Method for opening section emergency lane under heavy traffic flow

Abstract

The application provides a section emergency lane opening method under large traffic flow, comprising: when the target section traffic flow density reaches a first saturation threshold, starting an elastic lane dynamic decision engine, matching real-time traffic data with enabling criteria to generate an initial management and control decision and broadcast, triggering a first observation window; generating a resource configuration scheme containing an enabling gradient based on section occupancy, issuing an information sign update instruction and triggering a second window; reconstructing an optimization strategy according to the congestion coefficient and setting a decision cycle, generating a dynamic configuration command broadcast and triggering a third window; generating a final control instruction according to millimeter wave radar response data and triggering a fourth window; integrating the traffic efficiency index to determine the optimal elastic allocation scheme through multi-cycle iteration, returning to the strategy reconfiguration point after the strategy is reconstructed, and finally serving as the basis for execution. The application can improve the road network traffic efficiency under large traffic flow, and ensure regional collaborative control and optimal resource allocation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent traffic control, and more particularly, to a section emergency lane opening method under large traffic flow. BACKGROUND

[0002] In today's society, with the acceleration of urbanization and the continuous rise of motor vehicle ownership, traffic congestion has become one of the key factors restricting urban development and the improvement of residents' quality of life. Especially in the road section with large traffic flow, the traffic congestion phenomenon is more serious, not only leads to low vehicle traffic efficiency, but also may cause traffic accidents, further exacerbating the deterioration of traffic conditions. In order to alleviate this problem, the existing traffic management measures mainly focus on optimizing signal timing, increasing road capacity, and implementing traffic control, etc. Among them, some areas try to open emergency lanes for ordinary vehicles to use at certain times in order to alleviate traffic pressure by increasing available road resources. However, this emergency lane opening measure often lacks fine control means, usually based on manual judgment or simple flow threshold triggering, and is difficult to dynamically adjust according to real-time traffic situation, resulting in limited effect in actual application, and even may cause new traffic safety hazards.

[0003] The existing traffic flow control technology mainly relies on fixed traffic rules and pre-set lane allocation schemes, which are often not flexible enough to adapt to the rapid changes of traffic flow in the face of sudden traffic peaks or abnormal traffic events. In addition, although the existing traffic monitoring equipment can provide certain traffic data, there are still deficiencies in the real-time and accuracy of the data, which is difficult to meet the needs of fine traffic control. In terms of technical principles, the existing traffic control system mostly uses static models and algorithms, which lack dynamic adaptability in traffic flow prediction and control, and cannot fully consider the spatio-temporal characteristics of traffic flow and the mutual influence between different road sections.

[0004] Firstly, the existing emergency lane opening measures lack intelligent and dynamic control mechanisms and cannot accurately allocate lane resources according to real-time traffic conditions; secondly, the existing traffic monitoring equipment and data processing methods have deficiencies in data accuracy and real-time, which is difficult to meet the needs of fine traffic control; finally, the existing traffic control system lacks effective regional coordination control capability when facing complex traffic networks and variable traffic flow, and cannot achieve global optimization of the entire traffic system. SUMMARY

[0005] The present application provides a section emergency lane opening method under large traffic flow, comprising:

[0006] After detecting that the target section traffic flow density reaches the first saturation threshold, the elastic lane dynamic decision engine is started, real-time traffic situation data is matched with the elastic activation criterion, an initial management and control decision is generated and broadcast to the roadside perception unit, and a first observation window is triggered;

[0007] After the first observation window is closed, a lane resource configuration scheme is generated according to the section occupancy rate index reported by the roadside perception unit, a state update instruction is issued to the variable information sign, a second observation window is triggered, and the resource configuration scheme includes an elastic lane activation gradient;

[0008] After the second observation window is closed, the lane optimization strategy is reconstructed based on the real-time congestion coefficient, the current decision period is set to 1, the strategy reconfiguration point is entered, the dynamic lane configuration command is generated according to the optimization strategy and the decision period, and is broadcast to the millimeter wave radar, and a third observation window is triggered;

[0009] After the third observation window is closed, the final lane control instruction is generated according to the state response data of the millimeter wave radar, and is broadcast to the signal control machine, and a fourth observation window is triggered;

[0010] After the fourth observation window is closed, the optimal elastic allocation scheme is determined through multi-period iteration based on the integrated vehicle traffic efficiency index, and the optimization strategy is reconstructed to return to the strategy reconfiguration point in the iteration process. The optimal elastic allocation scheme is used as the basis for executing the dynamic lane control system.

[0011] Further, the lane resource configuration scheme is generated according to the section occupancy rate index reported by the roadside perception unit, including: verifying the effectiveness of the original data transmitted by the roadside perception unit, if the verification is passed, calculating the difference between the section occupancy rate and the output value of the dynamic baseline model to generate an elastic expansion coefficient;

[0012] When the elastic expansion coefficient breaks through the second saturation threshold, the elastic lane activation gradient is mapped to a segmented function of the expansion coefficient, the lane alignment optimization algorithm is used to calculate the channelization parameters corresponding to the gradient, and the resource configuration scheme including the activation gradient and the channelization parameters is generated;

[0013] When the elastic expansion coefficient does not break through the second saturation threshold, the elastic lane activation gradient is set to zero and a lane state maintenance instruction is generated.

[0014] Further, it further includes:

[0015] A short-term traffic situation deduction model is cooperatively run with adjacent section control nodes to jointly generate regional linkage control parameters as boundary constraint conditions for dynamic decision.

[0016] Further, the lane optimization strategy is reconstructed based on the real-time congestion coefficient, including: performing integrity verification on the section queue length data collected by the millimeter wave radar, and if the verification is passed, inputting the queue length into a congestion evaluation buffer;

[0017] If the congestion evaluation buffer meets a first activation condition, calculating a minimum activation interval of the elastic lane by a congestion diffusion model, taking the minimum activation interval as a core decision variable of the optimization strategy, and marking the congestion level as 1; the first activation condition is that the queue length growth rate in the continuous three observation windows exceeds a dynamic threshold;

[0018] If the first activation condition is not met, resetting the optimization strategy variable to a baseline value and marking the congestion level as 0.

[0019] Further, a dynamic lane configuration command is generated according to the optimization strategy and a decision cycle, including: using a rolling time domain optimization algorithm to solve the core decision variable in multiple objectives to generate a phase priority weight;

[0020] If the decision cycle is 1 and the congestion level is 1, encoding the core variable, the phase priority weight and the prediction result of the congestion diffusion model into a binary control instruction stream;

[0021] If the decision cycle is not 1 or the congestion level is not 1, only the phase priority weight is converted into a lane configuration command.

[0022] Further, a final lane control instruction is generated according to the state response data of the millimeter wave radar, including: performing compliance audit on the lane switching action completion rate returned by the millimeter wave radar;

[0023] When the audit fails, isolate abnormal data and activate the device self-checking process;

[0024] When the audit passes, if the state response data meets a second activation condition, perform deviation analysis on the effective switching rate and the expected index to generate a control instruction containing a phase difference compensation parameter; the second activation condition is that the number of effective data packets in a single decision cycle exceeds a confidence threshold;

[0025] If the second activation condition is not met, generate a lane state maintenance instruction.

