A lock station passing method based on laser edge detection

Abstract

This invention discloses a method for vehicle passage through lock stations based on laser edge detection, relating to the field of terminal automation technology. To address the problem that existing technologies struggle with the complex environmental information of lock station areas, leading to inaccurate route planning and correction when vehicles pass through lock stations, potentially causing vehicle vibration and traffic congestion when lock station access points are compressed, thus affecting the overall operational efficiency of the terminal, this invention precisely determines the edge position of the lock station by collecting laser points at the lock station edges in stages. It also eliminates interference such as intermediate guardrails based on vehicle chassis data, ensuring data accuracy and reliability. Utilizing positioning weights and a multi-route fusion scheme, it automatically optimizes the system when positioning is unstable, significantly reducing vehicle vibration when passing through lock stations, ensuring transportation safety, reducing container handling hazards, and enabling vehicles to pass through lock stations quickly through precise detection and correction, thereby increasing container turnover speed and effectively improving the overall operational efficiency of automated terminals.

Description

Technical Field

[0001] This invention relates to the field of automated port operations technology, and in particular to a method for navigating through lock stations based on laser edge detection. Background Technology

[0002] In automated container terminals, quay lock stations, as key facilities for container unloading and unlocking operations, are typically located beside the quay crane driveway or at essential intersections. When high-volume operations are conducted using multiple lines, the access points for lock stations are compressed. Currently, vehicles face numerous challenges when passing through lock stations. Existing autonomous driving technologies for lock station passage typically employ pure satellite positioning, visual recognition, or laser recognition; however, traditional lock station passage solutions are inefficient and struggle to meet the ever-increasing operational demands of the terminal. With compressed lock station access points, traffic congestion is easily caused, reducing container throughput and consequently impacting the overall operational efficiency of the terminal. Furthermore, existing technologies are insufficient in handling the complex environmental information of the lock station area, such as accurately identifying lock station edges and eliminating interference from redundant fencing. This results in inaccurate route planning and correction for vehicles passing through lock stations, potentially leading to driving vibrations or safety issues. Summary of the Invention

[0003] The purpose of this invention is to provide a lock station passage method based on laser edge detection. By employing strategies such as staged fitting, relative position filtering, and route weight stitching, the method can accurately identify lock station edges, eliminate interference items, effectively improve lock station passage efficiency, reduce vehicle dwell time, reduce passage jitter, ensure driving safety, and significantly improve the operational efficiency of automated terminals, thereby solving the problems mentioned in the background art.

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

[0005] A method for traversing a locked station based on laser edge detection, comprising:

[0006] Determine whether a vehicle has entered the locking station area and activate the locking station edge detection module for detection. Perform LiDAR scanning on the locking station based on a segmented detection strategy to selectively remove redundant interference items in the middle segment.

[0007] A stitching scheme combining positioning weights and multi-route fusion is used for route correction, outputting standard-compliant station edge lines to guide vehicles in route correction operations.

[0008] Furthermore, determining whether a vehicle has entered the station's locking area specifically includes:

[0009] The vehicle's location information is obtained in real time using a positioning device, and the obtained vehicle location information is compared with the pre-set locking station area range;

[0010] When the vehicle's location information is completely within the pre-defined locking station area, it is determined that the vehicle has entered the locking station area;

[0011] When the vehicle's location information is detected to be partially or entirely outside the pre-set locking station area, it is determined that the vehicle has not entered the locking station area.

[0012] If, during the process of determining whether a vehicle has not entered the locking zone, the distance between the vehicle's location and the boundary of the locking zone continuously decreases within the continuous monitoring period, and the decrease exceeds a preset threshold, it is determined that the vehicle's location has a tendency to move closer to the locking zone. The vehicle's location information is continuously monitored until the vehicle's location information is compared with the range of the locking zone again to re-determine whether the vehicle has entered the locking zone.

[0013] Furthermore, the segmented detection strategy includes three stages: entering the lock station, passing through the lock station, and exiting the lock station.

[0014] During the entry into the locking station phase: when the positioning device determines that the vehicle is gradually approaching the locking station area and its location information shows that it is about to enter the pre-set locking station area, the laser points on the inner edge of the right locking station scanned by the left-side lidar of the vehicle are saved to the left point set for left-side locking station edge detection, and the laser points on the left locking station scanned by the right-side lidar of the vehicle are saved to the right point set for right-side locking station edge detection.

