Anti-collision method, system and electronic equipment of all-weather ship loader

Abstract

The application discloses a collision avoidance method and system of an all-weather ship loader and electronic equipment, and relates to the field of collision detection. The application obtains point cloud data of the ship loader by adopting a millimeter wave radar, and establishes a three-dimensional space coordinate system based on cycle information of the ship loader. The point cloud data is unified in the three-dimensional space coordinate system through coordinate conversion and coordinate information. After normalization and clustering operations are respectively performed on the point cloud data, the required categories are obtained. The point cloud data corresponding to the categories matching the shape of the ship cabin is unified in the three-dimensional space coordinate system, and the ship cabin boundary is efficiently and accurately determined. A minimum bounding box is constructed according to the size of the roller and the platform where the roller is located, the minimum bounding box and the ship cabin boundary are projected onto the xy plane and the xz plane by using a projection method, a region of interest is set, and whether the projection of the minimum bounding box and the projection of the ship cabin boundary in the region of interest intersect or overlap is judged, so that the purpose of detecting and avoiding collision is quickly and accurately achieved.

Description

[0001] TECHNICAL FIELD

[0002] The present application relates to the field of collision detection, in particular to a collision prevention method and system for an all-weather ship loader and an electronic device. BACKGROUND

[0003] Coal is one of the main fossil energy sources and is widely used in the power, steel, and construction industries. At present, in order to meet market demand, a large number of cargo ships are waiting to load coal at northern ports every day. However, the low efficiency of coal loading operations has become an important problem restricting the development of the coal transportation industry, which not only leads to a shortage of coal supply, but also increases transportation costs and environmental pollution.

[0004] For coal loading operations, the ship loader is an important device, and once a collision accident occurs, it will cause serious damage and impact to goods, personnel, and equipment. Therefore, how to improve the anti-collision capability of the ship loader and improve efficiency under the premise of safety has become a problem to be solved. In the early days, manual operation was mainly used for safety protection of the ship loader, but due to frequent and complex ship loader operations, manual operation is prone to omissions, leading to frequent safety accidents. Therefore, advanced technologies are used to control the ship loader to improve its safety. The sensors commonly used in the collision prevention method of the ship loader today include cameras and laser radars. For fog, rain, and snow weather that may occur during coal transportation at sea ports and dust conditions that often occur during the operation process, optical devices and laser radars will be limited and cannot accurately and quickly realize collision detection of the ship loader, thereby bringing safety hazards to the working process of the ship loader. SUMMARY

[0005] To solve the above problems existing in the prior art, the present application provides an all-weather ship loader collision prevention method, system, and electronic device.

[0006] To achieve the above purpose, the present application provides the following solutions:

[0007] An all-weather ship loader collision prevention method, comprising:

[0008] Obtaining point cloud data of four directions of the ship loader during walking, and preprocessing the point cloud data; the point cloud data of four directions of the ship loader during walking is obtained by a millimeter wave radar;

[0009] Obtaining information of each collection period of the ship loader during walking and a large arm pitch angle of the ship loader; the information of each collection period includes coordinate information of an initial position of the ship loader and an installation position of the millimeter wave radar;

[0010] The information of each collection cycle and the luffing angle of the boom of the ship loader determine a coordinate origin, the sea side in the vertical direction of the ship loader is taken as the positive direction of the y-axis, and the sea side in the horizontal direction of the ship loader is taken as the positive direction of the x-axis, to establish a three-dimensional space coordinate system;

[0011] According to the running direction of the ship loader, the preprocessed point cloud data is classified, the classified point cloud data and the coordinate information of the ship loader are combined and unified in the three-dimensional space coordinate system; the running direction of the ship loader includes forward and backward running and left and right running;

[0012] The point cloud data unified in the three-dimensional space coordinate system is normalized to obtain a normalization result;

[0013] The DBSCAN algorithm is used to cluster the normalization result to obtain a clustering result;

[0014] Based on the clustering result, a category matching the shape of the ship cabin is determined, the point cloud data corresponding to the category matching the shape of the ship cabin is unified in the three-dimensional space coordinate system, and a ship cabin boundary is established;

[0015] According to the size of the chute and the platform where the chute is located, a minimum bounding box is constructed, the minimum bounding box and the ship cabin boundary are projected onto the xy plane and the xz plane by using the projection method, a region of interest is set, and it is judged whether the projection of the minimum bounding box and the projection of the ship cabin boundary in the region of interest intersect or overlap, to obtain a judgment result;

[0016] If the judgment result is yes, it is determined that there is a collision risk between the ship loader and the ship cabin, and a collision prompt information is generated;

[0017] If the judgment result is no, it is determined that there is no collision risk between the ship loader and the ship cabin.

