Traffic flow multi-dimensional control method based on mutual feedback adjustment under vehicle-road cooperation

Abstract

The invention relates to the technical field of intelligent traffic, in particular to a traffic flow multi-dimensional control method based on mutual feedback adjustment under vehicle-road cooperation, which comprises the following steps: collecting real-time traffic data of each intersection; after every preset time period, recommendation information is obtained in a preset mode; the preset mode comprises the following steps: respectively predicting the traffic flow of the entrance lane corresponding to the passing direction controlled by each signal phase in each intersection at the moment t; calculating the predicted vehicle waiting time of each entrance lane at the moment t based on the current queuing length and the signal phase state of each entrance lane; planning an optimal path for the target vehicle based on the predicted waiting time of each entrance lane and the road section driving time in the subsequent passing path; based on the predicted traffic flow of each entrance lane, determining the total duration of the signal period of the intersection, and optimizing the green light time ratio of each signal phase; and pushing the recommendation information to a vehicle navigation system corresponding to the target vehicle.

Description

Technical Field

[0001] The present invention relates to the field of intelligent transportation technology, and in particular to a multi-dimensional traffic flow control method based on mutual feedback regulation under vehicle-road collaboration. Background Art

[0002] To improve road efficiency and reduce congestion, traditional traffic management methods are gradually moving towards intelligent and networked approaches. As a key component of the next-generation intelligent transportation system, the Cooperative Vehicle Infrastructure System (CVIS) has become a crucial technology for improving urban traffic efficiency by enabling information exchange between vehicles and road infrastructure.

[0003] Existing V2X traffic control methods primarily employ signal control schemes based on fixed schedules or periodic adjustment strategies, lacking effective response to the real-time dynamics of traffic flow. This approach struggles to make timely optimization adjustments in the face of unexpected traffic incidents, rapidly changing traffic flows, or uneven local road loads, leading to problems such as signal timing lags, increased queue lengths, and reduced traffic efficiency. Furthermore, existing solutions often consider traffic control and route guidance separately, failing to achieve coordinated regulation between route guidance and signal control, thus limiting the overall effectiveness of V2X systems. Summary of the Invention

[0004] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides a multi-dimensional traffic flow control method based on mutual feedback regulation under vehicle-road collaboration.

[0005] In order to achieve the above objectives, the main technical solutions adopted by the present invention include:

[0006] An embodiment of the present invention provides a multi-dimensional traffic flow control method based on mutual feedback regulation under vehicle-road collaboration, comprising the following steps:

[0007] S1. Traffic sensing devices deployed at each intersection in the urban road network collect real-time traffic data at each intersection. The real-time traffic data includes: traffic flow at the intersection entrance, queue length, vehicle speed within the entrance, current state of the corresponding signal phase, and road load information for each entrance;

[0008] S2. After each preset time period, based on the collected real-time traffic data, obtain recommended information through a preset method; the preset method includes:

[0009] The weighted moving average method is used to predict the traffic flow at time t at the entrance lane corresponding to the traffic direction controlled by each signal phase at each intersection.

[0010] Calculate the estimated waiting time for vehicles at each entrance at time t based on the current queue length and signal phase state of each entrance;

[0011] Based on the estimated waiting time of each entrance lane and the travel time of the subsequent road segments, the optimal path is planned for each target vehicle within the perception area of ​​the intersection perception device;

[0012] Based on the predicted traffic flow of each entrance, the total signal cycle duration of the intersection is determined, and the green light time ratio of each signal phase is optimized;

[0013] The recommended information includes: route guidance suggestions, intersection signal status, and estimated waiting time for each entrance;

[0014] S3. Directly push the recommendation information to the vehicle navigation system corresponding to the target vehicle involved in the real-time traffic data through the Internet of Vehicles, so that it can perform route selection and dynamic adjustment; wherein, the target vehicle is a vehicle in the perception area of ​​the intersection perception device, and is associated and identified based on the vehicle's unique identifier and location data.

[0015] Preferably, at any intersection, the traffic flow Q at the entrance road corresponding to the direction of travel controlled by the i-th signal phase at time t is p (i, t) is calculated by the following weighted moving average formula (1);

[0016] The weighted moving average formula (1) is:

[0017]

[0018] Q(tk·Δt) represents the actual traffic flow observation value of the entrance lane corresponding to the traffic direction controlled by the i-th phase at the historical time tk·Δt;

[0019] γ∈(0,1) is the time series attenuation factor, which is used to control the influence weight of earlier historical data;

[0020] Δt is the time sampling interval;

[0021] N is the number of historical moments participating in the weighted average;

[0022] k represents the historical moment number.

