A real-time network-level traffic signal control method based on a connected vehicle (CV) information assisted prediction model

Abstract

The patent discloses a real-time network-level traffic signal control method based on connected vehicle (CV) information auxiliary prediction model. The method first collects traffic data on the road through sensors, including vehicle speed, vehicle density and traffic flow, etc., and uses the Wiedemann following model to preprocess and analyze these data with the assistance of connected vehicle information. Finally, the Kalman filtering method is used to predict the traffic state. The dynamic signal control strategy is adjusted based on the above data: 1. According to the real-time traffic state, the length and phase of the traffic signal are dynamically adjusted to realize adaptive signal control. 2. By coordinating the signal control of adjacent intersections, the continuity and coordination of traffic flow are realized, and the stagnation and delay of vehicles are reduced.

Description

Technical Field

[0001] This invention relates to the field of road traffic signal control, and more particularly to a real-time network-level traffic signal control method based on a prediction model assisted by connected vehicle (CV) information. Background Technology

[0002] With the continuous development of urban transportation, traffic congestion has become increasingly serious. To effectively alleviate traffic pressure and improve the efficiency of traffic flow management and control, researchers have proposed many traffic signal control systems based on real-time data. These systems dynamically adjust traffic signals by acquiring real-time traffic data on the road to optimize traffic flow. However, existing technologies have certain shortcomings in data acquisition, processing accuracy, and signal control strategies, making it difficult to achieve optimal traffic management results in complex urban traffic environments. Therefore, to solve these problems, it is necessary to update road traffic signal control methods. Summary of the Invention

[0003] The purpose of this invention is to address the shortcomings in the real-time signal data adjustment capabilities of the aforementioned background technologies, and to provide a real-time network-level traffic signal control method based on a vehicle-to-everything (CV) information-assisted prediction model. This method uses partial connected vehicle information and vehicle information collected by real-time detectors as preprocessed real-time basic data, and uses the Wiedemann vehicle following model to filter the data and predict future data, further enabling joint control and real-time decision-making for interconnected cross-traffic signals.

[0004] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0005] A real-time network-level traffic signal control method based on a prediction model assisted by connected vehicle (CV) information, the method specifically includes the following steps:

[0006] S1: Based on the continuous road signal control system network, a signal control road network is selected for a certain area during peak hours. Simultaneously, at the entrance to this area, the following are established: 1) Corresponding ring-shaped vehicle detectors are used to capture (UV) non-connected vehicle data. 2) (CV) connected vehicle sensors are used to collect connected vehicle data. Finally, based on the existing materials, real-time data is collected. UV vehicle data, due to its large volume and frequent data reception by the receiver, will be somewhat coarse. CV data, due to the smaller proportion of CV vehicles, can obtain more accurate real-time data within the area.

[0007] S2: Based on the above data, firstly, the UV data is arranged in time series as a large sample. Then, the CV data is interspersed within the UV data using the Wiedemann vehicle following model. Based on the interaction between the CV and adjacent vehicles, the UV data is corrected and filtered. The Wiedemann vehicle following model based on the CV data handles the interaction between the CV and UV vehicles.

[0008]

[0009] Among them, F t，j V represents the interaction force between vehicle i and vehicle j. i and v j Let d represent the speeds of vehicle i and vehicle j, respectively. ij This represents the distance between the two vehicles. a and b are model parameters that are determined by the real-time local road surface material and traffic density.

[0010] Furthermore, based on the mixed UV and CV data and the data modules between adjacent vehicles, adjustments are made to select sample data that conforms to the Wiedemann vehicle following model as the final processed real-time detection data.

[0011] S3: Further, based on the above data, the Kalman filter method is used to predict traffic conditions. The core formula of the Kalman filter method based on CV and UV data is as follows:

[0012] 1) Prediction steps:

[0013]

[0014] P k|k-1 =AP k-1|k-1 A T +Q (3)

[0015] 2) Update steps:

[0016] K k =P k|k-1 H T HP k|k-1 H T +R) -1 (4)

[0017]

[0018] P k|k =(IK k H)P k|k-1 (6) in For the state prediction at time k, P k|k-1 For the prediction error covariance matrix, Kk For Kalman gain, z k Let A, B, H, Q, and R be the observed values, and let A, B, H, Q, and R be the state transition matrix, control input matrix, observation matrix, process noise covariance matrix, and observation noise covariance matrix, respectively.