[0018] Optionally, the origin of the three-dimensional space coordinate system is (x0, y0*cos(β-90°), z0); wherein (x0, y0) is the coordinate information of the initial position of the ship loader, z0 is the installation position of the millimeter wave radar, and β is the luffing angle of the boom of the ship loader.

[0019] Optionally, according to the running direction of the ship loader, the preprocessed point cloud data is classified, the classified point cloud data and the coordinate information of the ship loader are combined and unified in the three-dimensional space coordinate system, and specifically includes:

[0020] According to the running direction of the ship loader, the preprocessed point cloud data is classified, the classified point cloud data and the coordinate information of the ship loader are combined and unified in the three-dimensional space coordinate system, and specifically includes:

[0021] Map the front and back side point cloud data from the radar coordinate system to the YZ plane in the three-dimensional space coordinate system;

[0022] Map the left and right side point cloud data from the radar coordinate system to the XZ plane in the three-dimensional space coordinate system.

[0023] Optionally, the method further comprises:

[0024] After obtaining the clustering result, the average y value of each class of the front and back side point cloud data and the average x value of each class are determined.

[0025] According to the average y value and the average x value, the class corresponding to each side of the cabin is determined.

[0026] According to the installation positions of the millimeter wave radars in the four directions, corresponding label values are added to the point cloud data and unified in the same coordinate system.

[0027] After clustering the point cloud data on the y axis and the point cloud data on the x axis in the same coordinate system by using the DBSCAN algorithm, the class matching the shape of the cabin is selected again.

[0028] Optionally, the preprocessing comprises straight-through filtering and standard deviation filtering.

[0029] Optionally, the preprocessing of the point cloud data comprises:

[0030] A window is determined, and in the window, the mean value of the X value and the mean value of the Y value in the point cloud data are determined, and the standard deviation corresponding to the X value and the standard deviation corresponding to the Y value are determined.

[0031] For each point in the point cloud data, the difference between the X value and the mean value of the X value and the difference between the Y value and the mean value of the Y value are determined.

[0032] It is judged whether the difference between the X value and the mean value of the X value is less than 1 times the standard deviation corresponding to the X value, and whether the difference between the Y value and the mean value of the Y value is less than 1 times the standard deviation corresponding to the Y value.

[0033] If the difference between the X value and the mean value of the X value is less than 1 times the standard deviation corresponding to the X value, and the difference between the Y value and the mean value of the Y value is less than 1 times the standard deviation corresponding to the Y value, the point is retained.

[0034] If the difference between the X value and the mean value of the X value is greater than or equal to 1 times the standard deviation corresponding to the X value, or the difference between the Y value and the mean value of the Y value is greater than or equal to 1 times the standard deviation corresponding to the Y value, the point is deleted.

[0035] According to the specific embodiments of the present application, the following technical effects are achieved:

[0036] The application obtains point cloud data of the ship loader by adopting the millimeter wave radar, and establishes a three-dimensional space coordinate system based on cycle information of the ship loader. The point cloud data is unified in the three-dimensional space coordinate system through coordinate conversion and coordinate information. After normalization and clustering operations are respectively performed on the point cloud data, the required categories are obtained. The point cloud data corresponding to the categories matching the shape of the ship cabin is unified in the three-dimensional space coordinate system, and the ship cabin boundary is efficiently and accurately determined. According to the size of the roller and the platform where the roller is located, a minimum bounding box is constructed, the minimum bounding box and the ship cabin boundary are projected onto the xy plane and the xz plane by using the projection method, the region of interest is set, and whether the projection of the minimum bounding box and the projection of the ship cabin boundary in the region of interest intersect or overlap is judged, so that the purpose of detecting anti-collision is quickly and accurately achieved. In addition, using the millimeter wave radar sensor as the detection device can avoid the influence of fog, rain and snow weather and industrial dust, and can detect all day long, which has strong stability.

