Construction zone traffic guidance and speed coordination control system

Abstract

The application belongs to the field of intelligent transportation, and particularly relates to a construction zone vehicle flow guiding and speed cooperative control system, which realizes accurate control of vehicle flow through the cooperative operation of a regional perception module, a distributed guiding module, a communication module and an emergency response module; the regional perception module is based on three-dimensional modeling and digital twin technology, fuses real-time traffic flow, vehicle state and environmental data, and constructs a three-dimensional simulation space of a target region; the distributed guiding module uses a distributed reinforcement learning algorithm to establish a flow guiding control model, combines three-level communication nodes of vehicle-road-cloud, dynamically optimizes congestion regulation and control strategies of a sub-region, and adjusts vehicle acceleration and speed in real time; the emergency response module quickly generates emergency instructions through an abnormality analysis algorithm and a strategy library and feeds back the emergency instructions to the guiding model to solve sudden abnormal events; the application integrates multi-source data real-time perception, distributed decision optimization and multi-level communication architecture, and realizes dynamic flow distribution and intelligent traffic control in a construction zone.

Description

Technical Field

[0001] This invention belongs to the field of intelligent transportation, and in particular relates to a traffic flow guidance and speed coordination control system for construction areas. Background Technology

[0002] In recent years, with the increasing frequency of urban road construction, traffic management in construction areas has faced severe challenges. Due to the confined space, mixed pedestrian and vehicle traffic, and complex dynamic changes in construction conditions, traditional management methods relying on manual command and static signage are ill-suited to real-time traffic flow fluctuations, easily leading to traffic congestion and safety hazards. In existing technologies, fixed speed limit signs and lane markings cannot flexibly adjust guidance strategies based on dynamic parameters in construction areas, causing vehicles to experience chain reactions of sudden braking due to sudden speed changes or disorderly lane changes in narrowing lane areas, further exacerbating traffic inefficiency and increasing accident risks. Simultaneously, traditional traffic signal control systems and vehicle navigation functions operate independently, lacking a collaborative optimization mechanism for traffic flow and speed, making it difficult to achieve smooth transition control of vehicles in construction areas. Furthermore, existing systems lack the real-time perception and processing capabilities for multi-source heterogeneous data, making rapid response difficult in the event of sudden accidents or severe weather. These technological bottlenecks have resulted in a long-standing dual dilemma of low efficiency and insufficient safety in traffic organization in construction areas, urgently requiring breakthroughs through intelligent collaborative control methods. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention proposes a traffic flow guidance and speed coordination control system for construction areas. This system achieves precise traffic flow control through the coordinated operation of a regional perception module, a distributed guidance module, a communication module, and an emergency response module. The regional perception module, based on 3D modeling and digital twin technology, integrates real-time traffic flow, vehicle status, and environmental data to construct a 3D simulation space of the target area. The distributed guidance module uses a distributed reinforcement learning algorithm to establish a traffic flow control model, and combines vehicle-road-cloud three-level communication nodes to dynamically optimize congestion control strategies in sub-areas, adjusting vehicle acceleration and speed in real time. The emergency response module uses anomaly analysis algorithms and a strategy library to quickly generate emergency commands and feed them back to the guidance model to resolve sudden abnormal events. This application integrates multi-source data real-time perception, distributed decision optimization, and a multi-level communication architecture to achieve dynamic traffic diversion and intelligent traffic control in construction areas.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] The construction area traffic flow guidance and speed coordination control system includes: an area perception module, a distributed guidance module, a communication module, and an emergency response module;

[0006] The area perception module, based on the area map layer and real-time collected traffic flow, vehicle status, area environment and target area entry and exit information, obtains a three-dimensional simulation space of the target area through a three-dimensional modeling algorithm combined with a twin algorithm.

[0007] Based on the three-dimensional simulation space of the target area, combined with the distributed diversion control model built into the distributed guidance module and the vehicle-road-cloud three-level communication node space built based on narrowband IoT technology in the communication module, the operating status of all vehicles in the target area is controlled in real time, and abnormal data in the target area is monitored in real time. At the same time, the abnormal data in the target area is fed back to the abnormal analysis algorithm and emergency handling strategy library built into the emergency response module to obtain emergency handling strategies. The emergency handling strategies are fed back to the distributed diversion control model to adjust the operating status of vehicles in abnormal state in real time.

[0008] The vehicle-road-cloud three-level communication node space is constructed by combining vehicle data, road distribution data, pedestrian flow data, environmental status data, and the relationship between the location of monitoring equipment and data transmission within the target area using a graph algorithm.

[0009] The distributed traffic control model is constructed by combining the congestion distribution of different target sub-regions, vehicle acceleration and speed, vehicle distance, probability distribution of anomalies between vehicles or in the environment, and vehicle travel paths with a distributed reinforcement learning algorithm.

[0010] Specifically, the area perception module includes a data acquisition unit, a 3D modeling unit, and a twin simulation unit;

[0011] The data acquisition unit, based on a preset data acquisition node network and a preset distributed filtering model, collects real-time information on the changes in the three-dimensional point cloud map of the target area, vehicle operation data, vehicle and pedestrian flow and road distribution information, and obtains a real-time status set of the target area.

[0012] The three-dimensional modeling unit obtains a static three-dimensional model of the target area by combining a subset of static information from the real-time state set of the target area with a three-dimensional modeling algorithm; the static information subset includes the building distribution, road distribution, road surface attributes, and infrastructure information of the target area.

[0013] The twin simulation unit, based on the static three-dimensional model of the target area and the dynamic information subset of the real-time state set of the target area, uses the twin simulation algorithm to simulate the real-time changes in vehicles, pedestrian flow and equipment in the target area, thereby obtaining the dynamic simulation information set of the target area.