[0026] Further, the optimal elastic ratio scheme is determined through multi-cycle iteration, including: constructing a traffic efficiency evaluation matrix according to historical decision logs, the matrix storing the mapping relationship between the traffic efficiency of each cycle and the lane activation gradient;

[0027] If the current decision cycle is 1, set the baseline traffic improvement rate to an industry standard value;

[0028] If the current decision period is not 1, a dynamic shunting algorithm is cooperatively executed with an adjacent section node, and a regional performance improvement threshold is jointly generated;

[0029] Based on the regional performance improvement threshold and the evaluation matrix, the elastic ratio corresponding to the extreme value of the passing efficiency is solved by the gradient descent method, and the optimized strategy is returned to the strategy reconfiguration point.

[0030] Further, when constructing the passing performance evaluation matrix, trend analysis is performed on the outlying data in the historical decision log.

[0031] When the analysis finds that the data is abnormal, the Kalman filter algorithm is used for data repair.

[0032] When the data is normal, the passing efficiency of each period is subjected to multivariate linear fitting analysis with the corresponding gradient, an efficiency-gradient response surface is generated and stored in the evaluation matrix.

[0033] Further, the optimal elastic ratio scheme is solved by the gradient descent method, including: if the evaluation matrix meets the third activation condition, the ratio scheme corresponding to the extreme point of the surface is taken as the optimal solution candidate set; the third activation condition is that the efficiency gain rate of adjacent two iterations is lower than the gradient convergence threshold;

[0034] If the passing efficiency corresponding to the candidate set exceeds the regional performance improvement threshold, the scheme is determined as the optimal elastic ratio scheme;

[0035] If the efficiency of the candidate set does not reach the threshold, the current decision period is incremented by 1, and the candidate set is returned to the strategy reconfiguration point as a new optimization strategy;

[0036] If the evaluation matrix does not meet the third activation condition, the decision period is incremented by 1, and the regional performance improvement threshold is used.

[0037] As a strategy parameter, the strategy reconfiguration point is returned.

[0038] Further, it further includes:

[0039] Real-time detection is performed to determine whether a regional cooperative control instruction of an adjacent section is received;

[0040] If the instruction is received, the optimal ratio scheme is coupled with the regional control strategy to generate a wide-area cooperative control code stream and broadcast it to the networked joint control device.

[0041] The above embodiments of the present application have at least the following beneficial effects:

[0042] 1. Based on the real-time traffic situation, the opening time of the emergency lane and the gradient to be used are dynamically decided, the management and control strategy is continuously optimized through a multi-level observation window closed-loop feedback mechanism, thereby accurately matching the fluctuation demand of traffic flow, effectively relieving road congestion during peak hours and improving the road network passing capacity.

[0043] 2. Combined with multi-traffic parameters such as cross-section occupancy and queue length, a rolling time domain optimization algorithm is used to generate a lane resource allocation scheme, and the optimal elastic allocation is determined through multi-cycle iteration calculation to realize fine control of lane resources and avoid resource mismatch caused by traditional fixed control mode.

[0044] 3. The regional traffic flow is optimized through a cooperative control mechanism with adjacent sections, the congestion propagation trend is predicted by using a short-term traffic situation deduction model, and the reliable execution of control instructions is ensured by cooperating with a device state monitoring and data verification system, thereby improving the coordination and stability of the entire road network system. BRIEF DESCRIPTION OF DRAWINGS

[0045] The above and other objects, features and advantages of the present application will become more apparent from the following detailed description read in conjunction with the accompanying drawings. In the drawings, several embodiments of the present application are shown by way of example and not limitation, in which:

[0046] Figure 1 A flowchart of a section emergency lane opening method under heavy traffic flow is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0047] The principles and spirits of the present application will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are given only to enable those skilled in the art to better understand and implement the present application, and do not limit the scope of the present application in any way. On the contrary, these embodiments are provided to make the present application more thorough and complete, and to fully convey the scope of the present application to those skilled in the art.

[0048] Those skilled in the art know that the embodiments of the present application can be implemented as a system, device, apparatus, method or computer program product. Therefore, the present application can be specifically implemented in the following forms: complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software.

[0049] It should be noted that any number of elements in the drawings is used for example and not limitation, and any naming is only used for distinction and does not have any limiting meaning.

[0050] The principles and spirits of the present application will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are given only to enable those skilled in the art to better understand and implement the present application, and do not limit the scope of the present application in any way. On the contrary, these embodiments are provided to make the present application more thorough and complete, and to fully convey the scope of the present application to those skilled in the art. Figure 1 , Figure 1 A flowchart of a section emergency lane opening method under heavy traffic flow is provided for an embodiment of the present application. As shown in Figure 1 , a section emergency lane opening method under heavy traffic flow includes:

[0051] S1, after detecting that the target cross-section traffic density reaches a first saturation threshold, starting the elastic lane dynamic decision engine, matching real-time traffic situation data with elastic activation criteria, generating an initial management and control decision and broadcasting to the roadside perception unit, and triggering a first observation window;

[0052] S2, after the first observation window is closed, generating a lane resource configuration scheme according to the cross-section occupancy rate index reported by the roadside perception unit, issuing a state update instruction to the variable information sign, triggering a second observation window, the resource configuration scheme including an elastic lane activation gradient;

[0053] S3, after the second observation window is closed, reconstructing the lane optimization strategy based on the real-time congestion coefficient, setting the current decision period to 1; entering the strategy reconfiguration point; generating a dynamic lane configuration command according to the optimization strategy and the decision period, and broadcasting to the millimeter wave radar, triggering a third observation window;

[0054] S4, after the third observation window is closed, generating a final lane control instruction according to the state response data of the millimeter wave radar, and broadcasting to the signal control machine, triggering a fourth observation window;

[0055] S5, after the fourth observation window is closed, determining the optimal elastic allocation scheme through multi-cycle iteration based on the integrated vehicle traffic efficiency index, returning to the strategy reconfiguration point after reconstructing the optimization strategy in the iteration process, and taking the optimal elastic allocation scheme as the basis for executing the dynamic lane control system.

[0056] It should be noted that when the target cross-section traffic density reaches the first saturation threshold, the elastic lane dynamic decision engine will be started. The target cross-section refers to a specific cross-sectional position on the road for monitoring traffic density; the first saturation threshold is a preset traffic density value, when the actual traffic density reaches or exceeds this value, it indicates that the traffic flow has approached or reached the carrying limit of the road, and measures need to be taken for dredging. At this time, the elastic lane dynamic decision engine will be activated, which is an intelligent decision system that can match real-time traffic situation data with elastic activation criteria, generate an initial management and control decision and broadcast to the roadside perception unit. The roadside perception unit is a sensor device installed on both sides of the road for collecting traffic data such as speed, traffic flow, etc. At the same time, the system will trigger the first observation window, which is a time window for collecting and analyzing traffic data within that time period to further assess traffic conditions and make appropriate decisions.