[0023] Preferably, the time series attenuation factor γ is in the range of 0.6-0.8 during the morning and evening traffic peak hours.

[0024] During the off-peak period, the timing attenuation factor γ is in the range of 0.3-0.5.

[0025] Preferably, at any intersection, the estimated waiting time W of the vehicle at the entrance lane corresponding to the traffic direction controlled by the i-th signal phase at time t is p (i, t) is calculated by formula (2):

[0026] The formula (2) is:

[0027]

[0028] L p (i, t) represents the queue length at time t at the entrance corresponding to the traffic direction controlled by the i-th signal phase;

[0029] C(i) represents the traffic capacity per unit time of the entrance lane corresponding to the traffic direction controlled by the i-th signal phase;

[0030] G(i, t) represents the current or upcoming green light duration of the entrance lane corresponding to the direction of travel controlled by the i-th phase;

[0031] R(i, t) represents the remaining time of the current red light on the entrance lane corresponding to the direction of travel controlled by the i-th phase;

[0032] δ is a pre-set adjustment factor;

[0033] ∈ is a small constant to avoid the denominator being zero.

[0034] Preferably, based on the expected waiting time of each entrance road and the travel time of the road section in its subsequent passage path, the optimal path is planned for each target vehicle in the sensing area, specifically including:

[0035] Based on the expected waiting time of each entrance road and the road travel time of each candidate path in its subsequent passage path, the path cost function C of the target vehicle in the perception area is obtained. r value;

[0036] Select the cost function C from all candidate paths r The smallest path is taken as the optimal path for the target vehicle.

[0037] Preferably, based on the expected waiting time of each entrance road and the road segment travel time of each candidate path in its subsequent pass path, the path cost function C of the target vehicle in the perception area is obtained using formula (3): r value;

[0038] The formula (3) is:

[0039]

[0040] d irepresents the travel time of the i-th candidate path in the subsequent travel path;

[0041] The road segment travel time of the i-th candidate path in the subsequent pass path is obtained by dividing the i-th candidate path in the subsequent pass path by the historical average speed of the target vehicle;

[0042] σ i (t) represents the traffic state fluctuation factor of the i-th candidate path at time t, which is defined as the ratio of the standard deviation of traffic flow to the average flow in the recent several periods;

[0043] path is the subsequent path of the target vehicle including all candidate paths;

[0044] α is the travel time weight coefficient; β is the waiting time weight coefficient; is the traffic fluctuation weight coefficient.

[0045] Preferably, when the target vehicle is a public transportation vehicle,

[0046] When the target vehicle type is a social vehicle,

[0047] When the queue length of the entrance exceeds the preset threshold, the preset adjustment factor δ is increased to 1.5 times the original value.

[0048] Preferably, based on the predicted traffic flow of each entrance, the total signal cycle time T of the intersection is determined using formula (4): c ;

[0049] The formula (4) is:

[0050]

[0051] T0 is the basic signal cycle time;

[0052] ε is the first adaptive correction coefficient;

[0053] is the second adaptive correction coefficient;

[0054] S total is the total saturation flow rate available for the entire intersection;

[0055] n is the total number of signal phases at the intersection.

[0056] Preferably, based on the predicted traffic flow of each entrance road, the green light time ratio of each signal phase of the corresponding intersection is optimized using formula (5);

[0057] The formula (5) is:

[0058]

[0059] L is the total loss time of each phase interval at the intersection;

[0060] G i Indicates the green light time of the i-th signal phase.

[0061] Preferably, the route guidance suggestion is generated by converting the optimal route into navigation instructions;

[0062] The intersection signal status is generated by converting the optimized green light time ratio of each signal phase into the remaining time information of the current signal light;

[0063] The estimated waiting time for each entrance lane is generated by integrating the estimated waiting time of vehicles according to the entrance lane.

[0064] The beneficial effects of the present invention are:

[0065] The present invention provides a multi-dimensional traffic flow control method based on mutual feedback regulation under vehicle-road collaboration. Since it adopts the deployment of traffic sensing equipment in the urban road network to collect real-time traffic data at each intersection, compared with the existing technology, it can fully grasp multi-dimensional traffic information such as intersection traffic flow, queue length, vehicle speed, signal status and road load, thereby achieving the effect of providing high-precision and real-time data support for subsequent traffic prediction and signal control.

[0066] Because the weighted sliding average method is used to predict the traffic flow in each signal phase control direction, and the vehicle waiting time is estimated in combination with the real-time queue length and signal status, compared with the existing technology, it can more accurately reflect the dynamic queuing and traffic trends of vehicles at the intersection, thereby achieving the effect of improving the timeliness and accuracy of traffic flow regulation.