[0015] During the vehicle crossing the lock station phase: when it is determined that the vehicle is already within the lock station area, the laser points on the left side of the lock station inner edge scanned by the laser on the left side of the vehicle are saved to the left point set for left lock station edge detection, and the laser points on the right side of the lock station scanned by the laser on the right side are saved to the right point set for right lock station edge detection.

[0016] During the exiting the lock station phase: When the system detects that the vehicle is about to leave the lock station area, it continues to follow the operation method of passing through the lock station phase, saving the lock station laser points scanned by the left and right LiDAR to the corresponding left point set and right point set respectively.

[0017] Furthermore, during the crossing of the lock station, the process also includes: acquiring vehicle chassis data based on the sensors mounted on the vehicle; determining the distance range of the intermediate guardrail based on the vehicle chassis data; filtering interference points on the acquired lock station laser points based on the distance range of the intermediate guardrail; determining the relationship between the lock station laser points, the lock station position, and the distance range of the intermediate guardrail; and removing laser points determined to belong to the intermediate guardrail based on the relationship.

[0018] Furthermore, the lock station is scanned using lidar, specifically including:

[0019] Receive raw scanning data output from the lidar equipment, including the three-dimensional coordinates and intensity information of the locking laser point;

[0020] The acquired raw scan data is preprocessed, and filtering conditions are set according to the specific size of the vehicle and the actual layout of the locking station area.

[0021] The vehicle's driving status is determined based on the real-time vehicle location information, and the filtering criteria are dynamically adjusted.

[0022] Real-time condition judgment is performed on each locking station laser point. Based on preset and dynamically adjusted filtering conditions, the collected scanning data is filtered, and valid locking station laser points are obtained based on the filtering results.

[0023] Furthermore, the lidar scanning of the locking station also includes: classifying the valid locking station lidar points based on the vehicle's reference frame, and saving the valid locking station lidar points located on the vehicle to the corresponding left and right point sets based on a segmented detection strategy.

[0024] Furthermore, the effective locking laser points are classified, specifically including:

[0025] Read the valid locking laser points from the left and right point sets, and determine the data type of the valid locking laser points based on the data characteristics of the valid locking laser points;

[0026] Match the corresponding laser scan data sample in the lock station laser scan database according to the valid lock station laser point data type;

[0027] Based on laser scanning data samples, determine the reasonable fluctuation range and distribution pattern of effective locking laser points in spatial coordinates, and generate constraints for data verification.

[0028] Based on the generated constraints, each valid locking laser point in the left and right point sets is verified, and valid locking laser points that do not meet the constraints are marked as outliers.

[0029] If the number and deviation of abnormal points in the left and right point sets are lower than the preset thresholds, the abnormal points are corrected according to the distribution pattern of the surrounding normal points.

[0030] If the number or deviation of abnormal points in the left and right point sets exceeds a preset threshold, the abnormal points are removed and an abnormality warning is issued.

[0031] Furthermore, the operation of targeted removal of redundant interference items in the middle segment also includes:

[0032] When filtering interference points for the laser points on the left and right sides of the locking station, the vehicle is initially assumed to be in front of the locking station. The number of locking station laser points on the left and right sides that are more than 7.5 meters in front of the vehicle is counted.

[0033] If more than half of the locking laser points on both the left and right sides are more than 7.5 meters in front of the vehicle, then two specific signs will be marked to indicate that the vehicle is not on the corresponding side of the locking station.

[0034] Conversely, if more than half of the laser points on both the left and right sides of the locking station are less than 7.5 meters in front of the vehicle, the vehicle is determined to be in a locking station, and two specific markers are set to indicate the vehicle's status on the corresponding side of the locking station.

[0035] Furthermore, after targeted removal of redundant interference terms in the middle segment, it also includes:

[0036] The valid station laser clouds stored in the left point set and the right point set are processed to obtain two segmented point cloud sets, namely the left segmented point cloud set and the right segmented point cloud set.

[0037] Straight line fitting operations are performed on the left and right segmented point cloud sets respectively to obtain the slope and intercept of the straight lines at the locking edge on the left and right sides respectively.

[0038] The slope and intercept of the straight lines at the edges of the left and right locking stations are verified. Based on the verification results, it is determined whether the slope and intercept of the straight lines at the edges of the left and right locking stations meet the standards, and the locking station edge line parameters that meet the standards are output.