[0014] Specifically, the distributed guidance module includes an edge computing unit and a guidance control unit; the communication module includes a communication construction unit and a communication transmission unit; and the distributed traffic control model includes a distributed edge optimization sub-model and a distributed edge control sub-model.

[0015] The edge computing unit is used to obtain the operation optimization parameters for each vehicle on all roads within the target area based on the dynamic simulation information set of the target area and the preset distributed edge optimization sub-model.

[0016] The operational optimization parameters for each vehicle include the corresponding operating speed, acceleration, and the distance between vehicles in front and behind, as well as the distribution of command transmission intervals.

[0017] The guidance control unit obtains the guidance control command corresponding to each vehicle based on the operation optimization parameters of each vehicle and the distributed edge control sub-model.

[0018] Simultaneously, based on the command transmission interval distribution, the guidance and control command corresponding to each vehicle is sent to the corresponding vehicle node through the vehicle-road-cloud three-level communication node space. Each vehicle node performs real-time control and adjustment of the vehicle's operating status and operating path through the received guidance and control command.

[0019] Specifically, the process of acquiring the real-time state set of the target area includes:

[0020] The static monitoring nodes in the data acquisition node network are defined as the location points corresponding to all static monitoring devices set up within the target area, and the dynamic monitoring nodes in the data acquisition node network are defined as the dynamic location points corresponding to all vehicles within the target area.

[0021] Simultaneously, a static monitoring connection sequence is constructed based on the data transmission path relationship between all static monitoring nodes, a first dynamic monitoring connection sequence is established based on the data transmission path relationship between all static monitoring nodes and dynamic monitoring nodes, and a second dynamic monitoring connection sequence is constructed based on the data transmission path relationship between all dynamic monitoring nodes.

[0022] A data acquisition node network is constructed based on all static monitoring nodes, dynamic monitoring nodes, and their corresponding static and dynamic monitoring connection sequences using a graph algorithm.

[0023] Specifically, the process of acquiring the real-time state set of the target area also includes:

[0024] Based on the monitorable distance corresponding to each static monitoring node, the target area is divided and the monitoring overlap area corresponding to all static monitoring nodes is obtained.

[0025] When any of the dynamic monitoring nodes is in any monitoring overlap area, the static monitoring node with the shortest signal transmission time corresponding to the current dynamic monitoring node is selected to establish the first dynamic monitoring connection.

[0026] When the distance between any dynamic monitoring node and any static monitoring node or dynamic monitoring node is greater than the corresponding monitorable distance, the first dynamic monitoring connection or the second dynamic monitoring connection of the corresponding dynamic monitoring node is disconnected.

[0027] Update the first or second dynamic monitoring connection of the current dynamic monitoring node based on the current location of the corresponding dynamic monitoring node and its distance relationship with any static or dynamic monitoring node.

[0028] When any dynamic monitoring node leaves the target area, the first and second dynamic monitoring connections corresponding to the current dynamic monitoring node are directly disconnected and the current dynamic monitoring node is deleted.

[0029] When a new dynamic monitoring node enters the target area, the process of building and updating the first and second dynamic monitoring connections described above is repeated to update the data acquisition node network.

[0030] Specifically, the process of acquiring the real-time state set of the target area also includes:

[0031] Based on Kalman filtering and the number of static monitoring nodes, a distributed filtering model is constructed and deployed to each static monitoring node in the data acquisition node network according to a distributed framework;

[0032] Based on the data acquisition node network, environmental data, vehicle operation data, road status data and pedestrian distribution data corresponding to the target area are collected in real time, and the vehicle operation data collected by each dynamic monitoring node is transmitted to the corresponding static monitoring node through the established first dynamic monitoring connection.

[0033] Each static monitoring node uses a deployed distributed filtering model to filter the monitoring data uploaded by itself and the dynamic monitoring nodes to obtain a real-time status set of the target area.

[0034] Specifically, the construction process of the vehicle-road-cloud three-level communication node space includes:

[0035] A guidance optimization subnet is constructed based on the static monitoring nodes and corresponding static monitoring connections in the data acquisition node network, and a guidance control subnet is constructed based on the dynamic monitoring nodes and corresponding second dynamic monitoring connections in the data acquisition node network.

[0036] Based on the first dynamic monitoring connection mapping corresponding to each static monitoring node, the optimized mapping connection between each guidance optimization node in the guidance optimization subnet and the corresponding guidance control node in the guidance control subnet is obtained.

[0037] Specifically, the process of constructing the command transmission interval distribution includes:

[0038] In the three-dimensional simulation space of the target area, the data acquisition node network is used to simulate and collect in real time the vehicle density, vehicle distance, vehicle speed and acceleration information corresponding to each path in the target area, the frequency of guidance and control command transmission under each vehicle density, and the probability of vehicle anomaly on each path under each guidance and control command transmission frequency, so as to obtain the real-time state set of the simulated target area.

[0039] Based on the frequency of guidance and control command transmission under each vehicle density and the probability of vehicle anomalies on each path under each guidance and control command transmission frequency, a command transmission interval distribution function is constructed by support vector machine.

[0040] The real-time state set of the simulated target area is input into the vehicle-road-cloud three-level communication node space. The command transmission interval is optimized by minimizing the anomaly probability corresponding to the command transmission interval distribution function as the optimization objective and by combining the optimization algorithm. The optimal command transmission interval distribution function and the distribution functions of vehicle density and speed corresponding to minimizing the anomaly probability are obtained for each vehicle density and vehicle speed.

[0041] Specifically, the construction process of the vehicle-road-cloud three-level communication node space also includes:

[0042] The optimal command transmission interval distribution function is embedded into the optimized mapping connection between each guidance optimization node and the corresponding guidance control node, and the vehicle density and vehicle speed distribution function is embedded into the distributed edge control sub-model in each guidance control node.