[0019] Furthermore, by setting the road intersection signal control data as the prediction step and applying Kalman filtering, it is necessary to define the state-space model of the system, including the state transition model: x k =Ax k-1 +Bu k +W k And observation model: z k =Hx k +V k At the initial moment of the Kalman filter, the state vector and covariance matrix are first initialized. Further, under time updates, prior state estimates are used to predict the current state and covariance matrix, thereby obtaining traffic data for the next time step. At each time step k, the time update and observation update steps are repeated to achieve optimal estimation of the system state. Through the above steps, this invention can predict traffic conditions in real time and dynamically adjust traffic signal control strategies based on the prediction results.

[0020] S4: Based on existing CV and UV data and predicted data, the data is converted into traffic state data. The duration and phase of traffic signals are dynamically adjusted according to the predicted traffic state. Specific adjustment strategies include:

[0021] Phase adjustment: Phase adjustment refers to dynamically adjusting the phase sequence of traffic lights based on the direction and density of traffic flow to optimize traffic efficiency. The specific implementation steps of this process are as follows:

[0022] 1. By collecting and processing real-time data, analyze the traffic flow direction at each intersection, including the proportion of vehicles going straight, turning left, and turning right. Understanding the traffic demand of vehicles in different directions helps optimize the phase sequence of traffic lights.

[0023] 2. Establish a phase optimization model based on traffic flow direction and density. The model considers the differences in traffic flow in each direction and uses mathematical algorithms to calculate the optimal traffic light phase sequence to maximize intersection efficiency. For example, during peak hours, priority is given to setting green light durations for directions with higher straight-through traffic to reduce congestion; during off-peak hours, green light durations are balanced across all directions to improve overall traffic efficiency.

[0024] 3. Based on the calculation results of the phase optimization model, the phase sequence of traffic lights is dynamically adjusted. The system monitors traffic conditions in real time, and when a significant increase in traffic flow in a certain direction is detected, the green light phase for that direction is automatically adjusted to ensure that vehicles can pass through the intersection quickly.

[0025] 4. The system incorporates a feedback mechanism for phase adjustment, monitoring traffic flow changes after signal adjustments to evaluate the effectiveness of the adjustments. If the adjusted traffic efficiency is found to be less than expected, the system will automatically recalculate and adjust the phase sequence to continuously optimize traffic signal control.

[0026] S5: Based on the traffic flow information exchange mechanism between adjacent intersections, this coordinates the dynamic interaction capabilities between intersections. Specifically, it refers to coordinating signal control at adjacent intersections through a central control platform to achieve traffic flow continuity and coordination, reducing vehicle stagnation and delays. The specific implementation steps of this process are as follows:

[0027] 1. The central control platform is the core of the entire traffic signal control system. It is responsible for receiving, processing and analyzing real-time traffic data from various intersections and coordinating the signal control at each intersection.

[0028] 2. The platform employs a coordinated control algorithm to calculate the optimal signal coordination scheme based on real-time traffic conditions at each intersection. The algorithm considers the distance between adjacent intersections, traffic flow, and the duration and phase of traffic lights to ensure effective coordination of signals at each intersection, reducing vehicle stagnation and waiting time.

[0029] 3. In coordinated control, green wave control technology is the key application. A green wave refers to a series of consecutive intersections where vehicles can encounter green lights sequentially during their journey, reducing the number of stops and waiting time, and improving traffic efficiency. By precisely calculating and adjusting the signal light duration and phase at each intersection, green wave control is achieved, allowing vehicles to pass through multiple intersections continuously on the main road.

[0030] 4. The central control platform monitors traffic conditions in real time and dynamically adjusts coordination and control strategies based on changes in traffic flow. For example, when traffic volume at a certain intersection increases significantly, the system will appropriately adjust the signal light duration and phase at that intersection and its adjacent intersections to ensure smooth and continuous traffic flow. Attached Figure Description

[0031] Figure 1 This is a flowchart illustrating the steps.

