffic safety. The system can learn and adjust itself based on real-time traffic flow data, adapting to traffic flow changes in different time periods and seasons, exhibiting strong self-adaptability. Attached Figure Description

[0057] Figure 1 Block diagram of the intelligent traffic control method of the present invention;

[0058] Figure 2 This is a plan view of an asymmetrical intersection according to the present invention;

[0059] Figure 3 This is a flowchart of the intelligent traffic control process of the present invention;

[0060] Figure 4 This is the traffic signal timing diagram for Example 3;

[0061] Figure 5 This is the traffic signal timing diagram for Example 4;

[0062] Figure 6 This is the traffic signal timing diagram for Example 5.

[0063] In the diagram: 1-Geomagnetic sensor, 2-Radar sensor, 3-High-definition camera, 4-Infrared thermal imaging sensor, 5-Image acquisition device. Detailed Implementation

[0064] To enable those skilled in the art to better understand the present invention, in conjunction with Figure 1~Figure 6 To further explain this application, the terms "upper," "lower," "left," "right," etc., used in this description indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly understood by those skilled in the art. They are used only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the components or parts referred to must have a specific orientation, or be constructed and operated in a specific orientation. The content mentioned in the embodiments is not intended to limit the present invention.

[0065] The asymmetric intelligent traffic control method for intersections of this invention is implemented through intersection traffic data collection, data processing and analysis, signal timing decision-making, signal control, monitoring and adjustment strategies, as described in [see details]. Figure 1A block diagram of an intelligent traffic control method. The specific scheme involves collecting traffic data from all directions at an intersection, processing and analyzing the data, and automatically adjusting signal timing decisions when there are no vehicles on the intersecting side roads before the main road's green light signal ends. This extends the main road's green light duration, improving the intersection's traffic efficiency.

[0066] Traffic data collection utilizes traffic flow detectors and vehicle detectors to detect vehicles entering waiting areas at intersections and pedestrians waiting to cross, monitoring vehicle flow, time distribution, traffic patterns, and vehicle type distribution on main roads and side roads. Real-time data on traffic flow and queue length on main roads and side roads is collected as a basis for signal timing.

[0067] Data processing and analysis involves transmitting the collected traffic data to the data processing unit. The system processes and analyzes this data, performing preprocessing, feature extraction, and parameter optimization on the main road and secondary road traffic flow data based on the traffic data from the data detection unit. Through Fourier transform, periodic features are extracted from the complex time-domain signal, analyzing the periodicity of traffic flow data, including peak hours, off-peak hours, weekday / weekend traffic differences, and seasonal / holiday traffic variations. The analysis results serve as the basis for intelligent signal adjustment. By analyzing traffic flow and vehicle queuing, the system determines the traffic conditions on main roads and secondary roads and sets the basic green light time for main roads. T mb Basic green light time for side streets T sb The system uses historical traffic data to train a regression model through self-learning, and employs an embedded system control algorithm to calculate the function. f ( Q h ), calculate the main road green light extension time Δ T It generates signal control commands and calculates intersection timing schemes and real-time signal adjustment parameters based on the data of vehicles and pedestrians waiting to cross the road in the waiting areas of each intersection.

[0068] Based on data analysis, the intelligent traffic control system makes signal timing decisions according to preset algorithms and strategies. If the system determines that there are no vehicles on the intersecting side roads before the main road's green light ends, it adjusts the main road's green light extension time by Δ. T .

[0069] Signal control: The intelligent traffic control system transmits the signal timing decision results to the signal control equipment, which controls the traffic lights on the main road and side roads. When there are no vehicles on the side roads intersecting with the main road, the system extends the green light time on the main road to allow vehicles on the main road sufficient time to pass, while the waiting time of the traffic lights on the side roads will be extended accordingly.

[0070] The intelligent traffic control system continuously monitors traffic flow and vehicle queuing, and makes adjustments based on real-time data. If the traffic conditions on the main road change or vehicles are waiting on the side roads, the system will re-make signal timing decisions based on the new data to ensure efficient and safe traffic.

[0071] The specific control method of this invention is as follows: when the main road has a basic green light time T mb Buffer time before end T bf Inside, there are still vehicles in the main road waiting area, and the system checks whether there are vehicles waiting in the side road waiting areas;

[0072] If there are vehicles waiting in the waiting area on the side road, the main road's green light time will be approximately [time missing]. T mb After the event, the main road light turns red, and the side road light turns green.

[0073] If there are no vehicles waiting in the side road waiting area, the system will automatically extend the green light time on the main road. ΔT ;

[0074] Extend the green light time on the main road ΔT Buffer time before end T bf Inside, there are still vehicles in the main road waiting area, and the system detects whether any vehicles from side roads have entered the waiting area;

[0075] If a vehicle enters the waiting area from a side road, the extended waiting time will continue on the main road. ΔT After the green light signal turns red, the basic green light time for the branch road begins. T sb Signal;

[0076] If the time is extended ΔT Buffer time before end T bf Inside, no vehicles have entered the waiting area on the side road, but there are still vehicles in the waiting area on the main road. The waiting time on the main road will be extended by another period. ΔT ;

[0077] During the second extended time ΔT If a vehicle enters the waiting area from the side road, the main road light turns yellow as a buffer signal, then red, and the side road then enters its basic green light period. T sb Signal;

[0078] If the second extension time ΔT No vehicles entered the waiting area on the middle branch road, and the extended time for the main road was extended. ΔT After a 3-5 second buffer period (from green to yellow), the light turns red, marking the start of the basic green light time for the secondary road. T sb Signal;

[0079] If the main road has a second extended time ΔT When the system detects pedestrians waiting to cross the main road, the main road light turns yellow for a buffer period, then turns red, and the secondary road enters its basic green light period. T sb Signal;

[0080] If all vehicles in the main road waiting area have left within the first or second extended time ΔT, the main road light turns red, and the secondary road light becomes essentially green. T sb Signal;

[0081] Basic green light time for side roads T sb After the green light ends, the main road will return to its basic green light period. T mb Moving on to the next cycle.

[0082] The basic green light time for the main road in this invention T mb Basic green light time for side roads T sb Extended green light time on main road Δ T Based on real-time traffic data from main roads and secondary roads, combined with historical traffic flow data, the system dynamically adjusts traffic signals. Urban traffic flow exhibits regular fluctuations over a certain period, such as daily cycles (morning peak, midday off-peak, evening peak, and nighttime off-peak), weekly cycles (weekday commuting patterns vs. weekend leisure patterns), and even annual cycles (seasonal and holiday influences). Different time periods exhibit drastically different traffic flow characteristics. This invention divides traffic fluctuation cycles based on macroscopic time series and establishes a historical traffic flow database to provide reference data for real-time dynamic adjustment of traffic signals.