[0039] Furthermore, a stitching scheme combining positioning weights and multi-route fusion is adopted for route correction, specifically including:

[0040] Determine whether the real-time vehicle location information is within the locking station area. When the vehicle is within the locking station area, read the configuration information of the corresponding area, obtain the relative offset of the locking station laser edge line, and calculate the absolute value of the difference between the actual offset and the relative offset. At the same time, obtain the absolute value of the slope of the locking station edge line.

[0041] When the absolute value of the difference between the actual offset and the relative offset and the absolute value of the slope of the station edge line are both less than the preset threshold, the station edge line is confirmed to meet the correction conditions, the station edge line parameters at the current time are saved, and the station edge line identification status flag is set to valid; otherwise, route correction is not performed.

[0042] Determine whether the current vehicle positioning status is stable. If the vehicle positioning status is unstable, calculate the current bias weight and adjust the value of the bias weight according to the calculation result.

[0043] When the lock station edge line identification status flag is valid, the route weight splicing judgment is performed to determine whether the curvature of the current route meets the straight line threshold. If the curvature exceeds the preset straight line threshold, the route point number with curvature exceeding the straight line threshold is recorded.

[0044] Traverse all route points from the current route to the curvature exceedance point number, and read the corresponding correction weight parameters. Based on the correction weights, perform weighted calculations on the original route and the station edge line to generate the final corrected route.

[0045] Compared with the prior art, the beneficial effects of the present invention are:

[0046] By collecting laser points at the edge of the locking station in stages, the edge position of the locking station is accurately determined. Interference items such as the middle guardrail are eliminated based on vehicle chassis data to ensure data accuracy and reliability. By using positioning weights and multi-route fusion schemes, the route is adjusted in combination with various factors. When the positioning is unstable, it is automatically optimized, which greatly reduces the vibration of vehicles passing through the locking station, ensures transportation safety, reduces the hidden dangers of container sliding. Precise detection and correction enable vehicles to pass through the locking station quickly, avoid congestion in the locking station area, and improve the container turnover speed. In the case of multi-line operations and narrow passages at the terminal, it effectively improves the overall operational efficiency of the automated terminal and provides strong support for the efficient operation of the terminal. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the station-crossing process based on laser edge detection according to the present invention;

[0048] Figure 2 This is a schematic diagram of the lidar device of the present invention;

[0049] Figure 3 This is a schematic diagram illustrating the removal of redundant interference terms in the middle segment according to the present invention;

[0050] Figure 4 This is a schematic diagram of the splicing route for the splicing scheme of the present invention;

[0051] Figure 5 This is a flowchart of the station-locking passage method based on laser edge detection of the present invention. Detailed Implementation

[0052] 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.

[0053] To address the technical issues arising from the inaccuracy of existing technologies in processing complex environmental information in lockout areas, leading to inaccurate route planning and correction for vehicles traversing lockouts, potentially causing vehicle vibration, traffic congestion, reduced container throughput, and ultimately impacting overall terminal operational efficiency when lockout access points are compressed, please refer to [link to relevant documentation]. Figures 1-5This embodiment provides the following technical solution:

[0054] A method for traversing a locked station based on laser edge detection, comprising:

[0055] Determine whether a vehicle has entered the locking station area and activate the locking station edge detection module for detection. Perform LiDAR scanning on the locking station based on a segmented detection strategy to reduce the amount of computation and selectively remove redundant interference items in the middle segment.

[0056] A stitching scheme combining positioning weights and multi-route fusion is used for route correction, outputting standard-compliant station edge lines to guide vehicles in route correction operations.

[0057] In this embodiment, determining whether a vehicle has entered the station's locking area specifically includes:

[0058] The vehicle's location information is obtained in real time using a positioning device, and the obtained vehicle location information is compared with the pre-set locking station area range;

[0059] When the vehicle's location information is completely within the pre-defined locking station area, it is determined that the vehicle has entered the locking station area;

[0060] When the vehicle's location information is detected to be partially or entirely outside the pre-set locking station area, it is determined that the vehicle has not entered the locking station area.

[0061] If, during the process of determining whether a vehicle has not entered the locking zone, the distance between the vehicle's location and the boundary of the locking zone continuously decreases within the continuous monitoring period, and the decrease exceeds a preset threshold, it is determined that the vehicle's location has a tendency to move closer to the locking zone. The vehicle's location information is continuously monitored until the vehicle's location information is compared with the range of the locking zone again to re-determine whether the vehicle has entered the locking zone.