[0037] Further, the application provides a full-time ship loader anti-collision system, which is used to implement the full-time ship loader anti-collision method provided above; the system comprises:

[0038] A point cloud data acquisition module is configured to acquire point cloud data of four directions of the ship loader in a walking process, and to pre-process the point cloud data; the point cloud data of the four directions of the ship loader in the walking process is acquired by a millimeter wave radar;

[0039] An information pitch angle acquisition module is configured to acquire information of each acquisition cycle of the ship loader in the walking process and a pitch angle of a large arm of the ship loader; the information of each acquisition cycle comprises coordinate information of an initial position of the ship loader and an installation position of the millimeter wave radar;

[0040] A space coordinate system construction module is configured to determine a coordinate origin according to the information of each acquisition cycle and the pitch angle of the large arm of the ship loader, to take a sea-facing side of a vertical direction of the ship loader as a positive direction of a y-axis, to take a sea-facing side of a horizontal direction of the ship loader as a positive direction of an x-axis, and to establish a three-dimensional space coordinate system;

[0041] A point cloud data mapping module is configured to classify the pre-processed point cloud data according to a walking direction of the ship loader, to combine and unify the classified point cloud data and coordinate information of the ship loader in the three-dimensional space coordinate system; the walking direction of the ship loader comprises forward and backward walking and left and right walking;

[0042] A point cloud data normalization module is configured to perform normalization processing on the point cloud data unified in the three-dimensional space coordinate system to obtain a normalization result;

[0043] A clustering processing module is configured to perform clustering on the normalization result by using a DBSCAN algorithm to obtain a clustering result.

[0044] a cabin boundary construction module, configured to determine a category matching the shape of the cabin based on the clustering result, unify point cloud data corresponding to the category matching the shape of the cabin into a three-dimensional spatial coordinate system, and establish a cabin boundary;

[0045] a collision judgment result, configured to construct a minimum bounding box according to the size of the chute and the platform where the chute is located, project the minimum bounding box and the cabin boundary onto xy and xz planes by using a projection method, set a region of interest, and judge whether the projection of the minimum bounding box and the projection of the cabin boundary in the region of interest intersect or overlap to obtain a judgment result; if the judgment result is yes, it is determined that there is a collision risk between the ship loader and the cabin, and collision prompt information is generated; if the judgment result is no, it is determined that there is no collision risk between the ship loader and the cabin.

[0046] Further, the present application further provides an electronic device, which comprises:

[0047] a memory, configured to store a computer program;

[0048] a processor, connected with the memory, configured to call and execute the computer program to implement the anti-collision method of the all-weather ship loader provided above.

[0049] The technical effects realized by the above-mentioned system and electronic device provided by the present application are the same as those realized by the anti-collision method of the all-weather ship loader provided by the present application, and therefore will not be described here in detail. BRIEF DESCRIPTION OF DRAWINGS

[0050] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments will be briefly introduced below. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0051] Figure 1 a flowchart of the anti-collision method of the all-weather ship loader provided by the present application;

[0052] Figure 2 an implementation flowchart of the anti-collision method of the all-weather ship loader provided by the embodiment of the present application;

[0053] Figure 3 a schematic diagram of the point cloud data after processing and clustering on the front side of the ship loader provided by the embodiment of the present application; wherein, Figure 3 (a) of the schematic diagram of the point cloud data after processing and clustering on the front side of the ship loader, Figure 3(b) is a three-dimensional schematic view of the point cloud data after processing and clustering on the front side of the ship loader;

[0054] Figure 4 The figure is a schematic view of the point cloud data after processing and clustering on the rear side of the ship loader provided by the embodiment of the application, wherein Figure 4 (a) is a plane schematic view of the point cloud data after processing and clustering on the rear side of the ship loader, Figure 4 (b) is a three-dimensional schematic view of the point cloud data after processing and clustering on the rear side of the ship loader;

[0055] Figure 5 The figure is a schematic view of the point cloud data after processing and clustering on the rear side of the ship loader provided by the embodiment of the application, wherein Figure 5 (a) is a plane schematic view of the point cloud data after processing and clustering on the left side of the ship loader, Figure 5 (b) is a three-dimensional schematic view of the point cloud data after processing and clustering on the left side of the ship loader;

[0056] Figure 6 The figure is a schematic view of the point cloud data after processing and clustering on the right side of the ship loader provided by the embodiment of the application, wherein Figure 6 (a) is a plane schematic view of the point cloud data after processing and clustering on the right side of the ship loader, Figure 6 (b) is a three-dimensional schematic view of the point cloud data after processing and clustering on the right side of the ship loader;

