Highway construction safety early warning method and system

Abstract

The invention discloses an expressway construction safety early warning method and system, which is operated on a central server, and dynamically determines a safety threshold rule set by acquiring meteorological data and static characteristics of a construction road section in real time; collecting traffic flow data in real time by using an upstream sensing unit array, querying a corresponding threshold value for each target vehicle according to the type of the target vehicle, and generating a third-level risk identifier by synthesizing the speed, the vehicle following distance and a lane changing intention based on depth analysis of course angle change trend, lateral vehicle distance and the like; generating a composite risk level through predefined fusion logic; for medium-risk and high-risk vehicles, an upstream roadside early warning unit is dynamically selected according to the positions, the vehicle types and the risk levels of the vehicles, and differentiated early warning instructions are generated and issued. According to the method, self-adaption and fusion analysis of meteorological conditions, road section characteristics, vehicle types and multi-dimensional driving behaviors are realized, and the accuracy, timeliness and effectiveness of safety early warning of the construction area are remarkably improved.

Description

A method and system for early warning of safety during highway construction Technical Field

[0001] This invention belongs to the field of big data analysis technology, specifically relating to a method and system for early warning of highway construction safety. Background Technology

[0002] With the continuous expansion and improvement of my country's expressway network, road maintenance, reconstruction, and other construction activities are becoming increasingly frequent, posing a severe challenge to traffic safety management in construction zones. Traditional construction safety early warning mainly relies on fixed traffic signs, markers, and cones for area isolation and alerts. This static and passive early warning method is difficult to adapt to the complex and ever-changing dynamic traffic environment of expressways. Especially under adverse weather conditions such as rain, fog, and snow with reduced visibility, or when facing different types of vehicles with significantly different reaction characteristics, fixed warning information and thresholds often lack specificity and cannot provide timely, effective, and differentiated risk warnings to vehicles about to enter the construction impact zone, making construction zones still high-incidence areas for traffic accidents.

[0003] Some existing technical solutions attempt to use sensing devices for safety monitoring and early warning in construction areas. However, these solutions have significant limitations: First, most early warning logic is relatively simple, usually based on a single vehicle speed or distance threshold, failing to fully consider the combined effects of different weather conditions on road surface adhesion coefficient, braking distance, and driver visibility, resulting in rigid threshold settings and a tendency to generate false alarms or missed alarms when the weather changes. Second, there is a general lack of fine differentiation and response for different types of highway vehicles. Different vehicles have fundamental differences in mass, inertia, braking performance, blind spots, and driver behavior patterns, and a uniform safety standard cannot meet the actual risk management needs. Third, the early warning strategies and risk assessment dimensions are too simplistic, failing to integrate and classify various risk behaviors such as speeding, following too closely, and abnormal lane-changing intentions. The early warning information generation and release mechanisms are crude, making it difficult to achieve accurate risk profiling and graded intervention, resulting in limited early warning effectiveness.

[0004] In summary, existing technologies struggle to provide a proactive construction safety protection method that can deeply integrate real-time weather conditions, inherent road characteristics, diverse vehicle types, and multi-dimensional driving behavior data to conduct dynamic, intelligent, and differentiated risk assessment and early warning. Therefore, a new early warning method is urgently needed to address the shortcomings of existing technologies, such as poor adaptability to environmental and objective differences, limited risk assessment dimensions, and insufficiently precise early warning strategies. This would improve the proactive safety protection level of highway construction areas and reduce accident risks. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for early warning of safety during highway construction, which can effectively solve the problems mentioned in the background art, such as poor accuracy, weak adaptability, and inaccurate risk intervention in complex traffic and environmental conditions due to the lack of comprehensive perception and dynamic response capabilities of existing early warning technologies for real-time weather, vehicle type, and multi-dimensional driving behavior.

[0006] According to a first aspect of the present invention, the present invention claims protection for a method for early warning of safety during highway construction, which operates on a central server of a highway management system and is communicatively connected to a sensing unit array deployed upstream of the construction section and multiple roadside early warning units, comprising: S1, acquiring current meteorological data and a construction section identifier, and retrieving corresponding static road section characteristic parameters according to the construction section identifier; S2, determining the current meteorological impact level based on the current meteorological data, and selecting a corresponding target basic safety threshold rule set from multiple preset basic safety threshold rule sets based on the current meteorological impact level and the static road section characteristic parameters; S3, receiving real-time streaming traffic flow data from the sensing unit array; S4, for each target vehicle, querying and obtaining the corresponding basic speed threshold, basic safe following distance threshold, and basic lane change risk coefficient threshold from the target basic safety threshold rule set according to the vehicle type; S5, respectively... Based on the comparison of the target vehicle's real-time speed with the corresponding basic speed threshold, the comparison of the real-time distance to the vehicle ahead with the corresponding basic safe following distance threshold, the real-time heading angle change trend, real-time speed, and real-time distance to vehicles in adjacent lanes, and combined with the corresponding basic lane change risk coefficient threshold, a first-level speed risk indicator, a second-level following distance risk indicator, and a third-level lane change intention risk indicator are generated; S6, a composite risk level for the target vehicle is generated by combining predefined risk indicator fusion logic; S7, for the target vehicle determined to be high-risk or medium-risk, the target roadside early warning unit to be activated is determined based on its real-time location coordinates and the construction section marker, and an instruction data packet containing the composite risk level, the vehicle type of the target vehicle, and a suggested early warning strategy is generated; S8, the instruction data packet is sent to the target roadside early warning unit to drive the execution of the corresponding early warning information release action.

[0007] Furthermore, S2 also includes: S21, parsing the current meteorological data and extracting the current precipitation intensity value, current visibility value, and current wind speed value; S22, preset a first precipitation intensity threshold, a second precipitation intensity threshold, a first visibility threshold, a second visibility threshold, a first wind speed threshold, and a second wind speed threshold; S23, determining the current meteorological impact level based on a comparison between the current precipitation intensity value, current visibility value, and current wind speed value and the first precipitation intensity threshold, the second precipitation intensity threshold, the first visibility threshold, the second visibility threshold, the first wind speed threshold, and the second wind speed threshold; the current meteorological impact level includes a first level, a second level, and a third level, where the first level corresponds to normal weather conditions, the second level corresponds to general severe weather conditions, and the third level corresponds to severe severe weather conditions.