[0043] Based on the path information of the corresponding positions of all guidance and control nodes at the current moment, combined with the guidance optimization subnet, guidance control subnet and optimization mapping connection, the vehicle-road-cloud three-level communication node space with differential control is obtained by combining graph algorithm with sleep wake-up algorithm and power control algorithm;

[0044] The static monitoring nodes correspond one-to-one with the guidance optimization nodes; the dynamic monitoring nodes correspond one-to-one with the guidance control nodes; and each guidance control node corresponds to each vehicle within the target area.

[0045] Specifically, the process of obtaining the guidance control command for each vehicle includes:

[0046] Based on the data acquisition node network, the vehicle density, vehicle distance, vehicle speed and acceleration information corresponding to each path in the target area, and vehicle entry and exit information are collected in real time to obtain the real-time status set of the target area.

[0047] Based on the real-time state set of the target area, the optimal command transmission frequency and vehicle speed guidance parameter sequence for each guidance optimization node are obtained through the guidance optimization subnet in the vehicle-road-cloud three-level communication node space.

[0048] Based on the optimal command transmission frequency, the vehicle speed guidance parameter sequence is transmitted to the corresponding guidance control node, and the distributed edge control sub-model in the corresponding guidance control node is used to generate guidance control commands to perform differentiated real-time control of the corresponding vehicle speed.

[0049] This invention addresses the shortcomings of existing technologies by achieving dynamic collaborative control of traffic flow in construction zones through multi-dimensional technology fusion. Specific technical effects are as follows: Based on a distributed filtering model and Kalman filtering algorithm using a data acquisition node network, static monitoring node data and dynamic vehicle data are fused in real time, eliminating environmental noise interference and improving the accuracy of traffic status perception; a static-dynamic coupled model of the target area is constructed through 3D modeling and twin algorithms, combined with graph algorithm optimization of the vehicle-road-cloud three-level communication node space, achieving low-latency transmission and differentiated control of vehicle guidance control commands; a distributed reinforcement learning algorithm is used to dynamically generate the optimal command transmission interval function based on vehicle density, vehicle spacing, and anomaly probability distribution, simultaneously optimizing vehicle speed and acceleration parameters to ensure smooth traffic flow transition; by coupling support vector machines with optimization algorithms, the vehicle density-vehicle speed distribution function is embedded into the edge control sub-model, adaptively adjusting the path planning strategy and enhancing the system's responsiveness to dynamic traffic environments; this method significantly improves the real-time performance, accuracy, and robustness of vehicle management in construction zones. Attached Figure Description

[0050] Figure 1 This is a structural diagram of the construction area traffic guidance and speed coordination control system module in Embodiment 1 of the present invention;

[0051] Figure 2 This is a schematic diagram of a vehicle-road-cloud three-level communication node in Embodiment 1 of the present invention. Detailed Implementation

[0052] Example 1

[0053] In recent years, with the increasing frequency of urban road construction, traffic management in construction areas has faced severe challenges. Due to the limited space, mixed pedestrian and vehicle traffic, and complex dynamic changes in construction conditions, traditional management methods relying on manual command and static signage are ill-suited to real-time traffic flow fluctuations, easily leading to traffic congestion and safety hazards. Therefore, developing an intelligent traffic management system for use in construction areas is essential. For this purpose, please refer to... Figure 1 The present invention provides an embodiment of a construction area traffic flow guidance and speed coordination control system, comprising: an area perception module, a distributed guidance module, a communication module and an emergency response module;

[0054] The area perception module, based on the area map layer and real-time collected traffic flow, vehicle status, area environment and target area entry and exit information, obtains a three-dimensional simulation space of the target area through a three-dimensional modeling algorithm combined with a twin algorithm.

[0055] Based on the three-dimensional simulation space of the target area, combined with the distributed diversion control model built into the distributed guidance module and the vehicle-road-cloud three-level communication node space built based on narrowband IoT technology in the communication module, the operating status of all vehicles in the target area is controlled in real time, and abnormal data in the target area is monitored in real time. At the same time, the abnormal data in the target area is fed back to the abnormal analysis algorithm and emergency handling strategy library built into the emergency response module to obtain emergency handling strategies. The emergency handling strategies are fed back to the distributed diversion control model to adjust the operating status of vehicles in abnormal state in real time.

[0056] It is important to note that the data transmission process in the vehicle-road-cloud three-level communication node space of this embodiment is based on narrowband IoT technology. Narrowband IoT is a cellular network technology specifically designed for low-power, wide-coverage, and high-connectivity scenarios. It is an important branch of IoT and solves the pain points of traditional IoT devices in terms of connectivity, power consumption, and cost by optimizing communication protocols and spectrum resources. It is particularly suitable for data collection and transmission in narrow areas such as the construction area corresponding to this embodiment. Furthermore, the narrowband IoT technology in this embodiment preferably utilizes a sleep-wake algorithm and a power control algorithm.

[0057] The vehicle-road-cloud three-level communication node space is constructed by combining vehicle data, road distribution data, pedestrian flow data, environmental status data, and the relationship between the location of monitoring equipment and data transmission within the target area using a graph algorithm.

[0058] The distributed traffic control model is constructed by combining the congestion distribution of different target sub-regions, vehicle acceleration and speed, vehicle distance, probability distribution of anomalies between vehicles or in the environment, and vehicle travel paths with a distributed reinforcement learning algorithm.

[0059] Furthermore, the area perception module includes a data acquisition unit, a 3D modeling unit, and a twin simulation unit;

[0060] The data acquisition unit, based on a preset data acquisition node network and a preset distributed filtering model, collects real-time information on the changes in the three-dimensional point cloud map of the target area, vehicle operation data, vehicle and pedestrian flow and road distribution information, and obtains a real-time status set of the target area.

[0061] The three-dimensional modeling unit obtains a static three-dimensional model of the target area by combining a subset of static information from the real-time state set of the target area with a three-dimensional modeling algorithm; the static information subset includes the building distribution, road distribution, road surface attributes, and infrastructure information of the target area.