[0067] Since the optimal route planning is carried out by combining the vehicle's expected waiting time with the total travel time of its route, compared with existing technologies, it can achieve coordinated optimization of route guidance and traffic signal status, thereby alleviating congestion in key sections and improving traffic efficiency.

[0068] Since it optimizes the total duration of the signal cycle and the ratio of green light time in each phase based on predicted traffic flow, compared with existing technologies, it can dynamically adjust the signal timing strategy, avoid uneven resource allocation and green light waste, and achieve the effect of improving the traffic capacity of intersections and the overall efficiency of the transportation system.

[0069] Since the recommended information such as route guidance suggestions, signal status and estimated waiting time is pushed to the target vehicle in the perception area through the Internet of Vehicles, compared with the existing technology, it can achieve information synchronization and mutual feedback of control strategies between the vehicle and road sides, thereby enhancing the collaborative perception and decision-making linkage between the vehicle and the road.

[0070] Since the target vehicle is accurately identified based on the vehicle's unique identification and location data, compared with existing technologies, it can ensure the targeted and real-time push of recommended information, thereby improving the efficiency of path guidance execution and user response rate.

[0071] In summary, the present invention constructs a dynamically adjusted and highly coordinated traffic flow control mechanism through the fusion of multi-dimensional perception, joint prediction, signal optimization and path guidance feedback, which significantly improves the intelligence level and operation efficiency of the urban road traffic system. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 This is a flow chart of a multi-dimensional traffic flow control method based on mutual feedback regulation under vehicle-road collaboration. DETAILED DESCRIPTION

[0073] In order to better explain the present invention and facilitate understanding, the present invention is described in detail below through specific implementation methods in conjunction with the accompanying drawings.

[0074] To better understand the above technical solutions, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. Instead, these embodiments are provided to enable a clearer and more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0075] Example 1

[0076] See also Figure 1 This embodiment provides a multi-dimensional traffic flow control method based on mutual feedback regulation under vehicle-road collaboration, including the following steps:

[0077] S1. Traffic sensing devices deployed at each intersection in the urban road network collect real-time traffic data at each intersection. The real-time traffic data includes: traffic flow at the intersection entrance, queue length, vehicle speed within the entrance, current state of the corresponding signal phase, and road load information for each entrance;

[0078] In this implementation, traffic sensing equipment is deployed at every intersection in the urban road network to collect real-time, multi-dimensional traffic information, including entrance traffic volume, queue lengths, vehicle speeds, signal status, and road load. This real-time data reflects current traffic conditions and serves as the foundation for subsequent dispatch control and route guidance.

[0079] S2. After each preset time period, based on the collected real-time traffic data, obtain recommended information through a preset method; the preset method includes:

[0080] The weighted moving average method is used to predict the traffic flow at time t at the entrance lane corresponding to the traffic direction controlled by each signal phase at each intersection.

[0081] Calculate the estimated waiting time for vehicles at each entrance at time t based on the current queue length and signal phase state of each entrance;

[0082] Based on the estimated waiting time of each entrance lane and the travel time of the subsequent road segments, the optimal path is planned for each target vehicle within the perception area of ​​the intersection perception device;

[0083] Based on the predicted traffic flow of each entrance, the total signal cycle duration of the intersection is determined, and the green light time ratio of each signal phase is optimized;

[0084] The recommended information includes: route guidance suggestions, intersection signal status, and estimated waiting time for each entrance;

[0085] S3. Directly push the recommendation information to the vehicle navigation system corresponding to the target vehicle terminal involved in the real-time traffic data through the Internet of Vehicles, so that it can perform route selection and dynamic adjustment; wherein, the target vehicle terminal is a vehicle in the perception area of ​​the intersection perception device, and is associated and identified based on the vehicle's unique identification and location data.

[0086] In this embodiment of the present invention, associating and identifying vehicles based on their unique identifiers and location data is a key step in achieving precise control of vehicle-road collaborative systems. Specifically, target vehicles within the perception area of ​​the intersection sensing device periodically broadcast information containing their unique identifiers, such as the OBU device ID, an encrypted license plate number, or an electronic tag ID, to the roadside unit (RSU) via their on-board unit (OBU). Simultaneously, the RSU uses deployed traffic sensing equipment, such as video recognition and millimeter-wave radar, to obtain real-time information on the vehicle's location, motion status, and other characteristic parameters on the entrance lane. The system uniquely identifies and precisely associates the target vehicle by matching the GPS location broadcast by the on-board unit with the vehicle's spatial position as sensed by the RSU, incorporating auxiliary information such as timestamps, lane numbers, and motion directions. This allows the roadside system to accurately identify each vehicle within the perception area, enabling targeted push of recommended information (such as route guidance suggestions and signal status) to the corresponding vehicle's navigation system, ensuring real-time and targeted routing and signal optimization.