[0008] Furthermore, S5 also includes: S51, calculating the heading angle change rate sequence of the target vehicle within the current time window based on the real-time position coordinates and real-time heading angle of the target vehicle uploaded by the sensing unit array at multiple consecutive timestamps; S52, performing statistical analysis on the heading angle change rate sequence to extract its mean, variance, and the proportion of time that continuously exceeds a preset small angle threshold; S53, obtaining the real-time speed of the target vehicle and the real-time distance of vehicles in adjacent lanes, wherein the real-time distance of vehicles in adjacent lanes includes the distance of the nearest vehicle in the adjacent lane parallel to the target vehicle, and the distance of the nearest vehicle in the adjacent lane within a predetermined range behind the target vehicle; S54, based on the vehicle type of the target vehicle, from the target basic safety threshold... Then, the corresponding basic lane change risk coefficient thresholds are centrally obtained. The basic lane change risk coefficient thresholds are multi-dimensional judgment benchmark vectors, including the mean threshold of the heading angle change rate, the variance threshold, the continuous deviation duration ratio threshold, the parallel vehicle distance threshold, and the following vehicle distance threshold; S55, the mean, variance, and continuous deviation duration ratio extracted in S52, and the parallel vehicle distance and following vehicle distance obtained in S53 are compared with the corresponding thresholds in the multi-dimensional judgment benchmark vector; S56, the third-level lane change intention risk identifier is generated according to the predefined comparison logic rules; S57, the different combinations of the four sub-judgment results of directional lane change trend, continuous lateral displacement intention, side collision risk, and rear-end collision risk are mapped to the preset lane change risk identifier enumeration value.

[0009] Furthermore, S6 also includes: S61, setting a preset risk indicator fusion decision matrix, using the first-level speed risk indicator, the second-level following distance risk indicator, and the third-level lane change intention risk indicator as input dimensions, and the composite risk level as output; S62, discretizing and encoding the first-level speed risk indicator, the second-level following distance risk indicator, and the third-level lane change intention risk indicator to form a risk feature vector; S63, querying the risk indicator fusion decision matrix, determining the output row matching the risk feature vector, and defining the composite risk level; wherein, the construction rules of the risk indicator fusion decision matrix include: if the third-level lane change intention risk indicator is an immediate high risk, then regardless of the values ​​of the first-level speed risk indicator and the second-level following distance risk indicator, the composite risk level is determined to be high risk; if the third-level lane change intention risk indicator is a highly potential risk, and the first-level speed risk indicator is speeding or the... If the second-level following distance risk indicator is "following too closely," the composite risk level is determined to be high risk. If the third-level lane change intention risk indicator is "high potential risk," and both the first-level speed risk indicator and the second-level following distance risk indicator are normal, the composite risk level is determined to be medium risk. If the third-level lane change intention risk indicator is "low potential risk" or "no risk," and both the first-level speed risk indicator (speeding) and the second-level following distance risk indicator (following too closely) occur simultaneously, the composite risk level is determined to be high risk. If the third-level lane change intention risk indicator is "low potential risk" or "no risk," and only one of the first-level speed risk indicator (speeding) or the second-level following distance risk indicator (following too closely) occurs, the composite risk level is determined to be medium risk. If all three levels—first-level speed risk indicator, second-level following distance risk indicator, and third-level lane change intention risk indicator—are normal or no risk, the composite risk level is determined to be low risk.

[0010] Further, S7 also includes: S71. Obtaining the preset location information of all roadside early warning units upstream of the construction section based on the construction section identifier; S72. Calculating the first distance between the real-time location coordinates of the target vehicle and the starting point of the construction area, and calculating the second distance between the real-time location coordinates of the target vehicle and the preset locations of each roadside early warning unit; S73. Searching for the corresponding early warning strategy template from a preset early warning strategy mapping table based on the vehicle type of the target vehicle and the composite risk level; S74. Determining the early warning lead time based on the first distance, the composite risk level, and the early warning priority; S75. Selecting a roadside early warning unit that meets the following conditions as the target roadside early warning unit: its preset location is located upstream of the direction of travel of the target vehicle, and the second distance is closest to the early warning lead time; S76. Replacing the placeholder in the early warning display content template with the vehicle type of the target vehicle and the composite risk level to generate specific early warning display content; S77. Encapsulating the identifier of the target roadside early warning unit, the early warning display content, the early warning display intensity level, and the trigger command to form the command data packet.

[0011] Furthermore, the construction logic of the warning strategy mapping table described in S73 is as follows: For target vehicles with a composite risk level of high risk, if the vehicle type is a large freight vehicle, the warning display content template in the corresponding warning strategy template is: [Large Truck] Please note, construction ahead, your current state is extremely risky, please slow down immediately, maintain the maximum safe distance, and avoid changing lanes! The warning display intensity level is the highest, and the warning priority is the highest. For target vehicles with a composite risk level of high risk, if the vehicle type is a small passenger vehicle, the corresponding warning display content template is: [Small Passenger Vehicle] Warning, construction ahead, high-risk driving behavior, please slow down immediately and increase the following distance! The warning display intensity level is the highest, and the warning priority is high. For target vehicles with a composite risk level of medium risk, if the vehicle type is a large freight vehicle, the corresponding warning display content template is: [Large Truck] Reminder, construction ahead, you are currently at risk, it is recommended to slow down in advance and increase the following distance. The warning display intensity level is medium, and the warning priority is high. For target vehicles with a composite risk level of medium risk, if the vehicle type is a small passenger vehicle, the corresponding warning display template is: [Small Passenger Vehicle] Attention, construction ahead, please maintain a safe speed and distance. The warning display intensity level is medium, and the warning priority is medium. For target vehicles with a composite risk level of low risk, a general construction warning message is used, with a warning display intensity level of low and a warning priority of low.

[0012] Furthermore, the method also includes a monitoring and feedback step for the status of the roadside warning unit: S91, after sending the instruction data packet to the target roadside warning unit, start a feedback timer; S92, before the feedback timer reaches a preset timeout threshold, if an execution confirmation signal is received from the target roadside warning unit, the warning instruction is marked as successfully executed; S93, if the execution confirmation signal is not received after the feedback timer reaches the preset timeout threshold, the target roadside warning unit is marked as having a communication abnormality; S94, for the target roadside warning unit marked as having a communication abnormality, a backup roadside warning unit is selected as the new target roadside warning unit based on the real-time location coordinates and warning lead of the target vehicle, and S77 and S8 are re-executed; S95, the communication abnormality information is recorded in the equipment operation and maintenance log, and a corresponding equipment maintenance prompt is generated.

[0013] Furthermore, the method also includes a post-warning effect evaluation step: S101, for each target vehicle that triggers a warning, after the warning command is issued, continuously receive subsequent streaming traffic flow data of the target vehicle from the sensing unit array for a preset observation window; S102, analyze the real-time speed of the target vehicle, the real-time distance to the vehicle in front, and the trend of heading angle change within the observation window; S103, based on the analysis results, determine whether the target vehicle exhibits driving behavior correction that meets the warning expectation within the observation window, the driving behavior correction including: speed decreasing to below the basic speed threshold, following distance increasing to above the basic safe following distance threshold, or termination of abnormal heading angle change trend; S104, based on whether the driving behavior correction occurs and the timeliness of the correction, rate the effect of the warning event, and record the rating result in the warning history database; S105, periodically statistically analyze the data in the warning history database to generate warning response effectiveness reports under different meteorological impact levels, different vehicle types, and different composite risk levels.