[0062] In this embodiment, the segmented detection strategy includes three stages: entering the lock station, passing through the lock station, and exiting the lock station.

[0063] During the locking station entry phase: When the positioning device determines that the vehicle is gradually approaching the locking station area and its position information shows that it is about to enter the pre-set locking station area, the laser points on the inner edge of the right locking station scanned by the left-side lidar of the vehicle (with their height compressed) are saved to the left point set for left-side locking station edge detection, and the laser points on the left locking station scanned by the right-side lidar of the vehicle (with their height compressed) are saved to the right point set for right-side locking station edge detection.

[0064] Vehicle chassis data, including vehicle speed, angular velocity, and steering information, is acquired using sensors mounted on the vehicle. The distance range of the median guardrail is determined based on the vehicle chassis data. This distance range is equal to the vehicle speed multiplied by one-fifth of a second. Interference points are filtered for the acquired locking station laser points based on the distance range of the median guardrail. The relationship between the locking station laser points, the locking station position, and the distance range of the median guardrail is determined. Laser points that are determined to belong to the median guardrail are removed based on the relationship.

[0065] During the vehicle crossing the lock station phase: when it is determined that the vehicle is already within the lock station area, the laser points on the left side of the lock station inner edge scanned by the laser on the left side of the vehicle are saved to the left point set for left lock station edge detection, and the laser points on the right side of the lock station scanned by the laser on the right side are saved to the right point set for right lock station edge detection.

[0066] During the exiting the lock station phase: When the system detects that the vehicle is about to leave the lock station area, it continues to follow the operation method of passing through the lock station phase, saving the lock station laser points scanned by the left and right LiDAR to the corresponding left point set and right point set respectively.

[0067] In this embodiment, the height of the laser points of the lock station detected by laser scanning is compressed.

[0068] In this embodiment, by accurately determining whether a vehicle enters the locking station area, a segmented detection strategy is employed. At different stages of the vehicle's entry, passage, and exit from the locking station, laser points at the locking station edge are collected using LiDAR, ensuring the integrity and accuracy of the locking station edge line. Precise identification of the locking station edge helps reduce the collision risk of vehicles passing through the locking station. When the vehicle's position information partially or completely exceeds the locking station area, by monitoring the trend of distance changes between the vehicle's position and the boundary of the locking station area, it is possible to predict whether the vehicle is approaching the locking station area, thereby achieving a more flexible and intelligent detection process.

[0069] In this embodiment, the locking station is scanned by lidar, specifically including:

[0070] Receive raw scanning data output from the lidar equipment, including the three-dimensional coordinates and intensity information of the locking laser point;

[0071] In this embodiment, three lidar devices are configured: one lidar installed at the front of the vehicle, which is responsible for scanning and detecting the area directly in front of the vehicle to obtain environmental information in front of the vehicle. During the process of the vehicle approaching, entering, and exiting the locking station, it provides data support for determining the position and shape of the locking station in front. Two lidars are installed on the sides of the vehicle to scan the area on the sides of the vehicle. They can detect the edge information of the locking stations on the left and right sides of the vehicle, including the internal edge of the locking station and information on obstacles that may exist on the side. Through the coordinated work of these three lidars, the vehicle can obtain more comprehensive point cloud data of the locking station and the surrounding environment.

[0072] The acquired raw scan data is preprocessed, including noise reduction, removal of outliers caused by equipment errors or environmental factors, time synchronization of the lock station laser points to ensure that all lock station laser point data correspond to the vehicle's position and status at the same time, and setting filtering conditions according to the specific size of the vehicle and the actual layout of the lock station area.

[0073] In this embodiment, the filtering conditions specifically include:

[0074] Points more than 25 meters in front of the vehicle are excluded as they are not very useful for guiding passage through the lock station.

[0075] Points located more than 2.5 meters from the side of the vehicle may be outside the designated area for vehicle locking.