[0032] Figure 2 To study road diagrams; Detailed Implementation

[0033] To clearly illustrate the present invention, the invention will be further described in conjunction with examples and accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0034] This plan selects three consecutive intersections in the actual road segment, such as... Figure 2As shown, various sensors, including loop detectors, cameras, and connected vehicle sensors, are installed on the main roads. These sensors collect real-time traffic data from non-connected vehicles (UV), including vehicle speed, traffic flow, vehicle density, and location. Combined with connected vehicle sensors, the precise location, speed, acceleration, and direction of travel of connected vehicles (CVs) can be obtained. This data is transmitted wirelessly to a central control platform for analysis and processing.

[0035] Furthermore, the three consecutive intersections are named A1, A2, and A3. First, signal sensor placement is performed at the beginning and end positions of the signal input and output segments for A1 and A3, respectively. Second, this data is used as preprocessed data, primarily as input to the Wiedemann vehicle-following model. At the instant vehicles i and j pass through the circular detection period, their speed and acceleration are recorded as microscopic vehicle states. Real-time vehicle information from connected vehicles (including real-time position dispersion, speed, and acceleration information) is used to screen for vehicles that do not conform to the connected vehicle environment at a macroscopic level. Simultaneously, joint control is implemented based on UV (unique vehicle) data from intersections A1, A2, and A3. The next step involves selecting peak hours with strong vehicle continuity.

[0036] Furthermore, in this invention, the Kalman filtering method is used for traffic state prediction, and its implementation process includes the following steps:

[0037] 1) First, a state-space model of the system needs to be defined, which includes a state transition model and an observation model: this model describes the change of the system state from the previous moment to the current moment. The state vector represents the state of the traffic system. The state transition matrix is ​​used to predict the state at the current moment, and the control input matrix is ​​used to integrate any external control inputs (e.g., changes in traffic lights). Furthermore, process noise reflects random variations or uncertainties in the system; this model describes the relationship between observations and the system state. The observation vector represents the actual measured traffic data, such as vehicle speed and position. The observation matrix maps the system state to the observation space, and the observation noise represents random errors in the measurements.

[0038] 2) Initialization: In the initial stage of Kalman filtering, the state vector and covariance matrix need to be initialized. The initial state estimate represents the best state estimate of the system at the initial time. This is usually based on historical data or empirical estimation.

[0039] 3) Time Update (Prediction Step): The time update step is used to predict the current state and covariance matrix using prior state estimates.

[0040] 4) Observation Update (Correction Step): The observation update step uses the observation data at the current time to correct the prior estimate, obtaining the posterior estimate: Kalman gain calculation (a weighting coefficient that determines the impact of the observation data on the state estimate. The calculation of the Kalman gain considers the prediction error covariance and observation noise).

[0041] 5) State Update: Update the state estimate based on the current observation data. This step combines the observation data with the prior state estimate to obtain the corrected optimal state estimate.

[0042] 6) Recursive execution: The above time update and observation update steps are executed recursively at each time step to continuously update the system's state estimate and covariance matrix.

[0043] Furthermore, based on predicted traffic conditions, the duration and phase of traffic signals are dynamically adjusted to achieve adaptive signal control. Specific adjustment strategies include:

[0044] Duration Adjustment: Based on real-time traffic flow data, the duration of green, red, and yellow lights for each traffic light is dynamically adjusted. For example, in areas with high traffic volume, the green light duration is extended to reduce vehicle queuing time and improve traffic efficiency; in areas with low traffic volume, the green light duration is appropriately shortened to avoid unnecessary waiting. Simultaneously, the yellow light duration is adjusted appropriately based on the average vehicle speed to ensure vehicles can safely pass through the intersection.

[0045] Phase adjustment: Based on the direction and density of traffic flow, the phase sequence of traffic lights is dynamically adjusted to optimize traffic flow efficiency. In practice, traffic flow direction analysis is generally used.

[0046] Traffic flow direction analysis: Through real-time data collection and processing, the traffic flow direction at each intersection is analyzed, including the proportion of vehicles going straight, turning left, and turning right. Understanding the traffic demand from different directions helps optimize the phase sequence of traffic lights.

[0047] Phase optimization model: A phase optimization model is established based on traffic flow direction and density. The model considers the differences in traffic volume in each direction and uses mathematical algorithms to calculate the optimal signal light phase sequence to maximize intersection efficiency. For example, during peak hours, green light time is prioritized for directions with higher through traffic volume to reduce congestion; during off-peak hours, green light time is balanced across all directions to improve overall traffic efficiency.