[0083] Historical traffic flow data is obtained by performing a Fourier Transform (DFT) on real-time collected main and secondary road traffic data. This transforms the traffic flow data from the time domain to the frequency domain, thereby detecting periodicity in the data. This allows for the analysis of the periodic characteristics of traffic flow data, including periodic variations during morning peak hours, evening peak hours, and off-peak periods, weekday and weekend traffic differences, and traffic variations across different seasons and holidays. The analysis results serve as the basis for intelligent signal adjustment. The Fourier Transform is a commonly used signal processing method that converts time-domain signals into frequency-domain signals, thereby extracting the periodic characteristics of the signal. The specific steps for calculating the traffic fluctuation period using the Fourier Transform are as follows:

[0084] 1) First, normalize the real-time collected traffic data for main roads and secondary roads, removing outliers and filling in missing values ​​before normalization. Scale the traffic data to the range [0, 1] to improve model convergence speed and prediction accuracy, using Min-Max Scaling, as shown in the following formula:

[0085] ,

[0086] in: Q ′ represents the raw traffic flow data. Q min It is the minimum value in the data. Q max It is the maximum value in the data;

[0087] 2) Normalized flow data Q Perform a discrete Fourier transform to obtain the frequency domain signal. F ( u The Fourier transform formula is:

[0088] ,

[0089] in: Q [ n [This refers to a specific point in time.] n Traffic flow data, N It is the total number of data points. u It is a frequency index, ( u =0,1,2,…,N-1). F ( u ) is the result of the frequency domain transformation;

[0090] 3) Calculate the amplitude of the frequency domain signal | F ( u Frequency domain signal F ( u ) is a complex number with magnitude AMP( u It can be obtained by calculating the modulus of the complex number. For complex numbers... F ( u )= a + b Its amplitude is:

[0091] ,

[0092] in: a yes F ( u The real part of ) b yes F ( u Find the imaginary part of the amplitude. Locate the frequencies corresponding to the first few peaks with the largest amplitudes.u When the peak appears u=k When the time is specified, the corresponding period is: T=N / k , k This represents the peak position indexed in the frequency domain. The periods corresponding to the first few peaks are selected as the candidate signal control periods.

[0093] The obtained period T It is measured in data point intervals. If you need to convert it to actual time units, you need to perform a conversion based on the sampling frequency. Specifically, the formula... T=N / k middle, N The total number of data points (time series length), expressed in units of data point intervals, can be... T Convert to actual time units. k For frequency domain peak index, the calculation result T This directly represents the cycle length of traffic flow (units: minutes, hours, or days). For example, 12 hours (720 minutes) from 7:00 AM to 7:00 PM, with a sampling interval of 5 minutes (300 seconds), is a common sampling interval for traffic flow analysis. N =144 (12-hour data) data points, which is a typical total number of data points for daytime traffic flow analysis. If k =4, based on the total number of data points N =144, period T It can be calculated using a formula. T = N / k =144 / 4=36, period T The interval is 36 data points, with each data point spaced 5 minutes apart, and the period is... T The representation in terms of time is: T Time = 36 × 5 minutes = 180 minutes. This means the traffic flow variation cycle is 180 minutes, or 3 hours. This cycle length indicates that within 12 hours of traffic flow data, there exists a periodic pattern that repeats approximately every 3 hours. This is related to the periodic changes during peak traffic periods, such as the periodic increase in traffic flow during morning, noon, and evening peak hours. The step of performing periodic calculations using Fourier transforms is only performed during the initial system setup and does not need to be frequently calculated under normal circumstances. When there are significant changes in the intersection's traffic conditions (such as the opening of other roads diverting traffic from the intersection, or road construction and traffic control increasing traffic flow at the intersection), the system detects that the intersection's traffic flow continuously deviates from the original traffic data. The system automatically performs Fourier transforms to perform periodic calculations and re-corrects the periodicity. T .

[0094] Different traffic fluctuation cycles were determined using Fourier transform. T For different traffic fluctuation cycles TThe detected traffic flow is used as historical traffic flow data. Q h This data is saved and used to extract matching data during traffic signal control, enabling dynamic adjustment of the basic green light time on the main road. T mb Basic green light time for side roads T sb Extended green light time on main road Δ T .

[0095] In practical applications, traffic flow is dynamic, especially in urban traffic, where it is affected by various factors such as traffic accidents, road construction, and special events. There are also situations where traffic flow may fluctuate drastically in a short period, such as due to sudden acceleration or deceleration of individual vehicles or pedestrians crossing the road. These short-term fluctuations may affect the actual observed traffic flow. Q ( t The instantaneous value of the signal may deviate significantly. In traffic signal control, adjusting the timing of traffic lights requires comprehensive consideration of both current and future traffic flow trends. This is achieved through prediction. Q h ( t This allows for a more accurate assessment of whether traffic light timings need adjustment, and to what extent, based on current traffic flow.

[0096] Historical traffic flow data in this application Q h The historical average model, a first-order exponential smoothing model, is used for calculation. This dynamic forecasting method combines historical and current observation data, smoothing data fluctuations through a weighted average to obtain a more stable and trend-reflecting forecast. This model is widely used in time series analysis, particularly in traffic flow forecasting, effectively handling both short-term fluctuations and long-term trends. The formula for predicting future traffic flow using historical traffic data is as follows:

[0097] Q h ( t )= c · Q h ( t -1)+(1- c )· Q ( t ),

[0098] in: t For time, Q ( t ) is in time t The actual observed flow rate cThis is a smoothing coefficient, ranging from [0,1], used to adjust the weights of historical data. It controls the weight balance between historical data and new observations. When c When the value is close to 1 (e.g., 0.9), the formula places greater trust in historical memory. Q h ( t -1), measured data Q ( t The impact is relatively small; when c When the value is close to 0 (0.1), the formula has greater confidence in the measured data. Q ( t Historical data has a relatively small impact. The smoothing coefficient in this application... c The value is dynamically adjusted according to traffic hours, during the morning peak (7:00~9:00) and evening peak (17:00~19:00). c =0.8, to enhance the weight of historical data; during off-peak hours (9:00~17:00), c =0.6; during the nighttime period (22:00~6:00), c =0.3, to improve the sensitivity of real-time data. Here, Q h ( t It is not simply a matter of predicting time. t Instead of calculating the actual traffic flow, the forecast uses a weighted average of historical data to smooth out changes in traffic flow, thus obtaining a more stable forecast that better reflects traffic flow trends.

[0099] In practical applications, the initial value of the first-order exponential smoothing model Q h (0) The initial value can be set based on the average of historical data or the first observation. If sufficient historical data is available, the average of historical data can be used as the initial value; if insufficient historical data is available, the first observation can be used as the initial value. At each time step... t In the middle, according to the formula Q h ( t )= c · Q h ( t -1)+(1- c )· Q ( t (Dynamically updated) Q h ( t This dynamic update mechanism ensures Q h ( t It can reflect the changing trends of traffic flow in a timely manner. Qh ( t -1) is in t The historical estimate calculated at time -1 (i.e., the previous time) can be understood as the "memory" or "summary" of all historical information up to the previous time.

[0100] Example 1: Suppose we design a traffic light system at an intersection and optimize traffic flow by dynamically adjusting the green light duration. We use a first-order exponential smoothing model to predict the traffic flow at the current moment. Q h The smoothing coefficient of a first-order exponential smoothing model. c Take 0.8. At time t=0, based on the historical data average (or the first observation). Q h (0) = 1600 (vehicles / hour);

[0101] Dynamic updates: in time t In the middle, it is dynamically updated according to the formula. Q h ( t ):

[0102] Q h ( t )= c · Q h ( t -1)+(1- c )· Q ( t ),

[0103] Assume that at time t=1, the actual observed traffic flow is... Q (1) = 2200 (vehicles / hour); therefore:

[0104] Q h (1) = 0.8×1600 + 0.2×2200 =1280+440 =1720 (vehicles / hour);

[0105] Assume that at time t=2, the actual observed traffic flow is... Q (2) = 1100 (vehicles / hour), therefore:

[0106] Q h (2) = 0.8×1600 + 0.2×1100 =1280+220 =1500 (vehicles / hour);

[0107] As can be seen from this embodiment, Q h( t It is not simply a matter of predicting time. t Instead of calculating the actual traffic flow, the forecast uses a weighted average of historical data to smooth out changes in traffic flow, thus obtaining a more stable forecast that better reflects traffic flow trends.

