A port route avoidance method based on multi-line laser radar

Abstract

A method for port route avoidance based on multi-line laser radar belongs to the field of container truck perception and positioning technology in port automatic driving. The method uses the GPS signal prediction algorithm to convert the point cloud data of the stack distribution information into the GPS signal strength distribution information of the road ahead, and decides whether the vehicle will take evasive action based on the strength of the GPS signal. This method avoids the problem that the existing methods only focus on processing complex scenes but cannot improve accuracy. It uses GPS signals to predict roads in advance and avoid roads that cannot be accurately controlled automatically, which significantly improves the port accident rate and vehicle efficiency of the port.

Description

Technical Field

[0001] The present invention relates to the field of container truck sensing and positioning technology in port autonomous driving, and in particular to a port route avoidance method based on multi-line laser radar. Background Art

[0002] Traditional outdoor positioning methods mainly use differential GPS to provide full high-precision initial positioning, and then achieve globally consistent high-precision positioning by integrating wheel speed meters and laser / visual odometer methods. However, differential GPS is easily affected by terrain. Common factors affecting GPS positioning accuracy include: multipath effect and occlusion. Due to the influence of production operations, the terrain of the port scene will change frequently. Therefore, it is an important part of the cognitive ability of container trucks in port autonomous driving to perceive and model the terrain around the driving route in advance.

[0003] The traditional multi-sensor fusion positioning method is mainly timely, that is, the container truck in the port automatic driving needs to achieve high-precision positioning through the Kalman filter at the same time as the GPS positioning state changes. However, in the port scenario, the scene changes caused by container transshipment, quay crane movement, and yard crane movement not only lead to a decrease in GPS positioning accuracy, but also a decrease in the accuracy of laser / visual odometers. Container trucks in the port automatic driving can usually only complete the lateral alignment in the lane through lane line recognition, and realize longitudinal alignment through auxiliary equipment such as reflective columns. According to the timely online fusion method, although it can cope with most scenarios, long-term testing has found that the reliability of the global positioning system decreases in sections with high container accumulation or under quay cranes / long bridges. In addition, the traditional method mainly focuses on solving how to deal with the impact of complex scenes, but does not consider avoiding such complex scenes. Although high-precision maps can annotate some static complex scenes through semantic editing, they cannot update scene changes in a timely manner. Therefore, the strategy mode hard-coded into the map often brings potential risks. Summary of the invention

[0004] In view of the shortcomings of the prior art, the object of the present invention is to provide a port route avoidance method based on multi-line laser radar.

[0005] The technical solution adopted by the invention is: a method for avoiding port routes based on multi-line laser radar, the technical key points of which are as follows:

[0006] Vehicles entering the container stacking area obtain point cloud data describing the distribution information of the containers around them in real time, and use the GPS signal prediction algorithm to convert the above point cloud data into the strength distribution information of the GPS signal on the road ahead, and decide whether the vehicle should take evasive action based on the strength of the GPS signal.

[0007] In the above solution, whether the vehicle enters the container yard area is determined based on the global high-precision map and the electronic fence of the container yard area.

[0008] In the above scheme, the stack box distribution information described by the point cloud data includes the number of stack box layers, thickness and the position of the field bridge in the field.

[0009] In the above scheme, the GPS signal prediction algorithm obtains the blocking angle indicating the strength of the GPS signal based on the blocking of the GPS signal on the top and both sides of the road where the stack boxes are located according to the stack box distribution information generated by the point cloud data.

[0010] In the above scheme, the calculation formula of the blocking angle is:

[0011]

[0012] Where P is a waypoint on the planned path, d1, d2, …, dm are the distances from the container bays of the same bay but different columns near point P to point P, L is the length of a single bay, h1, h2, …, hm represent the heights of the stacked containers in the bays, and the first occlusion angle of point P is calculated by traversing the stacking heights of containers in different columns, and the maximum calculated angle is taken as the occlusion angle of point P. The larger the occlusion angle, the greater the impact on the GPS signal and the lower the signal strength.

[0013] The beneficial effect of the present invention is that the port route avoidance method based on multi-line laser radar converts the point cloud data of the stack distribution information into the GPS signal strength distribution information of the road ahead using the GPS signal prediction algorithm, and decides whether the vehicle will take avoidance behavior based on the GPS signal strength. This method avoids the problem that the existing methods only focus on processing complex scenes and cannot improve accuracy. It uses GPS signals to predict roads in advance and avoids roads that cannot be accurately controlled automatically, so that the port accident rate and vehicle efficiency of the port are significantly improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0015] Figure 1 The present invention is a flow chart of the method for avoiding port routes based on multi-line laser radar. DETAILED DESCRIPTION

[0016] To make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the following is a brief description of the present invention in conjunction with the attached Figure 1 The present invention is further described in detail with reference to the accompanying drawings and specific embodiments.