[0014] Furthermore, the sensing unit array includes multiple fusion sensing devices consisting of millimeter-wave radar and video cameras deployed along the road. The vehicle type identification in S3 is achieved locally by the fusion sensing device in the following ways: S31, extracting the outline size, point cloud density distribution features, and motion echo features of the target vehicle from millimeter-wave radar point cloud data; S32, extracting the visual outline, vehicle structure features, and license plate information of the target vehicle from video image data; S33, fusing the outline size, point cloud density distribution features, motion echo features, visual outline, and vehicle structure features, and matching them with a pre-stored vehicle type feature template library; S34, classifying the target vehicle into multiple predefined vehicle types based on the matching results, wherein the multiple vehicle types at least distinguish between large freight vehicles, medium-sized passenger and freight vehicles, small passenger vehicles, and special vehicles.

[0015] According to a second aspect of the present invention, the present invention claims protection for a highway construction safety early warning system, comprising: one or more processors; and a memory having stored one or more programs thereon, which, when executed by the one or more processors, cause the one or more processors to implement a highway construction safety early warning method according to any one of claims 1 to 9.

[0016] Compared with existing technologies, this invention has the following advantages: By establishing a dynamic mapping relationship between meteorological impact levels and a set of basic safety threshold rules, this invention achieves adaptive adjustment of road risks under complex meteorological conditions, solving the problem of traditional fixed-threshold warnings easily failing during sudden weather changes; by introducing a vehicle type identification module and combining it with the inherent risk coefficient of road segments to differentiate the baseline safety parameters of various vehicle types, the invention significantly improves the pertinence and rationality of the warning strategy; by constructing a three-level risk identification system covering speed, following distance, and lane-changing intent, and using a fusion logic based on a decision matrix to generate composite risk levels, it achieves a refined quantitative assessment of multi-dimensional dangerous driving behaviors; by setting a dynamic calculation model for warning lead time and optimizing the timing and location of warning releases by combining vehicle type and risk level, it improves the timeliness and spatial matching accuracy of warning information; by establishing a status feedback and fault rerouting mechanism for roadside warning units, it ensures the operational reliability and fault tolerance of the warning system; and by implementing a post-warning effect evaluation process, it forms a closed-loop management mechanism of monitoring-warning-feedback-optimization, providing data support for continuous improvement of the warning algorithm. Overall, this invention significantly improves the level of intelligent safety early warning in highway construction areas, effectively reduces the incidence of traffic accidents in construction areas, and enhances the proactive prevention and control capabilities of traffic management systems. Attached Figure Description

[0017] Figure 1 is a flowchart of a highway construction safety early warning method claimed in an embodiment of the present invention; Figure 2 is a second flowchart of a highway construction safety early warning method claimed in an embodiment of the present invention; Figure 3 is a third flowchart of a highway construction safety early warning method claimed in an embodiment of the present invention; Figure 4 is a fourth flowchart of a highway construction safety early warning method claimed in an embodiment of the present invention; Figure 5 is a fifth flowchart of a highway construction safety early warning method claimed in an embodiment of the present invention; Figure 6 is a sixth flowchart of a highway construction safety early warning method claimed in an embodiment of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0019] The terms "first," "second," and "third" in this application are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first," "second," or "third" may explicitly or implicitly include at least one of those features. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified. All directional indications in the embodiments of this application, such as up, down, left, right, front, back, etc., are only used to explain the relative positional relationships and movements between components in a specific orientation as shown in the accompanying drawings. If the specific orientation changes, the directional indications will change accordingly. Furthermore, the terms "including" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0020] References to embodiments herein mean that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0021] According to a first embodiment of the present invention, the present invention claims protection for a method for early warning of highway construction safety, which runs on the central server of a highway management system and is communicatively connected to a sensing unit array deployed upstream of the construction section and multiple roadside early warning units. Referring to Figure 1, the method includes: S1, acquiring current meteorological data and construction section identification, and retrieving corresponding static road section characteristic parameters according to the construction section identification; S2, determining the current meteorological impact level based on the current meteorological data, and selecting a corresponding target basic safety threshold rule set from multiple preset basic safety threshold rule sets based on the current meteorological impact level and static road section characteristic parameters; S3, receiving real-time streaming traffic flow data from the sensing unit array; S4, for each target vehicle, querying the target basic safety threshold rule set to obtain the corresponding basic speed threshold, basic safe following distance threshold, and basic lane change risk coefficient threshold according to the vehicle type. S5. Based on the target vehicle's real-time speed and the corresponding basic speed threshold, the real-time distance to the vehicle ahead and the corresponding basic safe following distance threshold, the real-time heading angle change trend, real-time speed, and real-time distance to vehicles in adjacent lanes, and combined with the corresponding basic lane change risk coefficient threshold, calculate to generate a first-level speed risk label, a second-level following distance risk label, and a third-level lane change intention risk label; S6. Combine the predefined risk label fusion logic to generate a composite risk level for the target vehicle; S7. For target vehicles judged to be high-risk or medium-risk, based on their real-time location coordinates and construction section signs, determine the target roadside warning unit to be activated, and generate an instruction data packet containing the composite risk level, the target vehicle's vehicle type, and a suggested warning strategy; S8. Send the instruction data packet to the target roadside warning unit to drive the execution of the corresponding warning information release action.

[0022] In this embodiment, the early warning system of this invention is deployed upstream of a closed construction section of an eight-lane, two-way highway. The core of the system is a central server of the highway management center. Multiple sensing units are installed at specific intervals within a range of several kilometers upstream of the construction area. Each unit integrates millimeter-wave radar and a high-definition camera, forming a sensing unit array. Simultaneously, roadside early warning units equipped with variable message signs and audible and visual alarms are deployed at key locations along the route, such as above the lanes and on the roadside. All devices communicate with the central server in real time via a fiber optic network.

[0023] At the start of implementation, the central server first initializes the data. Administrators input a unique identifier for the construction section via a human-machine interface, such as GXX Expressway K100+200 to K100+500 northbound. Based on this identifier, the server retrieves the static road segment characteristic parameters from the expressway basic information database, including: the design maximum speed limit for the section, the average daily traffic volume based on historical statistics, the total number of lanes, the radius of curvature characteristic value representing the sharpness of curves, and the slope characteristic value reflecting the gradient. These parameters will serve as the static basis for subsequent risk calculations.