[0108] "cycle T "Analysis is the foundation of intelligent traffic signal control strategies, while 'signal control' is the specific execution action under this strategy. The signal control in this invention includes: the basic green light time on the main road." T mb Basic green light time for side roads T sb Extended green light time on main road Δ T Asymmetrical intersection plan view is shown below. Figure 2 , Figure 2 This is just a typical example of an asymmetrical intersection. Figure 2 Intersections 1 and 2 are main roads, while intersections 3 and 4 are side roads. The signal control time calculation for the intelligent traffic control system is as follows.

[0109] Basic green light time on main roads T mb Calculated using the following formula:

[0110] T mb = Q k · h+t s ,

[0111] in: Q k The maximum number of vehicles that need to pass through a green light is the number of vehicles queuing to pass through the main road during that signal cycle. It is the total number of vehicles queuing behind the stop line on the main road within a green light signal. This value is obtained in real time by the geomagnetic sensor 1 or the high-definition camera 3. h Saturation headway is the time interval between two consecutive vehicles passing through the same cross-section. It represents the average time interval between the headways of two consecutive vehicles passing through the same cross-section (such as the stop line at an intersection) when traffic flow is saturated (i.e., vehicles pass through one after another in a continuous and dense manner). The unit is (seconds / vehicle). Once the convoy begins to move, it represents the average time it takes for each vehicle to pass through the stop line. This is a relatively fixed value that depends on vehicle performance, driver habits, and road conditions, and is usually obtained through observation (e.g., the average is approximately 2~2.5 seconds / vehicle). t sThe starting loss (in seconds) is the total green light loss time accumulated by the first few vehicles (usually about 3) from the moment they start moving until they cross the stop line at the start of the green light. This formula describes the method for calculating the basic green light time for the main road (usually the direction with the highest traffic volume) in traffic signal control. Its core idea is that the green light time allocated to the main road should be just enough for all vehicles queuing in the current cycle to safely pass through the intersection. This ensures that the green light time can clear the queue of vehicles without causing the green light to remain on for too long (no cars at the intersection, but the green light is still on), thereby improving the efficiency of the intersection. Basic Green Light Time for Main Road T mb Includes the time for going straight and turning left at intersections 1 and 2. Calculation. T mb When choosing, select the number of vehicles that need to pass between intersection 1 and intersection 2. Q k Larger datasets. For example: the number of vehicles that need to pass through intersection 1 in a single cycle. Q k =30 vehicles, number of vehicles required to pass through intersection 2 in a single cycle Q k =26 vehicles, saturation headway h =2.4 seconds / vehicle t s Start-up loss time = 3 seconds (including yellow light time). The calculation is based on the number of vehicles required to pass through intersection 1 in a single cycle. T mb =30 (vehicles) × 2.4 (seconds / vehicle) + 3 (seconds) = 72 + 3 = 75 seconds, of which 25 seconds are for turning left, 5 seconds for the yellow light, and 45 seconds for going straight.

[0112] Basic green light time on main roads T mb Calculated using the above method, if the scenario matching error with similar dimensional features to historical traffic flow data is large, such as when the number of vehicles entering the waiting area on the main road during peak hours is 30-50% less than historical traffic flow data, it is possible that subsequent vehicles have not yet entered the waiting area. These vehicles may enter the waiting area when the main road light turns green. To reduce congestion on the main road, the basic green light time on the main road will be adjusted accordingly. T mb Traffic flow data can be used for time-based control. This is achieved by determining the cycle through Fourier transform. T For different periods T The detected traffic flow is used as historical traffic flow data. Q h This data is saved and used to extract matching data during traffic signal control, enabling dynamic adjustment of the basic green light time on the main road. T mb The mechanism.

[0113] Basic green light time for side roads T sb Calculated using the following formula:

[0114] T sb =max( T min , T d + T bf ),

[0115] in, T min The minimum green light time (in seconds) for a branch road. T d The time difference (in seconds) from the moment a vehicle on the side road was detected entering the waiting area to the present. T bf This is the buffer time (in seconds). The formula means: basic green light time for branch roads. T sb =Minimum green light time for branch roads T min The time difference from the start of vehicle detection T d Add a buffer time T bf Take the larger of the two values. Minimum green light time for the branch road. T min It is a preset fixed value, which is the minimum guarantee for the green light time of the side road, ensuring that vehicles that have entered the waiting area of ​​the side road have enough time to pass safely. If the green light time is too short, vehicles will have to brake as soon as they start moving, which is very dangerous. T d This is the time elapsed from the point when a vehicle was detected entering the waiting area on the side road to the current time (the moment the system performs the calculation). This parameter reflects the system's "memory" function, recording how long the vehicle on the side road has been waiting. If the main road traffic flow is continuous, the waiting time for the vehicle on the side road ( T d This waiting time accumulates, and the system will take it into account when allocating green light time next time, thus achieving fairness to some extent. This is a traffic light adaptive control algorithm, mainly used to calculate the green light time for side roads (usually directions with less traffic). Its core idea is to dynamically determine how long the green light should be based on the actual detected vehicle situation, ensuring both traffic efficiency and safety. For example: at an intersection, the main road is very busy, while the side road has less traffic. System parameter settings: T min = 15 seconds, T bf= 5 seconds. Scenario 1: A car enters the waiting area on a side road. The ground loop detector or camera detects the vehicle. At this time, T d = 0 (because it was just detected), the system begins calculation: T sb = max(15, 0 + 5) = max(15, 5) = 15 seconds. After the main road's green light ends, the system will give the side road a green light for at least 15 seconds. Because T min This is the dominant factor, ensuring the minimum passage time. Scenario 2: Multiple vehicles enter the waiting area on a side road; the first vehicle has been waiting for a long time. T d The time has accumulated to 20 seconds; the system is recalculating. T sb = max(15, 20 + 5) = max(15, 25) = 25 seconds. This time, the system will give the branch a green light for at least 25 seconds because... T d + T bf This has become the dominant factor, demonstrating the system's "adaptive" or "compensatory" capabilities. It allows directions with longer waiting times to have longer passage times, thereby optimizing overall intersection efficiency and reducing average vehicle delays. The system dynamically adjusts the base time of the main road's green light based on historical traffic flow data. T mb The mechanism is the same, calculating the basic green light time for branch roads. T sb When the scenario matching error is large for features similar to historical traffic flow data, historical traffic flow data will also be used. Q h Based on the matched data, the basic green light time of the branch road is dynamically adjusted. T sb .

[0116] Main road green light extension time Δ T Calculated using the following formula:

[0117] Δ T = α ·( Q h - Q t )+ β ,

[0118] in: Q hThe main road's historical traffic flow data (vehicles / unit time) is used. The unit (vehicles / unit time) is determined based on the unit of traffic flow detected by the data detection unit, and can be either "vehicles / hour" or "vehicles / minute". Q h This is historical traffic flow data containing multi-dimensional features, including vehicle flow, time distribution, traffic patterns, and vehicle type distribution information for both main roads and secondary roads. The main road green light extension time Δ in this application...