[0048] Dynamic phase adjustment: Based on the calculation results of the phase optimization model, the phase sequence of traffic lights is dynamically adjusted. The system monitors traffic conditions in real time, and when a significant increase in traffic flow in a certain direction is detected, the green light phase for that direction is automatically adjusted to ensure that vehicles can pass through the intersection quickly.

[0049] Phase Adjustment Feedback Mechanism: The system is equipped with a phase adjustment feedback mechanism to evaluate the effectiveness of the phase adjustment by monitoring traffic flow changes after signal adjustment. If the adjusted traffic efficiency is found to be less than expected, the system will automatically recalculate and adjust the phase sequence to continuously optimize traffic signal control.

[0050] In terms of coordinated control, a central control platform coordinates signal control at adjacent intersections to achieve continuity and coordination of traffic flow, reducing vehicle stagnation and delays. The specific implementation steps are as follows:

[0051] Coordination Control Algorithm: The platform employs a coordination control algorithm to calculate the optimal signal coordination scheme based on real-time traffic conditions at each intersection. The algorithm considers the distance between adjacent intersections, traffic flow, and the duration and phase of traffic lights to ensure effective coordination of signals at each intersection, reducing vehicle stagnation and waiting time.

[0052] Green wave control: Green wave control technology is a key application in coordinated traffic control. A green wave refers to a series of consecutive intersections where vehicles can sequentially encounter green lights during their journey, reducing the number of stops and waiting times, and improving traffic efficiency. By precisely calculating and adjusting the signal light duration and phase at each intersection, green wave control is achieved, allowing vehicles to pass through multiple intersections continuously on main roads.

[0053] Dynamic adjustment and optimization: The central control platform monitors traffic conditions in real time and dynamically adjusts the coordination and control strategy according to changes in traffic flow. For example, when the traffic volume at a certain intersection increases significantly, the system will appropriately adjust the signal light duration and phase at that intersection and its adjacent intersections to ensure smooth and continuous traffic flow.

[0054] Through the specific implementation methods described above, this invention enables intelligent and dynamic control of urban road traffic signals, improves the continuity and coordination of traffic flow, reduces vehicle stagnation and delays, significantly improves traffic conditions, and enhances road traffic efficiency. The system demonstrates good stability and adaptability in practical applications, and has broad promotional value and application prospects.

Claims

1. A real-time network-level traffic signal control system based on a networked vehicle (CV) information-assisted prediction model, comprising the following modules: S1: Data acquisition module, used to acquire and connect traffic data on the road in real time; S2: Data processing module, used to preprocess and analyze the collected traffic data and estimate the current traffic status; S3: Signal control module, which dynamically adjusts the duration and phase of traffic signals according to traffic conditions; S4: Central control platform, which coordinates the work of various modules to achieve dynamic control and coordination of traffic signals.

2. The system according to claim 1 detects the presence of unconnected vehicles by means of a loop detector, camera or other traffic sensing device (UV) and records relevant data, including the number of vehicles, speed and acceleration information.

3. The system according to claim 1 utilizes vehicle-to-everything (V2X) technology to acquire real-time data from connected vehicles (CVs). This data includes information such as the vehicle's precise position, speed, acceleration, and direction of travel. Real-time collection and updating of CV data is achieved through communication between the vehicle and the traffic control system. Simultaneously, a Wiedemann following model is employed to dynamically update the state of the UVs based on the relative position, speed, and distance between the CVs and UVs. The Wiedemann model refines and assists in optimizing all vehicle data by calculating the interactions between vehicles.

4. The system according to claim 1, in combination with refined data, uses the Kalman filter method to predict traffic conditions using the state estimate of the previous time step and the measurement value of the current time step.

5. The system according to claim 1, based on real-time traffic conditions and future traffic forecast data: 1) dynamically adjusts the duration and phase of traffic signals to achieve adaptive signal control; 2) coordinates signal control between adjacent traffic lights based on linkage between adjacent lights. This achieves continuity and coordination of traffic flow, optimizing the efficiency of vehicles passing through intersections.

Images