A highway maintenance work area electronic fence intelligent monitoring and early warning method

Abstract

The present application belongs to the technical field of highway maintenance operation early warning, and specifically discloses an intelligent monitoring and early warning method for an electronic fence in a highway maintenance operation area, wherein the fence boundary of the maintenance operation area is divided into three types of hard separation boundary, warning boundary and open boundary, and each boundary is bound with a corresponding risk area; when target intrusion is determined, the distance from the target to the specific boundary is directly used as the determination basis instead of whether the vehicle enters the macro area, so that the micro track of the target approaching the boundary can be continuously calculated, the accurate spatial corresponding relationship is realized, the response lag caused by the traditional overall area determination is avoided, and when the distance is used to determine the target approaching a certain boundary, the target intrusion risk is determined by combining the target speed vector direction and the intrusion direction angle of the normal line of the boundary, so that the intrusion intention of the target can be recognized in advance, the false alarm and the missed alarm caused by the single distance are avoided, and the accuracy and timeliness of the early warning are improved.

Description

Technical Field

[0001] This invention belongs to the field of early warning technology for highway maintenance operations, and specifically discloses an intelligent monitoring and early warning method for electronic fences in highway maintenance operation areas. Background Technology

[0002] Highway maintenance work is usually carried out under open traffic conditions. Although the work area and passing vehicles are separated by fences with cones, water-filled barriers and other facilities, these fences are mostly temporary and lack the physical isolation capability to withstand vehicle impacts. Unauthorized vehicles entering the work area is the main type of accident that causes injury or death to maintenance workers. In order to ensure the safety of maintenance work, a variety of proactive early warning methods have been developed in the existing technology.

[0003] For example, Chinese Patent Publication No. CN120183094A discloses a collaborative prevention and control system for highway maintenance safety operations that integrates four prevention and one avoidance technologies. The system divides the highway maintenance operation area along the driving direction into a warning zone, an upstream transition zone, a buffer zone, and a work zone. Remote audio-visual units, guardrail warning devices, and roadside anti-intrusion devices are deployed in each area. Its warning logic is mainly based on the section location and distance threshold: it detects the distance range of social vehicles entering a specific area by radar, or calculates the risk value based on distance and speed factors using a weighted average algorithm. When the parameters exceed the preset threshold, an audio-visual alarm is triggered.

[0004] The aforementioned scheme, through differentiated early warning systems across multiple regions, has improved the safety protection capabilities of maintenance work areas to some extent. However, the following problems still exist in practical applications: First, the scheme adopts a region division method based on segment location, with the boundaries between regions being merely virtual cross-sectional lines without clear geometric boundaries. This division method cannot establish a spatial correspondence between the real-time location of vehicles and specific boundary lines, resulting in an ambiguous judgment logic for vehicles entering a region. A risk is only identified and an early warning is triggered when a vehicle crosses the virtual cross-sectional line and enters a specific region. This coarse-grained judgment leads to a lack of continuous calculation of vehicle approach to the boundary, resulting in significant response lag.

[0005] Secondly, existing technologies use a single distance or risk threshold as the trigger condition. However, when a vehicle is driving normally in an adjacent lane, although the distance condition is met, its direction of movement is parallel to the boundary, and it does not actually pose a risk. In this case, triggering a warning leads to a false alarm. Conversely, when a vehicle approaches the work area at a small angle from the side, although the distance is relatively far, its direction of movement towards the work area also constitutes a disturbance threat. It is evident that existing technologies rely solely on distance thresholds and cannot identify intrusion intentions in advance, thus reducing the accuracy and timeliness of warnings. Summary of the Invention

[0006] In order to solve the above-mentioned technical problems, or at least partially solve the above-mentioned technical problems, the present invention provides a method for intelligent monitoring and early warning of electronic fences in highway maintenance work areas.

[0007] The objective of this invention can be achieved through the following technical solution: a method for intelligent monitoring and early warning of electronic fences in highway maintenance work areas, comprising: dividing the fence boundary of the maintenance work area into a hard isolation boundary, a warning boundary, and an open boundary.

[0008] Each boundary is bound to a corresponding risk area, and the distance from each boundary to the interior of its bound area is calculated as the area depth.

[0009] The system acquires target distance and angle data from the maintenance site in real time, calculates the two-dimensional plane coordinates of the target, and obtains the target velocity vector direction based on coordinate difference calculations at multiple consecutive time points.

[0010] The shortest distance from the target to each boundary is calculated based on the target's two-dimensional planar coordinates. The ratio of the shortest distance from the target to each boundary to the depth of the region bound by the boundary is used to identify the current boundary of interest. At the same time, the intrusion direction angle between the target's velocity vector direction and the normal of the current boundary of interest is obtained.