[0119] T The calculation is based on scenario matching with multi-dimensional features. It doesn't simply call up historical traffic flow data, but follows the principle of "scenario matching" from... Q h The system intelligently retrieves historical traffic flow data records from the database that are highly similar to the current time. This data is then used to calculate the extended time Δ. T At that time, it will be based on the following dimensions from Q h Retrieve “same or similar” vehicle traffic data: 1) Time distribution (season / month, date type, peak period), 2) Traffic pattern (traffic saturation, traffic flow ratio of main road to branch road), 3) Vehicle type distribution (proportion of large vehicles and small vehicles). Q t It is the traffic flow threshold (vehicles / unit of time), with the specific unit being consistent with the historical traffic flow data of the main road. It is a preset traffic flow threshold, which is the "threshold" for determining whether the signal timing needs to be adjusted. α and β It is an adjustment coefficient used to control Δ T The range. α Control the increase or decrease in extension time (seconds / (vehicles / unit hour)), based on historical traffic flow data of the main road. Q h The unit is "vehicles per hour". α The unit is "seconds / (vehicles / hour)". If the historical traffic flow data for the main road... Q h The unit is "vehicles / minute". α The unit is "seconds / (vehicles / minute)," which determines the flow difference. Q h - Q t This can be converted into how many seconds of green light time, which can be understood as "the additional time required for the flow to exceed the threshold", the "sensitivity" or "response intensity" of the control system. β It is the base extension time (in seconds), the "base" of the extension time, even if ( Q h - Qt ) is zero or very small, even Q h < Q t hour, ( Q h - Q t Even if the value is negative, there can still be a basic extension time. β (seconds), ensuring that at least the detected vehicle can pass.

[0120] The system uses Fourier transform to obtain the periodicity of traffic flow data, including the periodic changes during morning peak, evening peak, and off-peak hours, the differences in traffic flow between weekdays and weekends, and the changes in traffic flow in different seasons or holidays. The analysis results serve as the basis for intelligent signal adjustment. The system uses historical traffic data to self-learn and train a regression model for calculation. α , β and Q t The optimal value. Different seasons, different peak times. α , β , Q t The calculation of the main road green light extension time Δ varies depending on the value of Δ. T At that time, the system automatically selects the matching traffic scenario. α , β , Q t value.

[0121] Main road green light extension time Δ T It is calculated based on the real-time traffic flow detected in the main road waiting area, and is a dynamic time, representing the green light extension time Δ of the main road green light phase in each signal cycle. T They are different. In calculating Δ T hour, Q h ( t As a dynamically updated predictive value, it can reflect the current traffic flow trend, thus providing a basis for signal adjustment. It gives traffic lights "foresight." For example, during the morning rush hour on weekdays, even if there are not many cars at the moment, historical data tells the system that "there will be a large flow of traffic at this time." The system will tend to extend the green light duration more to cope with the upcoming traffic flow and prevent intersection congestion.

[0122] Example 2: Calculate the main road green light extension time Δ based on the predicted value calculated from the actual observed traffic flow. T The actual observed traffic flow and the calculated predicted values ​​are based on Example 1, using a dependent periodic system calculation. α , β , Qt The optimal value is the adjustment coefficient. α = 0.04, unit (seconds / (vehicles / hour)); adjustment factor β = 20, unit (seconds); flow threshold Q t =1000, unit (vehicles / hour);

[0123] Traffic flow actually observed at time t=1 Q When (1) = 2200 (vehicles / hour), the predicted value is... Q h (1) = 1720 (vehicles / hour), according to the formula Δ T = α ·( Q h - Q t )+ β Calculate the main road green light extension time Δ T :

[0124] Δ T = α ·( Q h - Q t )+ β =0.04×(1720-1000)+20=48.8 (seconds), rounded up to 49 seconds;

[0125] At time t=2, the actual observed traffic flow Q (2) When = 1100 (vehicles / hour), the predicted value is Q h (2) = 1500 (vehicles / hour), according to the formula Δ T = α ·( Q h - Q t )+ β Calculate the main road green light extension time Δ T :

[0126] Δ T = α ·( Q h - Q t )+ β =0.04×(1500-1000)+20=40 (seconds);

[0127] As can be seen from this embodiment, the green light extension time Δ T It is dynamic and changing.

[0128] This invention uses data detection units installed at intersections to detect real-time traffic conditions, including when the main road has a basic green light period. T mb Buffer time before end T bf Within 3-5 seconds, if there are still vehicles in the main road waiting area, the system checks if there are vehicles waiting in the side road waiting area. If there are vehicles waiting in the side road waiting area, the main road's green light time is basically over. T mb After the event ends, the main road light turns red, and the side road light turns green. If there are no vehicles waiting in the side road waiting area, the system automatically extends the green light time on the main road. ΔT Extend the green light time on the main road. ΔT Buffer time before end T bf Within 3-5 seconds, if there are still vehicles in the main road waiting area, the system will check if any vehicles from the side roads have entered the waiting area. If any vehicles from the side roads have entered the waiting area, the main road will continue to complete the extended waiting time. ΔT After the green light signal, it turns yellow for a buffer period (3-5 seconds), then turns red, and the basic green light time for the branch road begins. T sb Signal. If the time is extended. ΔT Buffer time before end T bf Within 3-5 seconds, no vehicles from the side road enter the waiting area, but there are still vehicles in the main road's waiting area. The main road then extends the waiting time by another period. ΔT In the second extended time ΔT If a vehicle enters the waiting area from a side road, the main road light will turn yellow for a 3-5 second buffer period before turning red, ending the second extended waiting time early. ΔT The basic green light time for switching to a side road T sb Signal. If no vehicles enter the waiting area from the side road during the second extended cycle, the extended time on the main road ends. ΔT After the green light signal turns red, the basic green light time for the branch road begins. T sb Signal. If the main road has a second extension time. ΔT When the system detects pedestrians waiting to cross the main road from the side road, the main road light turns yellow for a buffer period of 3-5 seconds before turning red, and the side road then enters its basic green light period. T sb The signal indicates that pedestrians should be allowed to cross the main road even if there are no vehicles on the side road. If all vehicles in the main road waiting area have left within the first or second extended time ΔT, the main road light turns red, and the side road light becomes essentially green. T sb Signal. Basic green light time for branch lines. Tsb After the green light ends, the main road will return to its basic green light period. T mb Proceed to the next cycle. See the flowchart of the intelligent traffic control process of this invention. Figure 3 .

[0129] The system can also adjust signal control according to actual conditions. For example, the basic green light time on the main road... T mb Extended green light time on inner or main roads ΔT Once all vehicles in the inner waiting area have left, and vehicles are still waiting on the side roads, the main road's yellow light will give a 3-5 second buffer period before turning red, at which point the side roads will have essentially a green light period. T sb Signals are used to prevent a situation where the main road remains green while vehicles on side roads are still waiting at red lights; similarly, if the side road's green light duration is generally short... T sb Is adopted T d + T bf Controlled minimum green light time for branch roads T min After the procession ends, all vehicles in the waiting area on the side road have left, and the green light time on the side road is basically complete. T sb The process isn't over yet. The system detects that vehicles have entered the waiting area on the main road. The system controls the side road's yellow light to give a 3-5 second buffer period before turning it red, thus resuming the main road's basic green light period. T mb This strategy avoids unnecessary green light time occupation and improves the traffic capacity of intersections. According to this strategy, the basic green light time on the main road is [not specified] within a peak time period (peak or off-peak hours). T mb It can be fixed, eliminating the need for frequent calculations of the basic green light time on the main road. T mb The steps involve switching signals using the aforementioned dynamic adjustment strategy. This dynamic adjustment strategy for switching signals is merely a control method and is not mandatory, nor is it a limitation of this application.

[0130] The system of the asymmetric intersection intelligent traffic control method of the present invention includes: a data detection unit, a data processing unit, a signal control unit, and a data storage unit.

[0131] The data detection unit is used to detect vehicles entering the waiting area at each intersection, as well as pedestrians waiting to cross. It collects real-time traffic data from main roads and side roads, monitoring vehicle flow, time distribution, traffic patterns, and vehicle type distribution. Traffic data includes traffic flow, vehicle queue length, vehicle presence information, and pedestrian waiting information. The data detection unit is the perception layer of the intelligent transportation system. Its core task is to continuously, objectively, and quantitatively perceive the status of vehicles and pedestrians in all directions at the intersection, and convert this physical information into digital signals (data) that can be processed and calculated by computers, providing factual basis for subsequent intelligent decisions (such as calculating green light times). The data detection unit is the cornerstone of the entire adaptive signal control system's "intelligence." The accuracy, reliability, and comprehensiveness of the data detection unit directly determine the effectiveness of all subsequent intelligent control algorithms.