[0024] Further, referring to Figure 2, S2 also includes: S21, parsing the current meteorological data and extracting the current precipitation intensity value, current visibility value, and current wind speed value; S22, preset a first precipitation intensity threshold, a second precipitation intensity threshold, a first visibility threshold, a second visibility threshold, a first wind speed threshold, and a second wind speed threshold; S23, based on the comparison between the current precipitation intensity value, current visibility value, and current wind speed value and the first precipitation intensity threshold, the second precipitation intensity threshold, the first visibility threshold, the second visibility threshold, the first wind speed threshold, and the second wind speed threshold, determining the current meteorological impact level; the current meteorological impact level includes a first level, a second level, and a third level, where the first level corresponds to normal weather conditions, the second level corresponds to general severe weather conditions, and the third level corresponds to severe severe weather conditions.

[0025] In this embodiment, the central server obtains current meteorological data from an authoritative meteorological service interface at fixed intervals; in a simulation test, the system obtains a set of data, including real-time observations of precipitation intensity, visibility, and wind speed.

[0026] The meteorological analysis module within the server parses this data. The module has multiple preset thresholds, such as a precipitation intensity threshold to distinguish between light rain and moderate rain, a visibility threshold to distinguish between light fog and dense fog, and a wind speed threshold to distinguish between light wind and strong wind.

[0027] The judgment logic is as follows: First, the system checks whether the current precipitation intensity exceeds the first light rain threshold, and simultaneously checks whether the visibility is below the first light fog threshold, and whether the wind speed exceeds the first light wind threshold. If all current observations are better than these first thresholds (i.e., precipitation less than or equal to, visibility greater than or equal to, and wind speed less than or equal to), the system determines the current meteorological impact level to be Level 1, representing good weather conditions. Next, the system checks whether the observations are between the first threshold and the more stringent second threshold.

[0028] For example, if the precipitation intensity is between the light rain and moderate rain thresholds, or the visibility is between the light fog and dense fog thresholds, or the wind speed is between light wind and strong wind, meeting any one of these conditions will result in a Level 2 assessment, representing general severe weather. Finally, if any observed value exceeds its corresponding strictest Level 2 threshold, such as reaching the levels of heavy rain, dense fog, or strong winds, then regardless of other indicators, it will be directly assessed as Level 3, representing severe severe weather. In one test of this embodiment, the system detected a continuous decrease in visibility, eventually falling below the dense fog threshold; therefore, the meteorological impact level was dynamically adjusted from Level 2 to Level 3.

[0029] Furthermore, S5 also includes: S51, calculating the heading angle change rate sequence of the target vehicle within the current time window based on the real-time position coordinates and real-time heading angle of the target vehicle uploaded by the sensing unit array at multiple consecutive timestamps; S52, performing statistical analysis on the heading angle change rate sequence to extract its mean, variance, and the proportion of time that continuously exceeds a preset small angle threshold; S53, obtaining the real-time speed of the target vehicle and the real-time distance of vehicles in adjacent lanes, including the distance of the nearest vehicle in the adjacent lane parallel to the target vehicle, and the distance of the nearest vehicle in the adjacent lane within a predetermined range behind the target vehicle; S54, obtaining the target basic safety threshold rule set based on the vehicle type of the target vehicle. Take the corresponding basic lane change risk coefficient threshold. The basic lane change risk coefficient threshold is a multi-dimensional judgment benchmark vector, which includes the mean threshold of the heading angle change rate, the variance threshold, the continuous deviation duration percentage threshold, the parallel distance threshold, and the following vehicle distance threshold; S55, compare the mean, variance, and continuous deviation duration percentage extracted in S52, and the parallel distance and following vehicle distance obtained in S53 with the corresponding thresholds in the multi-dimensional judgment benchmark vector; S56, generate a third-level lane change intention risk identifier according to the predefined comparison logic rules; S57, map the different combinations of the four sub-judgment results of directional lane change trend, continuous lateral displacement intention, side collision risk, and rear-end collision risk to the preset lane change risk identifier enumeration value.

[0030] In this embodiment, the server pre-stores baseline safety parameters for different vehicle types, such as passenger cars and large trucks, under various weather conditions. However, the key to this invention is that these baseline values ​​are not used directly, but are dynamically adjusted according to the characteristics of specific road sections to form the final set of basic safety threshold rules for application.

[0031] First, the server calculates an inherent risk coefficient for the current construction section based on its static parameters. This coefficient integrates two core factors: the design speed and the historical average traffic volume. The calculation process is not a simple addition; instead, traffic volume and design speed are first normalized to a standard range of zero to one using mathematical methods, yielding a traffic flow influence factor and a speed influence factor. Then, based on safety management strategies, different weights are assigned to these two factors. For example, assuming that the traffic flow factor has a slightly greater impact on the basic risk than the design speed, a weighted sum is performed to finally obtain an inherent risk coefficient for the section between zero and one. The higher this coefficient, the higher the inherent basic risk level of the section under the same conditions.

[0032] Subsequently, the system makes differentiated adjustments to the baseline parameters based on the currently determined meteorological impact level. For the first level of good weather, the baseline speed threshold, baseline safe following distance, and baseline lane change risk coefficient for each type of vehicle are multiplied by a first adjustment coefficient that is positively correlated with the inherent risk coefficient of the road segment. This means that even in good weather, on road segments with inherently high design speeds and heavy traffic, safety standards will automatically tighten, with a slight decrease in speed thresholds and a slight increase in following distance requirements.

[0033] For Level 2 general severe weather, the adjustment strategy is more targeted: the speed threshold is multiplied by a coefficient smaller than the first adjustment coefficient, which further reduces the speed limit; while the safe following distance and lane change risk coefficient are multiplied by a larger coefficient, which means that a greater following distance is required and more sensitivity to lane change behavior is required. For Level 3 severe severe weather, this adjustment range is further increased, the speed limit is the most stringent, the following distance requirement is the greatest, and the judgment of any lane change intention is the most cautious.

[0034] The system periodically, for example weekly, uses updated historical average traffic flow data to recalculate the inherent risk coefficient of the road segment and refresh the threshold rule set for all levels accordingly, thereby achieving adaptive adaptation of safety thresholds and long-term traffic flow characteristics of the road segment.