[0011] The risk of target intrusion is determined based on the intrusion direction angle and its change over time.

[0012] When a target intrusion risk is determined, the intrusion direction determined by the intrusion direction angle and the current boundary of concern are combined to determine the warning level and execute the corresponding warning action.

[0013] Combining all the above technical solutions, the positive effects of this invention are as follows: 1. This invention divides the boundary of the maintenance work area fence into three categories: hard isolation boundary, warning boundary and open boundary, and binds a corresponding risk area to each boundary. When determining the target intrusion area, the distance from the target to the specific boundary is directly used as the judgment basis, thereby enabling continuous calculation of the micro-trajectory of the target approaching area, realizing accurate spatial correspondence, avoiding the response delay caused by traditional overall area judgment, and maximizing the timeliness of early warning.

[0014] 2. When using distance to determine the approach of a target to a certain boundary, this invention further combines the intrusion direction angle between the target's velocity vector direction and the boundary normal to determine the target intrusion risk. This can identify the target's intrusion intention in advance, effectively distinguish between normal passage and potential intrusion behavior, and avoid false alarms and missed alarms caused by a single distance threshold. This improves the accuracy and timeliness of early warning and enhances the proactive safety protection capabilities for maintenance personnel. Attached Figure Description

[0015] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0016] Figure 1 This is a diagram illustrating the implementation steps of the method of the present invention;

[0017] Figure 2 This is a flowchart illustrating the implementation process of determining the target intrusion risk based on the intrusion direction angle and its change over time in this invention.

[0018] Figure 3 This is a schematic diagram illustrating how the warning level is determined in this invention by combining the intrusion direction angle and the current concern boundary type. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] This invention proposes an intelligent monitoring and early warning method for electronic fences in highway maintenance work areas.

[0021] To clarify the meaning of directional terms such as upstream and downstream in this manual, the directional reference is first defined as follows: the direction of traffic flow on the road section where the maintenance work area is located is the driving direction; the direction of oncoming traffic is upstream, and the direction of outgoing traffic is downstream. All statements regarding upstream and downstream in this manual shall be based on this reference.

[0022] See Figure 1 As shown, the implementation of the present invention specifically includes the following steps: S1, dividing the boundary of the maintenance work area fence into a hard isolation boundary, a warning boundary and an open boundary.

[0023] In highway maintenance operations, fences are typically erected to separate the work area from the traffic lanes. However, the risk of vehicles accidentally entering these areas varies across different zones: some zones are adjacent to the traffic lanes, and a misplaced entry would directly cause a collision; while other zones have a safe distance from the traffic lanes, providing a buffer zone for vehicles that accidentally enter. Using a uniform standard for protection would not only waste resources but could also lead to insufficient protection in high-risk areas and excessive protection in medium- and low-risk areas.

[0024] To address the aforementioned issues, it is necessary to classify the maintenance work area by risk. During the classification process, it is considered that the area is enclosed by boundaries with different functions. Therefore, this invention subdivides the fence boundary of the maintenance work area into hard isolation boundaries, warning boundaries, and open boundaries, which facilitates the matching of differentiated protective measures according to the area to which different boundaries belong.

[0025] In a preferred embodiment of the present invention, the composition of the fence boundary needs to be clarified before the fence boundary is defined: the fence boundary set outside the work area is usually formed by a crash buffer vehicle and a continuously deployed cone array. The crash buffer vehicle is a special protective vehicle with collision energy absorption function parked at the front of the work area. Its function is to absorb the impact energy by deforming itself when a social vehicle accidentally enters the area, thereby reducing the severity of the accident. The cone array is a linear isolation facility formed by a series of cone-shaped traffic signs continuously deployed at a certain interval. It is used to guide the vehicle's driving trajectory and separate the work area from the traffic lane.

[0026] In addition to the aforementioned physical protective facilities, to identify non-physical risks within the work area, such as the movement range of operating machinery and material storage areas, the work supervisor must use a handheld terminal to delineate these areas on-site. This handheld terminal is a portable smart device provided to the on-site work supervisor that can transmit the boundary coordinates of risk areas to the system in real time.

[0027] Furthermore, the fence boundary delineation specifically includes the following steps: S11, First, obtain the location of the crash buffer vehicle from the crash buffer vehicle positioning terminal as the starting anchor point of the hard isolation boundary, obtain the location of the deployed cones from the cone array deployment path as the intermediate anchor point, and obtain the ending anchor point from the end of the cone array. Connect the starting anchor point, intermediate anchor point and ending anchor point in sequence to generate a continuous broken line, which constitutes the hard isolation boundary. This isolation boundary relies on physical protection facilities such as crash buffer vehicles and cones, and reflects the most direct physical protection range of the work area.