[0132] The data processing unit, connected to the data detection unit, is used for traffic data processing and analysis, signal timing decisions, and the generation of control commands. Signal timing decisions include: based on the data analysis results, the intelligent traffic control system makes signal timing decisions according to preset algorithms and strategies. If the system determines that there are no vehicles on the intersecting side roads before the main road's green light signal ends, it extends the main road's green light time. The data processing unit is the core computing and decision-making hub of the entire intelligent traffic signal control system. It is connected to the data detection units distributed throughout the intersection via a high-speed data bus, acting as the system's "brain." Its core mission is to transform the raw, fragmented physical signals collected by the detection units into a deep understanding of the traffic state, ultimately forming precise and efficient control commands. The data processing unit realizes a shift from passively collecting data to proactive intelligent decision-making. Through continuous data processing and analysis, it endows traffic signal control with human-like "judgment" and "flexibility," dynamically adjusting signal timing according to the actual instantaneous needs of the intersection. This allows for the precise allocation of valuable green light time resources to the most needed directions, ultimately achieving an intelligent upgrade from "cars waiting for lights" to "lights watching cars."

[0133] The signal control unit, connected to the data processing unit, controls the switching of main and branch road traffic lights based on the timing scheme and real-time signal adjustment parameters from the data processing unit, and adjusts the main road green light extension time Δ in real time. TThe signal control unit controls the traffic lights on main roads and side roads, achieving intelligent traffic signal control. It displays the remaining time of the traffic lights via digital tubes or displays. The signal control unit is the physical layer control core and execution terminal of the entire system. As a bridge between the data processing unit and the physical equipment at the intersection (traffic lights, countdown displays), it is responsible for accurately, reliably, and in real-time translating the calculated intelligent timing scheme into specific changes in traffic light colors. It is the final link in realizing the adaptive control concept. As the system's "operator," the signal control unit's value lies in putting upstream intelligent decisions into practice without delay or error. Through its precise timing control, reliable hardware drive, and real-time dynamic adjustment capabilities, it ensures that the entire intelligent transportation system can transform from an ideal algorithm model into a real-world, perceptible, efficient, and safe optimization of traffic order.

[0134] The data storage unit, connected to the data processing unit, stores traffic flow data, system operation logs, and reports supporting data analysis and decision-making. It stores collected traffic flow data and system status information locally or in the cloud, analyzes the stored data to generate traffic flow reports, and provides decision support for traffic management. The data storage unit is the core database and historical information repository of the intelligent transportation control system. Connected to the data processing unit, it is responsible for the persistent storage, efficient management, and in-depth analysis of massive amounts of traffic data, transforming instantaneous traffic flow states into traceable and analyzable strategic assets, providing solid data support for real-time system decision-making, long-term optimization, and strategic planning. The data storage unit elevates the intelligent transportation system from a control system focused on real-time response to a strategic platform with learning, memory, and optimization capabilities. Through the accumulation and analysis of historical data, it not only enables the system to "perceive the present" but also to "learn from the past and predict the future," thus achieving a spiral progression from single-point adaptive control to regional adaptive coordination, and then to long-term strategic planning, ultimately continuously improving the operational efficiency and management level of the entire road network.

[0135] The data detection unit includes a geomagnetic sensor 1, a radar sensor 2, a high-definition camera 3, an infrared thermal imaging sensor 4, and an image acquisition unit 5. This unit collects real-time traffic data from main roads and secondary roads, including traffic flow, vehicle queue length, vehicle presence information, and pedestrian waiting information. This data forms the basis for intelligent control. Vehicle presence information is an instantaneous, binary signal indicating "vehicle present" or "no vehicle present," while pedestrian waiting information is the basis for ensuring pedestrian right-of-way. To achieve comprehensive and accurate perception of traffic conditions at intersections, the system deploys three different types of sensors: a geomagnetic sensor 1, a radar sensor 2, and a high-definition camera 3, forming a multi-source data fusion collaborative detection network. The geomagnetic sensor 1, radar sensor 2, and high-definition camera 3 are installed at the entrance and exit of the waiting area for each lane at each intersection. The geomagnetic sensor 1, buried under the road surface, accurately determines whether a vehicle is directly above it by sensing changes in the magnetic field caused by the metal body of a vehicle. By detecting vehicles entering and leaving the waiting area, the system calculates the number of vehicles within the waiting area for intelligent lane signal control. Radar sensor 2 measures vehicle motion parameters and space occupancy. By emitting electromagnetic waves and analyzing the echoes, radar can measure the instantaneous speed and direction of movement of targets, and can also sense the spatial range of vehicle queues by scanning, used to measure vehicle speed and queue length, and calculate the time required to exit the waiting area. Radar sensor 2 is installed directly above the lane, and its beam can cover the entire lane area extending back from the stop line. Through continuous scanning, it can track the precise position of the last vehicle in the queue in real time, thereby calculating the queue length. High-definition camera 3 captures high-quality video streams and extracts rich traffic information through embedded computer vision algorithms for license plate recognition and vehicle type classification. High-definition camera 3 also serves as a sensor for detecting vehicles entering and exiting the waiting area. High-definition camera 3 is installed at a high position, with a field of view covering the entire lane area. By fusing data through the high-reliability presence detection of geomagnetic sensor 1, the all-weather motion and spatial perception of radar sensor 2, and the high-density visual recognition of high-definition camera 3, the system constructs a redundant, complementary, and accurate perception system. This system can overcome the limitations of a single sensor (such as the influence of weather on cameras, the inability of geomagnetism to measure speed, and the limited accuracy of radar recognition), providing comprehensive, three-dimensional, and realistic traffic data for the back-end data processing unit, thereby laying a solid data foundation for the final intelligent signal control decision.

[0136] Infrared thermal imaging sensor 4 and image acquisition device 5 are installed on the posts of the pedestrian walkways at the main road intersection, at a height of no less than 2.5 meters to avoid pedestrians bumping their heads. The installation height and orientation ensure complete coverage of the pedestrian waiting area. These two sensors work together to form an intelligent pedestrian detection system. Infrared thermal imaging sensor 4 detects the presence of pedestrians by detecting the infrared heat radiation emitted by their bodies. It does not rely on visible light and can operate stably under complex lighting conditions such as nighttime, fog, and backlighting. Infrared thermal imaging sensor 4 continuously tracks the movement of heat sources within the area, sensing the presence and movement of pedestrians in the main road's pedestrian waiting area. The system has a pedestrian stagnation time threshold, ideally between 5 and 10 seconds. Image acquisition device 5 serves as a verification and supplement to infrared detection, using computer vision technology to accurately identify, locate, and classify pedestrian targets. YOLO is used for image segmentation, feature extraction, bounding box prediction, and non-maximum suppression. Through computer vision and deep learning algorithms, pedestrians are detected in real time. Image acquisition device 5 employs the YOLO object detection algorithm, with the following processing flow: 1) Image segmentation and feature extraction: The algorithm quickly divides the acquired video frame images into an S×S grid. Each grid cell is responsible for predicting whether a pedestrian target exists within it, and uses a deep convolutional neural network to extract deep features of the image (such as contour, pose, and texture). 2) Bounding box prediction and confidence scoring: For each grid, the algorithm predicts one or more "bounding boxes" containing the target and simultaneously provides two scores: category confidence: the probability that the object within the box is a "pedestrian"; and localization accuracy: the degree of overlap between the predicted box and the actual pedestrian position. 3) Non-maximum suppression: To avoid repeated detection of the same pedestrian, this step filters out redundant predicted boxes with high overlap but low confidence, retaining only the most accurate box to ensure that each pedestrian is counted only once.