[0035] Further, referring to Figure 3, S6 also includes: S61, setting a preset risk indicator fusion decision matrix, using the first-level speed risk indicator, the second-level following distance risk indicator, and the third-level lane change intention risk indicator as input dimensions, and the composite risk level as output; S62, discretizing and encoding the first-level speed risk indicator, the second-level following distance risk indicator, and the third-level lane change intention risk indicator to form a risk feature vector; S63, querying the risk indicator fusion decision matrix to determine the output row that matches the risk feature vector and defining the composite risk level; wherein, the construction rules of the risk indicator fusion decision matrix include: if the third-level lane change intention risk indicator is an immediate high risk, then regardless of the values ​​of the first-level speed risk indicator and the second-level following distance risk indicator, the composite risk level is determined to be high risk; if the third-level lane change intention risk indicator is a highly potential risk, and the first-level speed risk indicator is speeding. If the Level 2 following distance risk indicator is "following too closely," the composite risk level is determined to be high risk. If the Level 3 lane change intention risk indicator is "high potential risk," and both the Level 1 speed risk indicator and the Level 2 following distance risk indicator are normal, the composite risk level is determined to be medium risk. If the Level 3 lane change intention risk indicator is "low potential risk" or "no risk," and both the Level 1 speed risk indicator (speeding) and the Level 2 following distance risk indicator (following too closely) occur simultaneously, the composite risk level is determined to be high risk. If the Level 3 lane change intention risk indicator is "low potential risk" or "no risk," and only one of the Level 1 speed risk indicator (speeding) or the Level 2 following distance risk indicator (following too closely) occurs, the composite risk level is determined to be medium risk. If all three Levels—Level 1 speed risk indicator, Level 2 following distance risk indicator, and Level 3 lane change intention risk indicator—are normal or no risk, the composite risk level is determined to be low risk.

[0036] In this embodiment, the sensing unit array continuously processes the raw sensing data radar point cloud and video stream into structured streaming traffic flow data and sends it to the central server. In the data packet, each tracked target vehicle includes its type, such as being identified as a heavy semi-trailer truck, instantaneous speed, precise latitude and longitude coordinates, heading angle, and distances to vehicles in the same lane directly in front and to the nearest vehicles in the adjacent lanes on the left and right, calculated by an algorithm.

[0037] For each target vehicle that enters the construction impact area, the server starts an independent risk analysis thread. First, based on the identified vehicle type, such as a heavy semi-trailer truck, and the currently effective third-level threshold rule set, it queries the specific thresholds for this type of vehicle in dense fog weather: a lower base speed threshold, a larger base safe following distance threshold, and a set of strict base lane change risk coefficient thresholds. This set of thresholds includes multiple sub-items, such as the allowable range of heading angle change rate and the minimum lateral safe distance.

[0038] Speed ​​and following distance risk assessment: Level 1 and Level 2 indicators: The server compares the truck's real-time speed with the queried low base speed threshold. If it finds that the truck is speeding, a Level 1 speed risk indicator is generated as speeding. At the same time, the server compares the actual distance between the truck and the vehicle in front with the increased base safe following distance threshold. If it finds that the distance is insufficient, a Level 2 following distance risk indicator is generated as following too closely.

[0039] The system employs a three-level risk assessment system for lane-change intention, corresponding to claim 4: The server retrieves the truck's heading angle historical sequence over the past few seconds and calculates its rate of change. Analysis reveals that the mean rate of change is consistently positive and stable, with a very small variance, indicating that the vehicle is making a stable and continuous yaw to one side, such as the left, rather than a random sway. Simultaneously, statistics show that this yaw has persisted for more than a preset time percentage threshold. These heading analysis characteristics are consistent with the criteria for determining a clear directional lane-change trend and a continuous lateral displacement intention. Next, the system checks its lateral safety environment: it finds a parallel passenger car slightly ahead in the adjacent lane to the left of the target lane, with a lateral distance between them that is very close, less than the parallel vehicle distance threshold specified in the threshold rule set; simultaneously, another rapidly approaching vehicle is located relatively close behind in the adjacent lane, with a distance less than the following vehicle distance threshold.

[0040] Based on predefined logical rules, the system determines that the truck exhibits a clear directional lane-changing tendency, accompanied by a continuous lateral displacement intention. In this case, the risk of a lateral collision exists due to the close proximity to the parallel vehicle, and the risk of being rear-ended exists due to the proximity of vehicles approaching from the adjacent lane. Combining these four sub-judgments, and based on the mapping table, the system ultimately generates a Level 3 lane-changing intention risk identifier as the highest level of immediate high risk.

[0041] Furthermore, S7 also includes: S71. Obtaining the preset location information of all roadside warning units upstream of the construction section based on the construction section identification; S72. Calculating the first distance between the real-time location coordinates of the target vehicle and the starting point of the construction area, and calculating the second distance between the real-time location coordinates of the target vehicle and the preset locations of each roadside warning unit; S73. Searching for the corresponding warning strategy template from the preset warning strategy mapping table based on the vehicle type and composite risk level of the target vehicle; S74. Determining the warning lead time based on the first distance, composite risk level, and warning priority; S75. Selecting a roadside warning unit that meets the following conditions as the target roadside warning unit: its preset location is located upstream of the target vehicle's driving direction, and its second distance is closest to the warning lead time; S76. Replacing the placeholders in the warning display content template with the vehicle type and composite risk level of the target vehicle to generate specific warning display content; S77. Encapsulating the identifier of the target roadside warning unit, the warning display content, the warning display intensity level, and the trigger command to form an instruction data packet.

[0042] In this embodiment, the server currently holds three risk indicators: high-risk speeding, high-risk following too closely, and the highest-risk indicator of immediate lane-changing intent. The system invokes a predefined risk indicator fusion decision matrix for final determination. One of the core rules of this matrix is: if the third-level lane-changing intent risk indicator is immediately high-risk, then regardless of speed and following conditions, the composite risk level is directly determined to be high-risk. Therefore, although the individual speed and following risks already constitute a threat, the superimposed combination of a clearly intentional dangerous lane-changing behavior ultimately leads the system to determine the truck's composite risk level as the highest level of high risk.

[0043] Furthermore, the construction logic of the warning strategy mapping table in S73 is as follows: For target vehicles with a composite risk level of high risk, if the vehicle type is a large freight vehicle, the warning display content template in the corresponding warning strategy template is: [Large Truck] Please note, construction ahead, your current state is extremely risky, please slow down immediately, maintain the maximum safe distance, and avoid changing lanes! The warning display intensity level is the highest, and the warning priority is the highest. For target vehicles with a composite risk level of high risk, if the vehicle type is a small passenger vehicle, the corresponding warning display content template is: [Small Passenger Vehicle] Warning, construction ahead, high-risk driving behavior, please slow down immediately and increase the following distance! The warning display intensity level is the highest, and the warning priority is high. For target vehicles with a composite risk level of medium risk, if the vehicle type is a large freight vehicle, the corresponding warning display content template is: [Large Truck] Reminder, construction ahead, you are currently at risk, it is recommended to slow down in advance and increase the following distance. The warning display intensity level is medium, and the warning priority is high. For target vehicles with a composite risk level of medium risk, if the vehicle type is a small passenger vehicle, the corresponding warning display template is: [Small Passenger Vehicle] Attention, construction ahead, please maintain a safe speed and distance. The warning display intensity level is medium, and the warning priority is medium. For target vehicles with a composite risk level of low risk, a general construction warning message is used, with a warning display intensity level of low and a warning priority of low.