[0062] The twin simulation unit, based on the static three-dimensional model of the target area and the dynamic information subset of the real-time state set of the target area, uses the twin simulation algorithm to simulate the real-time changes in vehicles, pedestrian flow and equipment in the target area, thereby obtaining the dynamic simulation information set of the target area.

[0063] Furthermore, the distributed guidance module includes an edge computing unit and a guidance control unit; the communication module includes a communication construction unit and a communication transmission unit; the distributed traffic diversion control model includes a distributed edge optimization sub-model and a distributed edge control sub-model.

[0064] The edge computing unit is used to obtain the operation optimization parameters for each vehicle on all roads within the target area based on the dynamic simulation information set of the target area and the preset distributed edge optimization sub-model.

[0065] The operational optimization parameters for each vehicle include the corresponding operating speed, acceleration, and the distance between vehicles in front and behind, as well as the distribution of command transmission intervals.

[0066] Furthermore, the process of constructing the command transmission interval distribution includes:

[0067] In the three-dimensional simulation space of the target area, the data acquisition node network is used to simulate and collect in real time the vehicle density, vehicle distance, vehicle speed and acceleration information corresponding to each path in the target area, the frequency of guidance and control command transmission under each vehicle density, and the probability of vehicle anomaly on each path under each guidance and control command transmission frequency, so as to obtain the real-time state set of the simulated target area.

[0068] It is important to note that in this embodiment, a multimodal sensor network is deployed in the target area, including LiDAR, millimeter-wave radar, cameras, and RFID readers. A cellular grid layout ensures coverage without blind spots. Multi-source data is fused through time synchronization protocols and spatial alignment algorithms to calculate vehicle density, distance, speed, and acceleration in real time. Historical data and real-time events are combined to dynamically model the probability of vehicle anomalies, constructing a simulated real-time state set of the target area that reflects traffic conditions. It should be noted that the definition of abnormal events includes behaviors such as sudden braking and lane departure. Sliding window statistics and smoothing filtering are used to ensure the stability of probability calculations.

[0069] Based on the frequency of guidance and control command transmission under each vehicle density and the probability of vehicle anomalies on each path under each guidance and control command transmission frequency, a command transmission interval distribution function is constructed by support vector machine.

[0070] It should be further explained that this embodiment uses a support vector machine model to learn the correlation between instruction intervals and traffic conditions based on features such as vehicle density, vehicle speed, and anomaly probability. It should also be explained that the model in this embodiment supports online incremental learning, periodically injects new data to update parameters, and triggers retraining through error threshold detection to ensure adaptation to dynamic changes in road conditions.

[0071] The real-time state set of the simulated target area is input into the vehicle-road-cloud three-level communication node space, and the command transmission interval optimization simulation is carried out with the optimization objective of minimizing the abnormal probability corresponding to the command transmission interval distribution function and the distribution function of vehicle density and vehicle speed under each vehicle density and vehicle speed.

[0072] It should be further explained that the three-tier communication architecture constructed in this embodiment, consisting of vehicle-side, roadside, and cloud, involves the vehicle-side handling command execution and status reporting, the roadside aggregating regional data and performing edge computing, and the cloud globally optimizing historical data and strategies. With minimizing the anomaly probability as the core objective, a genetic algorithm is used to solve a multi-constraint optimization problem. Constraints include communication load, latency, and anomaly probability thresholds. The simulation platform verifies and outputs the optimal command interval curve and vehicle speed-density relationship table, guiding dynamic speed limiting and resource scheduling strategies.

[0073] The guidance control unit obtains the guidance control command corresponding to each vehicle based on the operation optimization parameters of each vehicle and the distributed edge control sub-model.

[0074] Simultaneously, based on the command transmission interval distribution, the guidance and control command corresponding to each vehicle is sent to the corresponding vehicle node through the vehicle-road-cloud three-level communication node space. Each vehicle node performs real-time control and adjustment of the vehicle's operating status and operating path through the received guidance and control command.

[0075] It should be further explained that, in this embodiment, the edge control sub-model generates acceleration, deceleration, or lane change commands based on the real-time vehicle status and the trajectory predicted by the LSTM network. The commands are issued according to safety level and vehicle priority, and multi-channel redundant transmission and data compression technology are used to ensure low-latency communication. The personalized strategy adjusts the command lead time according to the driver's style, and the roadside monitoring equipment ensures that the commands arrive reliably through the confirmation and retransmission mechanism, avoiding data loss caused by channel congestion.

[0076] This process, through multi-dimensional technical architecture and algorithm collaboration, constructs a three-dimensional technical system for traffic flow guidance and speed coordination control in the construction area. At the regional perception level, relying on the dynamic construction mechanism of the data acquisition node network and the distributed filtering model, it achieves real-time acquisition and noise filtering of multi-source information such as 3D point cloud changes, vehicle operation data, and pedestrian distribution in the target area, ensuring the integrity and accuracy of the real-time state set of the target area. Combining 3D modeling algorithms and twin simulation algorithms, static infrastructure information and dynamic traffic flow and equipment change information are mapped into a real-time interactive 3D simulation space, providing a precise digital twin carrier for subsequent control strategies. At the decision-making and control level, the distributed guidance module integrates distributed reinforcement learning algorithms through edge computing units, generating vehicle operation optimization parameters in real time based on the dynamic simulation information set, achieving refined modeling of traffic flow operation status. The vehicle operation optimization parameter package... This includes, but is not limited to, speed, acceleration, vehicle distance, and command transmission interval. A command transmission interval distribution function is constructed using support vector machines, and optimized through simulation with the goal of minimizing anomaly probability. This results in differentiated command transmission strategies adapted to different vehicle densities and speeds, effectively reducing communication congestion and control delay. The communication module constructs a three-level communication node space between vehicles, roads, and the cloud using graph algorithms. Static monitoring nodes and dynamic vehicle nodes are mapped to a guidance optimization subnet and a guidance control subnet, respectively. Layered data transmission is achieved through optimized mapping connections. Combined with the built-in optimal command transmission interval function, this ensures efficient and reliable transmission of control commands in complex traffic environments. At the anomaly response level, the emergency response module collects anomaly data in real time and matches it with the built-in anomaly analysis algorithm and strategy library to form a closed-loop feedback mechanism. Emergency strategies are injected into the distributed traffic control model in real time, enabling dynamic adjustment of vehicles in abnormal states. The overall technical solution constructs an intelligent collaborative system covering the entire process from perception and decision-making to execution through distributed collaboration of data acquisition, precise mapping of digital twins, localized decision-making of edge computing, dynamic optimization of communication networks, and closed-loop control of anomaly response. It provides multi-technology integration as the underlying support for dynamic guidance and speed coordination of traffic flow in construction areas, effectively improving the real-time performance, accuracy, and robustness of traffic control in complex scenarios.