[0087] This embodiment achieves multi-dimensional optimization of traffic signal control and route guidance through the deep integration of vehicle-road collaboration, data prediction and intelligent scheduling, thereby improving the operating efficiency and traffic capacity of the urban transportation system.

[0088] In practical applications, at any intersection, the traffic flow Q at the entrance lane corresponding to the direction of travel controlled by the i-th signal phase at time t is p (i, t) is calculated by the following weighted moving average formula (1);

[0089] The weighted moving average formula (1) is:

[0090]

[0091] Q(tk·Δt) represents the actual traffic flow observation value of the entrance lane corresponding to the traffic direction controlled by the i-th phase at the historical time tk·Δt;

[0092] γ∈(0,1) is the time series attenuation factor, which is used to control the influence weight of earlier historical data;

[0093] Δt is the preset time sampling interval;

[0094] N is the number of historical moments participating in the weighted average;

[0095] k represents the historical moment number.

[0096] In this embodiment, formula (1) assigns different weights to historical traffic data at different time points, making recent data have a greater impact on the prediction results, while the influence of older data gradually decreases. This method not only captures short-term fluctuations in traffic flow but also reflects its long-term trends, providing a reliable basis for traffic signal control and vehicle routing guidance. Specifically, the time series decay factor in formula (1) causes the weight of more recent traffic flow data to increase, while the weight of older historical data decays exponentially.

[0097] The high weighting of recent data ensures that forecasts quickly reflect changes in actual traffic conditions (such as sudden congestion or traffic surges), avoiding the lag inherent in traditional moving average methods. Because older data is weighted less, accidental detection errors or outliers have less impact on forecasts, improving model robustness. By requiring only limited historical data storage (e.g., the last N time periods), computational complexity is low, making this approach suitable for real-time execution on roadside edge computing devices.

[0098] For example, during the evening rush hour, traffic flow on a particular entrance can surge rapidly within a short period of time. The weighted moving average method can predict traffic growth trends by overweighting recent data, thereby triggering a longer green light at traffic lights and effectively reducing vehicle queues.

[0099] Specifically, during the morning and evening traffic peak hours, the time series attenuation factor γ is in the range of 0.6-0.8.

[0100] During the off-peak period, the timing attenuation factor γ is in the range of 0.3-0.5.

[0101] In this embodiment, the time series attenuation factor γ is set to 0.6-0.8 during morning and evening peak hours. This allows for rapid response to sudden changes in traffic flow. By increasing the weight of recent data, it promptly captures sudden congestion or traffic surges, ensuring dynamic optimization of signal timing. During off-peak hours, a γ value of 0.3-0.5 is used, prioritizing the stability of long-term trends. By reducing the weight of recent data, random fluctuations are smoothed, avoiding the waste of resources caused by frequent signal adjustments. This dynamic adjustment strategy not only improves traffic flow efficiency during peak hours, but also ensures smooth system operation during off-peak hours, achieving an optimal balance between prediction accuracy and computational efficiency.

[0102] At any intersection, the estimated waiting time W of the vehicle at the entrance lane corresponding to the traffic direction controlled by the i-th signal phase at time t is p (i, t) is calculated by formula (2):

[0103] The formula (2) is:

[0104]

[0105] Lp (i, t) represents the queue length at time t at the entrance channel corresponding to the traffic direction controlled by the i-th signal phase; C(i) represents the traffic capacity per unit time of the entrance channel corresponding to the traffic direction controlled by the i-th signal phase; G(i, t) represents the current or upcoming green light time length of the entrance channel corresponding to the traffic direction controlled by the i-th phase; R(i, t) represents the remaining time of the current red light of the entrance channel corresponding to the traffic direction controlled by the i-th phase; δ is a pre-set adjustment factor; ∈ is a small constant to avoid the denominator being zero.

[0106] In this embodiment, formula (2) comprehensively considers the queue length L q (i, t), communication capacity C(i), G(i, t), green light time and red light remaining time R(i, t), quantify the vehicle's expected waiting time. For example, the molecule Reflects the basic queue dissipation time, By dynamically correcting the impact of the red light phase and adjusting the factor δ, the system can flexibly adapt to different intersection characteristics (such as prioritizing main road traffic), while the small constant ∈ ensures the stability of the calculation. Formula (2) of this embodiment can not only accurately predict vehicle waiting time in real time, but also provide a dynamic adjustment basis for the signal control system (for example, automatically extending the subsequent green light phase when the red light time is long), thereby effectively reducing vehicle queues and overall delays. At the same time, the calculation of the formula relies only on the current cycle data, ensuring efficiency and practicality, enabling it to be seamlessly integrated into the vehicle-road cooperative system, providing reliable support for dynamic path planning and signal optimization.