[0028] S12. Obtain the vertex position of the warning line drawn by the work team leader based on the actual situation on site from the work team leader's handheld terminal. The vertex position is determined by the work team leader by clicking on the screen of the handheld terminal. It usually corresponds to the edge of the working machinery's activity range, the boundary of the material stacking area, etc. Connect each vertex in sequence to generate a continuous broken line to form the warning boundary.

[0029] S13. The intersection of the straight line perpendicular to the driving direction at the starting anchor point of the hard isolation boundary and the road shoulders on both sides of the road is taken as the two endpoints of the upstream open boundary. The upstream open boundary is generated by connecting the two endpoints. This boundary is located upstream of the hard isolation boundary and is connected to the road shoulders on both sides of the road. It is used to define the passage area in the upstream direction of the work area and provide upstream passage guidance for social vehicles.

[0030] S14. The intersection of the straight line perpendicular to the driving direction at the location of the termination anchor point of the hard isolation boundary and the road shoulders on both sides of the road is taken as the two endpoints of the downstream open boundary. The two endpoints are connected to generate the downstream open boundary. This boundary is located downstream of the hard isolation boundary and is connected to the road shoulders on both sides of the road. It is used to define the passage area in the downstream direction of the work area and provide downstream passage guidance for social vehicles.

[0031] S15. The boundaries generated by the above steps are as follows: the upstream open boundary is located at the upstreammost point, the hard isolation boundary starts upstream and ends downstream, and the downstream open boundary is located at the downstreammost point. After the boundaries are generated, the boundary priority weights are configured as follows: hard isolation boundary is the highest, warning boundary is the second highest, and open boundary is the lowest.

[0032] S2. Bind the corresponding risk area to each boundary and calculate the distance from each boundary to the interior of the bound area as the area depth.

[0033] Considering that the purpose of boundary delineation is to achieve risk-based management of the work area, it is necessary to further bind the specific risk areas corresponding to each boundary after completing the boundary delineation, in order to establish a mapping relationship between the boundary and the protected object. The principle of the binding logic is: the risk area bound to the boundary should be the area that the target will enter after crossing the boundary.

[0034] As one possible implementation method of this invention, binding a corresponding risk area to each boundary includes the following: Since a hard barrier boundary is the location of a physical barrier facility, its inner side directly encloses the activity space for maintenance personnel and machinery. Therefore, the internal area enclosed by the hard barrier boundary is defined as the work activity zone. Once a vehicle crosses the hard barrier boundary, it will directly enter this area, posing a threat to the workers. Therefore, binding a work activity zone to the hard barrier boundary is used to identify the most urgent intrusion risks.

[0035] The warning boundary, located between the upstream open boundary and the hard barrier boundary, serves as a transition zone for vehicles gradually approaching the work area from the normal traffic lane. Using the upstream open boundary as the upstream lateral boundary, the warning boundary as the downstream lateral boundary, and the extensions of the two hard barrier boundaries as lateral boundaries, the area between the upstream open boundary and the warning boundary is defined as the warning transition zone. When a vehicle crosses the warning boundary, it indicates that it has entered a buffer zone requiring slowing down and observation from the normal traffic area, and there is a tendency for it to further approach the hard barrier boundary. Therefore, the warning boundary is associated with the warning transition zone.

[0036] The downstream open boundary is located downstream of the hard barrier boundary. The hard barrier boundary extends from its starting anchor point along the driving direction to its ending anchor point. The area enclosed between the downstream end of the hard barrier boundary (the ending anchor point) and the downstream open boundary boundary is the area where maintenance work has been completed. Therefore, the downstream end section of the hard barrier boundary is designated as the upstream lateral boundary, the downstream open boundary as the downstream lateral boundary, and the road shoulders on both sides or the extension of the hard barrier boundary as lateral boundaries. The area between the downstream end of the hard barrier boundary and the downstream open boundary is defined as the safety zone. Under normal circumstances, vehicles should not enter this area from the downstream direction in reverse. Therefore, a safety zone is established for the downstream open boundary to identify the risk of reverse driving.

[0037] This invention implements area division by defining the boundaries of the maintenance work area fence and assigning a corresponding risk zone to each boundary. Compared to traditional area division based on segment location, this method provides clearer boundaries and can display the microscopic trajectory of a target approaching the area, rather than relying on the overall intrusion of the target to make a general risk assessment, thus greatly improving the timeliness of risk assessment. Simultaneously, as maintenance work progresses, the boundaries can be dynamically adjusted, allowing the area to automatically update as the work progresses. This avoids the problems of delayed area updates and disconnect from actual work in the traditional fixed-segment division model.