[0057] Figure 7 The figure is a point cloud graph of a required category selected by the embodiment of the application;

[0058] Figure 8 The figure is a point cloud graph after re-clustering provided by the embodiment of the application;

[0059] Figure 9 The figure is a schematic view of a ship cabin boundary and a chute and a platform thereof constructed by the embodiment of the application;

[0060] Figure 10 The figure is a schematic view of an image and an xy projection of a region of interest of a ship cabin boundary and a chute and a platform thereof constructed by the embodiment of the application;

[0061] Figure 11 The figure is a schematic view of an image and an xz projection of a region of interest of a ship cabin boundary and a chute and a platform thereof constructed by the embodiment of the application. DETAILED DESCRIPTION

[0062] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described, obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work belong to the scope of protection of the present application.

[0063] The present application aims to provide a collision avoidance method, system and electronic device for all-weather ship loaders

[0064] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0065] As Figure 1 shown, the present application provides a collision avoidance method for all-weather ship loaders, which comprises:

[0066] Step 100: Obtain point cloud data of four directions of the ship loader in the walking process, and pre-process the point cloud data. The point cloud data of four directions of the ship loader in the walking process is obtained by millimeter wave radar.

[0067] In actual application process, the method for pre-processing the point cloud data can be straight-through filtering and standard deviation filtering. According to the length and width of the surrounding cabin during loading operation, the RCS value of the millimeter wave radar collection point, the useless scanning point is removed by using the straight-through filtering mode. By using the statistical filtering based on mean and standard deviation, the outlier points are removed. Among them, the outlier points refer to the points far away from the operation environment, or the points outside the distance range selected in actual operation.

[0068] Based on the above description, the outlier removal process is:

[0069] Specifically, the following steps are included:

[0070] (1) According to the characteristics of the obtained point cloud data, a window is determined.

[0071] (2) In the determined window, the mean values of the X and Y values of the coordinate values in the point cloud data are calculated, as well as the corresponding standard deviations.

[0072] (3) For each point in the point cloud data, check whether the difference between its X and Y values and the corresponding mean values is less than 1 times the standard deviation.

[0073] (4) If it is less than, it is considered that the point is a normal point, which is retained. Otherwise, it is filtered out.

[0074] Repeat steps (2) and (3) to process all points in the point cloud data.

[0075] Step 101: Obtain the information of each collection cycle of the ship loader during walking and the boom pitch angle of the ship loader. The information of each collection cycle includes the coordinate information of the initial position of the ship loader and the installation position of the millimeter wave radar.

[0076] Step 102: The information of each collection cycle and the boom pitch angle of the ship loader determine the coordinate origin, take the sea side of the vertical direction of the ship loader as the positive direction of the y axis, take the sea side of the horizontal direction of the ship loader as the positive direction of the x axis, and establish a three-dimensional space coordinate system.

[0077] In actual application, the information (x0, y0) of the initial position of the ship loader and the boom pitch angle β are recorded, z0 is set as the installation position of the millimeter wave radar, and the value is 0. (x0, y0*(β-90°), z0) is taken as the coordinate origin (0, 0, 0), the sea side of the vertical direction of the ship loader is taken as the positive direction of the y axis, and the sea side of the horizontal direction is taken as the positive direction of the x axis. The mapping relationship of the relative displacement of each collection cycle of the ship loader is:

[0078]

[0079] In the formula, (x2, y2) represents the coordinate value of the ship loader after moving each collection cycle. (x1, y1) represents the information of each collection cycle.

[0080] Step 103: According to the walking direction of the ship loader, the preprocessed point cloud data is classified, the classified point cloud data and the coordinate information of the ship loader are combined and unified in the three-dimensional space coordinate system. The walking direction of the ship loader includes forward and backward walking and left and right walking.

[0081] In actual application, in order to facilitate modeling, the preprocessed point cloud data is divided into two categories according to the forward and backward walking and the forward and backward walking of the ship loader, the radar data during forward and backward walking (i.e. the front and back side point cloud data) is selected as the left and right side radar, and the radar data during left and right walking (i.e. the left and right side point cloud data) is selected as the front and back side. After combining the original point cloud data with the information of the ship loader obtained according to the displacement and stretching, it is unified in the three-dimensional space coordinate system obtained in step 102. For example, the front and back side point cloud data is projected onto the YZ plane of the three-dimensional space coordinate system, and the left and right side point cloud data is projected onto the XZ plane of the three-dimensional space coordinate system.