[0107] In this embodiment, when the queue length of the entrance lane exceeds a preset threshold, the preset adjustment factor δ is increased to 1.5 times the original value.

[0108] Specifically, based on the expected waiting time of each entrance and the travel time of the subsequent road sections, the optimal path is planned for each target vehicle in the sensing area, including:

[0109] Based on the expected waiting time of each entrance road and the road travel time of each candidate path in its subsequent passage path, the path cost function C of the target vehicle in the perception area is obtained. r value;

[0110] Select the cost function C from all candidate paths r The smallest path is taken as the optimal path for the target vehicle.

[0111] Among them, based on the expected waiting time of each entrance road and the road section travel time of each candidate path in its subsequent passing path, the path cost function C of the target vehicle in the perception area is obtained using formula (3): r value;

[0112] The formula (3) is:

[0113]

[0114] d i represents the road segment travel time of the i-th candidate path in the subsequent pass path; wherein, the road segment travel time of the i-th candidate path in the subsequent pass path is obtained by dividing the i-th candidate path in the subsequent pass path by the historical average speed of the target vehicle; σ i (t) represents the traffic state fluctuation factor of the i-th candidate path at time t, which is defined as the ratio of the standard deviation of traffic flow to the average flow in the recent period, reflecting traffic instability; path is the subsequent travel path of the target vehicle including all candidate paths; α is the travel time weight coefficient; β is the waiting time weight coefficient; is the traffic fluctuation weight coefficient.

[0115] Formula (3) in this embodiment evaluates the comprehensive travel cost of each candidate path by weighted summing three key parameters: 1) the travel time d of the road segment (calculated based on the historical average vehicle speed); 2) the expected waiting time at the entrance; and 3) the traffic state fluctuation factor (reflecting the degree of traffic flow fluctuation). Among them, α is the travel time weight coefficient; β is the waiting time weight coefficient; is the traffic fluctuation weight coefficient, which is used to adjust the influence of different factors on route selection.

[0116] In this embodiment, the target vehicle's subsequent travel path, including all candidate paths, is generated based on real-time road network topology and traffic conditions. The following steps are performed in advance to generate all feasible candidate paths: 1) Extract all possible routes from the vehicle's current location to the destination using electronic map data (such as OpenStreetMap); 2) Filter out invalid paths based on real-time traffic control information (such as construction closures and temporary restrictions); and 3) Further filter paths that meet traffic regulations based on vehicle type (such as trucks and buses). The resulting set of candidate paths covers all currently legal and accessible travel options, ensuring the global optimal path.

[0117] When the target vehicle type is a public transportation vehicle,

[0118] When the target vehicle type is a social vehicle,

[0119] Specifically, based on the predicted traffic flow of each entrance, the total signal cycle time T of the intersection is determined using formula (4). c ;

[0120] The formula (4) is:

[0121]

[0122] T0 is the basic signal cycle time; ε is the first adaptive correction coefficient; is the second adaptive correction coefficient; S total is the total saturation flow rate available at the entire intersection; n is the total number of signal phases at the intersection. total It refers to the maximum number of vehicles that can pass through all entrances of an intersection per unit time under ideal conditions (usually in "vehicles / hour"). Its core is the upper limit of traffic capacity determined by the physical space and signal timing of the intersection.

[0123] In this embodiment, based on the predicted traffic flow of each entrance road, formula (5) is used to optimize the green light time ratio of each signal phase at the intersection;

[0124] The formula (5) is:

[0125]

[0126] L is the total loss time of each phase interval at the intersection;

[0127] G i Indicates the green light time of the i-th signal phase.

[0128] In this embodiment, the green light time for each phase is not less than 10 seconds, and the pedestrian phase is not less than 20 seconds.

[0129] In this embodiment, formula (5) calculates the limited green light resources according to the real-time traffic ratio of each phase. Intelligent allocation and deduction of fixed loss time L can significantly improve the overall efficiency of the intersection.

[0130] In practical applications of this embodiment, the route guidance suggestion is generated by converting the optimal route into navigation instructions;

[0131] The intersection signal status is generated by converting the optimized green light time ratio of each signal phase into the remaining time information of the current signal light; the estimated waiting time of each entrance lane is generated by integrating the estimated waiting time of vehicles according to the entrance lane.