[0038] After binding each boundary to a corresponding risk area, given the differences in spatial scale among different risk areas, it is necessary to quantify the spatial buffering capacity of the target intrusion area. Therefore, this invention further calculates the distance from each boundary to the interior of its bound area as the area depth. Since the risk areas enclosed or associated by different types of boundaries have different geometries, the calculation method for area depth also differs.

[0039] In an optional embodiment of the present invention, the calculation process of the regional depth is as follows: a) For a hard isolation boundary, the work activity area enclosed by the hard isolation boundary is a closed area. After a social vehicle intrudes from any point on the boundary, the farthest point inside the area that may be threatened is located on the opposite boundary. Therefore, the distance between the two farthest points on the boundary is calculated as an approximate value of the regional depth. Specifically, all anchor points are selected from the hard isolation boundary as sampling points, and the maximum Euclidean distance from each sampling point to any point inside the area it is bound to is calculated as the depth value corresponding to that sampling point. Then, the maximum value among all the depth values ​​of the sampling points is taken as the regional depth of the work activity area, which can characterize the maximum distance from the boundary to the farthest point inside the area. The shorter the depth, the narrower the work activity area, the smaller the escape space for workers after a vehicle intrudes, and the higher the risk level.

[0040] (b) For the warning boundary, its associated warning transition zone is a strip-shaped area between the upstream open boundary and the warning boundary. Vehicles typically approach the warning boundary perpendicularly from the upstream direction. If the farthest point is used, for example, by connecting a vertex on the warning boundary diagonally to the far end of the upstream open boundary, this distance would be abnormally large and would not accurately reflect the buffer distance that a vehicle needs to traverse after entering via a typical path. Therefore, all vertices on the warning boundary are selected as sampling points, and the vertical distance from each sampling point to the upstream open boundary is calculated. If the perpendicular is not on the line segment of the upstream open boundary, the shortest distance from the sampling point to the two ends of the upstream open boundary is taken as the vertical distance. The vertical distances of all sampling points are summed and divided by the number of sampling points to obtain the average depth of the warning transition zone. This depth value reflects the average buffer distance that a vehicle needs to traverse from the upstream open boundary to touching the warning boundary.

[0041] c) For the downstream open boundary, the safety zone bound to it is located downstream of the hard isolation boundary and is also an open area. If a vehicle enters from the downstream in reverse, its direction of movement is perpendicular to the downstream open boundary and points towards the hard isolation boundary. If the farthest point is used, for example, connecting diagonally from a point on the downstream open boundary to the far end of the hard isolation boundary, this distance will deviate from the actual path length traversed by the vehicle, making it impossible to reasonably assess the risk of intrusion from the downstream. Therefore, two endpoints and the equally spaced points between the two endpoints are selected as sampling points on the downstream open boundary. The shortest distance from each sampling point to the hard isolation boundary is calculated. The sum of the shortest distances of all sampling points is divided by the number of sampling points to obtain the average depth of the safety zone. The shorter the average depth, the closer the vehicle will be to the hard isolation boundary after intrusion from the downstream, and the higher the risk.

[0042] S3. Real-time acquisition of target distance and angle data detected at the maintenance site, calculation of the target's two-dimensional plane coordinates, and obtaining the target velocity vector direction based on coordinate difference calculations at multiple consecutive time points.

[0043] After boundary delineation and risk zone binding are completed, the next step is to dynamically monitor targets at the maintenance work site. Traditional monitoring methods only focus on the distance between the target and the risk zone, triggering an alarm when the distance approaches a preset threshold. This method is prone to false alarms because it ignores the target's movement direction. To address this, this invention introduces angle into the target monitoring in addition to distance. By acquiring the distance and angle data of the target relative to the monitoring equipment in real time, the two-dimensional planar coordinates and velocity vector direction of the target are calculated. This allows for accurate determination of whether the target's movement trajectory points towards the risk zone, improving the accuracy of risk warnings.

[0044] The following is a description of a specific embodiment: S31, real-time acquisition of target distance and angle data from a millimeter-wave radar deployed on the roadside, wherein the distance data represents the straight-line distance between the target and the radar, and the angle data represents the azimuth angle of the target relative to the radar normal.

[0045] S32. Given that the distance and angle obtained above are relative position parameters relative to the radar, and not the absolute position of the target in the global environment, in order to convert the radar observation values ​​into absolute coordinates under a unified global coordinate system, it is necessary to determine a coordinate reference system. Therefore, a two-dimensional plane coordinate system is established with the location of the starting anchor point of the hard isolation boundary as the coordinate origin, the driving direction as the positive X-axis, and the direction perpendicular to the driving direction and pointing to the outside of the work area as the positive Y-axis.