[0077] Furthermore, the process of acquiring the real-time state set of the target area includes:

[0078] The static monitoring nodes in the data acquisition node network are defined as the location points corresponding to all static monitoring devices set up within the target area, and the dynamic monitoring nodes in the data acquisition node network are defined as the dynamic location points corresponding to all vehicles within the target area.

[0079] Simultaneously, a static monitoring connection sequence is constructed based on the data transmission path relationship between all static monitoring nodes, a first dynamic monitoring connection sequence is established based on the data transmission path relationship between all static monitoring nodes and dynamic monitoring nodes, and a second dynamic monitoring connection sequence is constructed based on the data transmission path relationship between all dynamic monitoring nodes.

[0080] Based on all static monitoring nodes, dynamic monitoring nodes, and their corresponding static and dynamic monitoring connection sequences, a data acquisition node network is constructed using a graph algorithm.

[0081] Based on the monitorable distance corresponding to each static monitoring node, the target area is divided and the monitoring overlap area corresponding to all static monitoring nodes is obtained.

[0082] When any of the dynamic monitoring nodes is in any monitoring overlap area, the static monitoring node with the shortest signal transmission time corresponding to the current dynamic monitoring node is selected to establish the first dynamic monitoring connection.

[0083] When the distance between any dynamic monitoring node and any static monitoring node or dynamic monitoring node is greater than the corresponding monitorable distance, the first dynamic monitoring connection or the second dynamic monitoring connection of the corresponding dynamic monitoring node is disconnected.

[0084] Update the first or second dynamic monitoring connection of the current dynamic monitoring node based on the current location of the corresponding dynamic monitoring node and its distance relationship with any static or dynamic monitoring node.

[0085] When any dynamic monitoring node leaves the target area, the first and second dynamic monitoring connections corresponding to the current dynamic monitoring node are directly disconnected and the current dynamic monitoring node is deleted.

[0086] When a new dynamic monitoring node enters the target area, the process of building and updating the first and second dynamic monitoring connections described above is repeated to update the data acquisition node network.

[0087] Based on Kalman filtering and the number of static monitoring nodes, a distributed filtering model is constructed and deployed to each static monitoring node in the data acquisition node network according to a distributed framework;

[0088] Based on the data acquisition node network, environmental data, vehicle operation data, road status data and pedestrian distribution data corresponding to the target area are collected in real time, and the vehicle operation data collected by each dynamic monitoring node is transmitted to the corresponding static monitoring node through the established first dynamic monitoring connection.

[0089] Each static monitoring node uses a deployed distributed filtering model to filter the monitoring data uploaded by itself and the dynamic monitoring nodes to obtain a real-time status set of the target area.

[0090] It is important to note that in this embodiment, fixed-location static monitoring nodes are deployed as infrastructure within the target area, while the real-time locations of all vehicles are used as dynamic monitoring nodes. Communication links between static nodes form a static monitoring connection sequence, links between static and dynamic nodes form a first dynamic monitoring connection sequence, and links between dynamic nodes form a second dynamic monitoring connection sequence. A graph algorithm is used to construct the network topology, dynamically adjusting connections based on node location and communication distance: when a dynamic node enters the overlapping monitoring area of ​​a static node, a connection is established with the static node exhibiting the least signal delay; when a dynamic node exceeds its communication range or leaves the area, the corresponding link is automatically disconnected and the network is updated. When a dynamic node is added or removed, the network topology is reconstructed in real time to ensure the continuity and integrity of data collection.

[0091] It is important to note that in this embodiment, each static monitoring node deploys an independent Kalman filter model to fuse sensor data from its own node and vehicle operation information uploaded by associated dynamic nodes. The dynamic nodes transmit data to the nearest static node in real time through the first dynamic monitoring connection sequence. The static nodes perform noise filtering, spatiotemporal alignment, and outlier correction on the raw data, and finally integrate environmental, vehicle, road, and pedestrian data to generate a highly consistent real-time regional status set. The filtering model adopts a distributed architecture, with each node performing independent calculations and sharing intermediate results, reducing the communication load of centralized processing.

[0092] Furthermore, the construction process of the vehicle-road-cloud three-level communication node space includes:

[0093] A guidance optimization subnet is constructed based on the static monitoring nodes and corresponding static monitoring connections in the data acquisition node network, and a guidance control subnet is constructed based on the dynamic monitoring nodes and corresponding second dynamic monitoring connections in the data acquisition node network.

[0094] The optimized mapping connection between each guidance optimization node in the guidance optimization subnet and the corresponding guidance control node in the guidance control subnet is obtained based on the first dynamic monitoring connection mapping corresponding to each static monitoring node.

[0095] The optimal command transmission interval distribution function is embedded into the optimized mapping connection between each guidance optimization node and the corresponding guidance control node, and the vehicle density and vehicle speed distribution function is embedded into the distributed edge control sub-model in each guidance control node.