[0076] LiDAR station-locking laser points within a height range of 0.2 meters to 1.0 meters, which may be located on the ground or low obstacles, as well as elevated structures or other unrelated obstacles;

[0077] Based on the real-time vehicle location information, the vehicle's driving status is determined, such as acceleration, deceleration, and turning. The filtering conditions are dynamically adjusted. For example, when the vehicle is accelerating, the filtering distance of the front locking laser point is adjusted according to the driving speed to ensure that the filtering conditions match the actual driving situation of the vehicle. When the vehicle is turning, the filtering range of the side locking laser point is appropriately expanded to avoid interference points outside the locking area from entering the subsequent processing process due to changes in vehicle posture.

[0078] Real-time condition judgment is performed on each locking station laser point. Based on the preset screening conditions and the dynamically adjusted screening conditions, the collected scanning data is filtered and the valid locking station laser points are obtained based on the screening results.

[0079] Based on the vehicle's reference frame, valid locking laser points are classified, and based on a segmented detection strategy, valid locking laser points located on the vehicle are saved to the corresponding left and right point sets.

[0080] In this embodiment, the vehicle's lidar collects omnidirectional point cloud data of the lock station and its surrounding environment, providing detailed information about the vehicle's front and side environments. This helps to more accurately determine the location and shape of the lock station. The filtering criteria are dynamically adjusted based on the vehicle's real-time location information and driving status, making the selection of lock station laser points more consistent with the vehicle's actual driving conditions. This enhances the system's adaptability and flexibility, improving not only the comprehensiveness and accuracy of environmental perception but also the quality of data processing through efficient data preprocessing and dynamic adaptability. Ultimately, this achieves accurate identification of lock station edges and precise vehicle guidance, significantly improving the safety and efficiency of vehicles passing through lock stations.

[0081] In this embodiment, the effective locking laser points are classified, specifically including:

[0082] Read the valid locking laser points from the left and right point sets, and determine the data type of the valid locking laser points based on the data characteristics of the valid locking laser points;

[0083] Based on the valid lock station laser point data type, match the corresponding laser scan data sample in the lock station laser scan database, such as normal lock station edge points, possible interference points (such as residual guardrail reflection points), etc.

[0084] Based on laser scanning data samples, determine the reasonable fluctuation range and distribution pattern of effective locking laser points in spatial coordinates (x, y, z directions), and generate constraints for data verification. Within a specific area, the variation range of the coordinates of the locking edge points in the horizontal direction (similar to the x direction) and the vertical direction (similar to the y direction) should meet certain standards. For suspected interference points, set corresponding exclusion rules, such as points with certain combinations of location and intensity characteristics being highly likely to be interference points.

[0085] Based on the generated constraints, each valid locking laser point in the left and right point sets is verified, and valid locking laser points that do not meet the constraints are marked as outliers.

[0086] If the number and deviation of abnormal points in the left and right point sets are lower than the preset thresholds, the abnormal points are corrected according to the distribution pattern of the surrounding normal points.

[0087] If the number or deviation of abnormal points in the left and right point sets exceeds a preset threshold, the abnormal points are removed and an abnormality warning is issued.

[0088] In this embodiment, the operation of directionally removing redundant interference items in the middle section, that is, filtering interference points of the obtained locking station laser points based on the distance range of the middle guardrail, further includes:

[0089] When filtering interference points for the laser points on the left and right sides of the locking station, the vehicle is initially assumed to be in front of the locking station. The number of locking station laser points on the left and right sides that are more than 7.5 meters in front of the vehicle is counted.

[0090] If more than half of the laser points on both sides of the locking station are more than 7.5 meters away from the front of the vehicle, then two specific markers will be used to indicate that the vehicle is not on the corresponding side of the locking station. This means that the vehicle may still be on its way to the locking station and has not yet actually entered the locking station area. These points that are far away from the front of the vehicle may be generated by the distant environment (rather than the locking station itself) and are not very meaningful for judging the edge of the locking station and the vehicle's position inside the locking station. They are more likely to be interference points.

[0091] Conversely, when the number of laser points on both sides of the locking station that are less than 7.5 meters in front of the vehicle exceeds half, the vehicle is determined to be in a locking station. Two specific markers are marked to indicate the vehicle's status on the corresponding side of the locking station, thus helping to identify interference points. When the vehicle is in a locking station, the closer laser points are more likely to be reflections of the locking station itself. At this time, those points that do not conform to the characteristics of the locking station (such as those located in the middle guardrail) are the interference points that need to be filtered out.