[0046] S33. The installation position of the millimeter-wave radar in a two-dimensional plane coordinate system is marked as a fixed coordinate, which is the horizontal and vertical coordinates. Based on the target distance and angle data acquired by the radar, the offset of the target relative to the radar is calculated using trigonometric functions. For example, the horizontal component of the offset is the cosine value of the distance data multiplied by the angle data, and the vertical component of the offset is the sine value of the distance data multiplied by the angle data. The offset reflects the relative displacement vector of the target in the radar's local coordinate system. The offset is added to the fixed coordinates of the radar to obtain the horizontal and vertical coordinates of the target in the two-dimensional plane coordinate system. With the target coordinates, a quantitative basis can be obtained for the target's approach distance to the boundary of the working area.

[0047] S34. By performing differential operations on the target coordinates at multiple consecutive times, the displacement vector of the target between adjacent sampling times can be calculated. The displacement vector points from the previous time to the next time, thus obtaining the direction of the target displacement vector, i.e. the direction of the target velocity vector. With the direction, there is a clear basis for determining whether the target's movement is pointing towards a risk area.

[0048] It should be noted that the above-mentioned monitoring of targets at the maintenance work site refers to social vehicles outside the work area that may pose an intrusion risk, rather than construction vehicles inside the maintenance work area.

[0049] S4. Calculate the shortest distance from the target to each boundary based on the target's two-dimensional plane coordinates. Use the ratio of the shortest distance from the target to each boundary to the depth of the region bound by the boundary to identify the current boundary of interest. At the same time, obtain the intrusion direction angle between the target's velocity vector direction and the normal of the current boundary of interest.

[0050] Once the target coordinates and velocity vector direction are determined, the distances and directions between the target and each boundary can be calculated. Due to significant differences in the depth of different risk areas, a fixed warning distance cannot accommodate the spatial scale of various regions. Therefore, this invention uses the regional depth as a reference benchmark. The specific implementation process is as follows: S41, Based on the real-time calculated target coordinates and the positions of each boundary, calculate the Euclidean distance from the target to the nearest point on each boundary, denoted as the nearest distance, and simultaneously obtain the regional depth of the risk area bound to each boundary.

[0051] S42. Calculate the ratio of the shortest distance from the target to each boundary to the depth of the boundary area. If the ratio of the shortest distance from the target to a certain boundary to the depth of the boundary area is less than the warning ratio, it is determined that the target has entered the vicinity area that needs to be monitored closely. At this time, the boundary is recorded as the current boundary of concern. The warning ratio can be configured according to the boundary type: For hard isolation boundaries, because their depth is relatively short, they need to be monitored earlier, so the warning ratio can be set to a smaller value, such as 10%; for warning boundaries, the warning ratio can be set to a moderate value, such as 20%; for open boundaries, the warning ratio can be set to a larger value, such as 30%, because their risk is lower.

[0052] Specifically, if multiple current concern boundaries are identified simultaneously, the boundary with the highest priority is selected as the final current concern boundary, according to the priority order of hard isolation boundary, warning boundary, and open boundary.

[0053] S43. Based on the direction of the line segment of the current boundary at the nearest point, take the direction perpendicular to the boundary direction and pointing inwards as the boundary normal direction. The inside of the boundary is the side facing the risk area bound to the boundary. The normal direction is chosen because the key to determining whether the target will invade the risk area inside the boundary is whether the target's movement direction has a component pointing inwards, rather than whether the target moves parallel to the boundary.

[0054] S44. Perform a vector dot product operation between the current target velocity vector direction and the normal direction of the current boundary of interest, and combine the magnitudes of the two vectors to calculate the angle between the target velocity vector direction and the boundary normal, which is denoted as the intrusion direction angle, and can quantify the intrusion intention.

[0055] When the intrusion direction angle is ≤90°: the projection of the target velocity vector onto the boundary normal is positive, indicating that the target's movement direction has a component pointing inwards towards the boundary, meaning the target is approaching the risk area. The smaller the angle, the more perpendicular the movement direction is to the boundary and points inwards, and the more direct the intrusion intention is. The larger the angle, the closer the movement direction is to being parallel to the boundary, the weaker the component pointing inwards, and the less obvious the intrusion intention is. When it is 90°, the target velocity vector is perpendicular to the normal, meaning the target's movement direction is parallel to the boundary and does not constitute an intrusion risk.