[0096] Based on the path information of the corresponding positions of all guidance and control nodes at the current moment, combined with the guidance optimization subnet, guidance control subnet and optimization mapping connection, the vehicle-road-cloud three-level communication node space with differential control is obtained by combining graph algorithm with sleep wake-up algorithm and power control algorithm;

[0097] It should be noted that the specific implementation process of the sleep-wake algorithm and power control algorithm in the vehicle-road-cloud three-level communication node space in this embodiment includes:

[0098] First, the topological connection relationship between the guidance optimization subnet and the guidance control subnet is analyzed using a graph algorithm. For static guidance optimization nodes, a dynamic periodic sleep strategy based on data traffic thresholds is adopted. Specifically, in this embodiment, the dynamic periodic sleep strategy involves initiating a sleep cycle of duration t when the data traffic is detected to be below threshold Q for M consecutive cycles. During the sleep period, only a low-power wake-up timer is retained, and communication is resumed via a fast handshake protocol after wake-up. Furthermore, this embodiment employs a closed-loop adaptive adjustment algorithm for static nodes, dynamically adjusting the transmit power based on real-time received RSRP (Reference Signal Received Power) and SINR (Signal-to-Interference-plus-Noise Ratio) feedback. For dynamic guidance control nodes, a fusion of GPS trajectory prediction and blind communication is used. The event-triggered sleep mechanism for area detection needs further explanation. Specifically, in this embodiment, the event-triggered sleep mechanism triggers a sleep mode and shuts down unnecessary radio frequency equipment when the vehicle speed is below 5 km / h and enters a preset blind zone for 30 seconds, or leaves the target area boundary. Further explanation is needed regarding the embedded optimal command transmission interval function f(λ) = μ × (1 + β × ρ) based on queuing theory in the optimized mapping connection, where λ is the current node queue length, μ is the service rate, ρ is the load rate, and β is an adjustment factor, specifically 0.5-1.2, linking the sleep timing and power level of the control node. Simultaneously, the distribution function of vehicle density N and vehicle speed v is implanted into the edge model of the guidance control node. When N > threshold N0, the sleep period is dynamically shortened. When the vehicle speed exceeds the preset speed threshold corresponding to the power, the power adjustment priority is increased, and the power parameters are adjusted in advance by predicting the position at the next moment through Kalman filtering, so as to achieve a dynamic balance between communication efficiency and energy consumption of the three-level nodes in high-density traffic scenarios.

[0099] The static monitoring nodes correspond one-to-one with the guidance optimization nodes; the dynamic monitoring nodes correspond one-to-one with the guidance control nodes; and each guidance control node corresponds to each vehicle within the target area.

[0100] For example, such as Figure 2In this document, A, B...N represent corresponding boot optimization nodes, A1, A2, and A3 represent boot control nodes corresponding to boot optimization node A, B1 and B2 represent boot control nodes corresponding to boot optimization node B, and N1 and N2 represent boot control nodes corresponding to boot optimization node N. The solid lines between A1, A2, and A3 and boot optimization node A represent corresponding optimization mapping connections. The solid lines between B1 and B2 and boot optimization node B and between N1 and N2 and boot optimization node N have the same meaning.

[0101] It is important to note that in this embodiment, a guidance optimization subnet is constructed based on static nodes, responsible for global policy calculation; a guidance control subnet is constructed based on dynamic nodes, executing vehicle-level commands. The optimization subnet and the control subnet are associated through mapping connections, with each mapping connection containing an optimal command emission interval function. Vehicle density-speed distribution models are embedded within the control subnet nodes. Graph algorithms are used to analyze node positions and path affiliations, and control logic is defined differently: optimization subnet nodes coordinate regional traffic flow, while control subnet nodes adjust commands according to the characteristics of their respective vehicles. This is achieved through mapping relationships connected by solid lines, for example... Figure 2 The A1 / A2 / A3 modules enable the precise distribution of optimization strategies to control commands, forming a three-tier architecture of "cloud decision-making - road collaboration - vehicle execution".

[0102] Furthermore, it's important to note that in this embodiment, when a vehicle enters the target area, its corresponding dynamic node automatically connects to the network and establishes a mapping connection with the nearest static node. When vehicle movement causes a deterioration in link quality, a connection switch to a better static node is triggered. When commands are issued, the optimized subnet nodes generate command intervals and content based on built-in functions and push them to the associated control subnet nodes through the mapping connection. After receiving the commands, the vehicle terminal adjusts the vehicle speed or route in real time. After the vehicle leaves the area, its dynamic nodes and associated connections are automatically cleared, releasing network resources. The system ensures the reliability and real-time performance of command transmission through heartbeat detection and link quality monitoring.

[0103] Furthermore, the process of acquiring the guidance control commands for each vehicle includes:

[0104] Based on the data acquisition node network, the vehicle density, vehicle distance, vehicle speed and acceleration information corresponding to each path in the target area, and vehicle entry and exit information are collected in real time to obtain the real-time status set of the target area.

[0105] Based on the real-time state set of the target area, the optimal command transmission frequency and vehicle speed guidance parameter sequence for each guidance optimization node are obtained through the guidance optimization subnet in the vehicle-road-cloud three-level communication node space.

[0106] Based on the optimal command transmission frequency, the vehicle speed guidance parameter sequence is transmitted to the corresponding guidance control node, and the distributed edge control sub-model in the corresponding guidance control node is used to generate guidance control commands to perform differentiated real-time control of the corresponding vehicle speed.