[0057] Specifically, the section occupancy rate reported by the roadside perception unit is an important parameter for measuring traffic flow. It represents the proportion of road occupied by vehicles at a certain section. When this indicator reaches a certain value, the system will generate a lane resource allocation scheme according to the pre-set rules. The flexible lane activation gradient here refers to the strategy of gradually activating or closing the emergency lane according to the change of traffic flow. For example, when traffic flow increases, part or all of the emergency lane can be gradually opened to ordinary vehicles; when traffic flow decreases, the original state of the emergency lane is gradually restored. This gradient can be flexibly adjusted according to the actual traffic situation to achieve the best traffic diversion effect. At the same time, the system will trigger a second observation window, which is used to further collect and analyze traffic data to optimize the lane resource allocation scheme. After the second observation window is closed, the system will reconstruct the lane optimization strategy based on the real-time congestion coefficient, which is a parameter reflecting the degree of traffic congestion and can be calculated based on the section queue length data collected by the millimeter wave radar. Millimeter wave radar is a sensor installed on the road surface that can detect vehicle passage and provide real-time traffic data to the system. Through these data, the system can more accurately assess the traffic congestion situation and adjust the lane optimization strategy accordingly.

[0058] Preferably, when generating the lane resource allocation scheme, the original data transmitted by the roadside perception unit will be verified for effectiveness. This means that the system will check whether the data is accurate and complete to ensure the reliability of subsequent decisions. If the verification is passed, the system will calculate the difference between the section occupancy rate and the dynamic baseline model output value to generate the flexible expansion coefficient. The dynamic baseline model here is a model established based on historical traffic data to predict the section occupancy rate under normal traffic conditions. By comparing the actual section occupancy rate with the model prediction value, the flexible expansion coefficient can be calculated, which reflects the deviation degree of the current traffic flow from the normal flow. When the flexible expansion coefficient exceeds the second saturation threshold, the system will map the flexible lane activation gradient to a segmented function of the expansion coefficient, and calculate the channelization parameters corresponding to the gradient using the lane line optimization algorithm. The channelization parameters here refer to parameters used to optimize lane layout and traffic flow lines, such as lane width, position of lane separation line, etc. Through the optimization of these parameters, the road capacity can be improved. At the same time, the system will trigger a third observation window, which is used to further collect and analyze traffic data to adjust the lane optimization strategy. After the third observation window is closed, the system will generate the final lane control instruction based on the state response data of the millimeter wave radar and broadcast it to the signal control machine. The state response data here refers to the data returned by the millimeter wave radar about the completion rate of lane switching actions, and the system will generate the final lane control instruction based on these data to achieve effective control of traffic flow.

[0059] In some embodiments, a lane resource configuration scheme is generated according to the cross-section occupancy rate index reported by the roadside perception unit, including: performing validity verification on the original data transmitted by the roadside perception unit, if the verification passes, calculating the difference between the cross-section occupancy rate and the output value of the dynamic baseline model to generate an elastic expansion coefficient;

[0060] When the elastic expansion coefficient breaks through the second saturation threshold, the elastic lane enabling gradient is mapped to a segmented function of the expansion coefficient, the lane line optimization algorithm is used to calculate the channelization parameters corresponding to the gradient, and a resource configuration scheme containing the enabling gradient and the channelization parameters is generated;

[0061] When the elastic expansion coefficient does not break through the second saturation threshold, the elastic lane enabling gradient is set to zero and a lane state maintenance instruction is generated.

[0062] It should be noted that the process of generating a lane resource configuration scheme according to the cross-section occupancy rate index reported by the roadside perception unit is a key link in the entire traffic flow control method. The cross-section occupancy rate index is an important parameter for measuring traffic flow density, which reflects the proportion of vehicles occupying the road on a specific road section. Through this index, the system can determine whether the current traffic flow has reached the level that requires adjustment of lane configuration. When generating the resource configuration scheme, the system will perform validity verification on the original data transmitted by the roadside perception unit to ensure the accuracy and reliability of the data. If the verification passes, the system will further calculate the elastic expansion coefficient, which is a parameter for measuring the degree of traffic flow exceeding the normal carrying capacity. When the elastic expansion coefficient breaks through the preset second saturation threshold, the system will determine the enabling gradient of the elastic lane according to this coefficient and generate the corresponding resource configuration scheme, including the enabling gradient and the channelization parameters, to optimize the lane layout and improve the road capacity.

[0063] Specifically, the roadside perception unit is a sensor device installed on both sides of the road, used to monitor traffic data such as traffic flow, speed, etc. in real time, and transmit these data to the control center. The section occupancy rate is an index obtained by analyzing and calculating these data, which is a value between 0 and 1, the higher the value, the greater the proportion of vehicles occupying the road. The elastic expansion coefficient is calculated by comparing the section occupancy rate with the dynamic baseline model output value, and the dynamic baseline model is established based on historical traffic data to predict the section occupancy rate under normal traffic conditions. The second saturation threshold is a preset threshold, when the elastic expansion coefficient exceeds this value, it indicates that the traffic flow has exceeded the normal carrying capacity of the road, and the elastic lane needs to be enabled to relieve traffic pressure. The elastic lane enabling gradient refers to the strategy of gradually enabling or closing the emergency lane according to the different elastic expansion coefficients. For example, when the elastic expansion coefficient is low, only part of the emergency lane is enabled; when the elastic expansion coefficient is high, all emergency lanes can be enabled. The channelization parameter refers to the parameter used to optimize the layout of the lane, such as the width of the lane, the position of the lane separation line, etc., the adjustment of these parameters can improve the efficiency of the road.

[0064] Preferably, when generating the lane resource configuration scheme, the system will use a series of algorithms and models to ensure the scientificity and effectiveness of the scheme. First, the system will verify the effectiveness of the original data transmitted by the roadside perception unit, which includes checking the integrity, accuracy and whether it meets the expected traffic model. If the verification is passed, the system will calculate the difference between the section occupancy rate and the dynamic baseline model output value to generate the elastic expansion coefficient. The construction process of the dynamic baseline model includes collecting historical traffic data such as traffic flow and speed under different time periods and weather conditions, and then establishing a model that can reflect the normal traffic conditions through data analysis and modeling techniques. When calculating the elastic expansion coefficient, the system will compare the current section occupancy rate with the normal section occupancy rate predicted by the model, and calculate the difference between the two. When the elastic expansion coefficient exceeds the second saturation threshold, the system will map the elastic lane enabling gradient to a segmented function of the expansion coefficient according to the preset rules. For example, the elastic expansion coefficient can be divided into several intervals, each interval corresponds to a different degree of elastic lane enabling. At the same time, the system will use a lane alignment optimization algorithm to calculate the channelization parameters corresponding to the gradient, which will consider the geometric shape of the road, traffic flow distribution, etc. through optimization calculation to obtain the optimal lane layout parameters. These parameters will be used to generate the final resource configuration scheme to achieve effective management and optimization of traffic flow.

[0065] In some embodiments, further comprising:

[0066] The short-time traffic situation deduction model cooperates with the adjacent section control nodes to jointly generate regional linkage control parameters as boundary constraint conditions for dynamic decision-making.

[0067] It should be noted that the short-time traffic situation deduction model mentioned in the method cooperates with the adjacent section control nodes to jointly generate regional linkage control parameters as boundary constraint conditions for dynamic decision-making, which is an important part of realizing regional traffic coordination optimization. The adjacent section control nodes here refer to other traffic control nodes that are adjacent to the current target section in geographical location in the traffic network, which together form a regional traffic control system. The short-time traffic situation deduction model is a model for predicting the trend of traffic flow changes in the future short time, which can predict the traffic conditions in the future period of time through analysis and processing of current traffic data. Regional linkage control parameters refer to parameters used to coordinate the actions of control nodes in regional traffic coordination control, which can ensure effective management and optimization of regional traffic flow. Boundary constraint conditions refer to limiting conditions that need to be considered when making dynamic decisions, which can ensure the feasibility and effectiveness of the decisions.