[0137] The system, combining infrared thermal imaging sensor 4 with continuous frame analysis from image acquisition unit 5, first defines a virtual "main road pedestrian waiting area." When a heat source (pedestrian) enters this area, a "pedestrian presence" signal is triggered. Stagnation time analysis (the core logic for identifying and eliminating false positives) is crucial for distinguishing between "waiting pedestrians" and "passing pedestrians." If a pedestrian enters the waiting area and stays for more than a threshold, they are initially identified as "a pedestrian intentionally waiting to cross the street." Conversely, if a pedestrian quickly passes through the area and stays for a short time (below the threshold), they are identified as "walking pedestrians," and the system will not record them. This effectively avoids misjudging pedestrians who are briefly stopping (such as waiting for someone or looking at a phone) as having a need to cross the street, greatly improving detection accuracy. The infrared thermal imaging sensor 4 and image acquisition unit 5 fuse data and collaborate in decision-making; infrared sensing and visual recognition do not work in isolation but are deeply integrated. Triggering and verification: The infrared sensor, with its rapid response characteristics, first detects potential pedestrian waiting and triggers the image acquisition unit for focused analysis. Information Complementarity: The system performs spatiotemporal matching of infrared data (heat source location, dwell time) with visual recognition results (pedestrian bounding boxes detected by YOLO, category). Final Judgment: Only when the infrared heat source stays in the waiting area for more than a time threshold, and the YOLO algorithm confirms the presence of a pedestrian target at the same location, does the system ultimately and reliably determine that the pedestrian has a need to wait to cross the main road. Control logic to guarantee pedestrian right-of-way: Once the system confirms that a pedestrian is waiting, it sends this signal as one of the highest priority passage requests to the data processing unit. When planning the next signal cycle, the data processing unit must allocate a green light time for pedestrians in an appropriate phase (usually after the vehicle phase with the least conflict with pedestrian crossing). The system dynamically adjusts the timing to ensure that pedestrians have sufficient and safe time to cross the road, truly realizing intelligent management from "vehicle-centric" to "pedestrian-vehicle collaboration."

[0138] The data processing unit is an embedded processing system. It is not a general-purpose computer, but a high-performance, high-reliability embedded processing system specifically designed for real-time traffic control. Employing a modular design, it integrates computing, interaction, communication, and self-diagnostic functions to ensure stable 24 / 7 operation. Its core architecture mainly includes the following four key modules: core control module, user interaction module, communication module, and fault detection and alarm module.

[0139] The core control module is the central hub of the entire traffic signal control system, responsible for the most critical intelligent decision-making tasks. It employs a multi-core processor, with high-performance cores running complex intelligent algorithms and real-time cores handling precise timing control. It performs real-time traffic flow analysis and signal timing decisions (calculation...). T mb , T sb , ΔT The core algorithm (etc.) dynamically adjusts the duration of traffic lights. It can process massive amounts of data from the data detection unit and make decisions within milliseconds. It monitors the system's operating status and takes measures when anomalies are detected to ensure stable system operation. The core control module, through a heterogeneous computing architecture, intelligent decision-making algorithms, millisecond-level response capabilities, and powerful self-monitoring functions, transforms raw traffic data into precise control commands. It evolves traffic signal control from rigid timing control into an intelligent entity capable of "thinking," "predicting," and "adapting," ultimately ensuring maximum intersection traffic efficiency and operational safety and reliability.

[0140] The user interaction module allows for system settings and adjustments via buttons, knobs, or a touchscreen. It uses an LCD or OLED display to show real-time system status and traffic light cycle information, providing an interface for maintenance personnel and administrators for system configuration and status monitoring. By combining the ease of use of the hardware interface with the comprehensiveness of the software functionality, the user interaction module significantly improves the maintainability and operability of the intelligent traffic control system. It eliminates the "blind box" nature of complex traffic control algorithms, allowing managers to gain a deep understanding of the system's operation through an intuitive interface and to perform precise and flexible configuration and optimization based on actual traffic characteristics, thereby ensuring the system always operates at its optimal state.

[0141] The communication module is responsible for data transmission and command interaction between systems, enabling synchronization and coordination between traffic light controllers, data transmission with the traffic management center, and supporting remote monitoring and management. The communication module is the "neural network" of the intelligent traffic signal control system, responsible for data exchange and command transmission between all internal components and between the system and the external environment. It employs various communication technologies and protocols to ensure the real-time performance, reliability, and security of information transmission, forming the foundation for intelligent collaborative control. By constructing a high-speed, reliable, and secure internal and external communication network, the communication module integrates isolated traffic signal equipment into an organic intelligent whole. It not only guarantees the real-time performance of intelligent control at individual intersections but also enables collaborative linkage among intersection groups and interconnection with the upper-level management center. It is a key infrastructure supporting the evolution of modern intelligent transportation systems from "single-point adaptive" to "regional adaptive" and "networked intelligence."

[0142] The fault detection and alarm module is responsible for monitoring the system's operational status, promptly detecting and handling anomalies, detecting key system parameters, and notifying maintenance personnel via audible and visual alarms or remote notification when anomalies occur. This module is a critical redundancy and safety assurance unit ensuring the continuous, stable, and reliable operation of the intelligent traffic signal control system. Like the system's "immune system," it proactively prevents, promptly detects, and quickly responds to various anomalies through continuous self-checking and environmental monitoring, minimizing the impact of faults on traffic operations. Through comprehensive monitoring, intelligent diagnostics, and hierarchical response, the fault detection and alarm module transforms traditional passive maintenance into proactive early warning and automatic fault tolerance, greatly improving the availability, reliability, and maintainability of the intelligent traffic signal control system. This ensures the system can cope with various emergencies and achieve stable 24 / 7 uninterrupted operation, thus providing a solid foundation for smooth and safe urban traffic.

[0143] Example 3: A crossroads where a main road and a side road meet; see the intersection plan. Figure 2 Intersections 1 and 2 are main roads, while intersections 3 and 4 are side roads. In the first phase, for left turns at intersections 1 and 2 on the main roads, the green light duration is 25 seconds, followed by a 5-second yellow light signal indicating the left-turn green light will turn red. In the second phase, due to higher traffic volume on the main roads, the green light duration for both straight and right turns is 45 seconds, with a buffer period before the 45-second green light ends. T bf Within the first phase (5 seconds), the system detects vehicles waiting in the waiting area of ​​the side road. After the main road's straight / right-turn green light ends, a yellow light signal indicates the signal will turn red after 5 seconds. In the third phase, since there are relatively few vehicles going straight, turning left, or turning right on the side road, a simultaneous straight / left / right-turn operation is adopted in both directions. Left-turning vehicles yield to straight-going vehicles. The green light duration is 25 seconds, and the 5-second yellow light indicates the end of the next signal cycle, starting the next cycle. See the traffic signal timing diagram for Example 3. Figure 4 .