[0082] Wherein, each side radar is divided into positive direction and oblique direction, the oblique cutting angle is α, and the mapping relationship of the radar point cloud data of the two installation angles on the left side from the radar coordinate system (x, y) to the three-dimensional space coordinate system is:

[0083]

[0084] In the formula, (x 11 , y11 , y 11 ) is the point cloud data of the radar with left side vertical installation mapped to the three-dimensional space coordinate system, (x 12 , y 12 , z 12 ) is the point cloud data of the radar with left side oblique installation mapped to the three-dimensional space coordinate system.

[0085] The mapping relationship of the radar point cloud data of the two installation angles on the right side from the radar coordinate system (x, y) to the three-dimensional space coordinate system is:

[0086]

[0087] In the formula, (x 21 , y 21 , z 21 ) is the point cloud data of the radar with right side vertical installation mapped to the three-dimensional space coordinate system, (x 22 , y 22 , z 22 ) is the point cloud data of the radar with right side oblique installation mapped to the three-dimensional space coordinate system.

[0088] The mapping relationship of the radar point cloud data of the two installation angles on the front side from the radar coordinate system (x, y) to the three-dimensional space coordinate system is:

[0089]

[0090] In the formula, (x 31 , y 31 , z 31 ) is the point cloud data of the radar with front side vertical installation mapped to the three-dimensional space coordinate system, (x 32 , y 32 , z 32 ) is the point cloud data of the radar with front side oblique installation mapped to the three-dimensional space coordinate system.

[0091] The mapping relationship of the radar point cloud data of the two installation angles on the back side from the radar coordinate system (x, y) to the three-dimensional space coordinate system is:

[0092]

[0093] In the formula, (x 41 , y 41 , z 41 ) is the point cloud data of the radar with back side vertical installation mapped to the three-dimensional space coordinate system, (x 42 , y 42 , z 42 ) is the point cloud data of the radar with back side oblique installation mapped to the three-dimensional space coordinate system.

[0094] Step 104: Normalizing the point cloud data unified into the three-dimensional space coordinate system to obtain a normalization result. The implementation process of this step can be:

[0095] (1) Find the maximum and minimum values of all coordinates in the point cloud data.

[0096] (2) Normalize each coordinate, the formula is:

[0097] Normalized coordinate = (original coordinate - minimum value) / (maximum value - minimum value).

[0098] Step 105: Clustering the normalization result using the DBSCAN algorithm to obtain a clustering result.

[0099] In actual application, the specific implementation process of the DBSCAN algorithm used is as follows:

[0100] (1) Determine a reasonable neighborhood radius ε and minimum sample number MinPts.

[0101] (2) Calculate the neighborhood of a point. For each point p (denoted as a point whose distance from p is less than the neighborhood radius), calculate the point set Nε(p) in its ε neighborhood. The neighborhood calculation formula is as follows:

[0102] Nε(p) = {q | dist(p, q) ≤ ε}

[0103] In the formula, dist(p, q) is the distance between points p and q.

[0104] (3) For each point p, if |Nε(p)| ≥ MinPts, mark point p as a core point.

[0105] (4) For each core point p, all points in its neighborhood are part of the same cluster. That is, point q in the neighborhood of point p is added to the cluster. For non-core points that do not belong to any cluster, mark them as noise points.

[0106] Step 106: Determine the category matching the shape of the cabin based on the clustering result, unify the point cloud data corresponding to the category matching the shape of the cabin into the three-dimensional space coordinate system, and establish the cabin boundary.

[0107] In actual application, according to the actual shape of the cabin, select the required category after clustering, and use the DBSCAN algorithm to cluster the y-axis of the front and rear point clouds and the x-axis of the left and right point clouds to filter out noise points. Select the point cloud data of the required category and unify it into the three-dimensional space coordinate system. Among them, clustering the y-axis of the front and rear point clouds is to extract the cross section in the y-axis, and clustering the x-axis of the left and right point clouds is to extract the cross section in the x-axis.

[0108] Step 107: Construct a minimum bounding box according to the size of the roller and the platform where the roller is located, project the minimum bounding box and the cabin boundary onto the xy plane and the xz plane using the projection method, set the region of interest, and determine whether the projections of the minimum bounding box and the cabin boundary in the region of interest intersect or overlap, to obtain a determination result.