[0044] In this embodiment, after generating a high-risk assessment, the server needs to decide how to issue a warning. First, it calculates the distance between the vehicle and the construction zone based on the vehicle's real-time location and the starting point of the construction zone. Considering that it is a high-risk large freight vehicle, the system retrieves the corresponding strategy from the warning strategy mapping table: the warning message template is a strong warning statement specifically for large trucks, with the highest display intensity (e.g., flashing red text plus a high-frequency alarm sound), and the highest processing priority. Based on this highest priority, the system calculates a longer warning lead time, aiming to alert the driver earlier.

[0045] Next, the server retrieves roadside warning units located ahead of the truck that meet the advance warning distance range. It selects the upstream variable message sign with the best distance match as the target roadside warning unit. The server replaces the placeholders in the information template with "heavy semi-trailer truck" and "high risk," generates the final display text, and packages it into an instruction data packet along with the display intensity command and unit identifier before sending it out.

[0046] Further, referring to Figure 4, the method also includes a monitoring and feedback step for the status of the roadside warning unit: S91, after sending the instruction data packet to the target roadside warning unit, start the feedback timer; S92, before the feedback timer reaches the preset timeout threshold, if an execution confirmation signal is received from the target roadside warning unit, the warning instruction is marked as successfully executed; S93, if no execution confirmation signal is received after the feedback timer reaches the preset timeout threshold, the target roadside warning unit is marked as having a communication abnormality; S94, for the target roadside warning unit marked as having a communication abnormality, a backup roadside warning unit is selected as the new target roadside warning unit based on the real-time location coordinates of the target vehicle and the warning lead time, and S7 and S8 are re-executed; S95, the communication abnormality information is recorded in the equipment operation and maintenance log, and a corresponding equipment maintenance prompt is generated.

[0047] In this embodiment, after the command is issued, the server starts a timer to wait for confirmation feedback from the information board. In this test, the information board normally received the command and displayed the warning information, and then returned an execution confirmation signal through the communication link. Upon receiving this signal, the server marked the command as successfully executed and reset the timer. In another simulated fault test, if the server did not receive a confirmation signal from a certain unit within a preset time, it marked it as a communication anomaly and automatically rerouted the warning command to another backup warning unit upstream of the anomaly unit for distribution, ensuring the reliability of the warning link. The abnormal status was recorded and a maintenance work order was generated.

[0048] Further, referring to Figure 5, the method also includes a post-warning effect evaluation step: S101, for each target vehicle that triggers the warning, after the warning command is issued, the subsequent streaming traffic flow data of the target vehicle from the sensing unit array is continuously received for a preset observation window; S102, the real-time speed of the target vehicle, the real-time distance to the vehicle in front, and the trend of heading angle change within the observation window are analyzed; S103, based on the analysis results, it is determined whether the target vehicle has exhibited driving behavior correction that meets the warning expectation within the observation window. Driving behavior correction includes: speed decreasing to below the basic speed threshold, following distance increasing to above the basic safe following distance threshold, or termination of abnormal heading angle change trend; S104, the effectiveness of the warning event is rated according to whether driving behavior correction occurs and the timeliness of the correction, and the rating result is recorded in the warning history database; S105, the data in the warning history database is periodically statistically analyzed to generate warning response effectiveness reports under different meteorological impact levels, different vehicle types, and different composite risk levels.

[0049] In this embodiment, the system tracks the post-event effects of all vehicles that trigger warnings. Taking the aforementioned high-risk truck as an example, the system continuously receives data during the observation period after the warning is issued. Analysis shows that after hearing the alarm and seeing the warning sign, the truck's speed begins to decrease significantly within seconds, eventually falling below the safety threshold; the distance to the vehicle in front gradually increases to above the safe range; simultaneously, the previously stable steering trend stops, the heading angle stabilizes, and the vehicle remains in its original lane. Based on these positive behavioral corrections, the system determines that the warning is timely and effective, and records this successful warning case, along with information such as the weather level (Level 3), the vehicle type (large truck), and the risk type (combined high risk), into the warning history database. Long-term statistical analysis of this data can generate reports. For example, in dense fog, the driver compliance rate for warnings regarding high-risk lane-changing behavior by large trucks reaches XX%, providing data support for evaluating and optimizing warning strategies.

[0050] Further, referring to Figure 6, the sensing unit array includes multiple fusion sensing devices consisting of millimeter-wave radar and video cameras deployed along the road. Vehicle type identification in S3 is achieved locally by the fusion sensing devices in the following ways: S31, extracting the target vehicle's contour size, point cloud density distribution features, and motion echo features from millimeter-wave radar point cloud data; S32, extracting the target vehicle's visual contour, vehicle structure features, and license plate information from video image data; S33, fusing the contour size, point cloud density distribution features, motion echo features, visual contour, and vehicle structure features, and matching them with a pre-stored vehicle type feature template library; S34, based on the matching results, classifying the target vehicle into multiple predefined vehicle types, at least distinguishing between large freight vehicles, medium-sized passenger and freight vehicles, small passenger vehicles, and special vehicles.

[0051] In this embodiment, accurate vehicle type identification is fundamental. The millimeter-wave radar of the sensing unit provides a 3D point cloud of the target. By analyzing the length, width, and height of the point cloud outline, it is possible to initially distinguish between different size categories such as cars and trucks. Simultaneously, the density distribution of the point cloud reflects the surface structure of the vehicle, such as the difference in reflection between the cargo box and the cab. Images provided by a high-definition camera, through visual analysis, can further confirm vehicle appearance features, such as the presence of a cargo box, trailer, and bus window layout, and provide important auxiliary information for license plate and vehicle type identification. A central server or an edge computing module built into the sensing unit fuses radar point cloud features and visual features, matching and comparing them with a pre-built rich vehicle feature template library. This reliably classifies targets into finer categories such as small passenger cars, large passenger cars, medium-sized trucks, heavy semi-trailer trucks, and hazardous materials transport vehicles, providing an accurate object classification basis for subsequent differentiated risk assessment and early warning.

[0052] To verify the effectiveness of the method of the present invention, a combination of test cases was constructed on a simulation test platform, including various weather scenarios such as sunny, rainy, and foggy, various vehicle types such as passenger cars and large trucks, and various dangerous driving behaviors such as speeding, following too closely, and illegal lane changing.

[0053] Test Scenario 1: A passenger car is driving normally in clear weather. The system determines the weather level to be Level 1 and uses a normal threshold. The vehicle speed and distance are normal, and there is no abnormal lane-changing intention. The system determines the composite risk level to be low and only issues a routine construction warning without triggering a strong alarm. The test results are in line with expectations, avoiding excessive warnings that could disrupt normal traffic flow.