[0144] Example 4: The intersection layout is the same as in Example 3. In Phase 1, for left turns at intersections 1 and 2 on the main road, the green light duration is 25 seconds, followed by a 4-second yellow light signal indicating the left-turn green light will turn red. In Phase 2, with more vehicles going straight on the main road, the green light duration for both straight and right turns is 45 seconds, with a buffer period before the 45-second green light ends. T bf Within 5 seconds, there are still vehicles in the main road waiting area, but the system detects no vehicles waiting in the side road waiting area; in the third phase, the system automatically extends the main road green light for straight-ahead and right-turn traffic by 30 seconds. ΔT The green light on the main road will be extended by 30 seconds. ΔT Buffer time before end T bfWithin 5 seconds, the system detects a vehicle entering the waiting area from the side road, and the main road extends its green light time by 30 seconds. ΔT The light then turns red; in the fourth phase, there are relatively few vehicles going straight, turning left, or turning right on the side road, so a simultaneous straight-ahead, left-turn, and right-turn operation is used. Left-turning vehicles yield to straight-ahead vehicles. The green light lasts for 25 seconds, and the yellow light lasts for 4 seconds. After this signal, one signal cycle ends and the next cycle begins. See the traffic signal timing diagram for Example 4. Figure 5 .

[0145] Example 5: The intersection layout is the same as in Example 3. Phase 1 and Phase 2 are the same as in Example 4. A buffer period of 45 seconds is included before the green light signal ends in Phase 2. T bf Within 5 seconds, the system detects that there are many vehicles in the straight-ahead and left-turn waiting areas of intersection 1, while there are few or no vehicles in the straight-ahead and left-turn waiting areas of the opposite intersection 2; in the third phase, the system automatically controls the green light signal for both straight-ahead and right-turn on the main road side of intersection 1 to be extended by 30 seconds. ΔT At intersection 2, the light turns red for both straight and left traffic, while the green light at intersection 1 is extended by 30 seconds. ΔT Buffer time before end T bf Within 5 seconds, the system detects a vehicle entering the waiting area from the side road, and intersection 1 completes a 30-second extension. ΔT After the green light signal, the yellow light lasts for 4 seconds before turning red; in the fourth phase, there are fewer vehicles going straight, turning left, and turning right on the side road, so a simultaneous straight, left, and right turn operation is adopted in both directions. Left-turning vehicles yield to straight-going vehicles. The green light signal lasts for 25 seconds, and the 4-second yellow light indicates the end of the next signal cycle and the start of the next cycle. See the traffic signal timing diagram for Example 5. Figure 6 .

[0146] In embodiments three to five above, right turns are controlled by signals. If the intersection has a dedicated right-turn lane, it can be set so that vehicles turning right on red have to yield to other vehicles and pedestrians before proceeding. In embodiments four and five, a one-second all-red signal is set after the yellow light, with all directions having red lights for a period of time to clear vehicles from the intersection. This is only one form of control and does not mean that it must be set, nor is it a limitation of this application.

[0147] The specification and drawings of this application are merely one specific embodiment and are not restrictive. Those skilled in the art can make many other modifications based on the teachings of this application without departing from the spirit and scope of this application, and all such modifications are within the scope of protection of this application.

Claims

1. A method for intelligent traffic control at an asymmetric intersection, characterized by: The intelligent traffic 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 on the main road of an intersection ends and there are no vehicles on the intersecting side road, the method automatically adjusts the signal timing to extend the green light signal duration on the main road and improves the traffic efficiency of the intersection. The traffic data collection uses traffic flow detectors and vehicle detectors to detect vehicles entering the waiting area at each intersection, as well as pedestrians waiting to cross the road at the intersection, and to monitor the vehicle flow, time distribution, traffic patterns, and vehicle type distribution on the main road and side roads; and to collect real-time traffic flow and vehicle queue length data on the main road and side roads as a basis for signal timing. The data processing and analysis involves transmitting the collected traffic data to the data processing unit. The system processes and analyzes the data, performing preprocessing, feature extraction, and parameter optimization on the main road and branch road traffic flow data based on the traffic data from the data detection unit. Through Fourier transform, periodic features are extracted from the complex time-domain signal, analyzing the periodicity of traffic flow data, including the periodic changes during morning peak, evening peak, and off-peak periods, the differences in traffic flow between weekdays and weekends, and the changes in traffic flow during different seasons or holidays. The analysis results serve as the basis for intelligent signal adjustment. By analyzing traffic flow and vehicle queuing, the system determines the traffic conditions of the main road and branch roads and sets the basic green light time for the main road. T mb Basic green light time for side streets T sb The system uses historical traffic data to train a regression model through self-learning, and employs an embedded system control algorithm to calculate the function. f ( Q h ), calculate the main road green light extension time Δ T It generates signal control commands and calculates intersection timing schemes and real-time signal adjustment parameters based on the data of vehicles and pedestrians waiting to cross the road in the waiting areas of each intersection. Different traffic fluctuation cycles were determined using Fourier transform. T For different traffic fluctuation cycles T The detected traffic flow is used as historical traffic flow data. Q h This data is saved and used to extract matching data during traffic signal control, enabling dynamic adjustment of the basic green light time on the main road. T mb Basic green light time for side roads T sb Extended green light time on main road Δ T ; The historical traffic flow data Q h The historical average model is used to predict future traffic flow using historical traffic data. The calculation formula is as follows: Q h ( t )= γ · Q h ( t -1)+(1- γ )· Q ( t )， in: t For time, Q ( t ) is in time t The actual observed flow rate; γ It is a smoothing coefficient, with a value range of [0,1], used to adjust the weight of historical data; The basic green light time on the main road T mb Calculated using the following formula: T mb = Q k · h+t s ， in: Q k The maximum number of vehicles that can pass through a green light, in vehicles; h Saturated headway, unit: seconds / vehicle, represents the average time interval between vehicles passing through the intersection under saturated conditions; t s Initiation loss, unit: seconds; The basic green light time of the branch road T sb Calculated using the following formula: T sb =max( T min , T d + T bf ), in, T min Minimum green light time for a branch road, in seconds. T d The time difference from the moment a vehicle was detected entering the waiting area on the side road to the present moment, in seconds. T bf Buffer time, in seconds; The main road green light extension time Δ T Calculated using the following formula: D T = α ·( Q h - Q t )+ β , in: Q h Historical traffic flow data for the main road, unit: vehicles / unit of time; Q t It is the traffic threshold, measured in vehicles per unit of time, used to determine whether the current traffic flow requires adjustment of the signal time. α and β It is an adjustment coefficient used to control Δ T Scope α Control the increase or decrease of the extension time, in seconds / (vehicle / unit time). β This is the base extension time, in seconds; The system uses historical traffic data to train a regression model through self-learning and calculation. α , β and Q t The optimal value varies depending on the season and peak time period. α , β , Q t The calculation of the main road green light extension time Δ varies depending on the value of Δ. T At that time, the system automatically selects the matching traffic scenario. α , β , Q t value; The signal timing decision is made based on data analysis results. The intelligent traffic control system performs signal timing decisions according to preset algorithms and strategies. If the system determines that there are no vehicles on the intersecting side roads before the main road green light signal ends, it adjusts the main road green light extension time Δ. T ; The signal control system transmits the signal timing decision results to the signal control equipment to control the traffic lights on the main road and the side roads. When there are no vehicles on the side roads intersecting with the main road, the system extends the green light time on the main road to allow vehicles on the main road enough time to pass, while the waiting time of the traffic lights on the side roads will be extended accordingly. The aforementioned monitoring and adjustment system continuously monitors traffic flow and vehicle queuing, and makes adjustments based on real-time data. If the traffic conditions on the main road change or vehicles are waiting on the side roads, the system will re-make signal timing decisions based on the new data to ensure efficient and safe traffic. The specific control method is as follows: when the main road's basic green light time... T mb Buffer time before end T bf Inside, there are still vehicles in the main road waiting area, and the system checks whether there are vehicles waiting in the side road waiting areas; If there are vehicles waiting in the waiting area on the side road, the main road's green light time will be approximately [time missing]. T mb After the event, the main road light turns red, and the side road light turns green. If there are no vehicles waiting in the side road waiting area, the system will automatically extend the green light time on the main road. ΔT ; Extend the green light time on the main road ΔT Buffer time before end T bf Inside, there are still vehicles in the main road waiting area, and the system detects whether any vehicles from side roads have entered the waiting area; If a vehicle enters the waiting area from a side road, the extended waiting time will continue on the main road. ΔT After the green light signal turns red, the basic green light time for the branch road begins. T sb Signal; If the time is extended ΔT Buffer time before end T bf Inside, no vehicles have entered the waiting area on the side road, but there are still vehicles in the waiting area on the main road. The waiting time on the main road will be extended by another period. ΔT ; During the second extended time ΔT If a vehicle enters the waiting area from the side road, the main road light turns yellow as a buffer signal, then red, and the side road then enters its basic green light period. T sb Signal; If the second extension time ΔT No vehicles entered the waiting area on the middle branch road, and the extended time for the main road was extended. ΔT After the green light signal turns yellow (a buffer period) and then red, it transitions to the basic green light time for the branch road. T sb Signal; If the main road has a second extended time ΔT When the system detects pedestrians waiting to cross the main road, the main road light turns yellow for a buffer period, then turns red, and the secondary road enters its basic green light period. T sb Signal; If all vehicles in the main road waiting area have left within the first or second extended time ΔT, the main road light turns red, and the secondary road light becomes essentially green. T sb Signal; Basic green light time for side roads T sb After the green light ends, the main road will return to its basic green light period. T mb Moving on to the next cycle.