[0068] Specifically, the short-time traffic situation deduction model is based on historical traffic data and real-time traffic data, which can predict the trend of traffic flow changes. The construction of this model requires the input of multiple parameters, including but not limited to current traffic density, speed, section occupancy rate, and other real-time traffic data, as well as parameters reflecting the traffic flow change law in historical traffic data. Through these parameters, the model can predict the trend of traffic flow changes in the future short time. Regional linkage control parameters are generated based on the prediction results of the deduction model, including but not limited to signal coordination parameters between control nodes, lane allocation parameters, etc., which can ensure the coordinated action of control nodes, thereby realizing the optimization of regional traffic flow. Boundary constraint conditions refer to limiting conditions that need to be considered when making dynamic decisions, including but not limited to road capacity limits, traffic regulation limits, etc., which can ensure the feasibility and effectiveness of the decisions.

[0069] Preferably, when running the short-time traffic situation inference model in cooperation with the adjacent section control node, the following steps can be taken: first, collect real-time traffic data of the current target section and the adjacent section control node, including traffic density, vehicle speed, section occupancy rate, etc. Then, input these real-time data into the short-time traffic situation inference model, which will predict the trend of traffic flow changes in the future short time according to historical traffic data and real-time traffic data. Then, according to the prediction results of the inference model, generate regional linkage control parameters, which will be used to coordinate the actions between control nodes. Finally, input these regional linkage control parameters as boundary constraints of dynamic decision-making into the elastic lane dynamic decision engine to ensure the scientificity and rationality of decision-making. In actual application, the parameters and algorithms of the model can be adjusted and optimized according to specific traffic conditions and needs to improve the prediction accuracy of the model and the effectiveness of decision-making.

[0070] In some embodiments, the lane optimization strategy is reconstructed based on real-time congestion coefficient, including: performing integrity verification on the section queue length data collected by the millimeter wave radar, and if the verification is passed, inputting the queue length into the congestion evaluation buffer area;

[0071] If the congestion evaluation buffer area meets the first activation condition, calculate the minimum activation interval of the elastic lane by the congestion diffusion model, take the minimum activation interval as the core decision variable of the optimization strategy, and mark the congestion level as 1; the first activation condition is that the queue length growth rate in the continuous three observation windows exceeds the dynamic threshold;

[0072] If the first activation condition is not met, reset the optimization strategy variable to the baseline value and mark the congestion level as 0.

[0073] It should be noted that the process of reconstructing the lane optimization strategy based on real-time congestion coefficient is an important link in the traffic flow control method, which is used to dynamically adjust the lane configuration to deal with traffic congestion. Real-time congestion coefficient is a parameter that reflects the current traffic congestion degree, which evaluates the congestion state of traffic flow by analyzing the section queue length data collected by the millimeter wave radar. In this process, the system will perform integrity verification on the collected data to ensure the accuracy and integrity of the data. If the verification is passed, the system will calculate the congestion coefficient according to the queue length data, and judge whether the lane optimization strategy needs to be adjusted according to the preset activation condition. The core of this process is to dynamically adjust the lane configuration through scientific models and algorithms to improve the efficiency of road traffic.

[0074] In particular, the millimeter wave radar is a sensor installed on the road surface, which can monitor the passing vehicles in real time, including the speed, number and queue length of the vehicles. The queue length data refers to the length of vehicles queuing at a certain section, which is an important basis for evaluating the degree of traffic congestion. The integrity verification is to check the collected queue length data to ensure that there is no missing or abnormal data. The congestion evaluation buffer is a temporary storage area for queue length data for further analysis and processing. The first activation condition is a preset condition for determining whether to start the adjustment of the lane optimization strategy. For example, if the growth rate of the queue length in the last three observation windows exceeds the preset dynamic threshold, it is considered that the first activation condition is met, and the lane optimization strategy needs to be adjusted. The congestion diffusion model is a model for predicting the diffusion trend of traffic congestion, which can calculate the minimum activation interval of the elastic lane according to the queue length data, which is the core decision variable of the optimization strategy. The congestion level is a parameter for marking the current traffic congestion state, which is marked as 1 when the first activation condition is met, indicating that the lane optimization strategy needs to be adjusted; otherwise, the congestion level is marked as 0, indicating that the current lane configuration is maintained.

[0075] Preferably, in the process of reconstructing the lane optimization strategy based on the real-time congestion coefficient, the following steps can be used for refinement. First, the integrity of the queue length data collected by the millimeter wave radar is verified, and the verification process includes checking the continuity, reasonableness and whether there are abnormal values. If the verification is passed, the queue length data is input into the congestion evaluation buffer. Then, it is judged whether the data in the congestion evaluation buffer meets the first activation condition, i.e. whether the growth rate of the queue length in the last three observation windows exceeds the preset dynamic threshold. If the condition is met, the minimum activation interval of the elastic lane is calculated by the congestion diffusion model. The construction of this model requires input of queue length data, road geometric parameters and traffic flow data, etc. By simulating the diffusion process of traffic congestion, the minimum activation interval is calculated. Then, the minimum activation interval is taken as the core decision variable of the optimization strategy, and the congestion level is marked as 1. If the first activation condition is not met, the optimization strategy variable is reset to the baseline value, and the congestion level is marked as 0. Finally, according to the congestion level and the core decision variable, the corresponding lane optimization strategy is generated to realize the effective management and optimization of traffic flow.

[0076] In some embodiments, according to the optimization strategy and the decision cycle, a dynamic lane configuration command is generated, including: using a rolling time domain optimization algorithm to solve the core decision variable for multiple objectives to generate a phase priority weight;

[0077] If the decision cycle is 1 and the congestion level is 1, the core variable, the phase priority weight and the prediction result of the congestion diffusion model are encoded into a binary control instruction stream.

[0078] If the decision period is not 1 or the congestion level is not 1, only the phase priority weight is converted into a lane configuration command.

[0079] It should be noted that the process of generating dynamic lane configuration commands according to the optimization strategy and the decision period is a key step in realizing dynamic control of traffic flow. This process solves multiple objectives for core decision variables through a rolling horizon optimization algorithm, generating phase priority weights. The rolling horizon optimization algorithm is a dynamic optimization method that can optimize traffic flow within a limited time range. Phase priority weight refers to the weight of the priority of passage given to traffic in different directions in traffic signal control. When the decision period is 1 and the congestion level is 1, the system encodes the core variables, phase priority weights, and prediction results of the congestion diffusion model into a binary control instruction stream to quickly respond to traffic congestion. If the decision period is not 1 or the congestion level is not 1, only the phase priority weight is converted into a lane configuration command to maintain the stability of traffic flow.

[0080] Specifically, the rolling horizon optimization algorithm is a dynamic optimization algorithm that optimizes traffic flow within a limited time range to achieve dynamic adjustment of traffic signals. Core decision variables refer to key parameters that need to be adjusted in the optimization process, such as the degree of activation of elastic lanes, the phase length of signal lights, etc. Phase priority weight is the priority weight of passage allocated to traffic in different directions according to traffic flow. The decision period refers to the time required for the system to make a complete optimization decision, which determines the response speed of the system to traffic flow changes. The congestion level is a parameter used to mark the current traffic congestion state, and when the congestion level is 1, it indicates that the traffic flow has reached the degree that requires emergency adjustment. The binary control instruction stream is a control instruction that encodes the optimization results in binary form, facilitating quick system execution. When the decision period is 1 and the congestion level is 1, the system encodes the core variables, phase priority weights, and prediction results of the congestion diffusion model into a binary control instruction stream to achieve quick adjustment of traffic flow. If the decision period is not 1 or the congestion level is not 1, only the phase priority weight is converted into a lane configuration command to maintain the stability of traffic flow.