[0017] The method for avoiding a port route based on a multi-line laser radar adopted in this embodiment includes the following steps:

[0018] Step 1: The autonomous driving container truck uses the global high-precision map and the electronic fence of the container yard to determine whether the vehicle has entered the container yard area;

[0019] Step 2: The container truck receives the surrounding point cloud data in real time in the autonomous driving system, and combines it with the electronic fence of the container stack area on the high-precision map to obtain the container stack distribution information described by the point cloud data, including the number of container stack layers, thickness, and the position of the yard bridge in the yard. The impact of the container stack on the strength of the GPS signal on the left and right sides of the road can be predicted based on the number of container stack layers and thickness, and the impact of the container stack on the strength of the GPS signal on the top of the road can be predicted based on the position of the yard bridge in the yard.

[0020] Step 3: The autonomous driving system converts the effective point cloud data into the GPS signal strength distribution information of the road ahead through the GPS signal prediction algorithm. The specific calculation formula is:

[0021]

[0022] Where P is a waypoint on the planned path, d1, d2, …, dm are the distances from the container bays of the same bay but different columns near point P to point P, L is the length of a single bay, h1, h2, …, hm represent the heights of the stacked containers in the bays. By traversing the stacking heights of containers in different columns, the angle of obstruction at point P is calculated, and the maximum value of the calculated angle is taken as the obstruction angle at point P. The larger the obstruction angle, the greater the impact on the GPS signal and the lower the signal strength.

[0023] Specifically, in this embodiment, the standard value of the container height is 2.6 meters, the width is 2.6 meters, and the length is 6 meters. The container is placed 2.4 meters away from the vehicle's driving track. Assuming that there are three rows of containers near a point P on the driving track, of which the first row close to the driving track has 4 containers, the second row has 4 containers, and the third row has 6 containers. According to the scoring calculation formula, the score of the first row is 77.04, the score of the second row is 29.235, and the score of the third row is 27.67. Therefore, it is concluded that the first row has the greatest impact on the GPS signal, so the impact of point P is worth 77.04.

[0024] Step 4: The autonomous driving decision module executes effective and safe decision-making behaviors based on the prediction results of the GPS signal of the road section ahead, including slowing down, reporting alarm information of weak GPS signal on the road section, changing lanes or re-planning the local route, and adjusting the positioning weight.

[0025] Specifically, in this embodiment, when the vehicle is in a lane between barriers and the lane detection result is normal, and the borrowed lane does not meet the lane change conditions, the vehicle slows down; when the GPS signal is predicted to be weak, an alarm signal is reported; when the vehicle is in a lane between barriers and there is a borrowed lane on the left side of the vehicle, combined with the target detection result, when the lane change conditions are met, the lane change strategy is preferentially adopted. The global positioning system is mainly composed of a differential GPS system and a laser SLAM system, in which the differential GPS is easily affected by occlusion, and the laser SLAM positioning accuracy is mainly affected by scene changes. In particular, in the port scene, the operation process will cause scene changes, which will lead to a decrease in the positioning accuracy of the laser SLAM system. The scene change perception system based on multi-line laser radar first gives a GPS signal occlusion estimation. When the GPS signal occlusion is small, the weight of the GPS positioning result is increased, and the GPS positioning result is used first. When the GPS signal occlusion is large, the GPS positioning result is considered unreliable, and the laser SLAM system positioning result is used first. At this time, according to the scene change estimation result given by the scene change perception system, when the scene changes greatly, the laser SLAM system positioning result is considered unreliable. At this time, the system positioning solution is downgraded from a global positioning system to a relative positioning system based on lane centering, and an early warning message of reduced global positioning accuracy is reported at the same time.

[0026] Step 5: The container truck in the autonomous driving system updates and shares the GPS signal strength distribution information with other autonomous driving vehicles.

[0027] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for avoiding port routes based on multi-line laser radar, characterized in that: The following steps are involved: Vehicles entering the container stacking area obtain point cloud data describing the distribution information of the containers around them in real time, and use the GPS signal prediction algorithm to convert the above point cloud data into the strength distribution information of the GPS signal on the road ahead, and decide whether the vehicle should take evasive action based on the strength of the GPS signal; The GPS signal prediction algorithm obtains the blocking angle indicating the strength of the GPS signal based on the blocking of the GPS signal on the top and both sides of the road where the stacking boxes are located according to the stacking box distribution information generated by the point cloud data; The calculation formula of the occlusion angle is: Where P is a waypoint on the planned path, d1, d2, …, dm are the distances from the container bays of the same bay but different columns near point P to point P, L is the length of a single bay, h1, h2, …, hm represent the heights of the stacked containers in the bays. By traversing the stacking heights of containers in different columns, the first occlusion angle of point P is calculated, and the maximum calculated angle is taken as the occlusion angle of point P. The larger the occlusion angle, the greater the impact on the GPS signal and the lower the signal strength.

2. The method for avoiding a port route based on a multi-line laser radar as claimed in claim 1, characterized in that: Determine whether the vehicle has entered the container yard area based on the global high-precision map and the electronic fence of the container yard.

3. The method for avoiding a port route based on a multi-line laser radar as claimed in claim 1, characterized in that: The stack box distribution information described by the point cloud data includes the number of stack box layers, thickness and the position of the field bridge in the field.

Images