[0132] In this embodiment, the route guidance suggestion is generated by: converting the optimal route for the target vehicle into navigation instructions containing heading, turn instructions, and distance information; generating a displayable route guidance suggestion based on real-time road network topology data; the real-time road network topology data is generated by using lane-level traffic status and turn restriction information collected in real time by traffic sensing equipment;

[0133] The intersection signal status is generated by: obtaining an optimized green light time ratio for each signal phase; generating visual signal status information including red / green light status and remaining time based on the current signal phase status and remaining time;

[0134] The estimated waiting time for each entrance lane is generated by: obtaining the estimated waiting time for vehicles in different lanes of each entrance lane; performing a weighted average calculation on the estimated waiting time for all lanes; and converting the calculation result into standardized waiting time prompt information.

[0135] Example 2

[0136] This embodiment provides a multi-dimensional traffic flow control method for an urban intelligent transportation system. Its core is to achieve dynamic optimization and control of road traffic flow through a vehicle-road collaborative mechanism. The overall system structure includes traffic sensing equipment deployed at major urban intersections, a signal light control module, an edge computing unit, a vehicle-to-vehicle communication module, and a vehicle-mounted navigation system.

[0137] In actual application, the control flow is set as follows:

[0138] Step 1: Deploy traffic sensing equipment such as video surveillance, radar sensors, and geomagnetic detectors at multiple intersections in the urban road network to collect real-time traffic parameters at each intersection. This includes, but is not limited to, the following data: traffic flow at each intersection entrance (number of vehicles passing per unit time); queue length (length of the vehicle queue extending from the stop line); vehicle speed (average speed of vehicles within the entrance area); current signal phase status (red, green, yellow, and remaining time); and road load information (vehicle density per unit area, traffic congestion indicators, etc.). This data is uploaded to a local edge computing node for subsequent processing.

[0139] Step 2: After each set time interval (e.g., 30 seconds), the system processes the collected traffic data and generates traffic control recommendations, which specifically includes the following sub-steps:

[0140] The weighted moving average method is used to predict the traffic flow at time t at the entrance lane corresponding to the traffic direction controlled by each signal phase at each intersection.

[0141] For example, the specific approach is to aggregate traffic flow data from a specific entrance to a certain intersection at different points in time and assign different weights to this historical data, so that recent data has a greater impact on the prediction results, while older data has a smaller impact. For example, in the three most recent records, the system can assign different weights to the traffic flow data at the current moment, the previous moment, and two moments ago, respectively, to form a "weighted sliding average" prediction. This method can more dynamically grasp the current traffic pressure in each signal phase control direction and provide a basis for subsequent adjustments to the signal light duration.

[0142] Based on the current queue length and signal phase status of each entrance lane, the estimated waiting time of vehicles on each entrance lane at time t is calculated; specifically, the queue length of vehicles on the intersection entrance lane, as well as the current signal light status (i.e., red light or green light) and its remaining time are collected in real time. If the current light is red, the system will estimate how many vehicles can pass after the green light comes on, while taking into account the time interval required for each vehicle to start. If the current number of vehicles in the queue can be fully released within the next green light cycle, then the estimated waiting time of the first vehicle will be the remaining time of the red light, and the waiting time of subsequent vehicles will be accumulated accordingly based on their position in the queue. If the current number of vehicles exceeds the release capacity of one cycle, some vehicles will need to wait for the next cycle. The system will also take this into account and calculate an average estimated waiting time accordingly. This time value serves as an important reference parameter for subsequent path recommendations.

[0143] Based on the estimated wait times for each approach and the travel times along the subsequent routes, the system plans the optimal route for each target vehicle within the intersection's perception area. Specifically, after determining the estimated wait times for each approach, the system further compares and analyzes multiple route options, taking into account the travel times along subsequent routes in the vehicle's target direction. For example, if a vehicle proceeds straight through the current approach, its total travel time may be longer. However, if it chooses to turn right or left and then detour to its destination, while the route may be slightly longer, the total travel time may be shorter. When performing this route comparison, the system comprehensively evaluates multiple routes based on real-time traffic data, traffic light status, and historical traffic efficiency, and automatically generates the optimal route recommendation for the target vehicle. This route recommendation is then sent to the vehicle's navigation system via the connected vehicle interface, guiding the driver to make adjustments or avoid congested areas.