[0081] Preferably, the following steps can be taken to refine the dynamic lane configuration command generation. First, based on real-time traffic data and historical traffic data, the initial values of the core decision variables are determined. Then, a multi-objective solution is obtained for the core decision variables using a rolling horizon optimization algorithm to generate phase priority weights. This process requires input of real-time traffic flow data, road geometry parameters, and traffic signal control parameters, etc. By simulating the trend of traffic flow, the optimal phase priority weights are calculated. Next, the current decision period and congestion level are determined. If the decision period is 1 and the congestion level is 1, the core variables, phase priority weights, and prediction results of the congestion diffusion model are encoded into a binary control instruction stream. The encoding process includes converting the optimization results into binary form and adding necessary control instruction headers and tails to facilitate quick identification and execution by the system. If the decision period is not 1 or the congestion level is not 1, only the phase priority weights are converted into lane configuration commands to maintain the stability of traffic flow. Finally, the generated dynamic lane configuration commands are sent to the corresponding traffic control devices, such as signal light controllers, lane indicators, etc., to achieve dynamic adjustment of traffic flow.

[0082] In some embodiments, generating the final lane control instruction according to the state response data of the millimeter wave radar includes: performing compliance auditing on the lane switching action completion rate returned by the millimeter wave radar;

[0083] When the audit fails, isolating abnormal data and activating a device self-checking process;

[0084] When the audit passes, if the state response data meets a second activation condition, performing deviation analysis on the effective switching rate and the expected index to generate a control instruction containing a phase difference compensation parameter; the second activation condition is that the number of effective data packets in a single decision period exceeds a confidence threshold;

[0085] If the second activation condition is not met, generating a lane state maintenance instruction.

[0086] It should be noted that the process of generating the final lane control instruction according to the state response data of the millimeter wave radar is a key link to ensure the effective execution of traffic flow control measures. This process first involves performing compliance auditing on the lane switching action completion rate returned by the millimeter wave radar to ensure the accuracy and reliability of the data. If the audit fails, the system will isolate abnormal data and activate a device self-checking process to exclude device faults or data transmission errors. When the audit passes, the system will further determine whether the state response data meets the second activation condition, i.e., whether the number of effective data packets in a single decision period exceeds the confidence threshold. If the condition is met, the system will perform deviation analysis on the effective switching rate and the expected index to generate a control instruction containing a phase difference compensation parameter; if the condition is not met, a lane state maintenance instruction is generated to maintain the current traffic flow configuration.

[0087] Specifically, the millimeter-wave radar is a sensor installed on the road surface for real-time monitoring of the completion of lane switching actions, such as whether vehicles correctly switch lanes according to instructions. The state response data refers to the data returned by the millimeter-wave radar about the completion rate of lane switching actions, which reflects the actual implementation effect of lane control measures. Compliance auditing is the inspection of these data to ensure that they meet the pre-set standards and specifications. If the audit fails, the system will isolate abnormal data and activate the device self-checking process to determine whether there is a device failure or data transmission error. The second activation condition is a pre-set condition for determining whether the state response data is reliable enough, such as whether the number of valid data packets in a single decision cycle exceeds the confidence threshold. The phase difference compensation parameter refers to a parameter used in traffic signal control to adjust the phase difference of traffic flow in different directions to optimize traffic flow. If the state response data meets the second activation condition, the system will perform deviation analysis on the effective switching rate and the expected indicator to generate control instructions containing phase difference compensation parameters to adjust the lane configuration; if the condition is not met, lane state maintenance instructions are generated to maintain the current traffic flow configuration.

[0088] Preferably, when generating the final lane control instructions, the following steps can be used for refinement. First, the state response data returned by the millimeter-wave radar is subjected to compliance auditing, which includes checking the integrity, accuracy and timeliness of the data. If the audit fails, the system will isolate abnormal data and activate the device self-checking process to rule out device failure or data transmission error. The self-checking process includes checking the working state of the sensor, the integrity of the data transmission link, etc. If the audit passes, the system will further determine whether the state response data meets the second activation condition, i.e. whether the number of valid data packets in a single decision cycle exceeds the confidence threshold. The confidence threshold is a pre-set value for determining the reliability of the data. If the condition is met, the system will perform deviation analysis on the effective switching rate and the expected indicator, calculate the deviation value, and generate control instructions containing phase difference compensation parameters based on the deviation value. The calculation of the phase difference compensation parameter involves a comprehensive analysis of the current traffic flow, lane switching rate, signal light phase duration, etc. to determine the optimal phase difference adjustment value. Finally, the generated control instructions are sent to the corresponding traffic control devices, such as signal light controllers, lane indicators, etc. to achieve dynamic adjustment of traffic flow.

[0089] In some embodiments, the optimal elastic ratio scheme is determined through multi-cycle iteration, including: constructing a traffic efficiency evaluation matrix according to a historical decision log, the matrix storing a mapping relationship between traffic efficiency and lane enabling gradient in each cycle;

[0090] If the current decision cycle is 1, set the baseline traffic improvement rate to the industry standard value;

[0091] If the current decision cycle is not 1, the dynamic flow distribution algorithm is executed in cooperation with the adjacent section nodes to jointly generate a regional performance improvement threshold;

[0092] Based on the regional performance improvement threshold and the evaluation matrix, the elastic proportion corresponding to the extreme value of the traffic efficiency is solved by the gradient descent method, and the strategy is returned to the strategy reconfiguration point after the optimization strategy is reconstructed.

[0093] It should be noted that the process of determining the optimal elastic proportion scheme through multi-cycle iteration is the core link in the entire traffic flow control method, which is used to dynamically adjust the lane configuration to achieve the best traffic relief effect. This process first involves constructing a traffic efficiency evaluation matrix based on historical decision logs, which stores the mapping relationship between traffic efficiency and lane activation gradient in each cycle. The traffic efficiency evaluation matrix is a data structure used to evaluate the traffic efficiency under different lane configuration schemes, which predicts the traffic effect under different configurations by analyzing historical data. When the decision cycle is 1, the system sets the baseline traffic improvement rate as the industry standard value as the initial performance indicator. For non-initial cycles, the system cooperates with adjacent section nodes to execute the dynamic flow distribution algorithm to jointly generate a regional performance improvement threshold to evaluate the effect of the current configuration scheme. Based on the regional performance improvement threshold and the evaluation matrix, the system solves the elastic proportion corresponding to the extreme value of the traffic efficiency by the gradient descent method, and returns the strategy to the strategy reconfiguration point after reconstructing the optimization strategy to achieve dynamic adjustment.