[0054] Test Scenario 2: A large truck was speeding and following too closely in rainy weather. The system determined the weather to be Level 2 and automatically applied stricter speed and following distance thresholds. It identified that the truck was simultaneously triggering speeding and following too closely risks. Although its course was stable and it showed no intention to change lanes, thus indicating no risk, the combined high risks of speed and following distance, according to the fusion decision matrix, resulted in a medium risk level. The system issued a warning specifically for the truck, emphasizing deceleration and increased following distance. The test showed that the system can effectively identify and respond to the comprehensive risks posed by large vehicles in severe weather.

[0055] Test Scenario 3: Dangerous Lane Change by a Passenger Car in Dense Fog, corresponding to the detailed implementation example described above. This is a highly complex scenario. The system determined the weather level to be Level 3 and applied the most stringent threshold. It successfully identified the passenger car's dangerous lane-changing intention under poor visibility conditions and, combined with its slightly excessive speed, ultimately determined it to be high-risk. The warning system successfully issued a high-intensity warning in advance. Simulation results showed that the warning prompted the driver to cancel the dangerous lane-changing behavior and slow down, verifying the system's timely intervention capability in high-risk, highly dynamic scenarios.

[0056] As can be seen from the detailed description and simulation test of the above specific embodiments, the method of the present invention constructs a safety threshold system that adapts to different weather and road conditions by dynamically integrating real-time meteorological data and static road segment characteristics; it achieves differentiated and accurate risk assessment for different types of vehicles by refining vehicle type identification and multi-dimensional behavior analysis of speed, distance, and deep lane change intentions; and it ensures that high-risk events can be intervened in the highest priority and most effective way by multi-level risk identification fusion decision-making and differentiated early warning strategy mapping. The entire method has a rigorous logical chain, clear steps, and is feasible, effectively overcoming the shortcomings of existing early warning methods such as rigidity, single dimension, and insufficient targeting, and significantly improving the intelligence level and active protection effectiveness of safety early warning in highway construction areas.

[0057] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0058] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units described above can be implemented in hardware or as software functional units. The above are merely embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

[0059] The specific embodiments of the invention have been described in detail above, but they are only examples, and this application is not limited to the specific embodiments described above. For those skilled in the art, any equivalent modifications or substitutions to the invention are also within the scope of this application. Therefore, all equivalent changes, modifications, and improvements made without departing from the spirit and principles of this application should be covered within the scope of this application.

Claims

1. A method for early warning of safety during highway construction, characterized in that, Running on the central server of the highway management system, it is connected to a sensing unit array deployed upstream of the construction section and multiple roadside early warning units, including: S1, acquiring current meteorological data and construction section identification, and retrieving corresponding static road section characteristic parameters based on the construction section identification; S2, determining the current meteorological impact level based on the current meteorological data, and selecting the corresponding target basic safety threshold rule set from multiple preset basic safety threshold rule sets based on the current meteorological impact level and the static road section characteristic parameters; S3, receiving real-time streaming traffic flow data from the sensing unit array; S4, for each target vehicle, querying the target basic safety threshold rule set to obtain the corresponding basic speed threshold, basic safe following distance threshold, and basic lane change risk coefficient threshold according to the vehicle type; S5, respectively based on the real-time speed of the target vehicle and the corresponding... The system compares the base speed threshold, the real-time distance to the vehicle ahead with the corresponding base safe following distance threshold, the real-time heading angle change trend, the real-time speed, and the real-time distance to vehicles in adjacent lanes, and calculates the risk level based on the corresponding base lane change risk coefficient threshold to generate a first-level speed risk indicator, a second-level following distance risk indicator, and a third-level lane change intention risk indicator; S6, it generates a composite risk level for the target vehicle by combining predefined risk indicator fusion logic; S7, for the target vehicle that is determined to be high-risk or medium-risk, it determines the target roadside early warning unit to be activated based on its real-time location coordinates and the construction section marker, and generates an instruction data packet containing the composite risk level, the vehicle type of the target vehicle, and a suggested early warning strategy; S8, it sends the instruction data packet to the target roadside early warning unit to drive the execution of the corresponding early warning information release action.

2. The highway construction safety early warning method according to claim 1, characterized in that, S2 further includes: S21, parsing the current meteorological data and extracting the current precipitation intensity value, current visibility value, and current wind speed value; S22, preset a first precipitation intensity threshold, a second precipitation intensity threshold, a first visibility threshold, a second visibility threshold, a first wind speed threshold, and a second wind speed threshold; S23, determining the current meteorological impact level based on a comparison between the current precipitation intensity value, current visibility value, and current wind speed value and the first precipitation intensity threshold, the second precipitation intensity threshold, the first visibility threshold, the second visibility threshold, the first wind speed threshold, and the second wind speed threshold; the current meteorological impact level includes a first level, a second level, and a third level, where the first level corresponds to normal weather conditions, the second level corresponds to general severe weather conditions, and the third level corresponds to severe severe weather conditions.

3. The highway construction safety early warning method according to claim 1, characterized in that, S5 further includes: S51, calculating the heading angle change rate sequence of the target vehicle within the current time window based on the real-time position coordinates and real-time heading angle of the target vehicle uploaded by the sensing unit array at multiple consecutive timestamps; S52, performing statistical analysis on the heading angle change rate sequence to extract its mean, variance, and the proportion of time that continuously exceeds a preset small angle threshold; S53, obtaining the real-time speed of the target vehicle and the real-time distance of vehicles in adjacent lanes, wherein the real-time distance of vehicles in adjacent lanes includes the distance of the nearest vehicle in the adjacent lane parallel to the target vehicle, and the distance of the nearest vehicle in the adjacent lane within a predetermined range behind the target vehicle; S54, based on the vehicle type of the target vehicle, selecting from the target basic safety threshold rule set... Obtain the corresponding basic lane change risk coefficient threshold, which is a multi-dimensional judgment benchmark vector, including the mean threshold of the heading angle change rate, the variance threshold, the continuous deviation duration percentage threshold, the parallel vehicle distance threshold, and the following vehicle distance threshold; S55, compare the mean, variance, and continuous deviation duration percentage extracted in S52, and the parallel vehicle distance and following vehicle distance obtained in S53 with the corresponding thresholds in the multi-dimensional judgment benchmark vector; S56, generate the third-level lane change intention risk identifier according to the predefined comparison logic rules; S57, map the different combinations of the four sub-judgment results of directional lane change trend, continuous lateral displacement intention, side collision risk, and rear-end collision risk to the preset lane change risk identifier enumeration value.