[0109] In actual application, the four-direction boundaries of the cabin are constructed according to the point cloud data obtained in step 106, the minimum bounding box is constructed according to the specific size of the roller and its platform, the above-constructed bounding box and boundary are projected onto the xy plane and the xz plane using the projection method, the region of interest is set, and whether the projection line segments intersect or overlap is checked according to the coordinate values of the projection line segments.

[0110] Step 108: If the determination result is yes, it is determined that there is a collision risk between the ship loader and the cabin, and a collision prompt information is generated.

[0111] Step 109: If the determination result is no, it is determined that there is no collision risk between the ship loader and the cabin.

[0112] The following provides an embodiment to describe the specific implementation process of the anti-collision method of the all-weather ship loader provided above.

[0113] In this embodiment, as shown in Figure 2 the implementation process is as follows:

[0114] Step one, collect millimeter wave radar point cloud data in four directions during the cross walking of the coal ship loader, and pre-process the data.

[0115] In this embodiment, a millimeter wave radar (such as a 77GHz millimeter wave radar) is used as the main detection device, and the millimeter wave radar is installed in four directions, front, back, left and right, on the platform of the ship loader, and the radar in each direction is divided into forward and oblique directions. Compared with the method of acquiring data by laser radar and camera, the millimeter wave radar has its own unique advantages, such as wide frequency band, short wavelength, small size, low power consumption and strong penetration, etc. Compared with laser and infrared measurement, it can work all-weather, avoid the influence of fog, rain and snow weather and industrial dust on the imaging effect of scanning, and is less affected by the natural environment and the working environment.

[0116] Since the length and width of most ship cabins are less than 20 meters during loading operation. And according to the RCS value of the scanning point required during the operation process. Through the straight-through filtering method, the x-axis range of the millimeter wave radar in each direction is limited to [0, -12] meters, the y-axis range is [-12, 12] meters, and the scanning point RCS is [-20, 20] dbm 2In this way, the scanning points of the irrelevant area are removed. And by using the statistical filtering based on the mean and standard deviation, some outliers are removed.

[0117] Step two, collect the information of the ship loader in each collection period during walking and the pitch angle of the ship loader boom, and establish a three-dimensional space coordinate system.

[0118] Since the ship loader needs to adjust the pitch angle of the boom according to the height of the ship during loading, the pitch angle β needs to be recorded. The information of the ship loader in each collection period during walking includes two: one is the distance value relative to the farthest end of the sea survey, denoted as x1, and the other is the extension distance value of the ship loader boom, denoted as y1.

[0119] According to the collected information, the information of the initial position of the ship loader (x0, y0) is recorded, and z0 is set as the installation position of the millimeter wave radar, which is 0. (x0, y0*cos(β-90°), z0) is taken as the coordinate origin (0, 0, 0), and the vertical direction of the ship loader is taken as the positive direction of the y-axis and the horizontal direction as the positive direction of the x-axis. The mapping relationship of the relative displacement of each collection period of the ship loader is described above.

[0120] Step three, the preprocessed data is divided into two categories according to the forward and backward walking of the ship loader, and the radar data during forward and backward walking is selected as the left and right side radar, and the radar data during left and right walking is selected as the front and rear side. After combining the original point cloud data with the coordinate information of the ship loader, it is unified in the coordinate system.

[0121] The cross walking of the ship loader is divided into two processes: one is the horizontal movement of the ship loader, and the other is the extension movement of the boom of the ship loader. In order to reduce the amount of data and facilitate modeling around the ship loader, the radar data of the left and right sides during the extension movement of the boom of the ship loader is selected. The radar data of the front and rear sides during the horizontal movement of the ship loader is selected. The radar data is combined with the horizontal movement and extension distance of the ship loader in the corresponding period. And the four around the ship loader is modeled by using the combined radar data. The mapping relationship of the radar point cloud data of each side from the radar coordinate system to the three-dimensional space coordinate system is described above.

[0122] Step four, project the spliced point cloud of the front and rear sides to the YZ plane, and project the spliced point cloud of the left and right sides to the XZ plane. Normalize the data respectively, and use DBSCAN algorithm for clustering after normalization.

[0123] The selected data corresponds to the two translational movements of the ship loader, and the spliced point cloud is projected to the corresponding plane. By normalizing, the scale difference between the point cloud data is eliminated, which facilitates the clustering processing of the point cloud data. Clustering is performed by using DBSCAN algorithm.