[0094] Specifically, the traffic efficiency evaluation matrix is a multi-dimensional data structure that records the traffic efficiency data of each cycle under different lane configuration schemes. The lane activation gradient refers to the degree of activation of the emergency lane under different traffic flow conditions, such as different stages from not activating at all to fully activating. The baseline traffic improvement rate is a preset performance indicator used to measure the effectiveness of traffic relief measures, which is usually based on industry standards or historical data. The dynamic flow distribution algorithm is an algorithm used to optimize traffic flow distribution, which dynamically adjusts the traffic flow distribution of each section node based on current traffic conditions and historical data. The regional performance improvement threshold is a parameter used to evaluate the effect of regional traffic improvement, which compares the traffic efficiency under the current configuration scheme with historical data or standard values to determine whether the lane configuration needs to be adjusted. The gradient descent method is an optimization algorithm used to find the optimal lane configuration scheme in the multi-cycle iteration process, which adjusts the elastic proportion step by step to maximize the traffic efficiency.

[0095] Preferably, when determining the optimal elastic proportioning scheme through multi-cycle iteration, the following steps can be adopted for refinement. First, according to the data in the historical decision log, a traffic efficiency evaluation matrix is constructed. This includes collecting and organizing the traffic efficiency data of each cycle and the corresponding lane activation gradient data. Then, trend analysis is performed on the outlier data in the evaluation matrix to identify and process abnormal data. If data anomalies are found, Kalman filter algorithm is used for data repair to ensure the accuracy and reliability of the data. Next, according to whether the current decision cycle is 1, the baseline traffic improvement rate is set or the dynamic split algorithm is executed in coordination with the adjacent section node to generate the regional efficiency improvement threshold. In each iteration cycle, based on the regional efficiency improvement threshold and the evaluation matrix, the traffic efficiency extremum corresponding to the elastic proportioning is solved by gradient descent method. The implementation process of gradient descent method includes calculating the traffic efficiency gradient under the current configuration scheme, adjusting the elastic proportioning according to the gradient, and until the preset convergence condition or regional efficiency improvement threshold is reached. Finally, the optimal elastic proportioning scheme is taken as the basis for the execution of the dynamic lane control system to realize the dynamic optimization management of traffic flow.

[0096] In some embodiments, when constructing the traffic efficiency evaluation matrix, trend analysis is performed on the outlier data in the historical decision log.

[0097] When data anomalies are found through analysis, Kalman filter algorithm is used for data repair.

[0098] When the data is normal, multivariate linear fitting analysis is performed on the traffic efficiency of each cycle and the corresponding activation gradient to generate an efficiency-gradient response surface and store it in the evaluation matrix.

[0099] It should be noted that when constructing the traffic efficiency evaluation matrix, trend analysis is performed on the outlier data in the historical decision log, which is a key step to ensure data quality. This process aims to identify and process inaccurate data caused by abnormal conditions, thereby improving the reliability and effectiveness of the evaluation matrix. When data anomalies are found through analysis, Kalman filter algorithm is used for data repair, which is an effective data smoothing technique that can predict and correct abnormal data based on historical data and current observations. When the data is normal, multivariate linear fitting analysis is performed on the traffic efficiency of each cycle and the corresponding activation gradient to generate an efficiency-gradient response surface and store it in the evaluation matrix. This process not only reflects the traffic efficiency trend under different lane configurations, but also provides data support for subsequent optimization strategies.

[0100] Specifically, the historical decision log refers to all the decision data recorded by the system during its past operation, including lane configuration schemes, traffic efficiency, and other information. Outlier data refers to data points that deviate significantly from the normal range, caused by equipment failure, data transmission errors, or abnormal traffic events. Trend analysis is a data analysis method used to identify long-term trends and patterns in data by calculating the moving average or regression line of the data to determine if it is abnormal. Kalman filter algorithm is a recursive filter that estimates the state of a system by combining prediction and observation data, effectively handling noise and abnormal data. Multivariate linear fitting analysis is a statistical method used to establish the linear relationship between multiple independent variables and dependent variables, fitting the best linear model through methods such as least squares. The efficiency-gradient response surface is a three-dimensional graph representing the relationship between traffic efficiency and lane activation gradient and other parameters, which can visually display the changes in traffic efficiency under different configurations.

[0101] Preferably, when constructing the traffic performance evaluation matrix, the following steps can be used for refinement. First, collect and organize the data in the historical decision log, including the lane configuration scheme, traffic efficiency, and other information for each decision period. Then, perform trend analysis on these data, calculate the moving average or regression value of each data point to identify possible outliers. If outliers are found, use the Kalman filter algorithm to repair the data by combining historical data and current observations to predict and correct abnormal data points. In the case of normal data, use multivariate linear fitting analysis to fit the best linear model with traffic efficiency as the dependent variable and lane activation gradient and other related parameters as the independent variables. According to the fitting result, generate the efficiency-gradient response surface and store it in the evaluation matrix. This process not only ensures the data quality of the evaluation matrix, but also provides accurate data support for subsequent optimization strategies, thereby achieving dynamic optimization management of traffic flow.

[0102] In some embodiments, the optimal elastic allocation scheme is solved by gradient descent method, including: if the evaluation matrix satisfies the third activation condition, the allocation scheme corresponding to the extreme point of the surface is taken as the optimal solution candidate set; the third activation condition is that the efficiency gain rate of adjacent two iterations is lower than the gradient convergence threshold;

[0103] If the traffic efficiency corresponding to the candidate set exceeds the regional performance improvement threshold, the scheme is determined as the optimal elastic allocation scheme;

[0104] If the efficiency of the candidate set does not reach the threshold, increment the current decision period by 1, and return to the strategy reconfiguration point with the candidate set as the new optimization strategy;

[0105] If the evaluation matrix does not satisfy the third activation condition, increment the decision period by 1, and return to the strategy reconfiguration point with the regional performance improvement threshold as the strategy parameter.

[0106] It should be noted that the process of solving the optimal elastic allocation scheme by gradient descent method is a key link to realize the dynamic optimization of traffic flow. This process first judges whether the evaluation matrix meets the third activation condition, i.e. whether the efficiency gain rate of adjacent two iterations is lower than the gradient convergence threshold. If the condition is met, the allocation scheme corresponding to the extreme point of the surface is taken as the optimal solution candidate set. The extreme point of the surface here refers to the point on the efficiency-gradient response surface where the traffic efficiency reaches the maximum value, and the corresponding allocation scheme is the possible optimal scheme. If the traffic efficiency of the candidate set exceeds the regional efficiency improvement threshold, the scheme is determined as the optimal elastic allocation scheme; if the threshold is not reached, the decision period is increased and the candidate set is taken as the new optimization strategy to return to the strategy reconfiguration point. If the evaluation matrix does not meet the third activation condition, the decision period is also increased and the regional efficiency improvement threshold is taken as the strategy parameter to return to the strategy reconfiguration point, and the optimization process continues.

[0107] Specifically, the gradient descent method is an optimization algorithm used to find the optimal lane allocation scheme in multiple period iterations by gradually adjusting the elastic allocation to maximize traffic efficiency. The evaluation matrix is a data structure that stores the mapping relationship between traffic efficiency and lane activation gradient in each period, used to evaluate the effect of different allocation schemes. The third activation condition is a preset condition used to judge whether the iteration process converges, i.e. whether the efficiency gain rate of adjacent two iterations is lower than a set gradient convergence threshold. The extreme point of the surface refers to the point on the efficiency-gradient response surface where the traffic efficiency reaches the maximum value, and the corresponding allocation scheme is the possible optimal scheme. The regional efficiency improvement threshold is a parameter used to evaluate the effect of regional traffic improvement, which is determined by comparing the traffic efficiency under the current allocation scheme with historical data or standard value to determine whether the lane allocation needs to be adjusted. The decision period refers to the time required for the system to make a complete optimization decision, which determines the response speed of the system to traffic flow changes.