4. The highway construction safety early warning method according to claim 1, characterized in that, S6 further includes: S61, setting a preset risk indicator fusion decision matrix, using the first-level speed risk indicator, the second-level following distance risk indicator, and the third-level lane change intention risk indicator as input dimensions, and the composite risk level as output; S62, discretizing and encoding the first-level speed risk indicator, the second-level following distance risk indicator, and the third-level lane change intention risk indicator to form a risk feature vector; S63, querying the risk indicator fusion decision matrix, determining the output row matching the risk feature vector, and defining the composite risk level; wherein, the construction rules of the risk indicator fusion decision matrix include: if the third-level lane change intention risk indicator is an immediate high risk, then regardless of the values ​​of the first-level speed risk indicator and the second-level following distance risk indicator, the composite risk level is determined to be high risk; if the third-level lane change intention risk indicator is a highly potential risk, and the first-level speed risk indicator is speeding or the second-level... If the following distance risk indicator is "following too closely," the composite risk level is determined to be high risk. If the third-level lane change intention risk indicator is "high potential risk," and both the first-level speed risk indicator and the second-level following distance risk indicator are normal, the composite risk level is determined to be medium risk. If the third-level lane change intention risk indicator is "low potential risk" or "no risk," and both the first-level speed risk indicator (speeding) and the second-level following distance risk indicator (following too closely) occur simultaneously, the composite risk level is determined to be high risk. If the third-level lane change intention risk indicator is "low potential risk" or "no risk," and only one of the first-level speed risk indicator (speeding) or the second-level following distance risk indicator (following too closely) occurs, the composite risk level is determined to be medium risk. If the first-level speed risk indicator, the second-level following distance risk indicator, and the third-level lane change intention risk indicator are all normal or no risk, the composite risk level is determined to be low risk.

5. The highway construction safety early warning method according to claim 1, characterized in that, S7 further includes: S71. Obtaining the preset location information of all roadside early warning units upstream of the construction section based on the construction section identifier; S72. Calculating the first distance between the real-time location coordinates of the target vehicle and the starting point of the construction area, and calculating the second distance between the real-time location coordinates of the target vehicle and the preset locations of each roadside early warning unit; S73. Searching for the corresponding early warning strategy template from a preset early warning strategy mapping table based on the vehicle type of the target vehicle and the composite risk level; S74. Determining the early warning lead time based on the first distance, the composite risk level, and the early warning priority; S75. Selecting a roadside early warning unit that meets the following conditions as the target roadside early warning unit: its preset location is located upstream of the direction of travel of the target vehicle, and the second distance is closest to the early warning lead time; S76. Replacing the placeholder in the early warning display content template with the vehicle type of the target vehicle and the composite risk level to generate specific early warning display content; S77. Encapsulating the identifier of the target roadside early warning unit, the early warning display content, the early warning display intensity level, and the trigger command to form the command data packet.

6. The highway construction safety early warning method according to claim 5, characterized in that, The construction logic of the early warning strategy mapping table described in S73 is as follows: For a target vehicle with a composite risk level of high risk, if its vehicle type is a large freight vehicle, the corresponding early warning strategy template will display the following content: [Large Truck] Please note that there is construction ahead. Your current risk level is extremely high. Please slow down immediately, maintain the maximum safe distance, and avoid changing lanes! The warning display intensity level is the highest level, and the warning priority is the highest. For target vehicles with a composite risk level of high risk, if the vehicle type is a small passenger vehicle, the corresponding warning display template will be: [Small Passenger Vehicle] Warning: Construction ahead, high-risk driving behavior, please slow down immediately and increase following distance! The warning intensity level will be the highest, and the warning priority will be high. For target vehicles with a composite risk level of medium risk, if the vehicle type is a large freight vehicle, the corresponding warning display template will be: [Large Truck] Reminder: Construction ahead, you are currently at risk, it is recommended to slow down in advance and increase following distance. The warning intensity level will be medium, and the warning priority will be high. For target vehicles with a composite risk level of medium risk, if the vehicle type is a small passenger vehicle, the corresponding warning display template is: [Small Passenger Vehicle] Attention, construction ahead, please maintain a safe speed and distance. The warning display intensity level is medium, and the warning priority is medium. For target vehicles with a composite risk level of low risk, a general construction warning message is used, with a warning display intensity level of low and a warning priority of low.

7. The highway construction safety early warning method according to claim 1, characterized in that, The method further includes a monitoring and feedback step for the status of the roadside warning unit: S91, after sending the instruction data packet to the target roadside warning unit, a feedback timer is started; S92, before the feedback timer reaches a preset timeout threshold, if an execution confirmation signal is received from the target roadside warning unit, the warning instruction is marked as successfully executed; S93, if the execution confirmation signal is not received after the feedback timer reaches the preset timeout threshold, the target roadside warning unit is marked as having a communication abnormality; S94, for the target roadside warning unit marked as having a communication abnormality, a backup roadside warning unit is selected as the new target roadside warning unit based on the real-time location coordinates and warning lead of the target vehicle, and S77 and S8 are re-executed; S95, the communication abnormality information is recorded in the equipment operation and maintenance log, and a corresponding equipment maintenance prompt is generated.

8. The method for early warning of highway construction safety according to claim 1, characterized in that, The method further includes a post-warning effect evaluation step: S101, for each target vehicle that triggers the warning, after the warning command is issued, the subsequent streaming traffic flow data of the target vehicle from the sensing unit array is continuously received for a preset observation window; S102, the real-time speed of the target vehicle, the real-time distance to the vehicle in front, and the trend of the heading angle change within the observation window are analyzed. S103. Based on the analysis results, determine whether the target vehicle has exhibited driving behavior correction that meets the warning expectation within the observation window. The driving behavior correction includes: speed decreasing to below the basic speed threshold, following distance increasing to above the basic safe following distance threshold, or termination of abnormal heading angle change trend. S104. Based on whether the driving behavior correction occurred and the timeliness of the correction, rate the effectiveness of the warning event and record the rating result in the warning history database; S105. Periodically analyze the data in the warning history database and generate warning response effectiveness reports under different meteorological impact levels, different vehicle types, and different composite risk levels.

9. The method for early warning of highway construction safety according to claim 1, characterized in that, The sensing unit array includes multiple fusion sensing devices consisting of millimeter-wave radar and video cameras deployed along the road. The vehicle type identification mentioned in S3 is achieved locally by the fusion sensing device in the following ways: S31, extracting the outline size, point cloud density distribution features, and motion echo features of the target vehicle from millimeter-wave radar point cloud data; S32, extracting the visual outline, vehicle structure features, and license plate information of the target vehicle from video image data; S33, fusing the outline size, point cloud density distribution features, motion echo features, visual outline, and vehicle structure features, and matching them with a pre-stored vehicle type feature template library; S34, classifying the target vehicle into multiple predefined vehicle types based on the matching results, wherein the multiple vehicle types at least distinguish between large freight vehicles, medium-sized passenger and freight vehicles, small passenger vehicles, and special vehicles.

10. A highway construction safety early warning system, characterized in that, include: One or more processors; A memory having stored one or more programs, which, when executed by one or more processors, cause the one or more processors to implement a highway construction safety early warning method according to any one of claims 1 to 9.

Images