[0144] Based on the predicted traffic flow of each entrance lane, the total signal cycle duration of the intersection is determined, and the green light time ratio of each signal phase is optimized. During each preset signal cycle refresh period, the system will determine the overall traffic load of the current intersection based on the predicted traffic flow of each entrance lane, and dynamically adjust the overall cycle length of the signal light and the green light time ratio of each phase accordingly. For example, when the predicted traffic flow of a certain entrance lane is significantly higher than that of other directions, the system can appropriately extend the green light duration in that direction to improve its traffic efficiency; the green light duration of directions with relatively low traffic flow can be appropriately compressed, thereby achieving rational allocation of resources and dynamic relief of traffic bottlenecks. In addition, during particularly congested periods, the system can also adjust the total duration of the signal cycle according to preset rules to more flexibly adapt to traffic fluctuations.

[0145] The recommended information includes: route guidance suggestions, intersection signal status, and estimated waiting time for each entrance;

[0146] Step 3: The generated recommendation information is pushed to the target vehicle within the sensing area via the connected vehicle (e.g., C-V2X) communication system. The vehicle dynamically adjusts its driving path and behavior strategy based on the information received by its navigation system. Target vehicles are identified by matching their unique identifier (e.g., license plate or OBU ID) with their location information. During the push process, the system continuously and dynamically updates its recommended path based on parameters such as the vehicle's current location and speed, enabling real-time traffic flow guidance and coordinated signal adjustments at multiple intersections.

[0147] Through the above implementation methods, this embodiment effectively realizes the information feedback mechanism between vehicles and road infrastructure, while improving the traffic efficiency of the road network, reducing the probability of traffic congestion and improving the intelligence level of the city's overall transportation system.

[0148] In the description of the present invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0149] In the present invention, unless otherwise expressly specified or limited, the terms "mounted," "connected," "connect," "fixed," etc. should be understood broadly. For example, they may refer to fixed connection, detachable connection, or integration; mechanical connection or electrical connection; direct connection or indirect connection through an intermediate medium; and internal communication between two components or interaction between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0150] In the present invention, unless otherwise expressly specified or limited, when a first feature is "above" or "below" a second feature, it may mean that the first and second features are in direct contact, or that the first and second features are in indirect contact through an intermediate medium. Furthermore, when a first feature is "above," "above," or "above" a second feature, it may mean that the first feature is directly above or obliquely above the second feature, or simply means that the first feature is at a higher level than the second feature. When a first feature is "below," "below," or "below" a second feature, it may mean that the first feature is directly below or obliquely below the second feature, or simply means that the first feature is at a lower level than the second feature.

[0151] In the description of this specification, the terms "one embodiment", "some embodiments", "embodiments", "examples", "specific examples" or "some examples" refer to the specific features, structures, materials or characteristics described in conjunction with the embodiment or example and included in at least one embodiment or example of the present invention. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and features of different embodiments or examples, unless they are mutually inconsistent.

[0152] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may alter, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A multi-dimensional traffic flow control method based on mutual feedback regulation under vehicle-road collaboration, characterized in that: The following steps are involved: S1. Traffic sensing devices deployed at various intersections in the urban road network collect real-time traffic data at each intersection; The real-time traffic data includes: traffic flow at the intersection entrance, queue length, vehicle speed within the entrance, current status of corresponding signal phases, and road load information for each entrance; S2. After each preset time period, based on the collected real-time traffic data, obtain recommended information through a preset method; the preset method includes: The weighted moving average method is used to predict the traffic flow at time t at the entrance lane corresponding to the traffic direction controlled by each signal phase at each intersection. Calculate the estimated waiting time for vehicles at each entrance at time t based on the current queue length and signal phase state of each entrance; Based on the estimated waiting time of each entrance lane and the travel time of the subsequent road segments, the optimal path is planned for each target vehicle within the perception area of ​​the intersection perception device; Based on the predicted traffic flow of each entrance, the total signal cycle duration of the intersection is determined, and the green light time ratio of each signal phase is optimized; The recommended information includes: route guidance suggestions, intersection signal status, and estimated waiting time for each entrance; S3. Directly push the recommendation information to the vehicle navigation system corresponding to the target vehicle involved in the real-time traffic data through the Internet of Vehicles, so that it can perform route selection and dynamic adjustment; wherein, the target vehicle is a vehicle in the perception area of ​​the intersection perception device, and is associated and identified based on the vehicle's unique identifier and location data.

2. The multi-dimensional traffic flow control method based on mutual feedback regulation under vehicle-road collaboration according to claim 1 is characterized in that: At any intersection, the traffic flow Q at the entrance lane corresponding to the direction of travel controlled by the i-th signal phase at time t is p (i, t) is calculated by the following weighted moving average formula (1); The weighted moving average formula (1) is: Q(tk·Δt) represents the actual traffic flow observation value of the entrance lane corresponding to the traffic direction controlled by the i-th phase at the historical time tk·Δt; γ∈(0,1) is the time series attenuation factor, which is used to control the influence weight of earlier historical data; Δt is the time sampling interval; N is the number of historical moments participating in the weighted average; k represents the historical moment number.