[0107] This process achieves multi-dimensional technical improvements by constructing a dynamically adaptive data acquisition node network and a vehicle-road-cloud hierarchical communication architecture. First, based on a collaborative networking mechanism between static monitoring nodes and dynamic vehicle nodes, and through a dynamic construction and real-time update algorithm for dynamic monitoring connection sequences, it effectively solves the problems of coverage blind spots and mobile node disconnection in traditional fixed monitoring networks. The topology optimization capability of graph algorithms ensures network connectivity and data transmission timeliness, while node-level data processing using a distributed filtering model significantly reduces monitoring data latency and noise interference. Second, by spatially mapping vehicle dynamic location information with static monitoring nodes, a guidance and control subnet with optimized mapping connections is constructed, achieving three-level differentiated control of communication node space. Relying on a built-in optimal command transmission interval distribution function and edge computing model, the spatiotemporal distribution parameters of vehicle density and speed can be embedded into the control link in real time, achieving deep integration of cloud-based decision-making and edge execution. Third, based on an adaptive feedback mechanism using real-time state sets, by dynamically adjusting the command transmission frequency and parameter sequence, combined with a distributed filtering model... The localized computing capabilities of the distributed edge control sub-model effectively overcome the communication bottleneck of centralized control systems. While ensuring the real-time performance of vehicle speed guidance control commands, it achieves a dynamic balance between vehicle-level precise control and road network-level collaborative optimization, significantly improving the spatiotemporal adaptability of control commands in complex traffic scenarios. Furthermore, through global collaboration of the vehicle-road-cloud architecture, the system can optimize regional road network load balancing in real time, effectively reducing the overall congestion index and improving road resource utilization. Based on digital twin dynamic simulation and reinforcement learning decision-making, it supports predictive management of emergencies, greatly shortening response time and reducing accident rates. Through differentiated command distribution and dynamic scheduling of edge computing resources, it enables concurrent control of thousands of vehicles. Simultaneously, adaptive path planning and vehicle speed guidance strategies can reduce conflicts caused by sudden braking and lane changes, contributing to the construction of green and intelligent transportation. This solution provides a highly reliable, low-latency, and adaptive intelligent management and control system for complex construction scenarios.

[0108] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art, under the guidance of the present invention, can make changes, modifications, substitutions and variations to the above embodiments without departing from the spirit and scope of the claims. All of these variations are within the protection scope of the present invention.

Claims

1. A traffic flow guidance and speed coordination control system for construction areas, characterized in that, include: Area perception module, distributed guidance module, communication module, and emergency response module; The area perception module, based on the area map layer and real-time collected traffic flow, vehicle status, area environment and target area entry and exit information, obtains a three-dimensional simulation space of the target area through a three-dimensional modeling algorithm combined with a twin algorithm. Based on the three-dimensional simulation space of the target area, combined with the distributed diversion control model built into the distributed guidance module and the vehicle-road-cloud three-level communication node space built based on narrowband IoT technology in the communication module, the operating status of all vehicles in the target area is controlled in real time, and abnormal data in the target area is monitored in real time. At the same time, the abnormal data in the target area is fed back to the abnormal analysis algorithm and emergency handling strategy library built into the emergency response module to obtain emergency handling strategies. The emergency handling strategies are fed back to the distributed diversion control model to adjust the operating status of vehicles in abnormal state in real time. The vehicle-road-cloud three-level communication node space is constructed by combining vehicle data, road distribution data, pedestrian flow data, environmental status data, and the relationship between the location of monitoring equipment and data transmission within the target area using a graph algorithm. The distributed traffic control model is constructed by combining the congestion distribution of different target sub-regions, vehicle acceleration and speed, vehicle distance, probability distribution of anomalies between vehicles or in the environment, and vehicle travel paths with a distributed reinforcement learning algorithm.

2. The construction area traffic guidance and speed coordination control system as described in claim 1, characterized in that, The region perception module includes a data acquisition unit, a 3D modeling unit, and a twin simulation unit; The data acquisition unit, based on a preset data acquisition node network and a preset distributed filtering model, collects real-time information on the changes in the three-dimensional point cloud map of the target area, vehicle operation data, vehicle and pedestrian flow and road distribution information, and obtains a real-time status set of the target area. The three-dimensional modeling unit obtains a static three-dimensional model of the target area by combining a subset of static information from the real-time state set of the target area with a three-dimensional modeling algorithm; the static information subset includes the building distribution, road distribution, road surface attributes, and infrastructure information of the target area. The twin simulation unit, based on the static three-dimensional model of the target area and the dynamic information subset of the real-time state set of the target area, uses the twin simulation algorithm to simulate the real-time changes in vehicles, pedestrian flow and equipment in the target area, thereby obtaining the dynamic simulation information set of the target area.

3. The construction area traffic guidance and speed coordination control system as described in claim 2, characterized in that, The distributed guidance module includes an edge computing unit and a guidance control unit; the communication module includes a communication construction unit and a communication transmission unit; the distributed traffic control model includes a distributed edge optimization sub-model and a distributed edge control sub-model. The edge computing unit is used to obtain the operation optimization parameters for each vehicle on all roads within the target area based on the dynamic simulation information set of the target area and the preset distributed edge optimization sub-model. The operational optimization parameters for each vehicle include the corresponding operating speed, acceleration, and the distance between vehicles in front and behind, as well as the distribution of command transmission intervals. The guidance control unit obtains the guidance control command corresponding to each vehicle based on the operation optimization parameters of each vehicle and the distributed edge control sub-model. Simultaneously, based on the command transmission interval distribution, the guidance and control command corresponding to each vehicle is sent to the corresponding vehicle node through the vehicle-road-cloud three-level communication node space. Each vehicle node performs real-time control and adjustment of the vehicle's operating status and operating path through the received guidance and control command.