[0056] When the intrusion direction angle is >90°: the projection of the target velocity vector onto the normal direction is negative, indicating that the target's movement direction is away from the inside of the boundary, that is, the target is moving away from the risk area and there is no risk of intrusion.

[0057] S5. Determine the target intrusion risk based on the intrusion direction angle and its change over time.

[0058] Once the intrusion risk angles of the target and the boundary are determined, the potential intrusion risk posed by the target to the boundary of the work area can be assessed.

[0059] See Figure 2 As shown, in one embodiment of the present invention, the target intrusion risk determination is carried out as follows: Considering that the intrusion direction angle at a single moment is difficult to fully reflect the target's movement trend and is easily affected by instantaneous state fluctuations, dynamic change analysis is introduced to obtain the intrusion direction angle between the target velocity vector direction at the previous moment and the current boundary of concern, and to calculate the change in the intrusion direction angle at the current moment compared with the previous moment.

[0060] A target intrusion risk is determined when the following conditions are met simultaneously: a) The intrusion direction angle at the current moment is less than the allowable threshold, which reflects that the target's current movement direction has a component pointing inward to the boundary and is on a potential intrusion path. An example value of the allowable threshold is 90°.

[0061] b) The change in the intrusion direction angle at the current moment compared to the previous moment is negative, indicating that the intrusion direction angle is decreasing, reflecting that the target's movement direction is converging towards the boundary normal, and the intrusion intention is increasing.

[0062] The above-mentioned determination of target intrusion risk effectively filters out misjudgments caused by slight target swaying or measurement noise through dual verification of static angle threshold and dynamic angle change, thus improving the accuracy of risk determination.

[0063] S6. When a target intrusion risk is determined, the warning level is determined by combining the intrusion direction angle and the current concern boundary type, and the corresponding warning action is executed.

[0064] When determining that there is a risk of target intrusion, since the current boundary of concern has already been identified in the previous step S4 by the ratio of the shortest distance from the target to the boundary to the depth of the area, this ratio directly reflects the relative urgency of the risk area bound to the boundary by the target. Therefore, subsequent risk warnings do not need to consider the distance factor again, but should focus on the direction of the target's movement, i.e. the direction of intrusion.

[0065] The determination of target intrusion risk using the intrusion direction angle in step S5 is intended to confirm whether the target truly intends to intrude, rather than to classify warning levels. Once an intrusion risk is confirmed, a warning response needs to be initiated based on the intrusion direction. Furthermore, different types of boundaries bind risk areas with varying security sensitivities; therefore, a combined determination of intrusion direction and boundary type is necessary to achieve accurate, tiered warnings. The specific implementation is as follows: See [link / reference] Figure 3 As shown, firstly, the invasion direction is determined based on the current invasion direction angle: when the invasion direction angle is less than or equal to the positive threshold, such as 45°, it is determined to be a positive invasion, indicating that the target is moving inward almost perpendicular to the boundary, reflecting that the target has the most direct and strongest invasion intention.

[0066] When the intrusion direction angle is between the positive threshold and the allowable threshold, it is judged as a lateral intrusion. This indicates that the target's movement direction has an inward component, but also a movement component along the boundary direction, reflecting that the target may cross obliquely. The intrusion intention is relatively clear but not as direct as a positive intrusion.

[0067] If the target's historical trajectory indicates that it is moving from downstream to the open boundary, and the intrusion direction angle is less than the allowable threshold, then it is determined to be a reverse intrusion.

[0068] Then, the warning level is determined based on different boundaries.

[0069] i) The current boundary of concern is a hard isolation boundary.

[0070] The hard boundary directly encloses the cooperative's operational area, which is extremely short. Once intruded, other vehicles will immediately threaten the lives of the workers. Therefore, whether the intrusion is from the front or the side, it will trigger an emergency warning, and the actions will include: remotely activating the audible and visual warning devices of the crash buffer vehicle and issuing an emergency avoidance instruction to the upstream variable message sign.

[0071] ii) The current focus boundary is the warning boundary.

[0072] The risk area bound by the warning boundary is the warning transition zone. This area itself has a certain buffer space, and even after a vehicle intrudes, there is still a safe distance from the hard isolation boundary. Therefore, a tiered warning is issued based on the direction of intrusion: Forward intrusion: The target approaches the warning boundary perpendicularly, indicating a strong intent to intrude, requiring immediate high alert from personnel, thus triggering a warning. Actions include: sending a directional broadcast instructing the target vehicle to slow down and avoid it via the roadside unit, and issuing real-time warning information to the variable message sign upstream of the target vehicle's lane.