3. The multi-dimensional traffic flow control method based on mutual feedback regulation under vehicle-road collaboration according to claim 2 is characterized in that: During the morning and evening traffic peak hours, the time series attenuation factor γ is in the range of 0.6-0.

8. During the off-peak period, the timing attenuation factor γ is in the range of 0.3-0.

5.

4. The multi-dimensional traffic flow control method based on mutual feedback regulation under vehicle-road collaboration according to claim 3 is characterized in that: At any intersection, the estimated waiting time W of the vehicle at the entrance lane corresponding to the traffic direction controlled by the i-th signal phase at time t is p (i, t) is calculated by formula (2): The formula (2) is: L q (i, t) represents the queue length at time t at the entrance corresponding to the traffic direction controlled by the i-th signal phase; C(i) represents the traffic capacity per unit time of the entrance lane corresponding to the traffic direction controlled by the i-th signal phase; G(i, t) represents the current or upcoming green light duration of the entrance lane corresponding to the direction of travel controlled by the i-th phase; R(i, t) represents the remaining time of the current red light on the entrance lane corresponding to the direction of travel controlled by the i-th phase; δ is a pre-set adjustment factor; ∈ is a small constant to avoid the denominator being zero.

5. The multi-dimensional traffic flow control method based on mutual feedback regulation under vehicle-road collaboration according to claim 4 is characterized in that: Based on the estimated waiting time of each entrance and the travel time of the subsequent road segments, the optimal path is planned for each target vehicle in the sensing area, including: Based on the expected waiting time of each entrance road and the road travel time of each candidate path in its subsequent passage path, the path cost function C of the target vehicle in the perception area is obtained. r value; Select the cost function C from all candidate paths r The smallest path is taken as the optimal path for the target vehicle.

6. The multi-dimensional traffic flow control method based on mutual feedback regulation under vehicle-road collaboration according to claim 5, It is characterized by: Among them, based on the expected waiting time of each entrance road and the road section travel time of each candidate path in its subsequent passing path, the path cost function C of the target vehicle in the perception area is obtained using formula (3): r value; The formula (3) is: d i represents the travel time of the i-th candidate path in the subsequent travel path; The road segment travel time of the i-th candidate path in the subsequent pass path is obtained by dividing the i-th candidate path in the subsequent pass path by the historical average speed of the target vehicle; σ i (t) represents the traffic state fluctuation factor of the i-th candidate path at time t, which is defined as the ratio of the standard deviation of traffic flow to the average flow in the recent several periods; path is the subsequent path of the target vehicle including all candidate paths; α is the travel time weight coefficient; β is the waiting time weight coefficient; is the traffic fluctuation weight coefficient.

7. The multi-dimensional traffic flow control method based on mutual feedback regulation under vehicle-road collaboration according to claim 6 is characterized in that: When the target vehicle type is a public transportation vehicle, When the target vehicle type is a social vehicle, When the queue length of the entrance exceeds the preset threshold, the preset adjustment factor δ is increased to 1.5 times the original value.

8. The multi-dimensional traffic flow control method based on mutual feedback regulation under vehicle-road collaboration according to claim 7 is characterized in that: Based on the predicted traffic flow of each entrance, the total signal cycle time T of the intersection is determined using formula (4): c ; The formula (4) is: T0 is the basic signal cycle time; ε is the first adaptive correction coefficient; θ is the second adaptive correction coefficient; S total is the total saturation flow rate available for the entire intersection; n is the total number of signal phases at the intersection.

9. The multi-dimensional traffic flow control method based on mutual feedback regulation under vehicle-road collaboration according to claim 8 is characterized in that: Based on the predicted traffic flow of each entrance, the green light time ratio of each signal phase at the intersection is optimized using formula (5); The formula (5) is: L is the total loss time of each phase interval at the intersection; G i Indicates the green light time of the i-th signal phase.

10. The multi-dimensional traffic flow control method based on mutual feedback regulation under vehicle-road collaboration according to claim 9 is characterized in that: The route guidance suggestion is generated by converting the optimal route into navigation instructions; The intersection signal status is generated by converting the optimized green light time ratio of each signal phase into the remaining time information of the current signal light; The estimated waiting time for each entrance lane is generated by integrating the estimated waiting time of vehicles according to the entrance lane.

Images