4. The construction area traffic flow guidance and speed coordination control system as described in claim 3, characterized in that, The process of acquiring the real-time status set of the target area includes: The static monitoring nodes in the data acquisition node network are defined as the location points corresponding to all static monitoring devices set up within the target area, and the dynamic monitoring nodes in the data acquisition node network are defined as the dynamic location points corresponding to all vehicles within the target area. Simultaneously, a static monitoring connection sequence is constructed based on the data transmission path relationship between all static monitoring nodes, a first dynamic monitoring connection sequence is established based on the data transmission path relationship between all static monitoring nodes and dynamic monitoring nodes, and a second dynamic monitoring connection sequence is constructed based on the data transmission path relationship between all dynamic monitoring nodes. A data acquisition node network is constructed based on all static monitoring nodes, dynamic monitoring nodes, and their corresponding static and dynamic monitoring connection sequences using a graph algorithm.

5. The construction area traffic flow guidance and speed coordination control system as described in claim 4, characterized in that, The process of acquiring the real-time status set of the target area also includes: Based on the monitorable distance corresponding to each static monitoring node, the target area is divided and the monitoring overlap area corresponding to all static monitoring nodes is obtained. When any of the dynamic monitoring nodes is in any monitoring overlap area, the static monitoring node with the shortest signal transmission time corresponding to the current dynamic monitoring node is selected to establish the first dynamic monitoring connection. When the distance between any dynamic monitoring node and any static monitoring node or dynamic monitoring node is greater than the corresponding monitorable distance, the first dynamic monitoring connection or the second dynamic monitoring connection of the corresponding dynamic monitoring node is disconnected. Update the first or second dynamic monitoring connection of the current dynamic monitoring node based on the current location of the corresponding dynamic monitoring node and its distance relationship with any static or dynamic monitoring node. When any dynamic monitoring node leaves the target area, the first and second dynamic monitoring connections corresponding to the current dynamic monitoring node are directly disconnected and the current dynamic monitoring node is deleted. When a new dynamic monitoring node enters the target area, the process of building and updating the first and second dynamic monitoring connections described above is repeated to update the data acquisition node network.

6. The construction area traffic flow guidance and speed coordination control system as described in claim 5, characterized in that, The process of acquiring the real-time status set of the target area also includes: Based on Kalman filtering and the number of static monitoring nodes, a distributed filtering model is constructed and deployed to each static monitoring node in the data acquisition node network according to a distributed framework; Based on the data acquisition node network, environmental data, vehicle operation data, road status data and pedestrian distribution data corresponding to the target area are collected in real time, and the vehicle operation data collected by each dynamic monitoring node is transmitted to the corresponding static monitoring node through the established first dynamic monitoring connection. Each static monitoring node uses a deployed distributed filtering model to filter the monitoring data uploaded by itself and the dynamic monitoring nodes to obtain a real-time status set of the target area.

7. The construction area traffic guidance and speed coordination control system as described in claim 6, characterized in that, The construction process of the vehicle-road-cloud three-level communication node space includes: A guidance optimization subnet is constructed based on the static monitoring nodes and corresponding static monitoring connections in the data acquisition node network, and a guidance control subnet is constructed based on the dynamic monitoring nodes and corresponding second dynamic monitoring connections in the data acquisition node network. Based on the first dynamic monitoring connection mapping corresponding to each static monitoring node, the optimized mapping connection between each guidance optimization node in the guidance optimization subnet and the corresponding guidance control node in the guidance control subnet is obtained.

8. The construction area traffic flow guidance and speed coordination control system as described in claim 7, characterized in that, The process of constructing the command transmission interval distribution includes: In the three-dimensional simulation space of the target area, the data acquisition node network is used to simulate and collect in real time the vehicle density, vehicle distance, vehicle speed and acceleration information corresponding to each path in the target area, the frequency of guidance and control command transmission under each vehicle density, and the probability of vehicle anomaly on each path under each guidance and control command transmission frequency, so as to obtain the real-time state set of the simulated target area. Based on the frequency of guidance and control command transmission under each vehicle density and the probability of vehicle anomalies on each path under each guidance and control command transmission frequency, a command transmission interval distribution function is constructed by support vector machine. The real-time state set of the simulated target area is input into the vehicle-road-cloud three-level communication node space. The command transmission interval is optimized by minimizing the anomaly probability corresponding to the command transmission interval distribution function as the optimization objective and by combining the optimization algorithm. The optimal command transmission interval distribution function and the distribution functions of vehicle density and speed corresponding to minimizing the anomaly probability are obtained for each vehicle density and vehicle speed.

9. The construction area traffic guidance and speed coordination control system as described in claim 8, characterized in that, The construction process of the vehicle-road-cloud three-level communication node space also includes: The optimal command transmission interval distribution function is embedded into the optimized mapping connection between each guidance optimization node and the corresponding guidance control node, and the vehicle density and vehicle speed distribution function is embedded into the distributed edge control sub-model in each guidance control node. Based on the path information of the corresponding positions of all guidance and control nodes at the current moment, combined with the guidance optimization subnet, guidance control subnet and optimization mapping connection, the vehicle-road-cloud three-level communication node space with differential control is obtained by combining graph algorithm with sleep wake-up algorithm and power control algorithm; The static monitoring nodes correspond one-to-one with the guidance optimization nodes; the dynamic monitoring nodes correspond one-to-one with the guidance control nodes; and each guidance control node corresponds to each vehicle within the target area.

10. The construction area traffic flow guidance and speed coordination control system as described in claim 9, characterized in that, The process of obtaining the guidance and control command corresponding to each vehicle includes: Based on the data acquisition node network, the vehicle density, vehicle distance, vehicle speed and acceleration information corresponding to each path in the target area, and vehicle entry and exit information are collected in real time to obtain the real-time status set of the target area. Based on the real-time state set of the target area, the optimal command transmission frequency and vehicle speed guidance parameter sequence for each guidance optimization node are obtained through the guidance optimization subnet in the vehicle-road-cloud three-level communication node space. Based on the optimal command transmission frequency, the vehicle speed guidance parameter sequence is transmitted to the corresponding guidance control node, and the distributed edge control sub-model in the corresponding guidance control node is used to generate guidance control commands to perform differentiated real-time control of the corresponding vehicle speed.

Images