[0092] In this embodiment, as Figure 3 As shown, the relative position of the locking station is P, and the distance from the vehicle to the locking station is P. x The intermediate position is calculated as P based on the station locking parameters. x +L, P x The calculation is G x -v*t, G x Let P represent the vehicle's current position coordinates, v represent the vehicle's current speed, and t represent the time interval, i.e., the program's frame rate is 0.2 seconds. x =G x -0.2v.

[0093] In this embodiment, after directional removal of redundant interference items in the middle segment, the method further includes:

[0094] The valid locking station laser clouds stored in the left point set and the right point set are processed to obtain two segmented point cloud sets, namely the left segmented point cloud set and the right segmented point cloud set. A specific direction is set and the laser point clouds in the records of the left and right locking station related information lists are segmented to obtain two new segmented point cloud sets.

[0095] Line fitting operations are performed on the left and right segmented point cloud sets respectively to obtain the slope and intercept of the left and right locking station edge lines respectively. The slope and intercept of the left locking station edge line and the right locking station edge line are calculated by the corresponding fitting algorithm.

[0096] The slope and intercept of the straight lines at the left and right sides of the station are verified. Based on the verification results, it is determined whether the slope and intercept of the straight lines at the left and right sides of the station meet the standards. The station edge line parameters that meet the standards are output, namely the average intercept (i.e. (left straight line intercept + right straight line intercept) / 2) and the average slope (i.e. (left straight line slope + right straight line slope) / 2), which are used for subsequent route correction.

[0097] In this embodiment, when the slope difference between the left and right straight lines is within ±0.08 and the intercept difference between the left and right straight lines minus 3.5 meters (which can be adjusted according to the actual situation) is within ±0.4 meters, the straight line parameters of the locking station edge are deemed to meet the standard. If the conditions are not met, no correction operation is performed, and the self-driving vehicle drives normally according to satellite positioning.

[0098] In this embodiment, by processing data and marking outliers, and issuing warnings when the number of outliers or the degree of deviation exceeds a threshold, the impact of erroneous data on the system is effectively reduced, enhancing the system's robustness. By selectively removing redundant interference items in the middle section, unnecessary data processing volume is reduced, the data processing flow is optimized, and the system's operating efficiency is improved. By fitting straight lines to the laser point clouds of the left and right lock stations, accurate lock station edge straight line parameters are obtained, providing accurate data support for vehicle route correction and enhancing the safety of the vehicle during lock station passage and the overall performance of the autonomous driving system.

[0099] In this embodiment, a stitching scheme combining positioning weights and multi-route fusion is used for route correction, specifically including:

[0100] Determine whether the real-time vehicle location information is within the locking station area. When the vehicle is within the locking station area, read the configuration information of the corresponding area, obtain the relative offset of the locking station laser edge line, which is the distance between the vehicle's preset original route centerline and the locking station edge line, and calculate the absolute value of the difference between the actual offset and the relative offset. At the same time, obtain the absolute value of the slope of the locking station edge line.

[0101] When the absolute value of the difference between the actual offset and the relative offset and the absolute value of the slope of the station edge line are both less than the preset threshold, the station edge line is confirmed to meet the correction conditions, the station edge line parameters at the current time are saved, and the station edge line identification status flag is set to valid; otherwise, route correction is not performed.

[0102] In this embodiment, when the absolute value of the relative offset of the locking station laser edge line is less than 0.3 meters and the absolute value of the slope is less than 0.03, the corresponding value at the current moment is saved, and the flag indicating the locking station edge line recognition status is set to true. When the flag is true, it means that the vehicle is located within the locking station area and the deviation between the detected locking station edge line and the preset reference line is within an acceptable range. If it is not within this range, it means that the deviation from the actual value is too large. If forced to accept and correct it, it will cause huge vibrations in the vehicle, so no correction is made.

[0103] The system determines whether the current vehicle positioning status is stable. The positioning stability threshold is 0.05, which is set based on the experience value of the positioning system under stable road conditions. It is used to evaluate the stability of the positioning status. If the vehicle positioning status is unstable, the current bias weight is calculated and the value of the bias weight is adjusted according to the calculation result. The value range is (0, 1). The system determines whether it exceeds the set maximum value. If it exceeds the maximum value, the maximum value is taken and the bias weight parameter at the current time is saved.

[0104] When the lock station edge line identification status flag is valid, the route weight splicing judgment is performed to determine whether the curvature of the current route meets the straight line threshold. If the curvature exceeds the preset straight line threshold, the route point number with curvature exceeding the straight line threshold is recorded. The route curvature threshold is 0.004, which is set based on port route experience and is used to determine whether the route meets the straight line condition.