[0073] Lateral intrusion: The target approaches at an angle, and the intrusion intent is weaker than that of a frontal intrusion, but it is still necessary to remind the driver to correct the driving trajectory, so an information warning is triggered. The actions include: sending an activation signal to the roadside intelligent warning light strip, and displaying the position and speed information of the approaching target on the work team leader's handheld terminal.

[0074] iii) The current boundary of concern is an open boundary.

[0075] The risk area bound by the open boundary is the safe zone. If it is a reverse intrusion, it is very easy to have a head-on collision with the work vehicle or personnel. Therefore, an emergency warning is triggered. The actions are the same as the emergency warning in the hard isolation boundary scenario. At the same time, continuous tracking of reverse-moving vehicles is added to remind the work personnel to take emergency avoidance.

[0076] If it is a non-reverse intrusion, it means that the target is in a normal passage state and does not pose a threat to the work area, so no warning is triggered.

[0077] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0078] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0079] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0080] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0081] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent monitoring and early warning of electronic fences in highway maintenance work areas, characterized in that, include: The maintenance work area fence boundary is divided into hard isolation boundary, warning boundary and open boundary; Each boundary is bound to a corresponding risk area, and the distance from each boundary to the interior of its bound area is calculated as the area depth. The system acquires target distance and angle data detected at the maintenance site in real time, calculates the two-dimensional plane coordinates of the target, and obtains the target velocity vector direction based on coordinate difference calculations at multiple consecutive time points. The shortest distance from the target to each boundary is calculated based on the target's two-dimensional planar coordinates. The ratio of the shortest distance from the target to each boundary to the depth of the region bound by the boundary is used to identify the current boundary of interest. At the same time, the intrusion direction angle between the target's velocity vector direction and the normal of the current boundary of interest is obtained. Target intrusion risk is determined based on the intrusion direction angle and its change over time. When a target intrusion risk is determined, the intrusion direction determined by the intrusion direction angle and the current boundary of concern are combined to determine the warning level and execute the corresponding warning action.

2. The intelligent monitoring and early warning method for electronic fences in highway maintenance work areas as described in claim 1, characterized in that: The division of the maintenance work area fence boundary into hard isolation boundary, warning boundary and open boundary includes the following: The location of the crash buffer vehicle is obtained from the crash buffer vehicle positioning terminal as the starting anchor point of the hard isolation boundary, the position of the deployed cones is obtained from the cone array deployment trajectory as the intermediate anchor point, and the termination anchor point is obtained from the end of the cone array. Connecting the starting anchor point, intermediate anchor point, and ending anchor point in sequence generates a continuous polyline, forming a hard isolation boundary; The positions of the vertices of the warning lines drawn by the work team leader are obtained from the work team leader's handheld terminal, and the vertices are connected in sequence to generate a continuous polyline, which forms the warning boundary. The intersection of the straight line perpendicular to the driving direction at the starting anchor point of the hard isolation boundary and the road shoulders on both sides is taken as the two endpoints of the upstream open boundary. The upstream open boundary is generated by connecting the two endpoints. The intersection of the straight line perpendicular to the driving direction at the location of the hard isolation boundary termination anchor point and the road shoulders on both sides is taken as the two endpoints of the downstream open boundary. The downstream open boundary is generated by connecting the two endpoints.

3. The intelligent monitoring and early warning method for electronic fences in highway maintenance work areas as described in claim 1, characterized in that: The risk area bound to each boundary includes the following: The area enclosed by the hard isolation boundary is defined as the work activity area, and the work activity area is bound to the hard isolation boundary. The area between the upstream open boundary and the warning boundary is defined as the warning transition zone, with the upstream open boundary as the upstream side boundary, the warning boundary as the downstream side boundary, and the extension lines of the hard isolation boundaries on both sides as the side boundaries. The warning transition zone is then bound to the warning boundary. The downstream end section of the hard isolation boundary is taken as the upstream side boundary, the downstream open boundary is taken as the downstream side boundary, and the road shoulders on both sides or the extension line of the hard isolation boundary are taken as the side boundaries. The area between the downstream end of the hard isolation boundary and the downstream open boundary is defined as the safety zone, and the safety zone is bound to the downstream open boundary.

4. The intelligent monitoring and early warning method for electronic fences in highway maintenance work areas as described in claim 1, characterized in that: The calculation of the distance from each boundary to the interior of its bounded region as the region depth includes the following: For hard isolation boundaries, all anchor points on the hard isolation boundaries are selected as sampling points. The maximum Euclidean distance from each sampling point to any point inside the area it is bound to is calculated as the depth value corresponding to the sampling point. Then, the maximum value among all the depth values ​​of the sampling points is taken as the depth of the work activity area. For the warning boundary, all vertices on the warning boundary are selected as sampling points. The vertical distance from each sampling point to the upstream open boundary is calculated. The vertical distances of all sampling points are summed and divided by the number of sampling points to obtain the regional depth of the warning transition zone. For the downstream open boundary, two endpoints and the equally spaced points between the two endpoints are selected as sampling points. The shortest distance from each sampling point to the hard isolation boundary is calculated. The shortest distances of all sampling points are summed and divided by the number of sampling points to obtain the regional depth of the safe zone.