[0108] Preferably, when solving the optimal elastic allocation scheme by gradient descent method, the following steps can be used for refinement. First, according to the data in the evaluation matrix, the efficiency gain rate of adjacent two iterations is calculated to determine whether the third activation condition is met. If the condition is met, the extreme point on the efficiency-gradient response surface is found, and the corresponding allocation scheme is taken as the optimal solution candidate set. Next, the traffic efficiency corresponding to the candidate set is compared with the regional performance improvement threshold. If the traffic efficiency of the candidate set exceeds the regional performance improvement threshold, the scheme is determined as the optimal elastic allocation scheme; if the threshold is not reached, the decision cycle is increased, and the candidate set is taken as the new optimization strategy to return to the strategy reconfiguration point to continue the optimization process. If the evaluation matrix does not meet the third activation condition, the decision cycle is also increased, and the regional performance improvement threshold is taken as the strategy parameter to return to the strategy reconfiguration point to continue the optimization process. In the whole process, the gradient descent method updates the optimization strategy according to the efficiency gain rate of each iteration by gradually adjusting the elastic allocation until the optimal lane configuration scheme is found.

[0109] In some embodiments, further comprising:

[0110] Real-time detection of whether a regional coordination control instruction of an adjacent section is received;

[0111] If the instruction is received, the optimal allocation scheme is coupled with the regional control strategy to generate a wide-area coordinated control code stream and broadcast it to the networked joint control equipment.

[0112] It should be noted that the method involves real-time detection of whether a regional coordination control instruction of an adjacent section is received, and when the instruction is received, the optimal allocation scheme is coupled with the regional control strategy to generate a wide-area coordinated control code stream and broadcast it to the networked joint control equipment. The regional coordination control instruction here refers to the control instruction from the adjacent section, which is used to coordinate the traffic flow between different sections to achieve the optimization management of regional traffic. The coupling operation refers to the combination of the optimal allocation scheme and the regional control strategy, through calculation and adjustment to generate a coordinated control strategy applicable to the entire region. The wide-area coordinated control code stream is an instruction stream for controlling the networked joint control equipment, which contains the optimized traffic control strategy and can realize the coordinated control of multiple sections.

[0113] Specifically, the regional coordination control instruction is an instruction for coordinating traffic flow on different road segments, which contains traffic condition information and control requirements of adjacent road segments. The coupling operation is a process of combining different control strategies, which fuses the optimal matching scheme with the regional control strategy through mathematical models and algorithms to generate a coordinated control strategy applicable to the entire region. The wide-area coordinated control code stream is a control instruction stream that contains the optimized traffic control strategy and can achieve coordinated control of multiple road segments. The networked control equipment refers to traffic control equipment such as signal light controllers and lane indicators that can receive and execute the wide-area coordinated control code stream. These devices are connected through a network and can achieve coordinated control of regional traffic.

[0114] Preferably, when implementing the method, the following steps can be used for refinement. First, the system detects in real time whether to receive the regional coordination control instruction of the adjacent road segment. If the instruction is received, the system will analyze the instruction content to obtain the traffic condition information and control requirements of the adjacent road segment. Then, the optimal matching scheme is coupled with the regional control strategy. This step includes establishing a mathematical model, inputting the optimal matching scheme and regional control strategy as parameters, and generating a coordinated control strategy applicable to the entire region through calculation and adjustment. Finally, the generated wide-area coordinated control code stream is broadcast to the networked control equipment, which will adjust the traffic signals and lane configuration according to the instruction to achieve coordinated control of regional traffic. Throughout the process, the system needs to ensure the real-time and accuracy of the instruction to achieve effective management of regional traffic.

[0115] The above-mentioned various embodiments of the present application have the following beneficial effects:

[0116] 1. The opening time and gradient of the emergency lane can be dynamically decided based on real-time traffic situation, and the control strategy is continuously optimized through a closed-loop feedback mechanism of multiple observation windows, so as to accurately match the fluctuation demand of traffic flow, effectively relieve road congestion during peak hours, and improve the road network capacity.

[0117] 2. The lane resource allocation scheme can be generated by combining multiple traffic parameters such as section occupancy rate and queue length, using a rolling time domain optimization algorithm, and determining the optimal elastic matching through multi-cycle iterative calculation, to realize fine management and control of lane resources and avoid resource mismatching problems caused by traditional fixed control mode.

[0118] 3. The regional traffic flow optimization can be achieved through the coordinated control mechanism with adjacent road segments, the congestion propagation trend can be predicted using a short-term traffic situation deduction model, and the reliable execution of control instructions can be ensured through the device state monitoring and data verification system, thereby improving the coordination and stability of the entire road network system.

[0119] In terms of intelligent expansion of expressways, the following can be done:

[0120] I. Intelligent Upgrading of Highway Infrastructure, including:

[0121] Pavement Sensing System: Advanced sensor technology is used to monitor road conditions in real time, such as traffic volume, speed, and accident warnings.

[0122] License Plate Recognition System: Install license plate recognition equipment at highway entrances and exits to achieve automatic vehicle identification, classification, and statistics.

[0123] Highway Lighting System: Intelligent lighting technology is used to automatically adjust the brightness of streetlights based on traffic volume and weather conditions, improving road lighting effects.

[0124] Highway Signs and Induction System: Use intelligent display screens to display real-time traffic information, route guidance, and accident warnings.

[0125] II. Intelligent Traffic Management System Construction, including:

[0126] Traffic Signal Control System: Dynamically adjust traffic signal timing based on real-time traffic volume to optimize traffic flow.

[0127] Electronic Police System: Use video monitoring and intelligent analysis technology to automatically identify and punish illegal behavior.

[0128] Intelligent Dispatching System: Through data analysis, achieve reasonable scheduling of highway operation resources to improve operational efficiency.

[0129] Intelligent Monitoring System: Real-time monitoring of key highway locations ensures smooth and safe traffic.

[0130] III. Intelligent Traffic Information Services and Decision Support, including:

[0131] Real-time Traffic Information Services: Provide real-time traffic information and route planning services to users through mobile apps, in-car navigation systems, and other channels.

[0132] Intelligent Travel Suggestions: Provide personalized travel suggestions based on user travel needs, such as optimal travel times and routes.

[0133] Traffic Big Data Analysis: Collect and analyze highway operation data to provide scientific basis for decision-makers.

[0134] Prediction and Early Warning System: Based on big data analysis, predict future highway operation conditions and issue early warning information in advance.

[0135] IV. Intelligent Traffic Operation and Management, including:

[0136] Resource management system: unified management of highway infrastructure, equipment and personnel resources to improve operation and maintenance efficiency.

[0137] Fault monitoring and prediction: real-time monitoring of infrastructure operation through intelligent monitoring devices to detect and handle faults in advance.

[0138] Intelligent maintenance decision: based on data analysis, develop scientific maintenance plan to improve road service life.

[0139] Personnel training and assessment: use virtual reality, online training and other means to improve the skills of operation and maintenance personnel to ensure the safe operation of the highway.