[0105] Traverse all route points from the current route to the curvature over-limit point number and read the corresponding correction weight parameters. Based on the correction weight, perform weighted calculations on the original route and the station edge line to generate the final corrected route. The splicing formula is: y-coordinate of the new route point = y-coordinate of the original route point * offset weight + (1 - offset weight) * y-coordinate of the original route point.

[0106] In this embodiment, a stitching scheme combining positioning weights and multi-route fusion is used for route correction. Retaining a certain portion of the route can greatly reduce the jitter caused by correction during travel. By accurately calculating the difference between the actual offset and the relative offset, and combining it with the slope of the lock station edge line, the accuracy of route correction is ensured. Setting identification status flags and stability thresholds avoids over-correction and improves stability. The weighted stitching formula achieves smooth route stitching, which improves the overall driving performance and safety of autonomous vehicles in the lock station area, provides drivers with a continuous and smooth driving experience, ensures smooth driving, and reduces the safety hazards of container trolleys.

[0107] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for traversing a locked station based on laser edge detection, characterized in that, include: Determine whether a vehicle has entered the locking station area and activate the locking station edge detection module for detection. Perform LiDAR scanning on the locking station based on a segmented detection strategy to selectively remove redundant interference items in the middle segment. A stitching scheme combining positioning weights and multi-route fusion is used for route correction, outputting standard-compliant station edge lines to guide vehicles in route correction operations. The corrective procedures specifically include: Determine whether the real-time vehicle location information is within the locking station area. When the vehicle is within the locking station area, read the configuration information of the corresponding area, obtain the relative offset of the locking station laser edge line, and calculate the absolute value of the difference between the actual offset and the relative offset. At the same time, obtain the absolute value of the slope of the locking station edge line. When the absolute value of the difference between the actual offset and the relative offset and the absolute value of the slope of the station edge line are both less than the preset threshold, the station edge line is confirmed to meet the correction conditions, the station edge line parameters at the current time are saved, and the station edge line identification status flag is set to valid; otherwise, route correction is not performed. Determine whether the current vehicle positioning status is stable. If the vehicle positioning status is unstable, calculate the current bias weight and adjust the value of the bias weight according to the calculation result. When the lock station edge line identification status flag is valid, the route weight splicing judgment is performed to determine whether the curvature of the current route meets the straight line threshold. If the curvature exceeds the preset straight line threshold, the route point number with curvature exceeding the straight line threshold is recorded. Traverse all route points from the current route to the curvature exceedance point number, and read the corresponding correction weight parameters. Based on the correction weights, perform weighted calculations on the original route and the station edge line to generate the final corrected route.

2. The method for traversing a locked station based on laser edge detection as described in claim 1, characterized in that, Determining whether a vehicle has entered the station's locking zone specifically includes: The vehicle's location information is obtained in real time using a positioning device, and the obtained vehicle location information is compared with the pre-set locking station area range; When the vehicle's location information is completely within the pre-defined locking station area, it is determined that the vehicle has entered the locking station area; When the vehicle's location information is detected to be partially or entirely outside the pre-set locking station area, it is determined that the vehicle has not entered the locking station area. If, during the process of determining whether a vehicle has not entered the locking zone, the distance between the vehicle's location and the boundary of the locking zone continuously decreases within the continuous monitoring period, and the decrease exceeds a preset threshold, it is determined that the vehicle's location has a tendency to move closer to the locking zone. The vehicle's location information is continuously monitored until the vehicle's location information is compared with the range of the locking zone again to re-determine whether the vehicle has entered the locking zone.

3. The method for traversing a locked station based on laser edge detection as described in claim 2, characterized in that, The segmented detection method includes: During the entry into the locking station phase: when the positioning device determines that the vehicle is gradually approaching the locking station area and its location information shows that it is about to enter the pre-set locking station area, the laser points on the inner edge of the right locking station scanned by the left-side lidar of the vehicle are saved to the left point set for left-side locking station edge detection, and the laser points on the left locking station scanned by the right-side lidar of the vehicle are saved to the right point set for right-side locking station edge detection. During the vehicle crossing the lock station phase: when it is determined that the vehicle is already within the lock station area, the laser points on the left side of the lock station inner edge scanned by the laser on the left side of the vehicle are saved to the left point set for left lock station edge detection, and the laser points on the right side of the lock station scanned by the laser on the right side are saved to the right point set for right lock station edge detection. During the exiting the lock station phase: When the system detects that the vehicle is about to leave the lock station area, it continues to follow the operation method of passing through the lock station phase, saving the lock station laser points scanned by the left and right LiDAR to the corresponding left point set and right point set respectively.