2. The intelligent traffic control method for asymmetric intersections according to claim 1, characterized in that: The basic green light time on the main road T mb Basic green light time for side roads T sb Extended green light time on main road Δ T The system dynamically adjusts its signals based on real-time traffic data from main roads and secondary roads, combined with historical traffic flow data. Historical traffic flow data is obtained by performing a Fourier transform on the real-time collected main road and secondary road traffic data, converting the traffic flow data from the time domain to the frequency domain. This detects the periodicity in the data and is used to analyze the periodic characteristics of traffic flow data, including periodic changes during morning peak hours, evening peak hours, and off-peak periods, differences in traffic flow between weekdays and weekends, and changes in traffic flow during different seasons or holidays. The analysis results serve as the basis for intelligent signal adjustment. The specific steps for calculating the traffic fluctuation cycle using Fourier transform are as follows: 1) Data normalization scales the traffic data to the range [0, 1] to improve the model's convergence speed and prediction accuracy. The formula is as follows: ， in: Q ′ represents the raw traffic flow data. Q min It is the minimum value in the data. Q max It is the maximum value in the data; 2) Normalized flow data Q Perform a discrete Fourier transform to obtain the frequency domain signal. F ( u The Fourier transform formula is: ， in: Q [ n [This refers to a specific point in time.] n Traffic flow data, N It is the total number of data points. u It is a frequency index. u =0,1,2,…,N-1 F ( u ) is the result of the frequency domain transformation; 3) Calculate the amplitude of the frequency domain signal | F ( u Find the frequencies corresponding to the first few peak values ​​with the largest amplitude. u When the peak appears u=k When the time is specified, the corresponding period is: T=N / k , k The peak position is indexed in the frequency domain; the periods corresponding to the first few peaks are selected as the control periods for the candidate signals.

3. A system for the intelligent traffic control method for asymmetric intersections as described in claim 1, characterized in that: the system include: Data detection unit, data processing unit, signal control unit, data storage unit; The data detection unit, This system is used to detect vehicles entering the waiting area at various intersections, as well as pedestrians waiting to cross the road. It collects traffic data on main roads and side roads in real time and monitors vehicle flow, time distribution, traffic patterns, and vehicle type distribution on main roads and side roads. The traffic data includes traffic flow, vehicle queue length, vehicle presence information, and pedestrian waiting information. The data processing unit, Connected to the data detection unit, it is used for traffic data processing and analysis, signal timing decision-making, and generation of control commands; wherein the signal timing decision-making includes, based on the results of data analysis, the intelligent traffic control system makes signal timing decisions according to preset algorithms and strategies, and if the system determines that there are no vehicles on the intersecting side roads before the green light signal of the main road ends, it extends the green light time of the main road. The signal control unit Connected to the data processing unit, it is used to control the switching of main road and branch road traffic lights according to the timing scheme and real-time signal adjustment parameters of the data processing unit, and to adjust the green light extension time Δ of the main road in real time. T It controls the traffic lights on the main road and branch roads, realizing intelligent control of traffic signals, and displays the remaining time of the traffic lights through digital tubes or displays. The data storage unit Connected to the data processing unit, it is used to store traffic flow data, system operation logs, and reports that support data analysis and decision-making; it 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.

4. The system of the intelligent traffic control method for asymmetric intersections according to claim 3, characterized in that: The data detection unit includes a geomagnetic sensor (1), a radar sensor (2), a high-definition camera (3), an infrared thermal imaging sensor (4), and an image acquisition device (5). The data detection unit is used to collect traffic data of the main road and branch roads in real time. The traffic data includes traffic flow, vehicle queue length, vehicle presence information, and pedestrian waiting information. The geomagnetic sensor (1), radar sensor (2), and high-definition camera (3) are set at the entrance and exit of the waiting area of ​​each lane at each intersection. The geomagnetic sensor (1) is used to detect the presence of vehicles. By detecting vehicles entering the waiting area and vehicles leaving the waiting area, the system calculates the vehicles in the waiting area for intelligent lane signal control. The radar sensor (2) is used to measure vehicle speed and queue length and calculate the time required to leave the waiting area; The high-definition camera (3) is used for license plate recognition and vehicle type classification. The high-definition camera (3) also serves as a sensor for detecting vehicles entering and leaving the waiting area. The infrared thermal imaging sensor (4) and the image acquisition device (5) are set on the sidewalk at the intersection of the main road, covering the waiting area for pedestrians waiting to cross the main road. The infrared thermal imaging sensor (4) senses the presence of pedestrians in the main road waiting area and sets a threshold for pedestrian stagnation time. The infrared thermal imaging sensor (4) combines the continuous frame analysis of the image acquisition device (5). If the pedestrian stays longer than the threshold and is located in the main road waiting area, it is determined to be a pedestrian waiting to cross the road. If the stagnation time is less than the threshold, it is determined to be a pedestrian walking and is not recorded as a pedestrian waiting to cross the road. The image acquisition device (5) uses YOLO for image segmentation, feature extraction, bounding box prediction, and non-maximum suppression. Through computer vision and deep learning algorithms, it detects pedestrians in real time and identifies pedestrian targets waiting to cross the main road.

5. The system of the intelligent traffic control method for asymmetric intersections according to claim 3, characterized in that: 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 central hub of the entire traffic signal control system. It receives traffic flow data from the data detection unit, analyzes and processes it, dynamically adjusts the duration of traffic lights based on the traffic flow data and preset control algorithms, monitors the system's operating status, and takes measures when anomalies are detected to ensure stable system operation. The user interaction module allows for system settings and adjustments via buttons, knobs, or touchscreens. It uses an LCD or OLED display to show system status and signal light cycle information in real time, providing an operating interface for maintenance personnel and administrators for system configuration and status monitoring. The communication module is responsible for data transmission and command interaction between systems, synchronization and coordination between traffic light controllers, data transmission with the traffic management center, and support for remote monitoring and management. The fault detection and alarm module is responsible for monitoring the system's operating status, promptly detecting and handling anomalies, detecting key system parameters, and notifying maintenance personnel via audible and visual alarms or remote notification when an anomaly occurs.

Images