5. The intelligent monitoring and early warning method for electronic fences in highway maintenance work areas as described in claim 1, characterized in that: The real-time acquisition of target distance and angle data detected at the maintenance work site is as follows: The distance and angle data of the target are acquired in real time from the millimeter-wave radar deployed on the roadside. The distance data represents the straight-line distance between the target and the radar, and the angle data represents the azimuth angle of the target relative to the radar normal.

6. The intelligent monitoring and early warning method for electronic fences in highway maintenance work areas as described in claim 5, characterized in that: The calculation of the target's two-dimensional plane coordinates, and the obtaining of the target's velocity vector direction based on coordinate difference calculations at multiple consecutive time points, are described below: A two-dimensional plane coordinate system is established with the location of the starting anchor point of the hard isolation boundary as the origin, the direction of travel as the positive X-axis, and the direction perpendicular to the direction of travel and pointing outward from the work area as the positive Y-axis. The installation position of the millimeter-wave radar in the two-dimensional plane coordinate system is marked as a fixed coordinate. Based on the target distance and angle data acquired by the radar, the offset of the target relative to the radar is calculated using trigonometric functions. This offset is added to the fixed coordinate of the radar to obtain the horizontal and vertical coordinates of the target in the two-dimensional plane coordinate system. The velocity vector direction of the target is obtained by performing a difference operation on the two-dimensional plane coordinates of the target at multiple consecutive time points.

7. The intelligent monitoring and early warning method for electronic fences in highway maintenance work areas as described in claim 1, characterized in that: The process for identifying the current focus boundary is as follows: Based on the real-time calculated target coordinates and the location of each boundary, the Euclidean distance from the target to the nearest point on each boundary is calculated and recorded as the nearest distance. At the same time, the regional depth of the risk area bound to each boundary is obtained. Calculate the ratio of the nearest distance from the target to each boundary to the depth of the boundary region. If the ratio of the nearest distance from the target to a certain boundary to the depth of the boundary region is less than the warning ratio, then the boundary is recorded as the current boundary of concern.

8. The intelligent monitoring and early warning method for electronic fences in highway maintenance work areas as described in claim 1, characterized in that: The invasion direction angle is obtained through the following process: Based on the direction of the line segment at the nearest point of the current concern boundary, the direction perpendicular to the boundary direction and pointing inwards from the boundary is taken as the boundary normal direction, where the inside of the boundary is the side facing the risk area bound to the boundary; Perform a vector dot product operation between the current target velocity vector direction and the normal direction of the current boundary of interest, and combine the magnitudes of the two vectors to calculate the angle between the target velocity vector direction and the boundary normal, which is denoted as the intrusion direction angle.

9. The intelligent monitoring and early warning method for electronic fences in highway maintenance work areas as described in claim 1, characterized in that: The target intrusion risk assessment is implemented as follows: Obtain the angle between the target velocity vector direction at the previous moment and the intrusion direction angle of the current boundary of interest, and calculate the change in the intrusion direction angle at the current moment compared to the previous moment; A risk of target intrusion is determined to exist when all of the following conditions are met: a) The current intrusion direction angle is less than or equal to the allowable threshold; b) The change in the intrusion direction angle at the current moment compared to the previous moment is negative.

10. The intelligent monitoring and early warning method for electronic fences in highway maintenance work areas as described in claim 1, characterized in that: The determination of the warning level includes the following: Determine the invasion direction based on the magnitude of the current invasion direction angle: When the intrusion direction angle is less than or equal to the positive threshold, it is determined to be a positive intrusion. When the intrusion direction angle is between the positive threshold and the allowable threshold, it is determined to be a lateral intrusion. If the target is determined to be moving from downstream to the open boundary in reverse direction based on the target's historical trajectory, and the intrusion direction angle is less than the allowable threshold, it is determined to be a reverse intrusion. When the current boundary of concern is a hard isolation boundary, both forward and lateral intrusion will trigger an emergency warning; When the current boundary of concern is the warning boundary, if it is a forward intrusion, an alert will be triggered; if it is a lateral intrusion, an information alert will be triggered. When the current boundary of concern is an open boundary, an emergency warning will be triggered if it is a reverse intrusion, and no warning will be triggered if it is a non-reverse intrusion.

Images