[0140] Further, the storage medium of the embodiments of the present application stores program instructions capable of implementing all the methods described above, wherein the program instructions can be stored in the storage medium in the form of a software product, including a number of instructions to make a computer device (which can be a personal computer, server, or network device, etc.) or processor execute all or part of the steps of the methods described in various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk, and various media that can store program codes, or terminal devices such as computers, servers, mobile phones, tablets, etc.

[0141] The above description is only some of the preferred embodiments of the present application and an explanation of the technical principles applied. Those skilled in the art should understand that the scope of the application involved in the embodiments of the present application is not limited to the technical solutions formed by the specific combinations of the above technical features, and should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the above features and technical features with similar functions disclosed in the embodiments of the present application (but not limited to) are replaced with each other to form a technical solution.

Claims

1. A method for opening an emergency lane in a section under heavy traffic, characterized by, The application relates to a flexible lane control core applied to a road network management and control platform. After detecting that a target section traffic flow density reaches a first saturation threshold, a flexible lane dynamic decision engine is started, real-time traffic situation data is matched with a flexible starting criterion, an initial management and control decision is generated, and a road side perception unit is broadcasted, and a first observation window is triggered; After the first observation window is closed, a section occupancy rate index reported by the road side perception unit is used to generate a lane resource configuration scheme, a state update instruction is issued to a variable information sign, a second observation window is triggered, and the resource configuration scheme comprises a flexible lane starting gradient; After the second observation window is closed, a lane optimization strategy is reconstructed based on a real-time congestion coefficient, section queue length data collected by a millimeter wave radar is verified for integrity, and if the verification is passed, the queue length is input into a congestion evaluation buffer area; If the congestion evaluation buffer area meets a first activation condition, a minimum activation interval of the flexible lane is calculated through a congestion diffusion model, the minimum activation interval is used as a core decision variable of the optimization strategy, and a congestion level is marked as 1; the first activation condition is that a queue length growth rate in the continuous three observation windows exceeds a dynamic threshold; If the first activation condition is not met, an optimization strategy variable is reset to a benchmark value, and the congestion level is marked as 0; A current decision period is set as 1; a strategy reconfiguration point is entered; A rolling horizon optimization algorithm is used to solve the core decision variable in multiple targets, and a phase priority weight is generated; If the decision period is 1 and the congestion level is 1, a core variable, a phase priority weight and a prediction result of the congestion diffusion model are coded into a binary control instruction stream; If the decision period is not 1 or the congestion level is not 1, only the phase priority weight is converted into a lane configuration command; According to the optimization strategy and the decision period, a dynamic lane configuration command is generated and broadcasted to the millimeter wave radar, and a third observation window is triggered; After the third observation window is closed, a final lane control instruction is generated according to state response data of the millimeter wave radar, and is broadcasted to a signal control machine, and a fourth observation window is triggered; According to the state response data of the millimeter wave radar, the final lane control instruction is generated, which comprises: compliance auditing is performed on a lane switching action completion rate returned by the millimeter wave radar; When the auditing is not passed, abnormal data is isolated and a device self-checking process is activated; When the auditing is passed, if the state response data meets a second activation condition, deviation analysis is performed on an effective switching rate and an expected index, and a control instruction containing a phase difference compensation parameter is generated; the second activation condition is that the number of effective data packets in a single decision period exceeds a confidence threshold; If the second activation condition is not met, a lane state maintaining instruction is generated; After the fourth observation window is closed, an optimal flexible matching scheme is determined through multi-period iteration based on integrated vehicle traffic efficiency indexes, and the optimal flexible matching scheme is used as an execution basis of a dynamic lane control system after the optimization strategy is reconstructed in the iteration process and the strategy reconfiguration point is returned. Wherein, the optimal elastic allocation scheme is determined through multi-period iteration, including: constructing a traffic efficiency evaluation matrix according to historical decision logs, the matrix storing the mapping relationship between the traffic efficiency of each period and the lane activation gradient; If the current decision period is 1, set the reference traffic improvement rate as the industry standard value; If the current decision period is not 1, perform a dynamic distribution algorithm in cooperation with adjacent section nodes to jointly generate a regional efficiency improvement threshold; Based on the efficiency threshold and the evaluation matrix, the optimal elastic allocation scheme is solved by gradient descent method, and the optimized strategy is returned to the strategy reconfiguration point; The traffic efficiency evaluation matrix refers to a data structure for evaluating the traffic efficiency under different lane configuration schemes, which predicts the traffic effect under different configurations by analyzing historical data; The regional efficiency improvement threshold refers to a parameter for evaluating the improvement effect of regional traffic, which determines whether the lane configuration needs to be adjusted by comparing the traffic efficiency under the current configuration scheme with historical data or standard values.

2. The method of claim 1, wherein, According to the section occupancy rate index reported by the roadside sensing unit, a lane resource allocation scheme is generated, including: verifying the effectiveness of the original data transmitted by the roadside sensing unit, if the verification is passed, calculating the difference between the section occupancy rate and the output value of the dynamic baseline model to generate an elastic expansion coefficient; When the elastic expansion coefficient breaks through the second saturation threshold, the elastic lane activation gradient is mapped to a segmented function of the expansion coefficient, the channelization parameters corresponding to the gradient are calculated by using the lane alignment optimization algorithm to generate a resource allocation scheme containing the activation gradient and the channelization parameters; When the elastic expansion coefficient does not break through the second saturation threshold, set the elastic lane activation gradient to zero and generate a lane state maintenance instruction; The dynamic baseline model refers to a model established according to historical traffic data, which is used to predict the section occupancy rate under normal traffic conditions; The channelization parameter refers to a parameter for optimizing lane layout and traffic flow line.

3. The method of claim 1, wherein, Also includes: Run the short-term traffic situation deduction model in cooperation with the adjacent section control node to jointly generate regional linkage control parameters as boundary constraint conditions for dynamic decision.

4. The method of claim 1, wherein, When constructing the traffic efficiency evaluation matrix, including: trend analysis of the outlying data in the historical decision logs; When data anomalies are found through analysis, use Kalman filtering algorithm for data repair; When the data is normal, perform multiple linear fitting analysis on the traffic efficiency of each period and the corresponding activation gradient to generate an efficiency-gradient response surface and store it in the evaluation matrix.

5. The method of claim 1, wherein, Solving the optimal elastic allocation scheme by gradient descent method, including: if the evaluation matrix meets the third activation condition, the allocation scheme corresponding to the extreme point of the surface is taken as the optimal solution candidate set; the third activation condition is that the efficiency gain rate of adjacent two iterations is lower than the gradient convergence threshold; If the traffic efficiency corresponding to the candidate set exceeds the regional efficiency improvement threshold, determine that the scheme is the optimal elastic allocation scheme; If the efficiency of the candidate set does not reach the threshold, increase the current decision period by 1, and return to the strategy reconfiguration point with the candidate set as the new optimization strategy; If the evaluation matrix does not meet the third activation condition, increase the decision period by 1, and return to the strategy reconfiguration point with the regional efficiency improvement threshold as the strategy parameter.

6. The method of claim 5, wherein, Also includes: Real-time detection of whether the regional coordinated control instruction of the adjacent road section is received; If the instruction is received, the optimal matching scheme is coupled with the regional control strategy for operation, a wide-area coordinated control code stream is generated and broadcast to the networked joint control equipment.

Images