4. The method for traversing a locked station based on laser edge detection as described in claim 3, characterized in that, The process of passing through the lock station also includes: acquiring vehicle chassis data based on the sensors mounted on the vehicle; determining the distance range of the intermediate guardrail based on the vehicle chassis data; filtering interference points on the acquired lock station laser points based on the distance range of the intermediate guardrail; determining the relationship between the lock station laser points, the lock station position, and the distance range of the intermediate guardrail; and removing laser points determined to belong to the intermediate guardrail based on the relationship.

5. The method for traversing a locked station based on laser edge detection as described in claim 1, characterized in that, The lock station is scanned by lidar, specifically including: Receive raw scanning data output from the lidar equipment, including the three-dimensional coordinates and intensity information of the locking laser point; The acquired raw scan data is preprocessed, and filtering conditions are set according to the specific size of the vehicle and the actual layout of the locking station area. The filtering criteria are dynamically adjusted based on the vehicle's real-time location information to determine the vehicle's driving status. Real-time condition judgment is performed on each locking station laser point. Based on preset and dynamically adjusted filtering conditions, the collected scanning data is filtered, and valid locking station laser points are obtained based on the filtering results.

6. The method for traversing a locked station based on laser edge detection as described in claim 5, characterized in that, The lidar scanning of the locking station also includes: classifying the valid locking station lidar points based on the vehicle's reference frame, and saving the valid locking station lidar points located on the vehicle to the corresponding left and right point sets based on a segmented detection strategy.

7. The method for traversing a locked station based on laser edge detection as described in claim 6, characterized in that, The effective locking laser points are classified as follows: Read the valid locking laser points from the left and right point sets, and determine the data type of the valid locking laser points based on the data characteristics of the valid locking laser points; Match the corresponding laser scan data sample in the lock station laser scan database according to the valid lock station laser point data type; Based on laser scanning data samples, determine the reasonable fluctuation range and distribution pattern of effective locking laser points in spatial coordinates, and generate constraints for data verification. Based on the generated constraints, each valid locking laser point in the left and right point sets is verified, and valid locking laser points that do not meet the constraints are marked as outliers. If the number and deviation of abnormal points in the left and right point sets are lower than the preset thresholds, the abnormal points are corrected according to the distribution pattern of the surrounding normal points. If the number or deviation of abnormal points in the left and right point sets exceeds a preset threshold, the abnormal points are removed and an abnormality warning is issued.

8. The method for traversing a locked station based on laser edge detection as described in claim 4, characterized in that, The operation of targeted removal of redundant interference items in the middle segment also includes: When filtering interference points for the laser points on the left and right sides of the locking station, the vehicle is initially assumed to be in front of the locking station. The number of locking station laser points on the left and right sides that are more than 7.5 meters in front of the vehicle is counted. If more than half of the locking laser points on both the left and right sides are more than 7.5 meters in front of the vehicle, then two signs will be marked to indicate that the vehicle is not on the corresponding side of the locking station. Conversely, if more than half of the laser points on both the left and right sides of the vehicle lock station are less than 7.5 meters in front of the vehicle, the vehicle is determined to be in a lock station, and two markers are displayed to indicate the vehicle's status on the corresponding side of the lock station.

9. The method for traversing a locked station based on laser edge detection as described in claim 8, characterized in that, After targeted removal of redundant interference terms in the middle section, it also includes: The valid station laser clouds stored in the left point set and the right point set are processed to obtain two segmented point cloud sets, namely the left segmented point cloud set and the right segmented point cloud set. Straight line fitting operations are performed on the left and right segmented point cloud sets respectively to obtain the slope and intercept of the straight lines at the locking edge on the left and right sides respectively. The slope and intercept of the straight lines at the edges of the left and right locking stations are verified. Based on the verification results, it is determined whether the slope and intercept of the straight lines at the edges of the left and right locking stations meet the standards, and the locking station edge line parameters that meet the standards are output.

Images