[0124] For the two categories of point cloud data features are different, the front and rear side point cloud data selects the neighborhood radius ε = 0.05 and the minimum sample number MinPts = 10, and the left and right side point cloud data selects the neighborhood radius ε = 0.03 and the minimum sample number MinPts = 10. After obtaining the classification diagram of the four directions, the classified point cloud label is assigned to the original point cloud data, and the classification result is observed from the three-dimensional perspective, which is convenient for subsequent selection of the required category. As shown in Figures 3 to 6 Figure 3 Figure 4 In

[0125] Step five, after selecting the required category after clustering, the y-axis of the front and rear side point cloud is clustered by DBSCAN algorithm, and the x-axis of the left and right side point cloud is clustered, and the noise points are filtered out. The point cloud data of the selected category is unified into a three-dimensional coordinate system.

[0126] After obtaining the clustered point cloud data, the average y value of each category of the front and rear side is calculated, and the average x value of each category of the front and rear side is calculated. According to the result, the required category of each side is selected. According to the position of the millimeter wave radar installed in the four directions, the point cloud is given a value respectively and unified in a coordinate system, as shown in Figure 7 Figure 8 The y-axis and x-axis are clustered respectively by using DBSCAN algorithm. When using DBSCAN algorithm on a single axis, the data on the axis needs to be converted into a column vector. The neighborhood radius of the front and rear side and the left and right side is ε = 0.03 and the minimum sample number MinPts = 10. After clustering each side, the required category is selected again, so as to determine the boundary, as shown in

[0127] Step six, according to the point cloud data obtained in step five, the boundaries of the four directions of the cabin are constructed, the minimum bounding box is constructed based on the specific size of the roller and its platform, the projection method is used to project the object to the xy plane and the xz plane, the region of interest is set, and it is checked whether the projection line segment intersects or overlaps.

[0128] In this embodiment, the specific size of the roller and its platform is about 5.9 meters left and right, 4.0 meters front and rear, and 5 meters high. According to the specific size, the corresponding three-dimensional minimum bounding box is constructed. The three-dimensional minimum bounding box is a smallest cube that can tightly enclose a given three-dimensional object. The bounding box and the cabin boundary are placed in a unified coordinate system, as shown in Figure 9 Figure 9 ​​​​The middle dotted line part is the minimum bounding box of the chute and its platform, and the solid line part is the cabin boundary. The formed cabin boundary and the minimum bounding box of the chute and its platform are projected onto the xy plane and the xz plane by using the projection method. A black dash-dot line region of interest is set on the xy plane, and the length and height of the region of interest are reduced by 2 meters compared with the cabin boundary, as shown in FIG. 2. Figure 10 Three dash-dot line segments are set on the xz plane, and the length of the left and right two line segments is reduced by 2 meters compared with the left and right boundaries of the cabin, and there is a middle one, as shown in FIG. 3. Figure 11 If intersection or overlap occurs in the two projection maps, it is considered that a collision will occur, and a prompt will be output to prompt the driver. Through this method, the purpose of preventing collision can be achieved.

[0129] Further, the present application provides a collision prevention system for all-weather ship loader. The system is used to implement the all-weather ship loader collision prevention method provided above. The system comprises:

[0130] A point cloud data acquisition module is configured to acquire point cloud data in four directions of the ship loader during walking and to pre-process the point cloud data. The point cloud data in four directions of the ship loader during walking is collected by a millimeter wave radar.

[0131] An information pitch angle acquisition module is configured to acquire information of each collection period of the ship loader and a pitch angle of a large arm of the ship loader. The information of each collection period includes coordinate information of an initial position of the ship loader and an installation position of the millimeter wave radar.

[0132] A space coordinate system construction module is configured to determine a coordinate origin based on the information of each collection period and the pitch angle of the large arm of the ship loader, to take a sea side of a vertical direction of the ship loader as a positive direction of a y axis, to take a sea side of a horizontal direction of the ship loader as a positive direction of an x axis, and to establish a three-dimensional space coordinate system.

[0133] A point cloud data mapping module is configured to classify the pre-processed point cloud data according to a walking direction of the ship loader, to combine and unify the classified point cloud data and coordinate information of the ship loader in the three-dimensional space coordinate system. The walking direction of the ship loader includes forward and backward walking and left and right walking.

[0134] A point cloud data normalization module is configured to perform normalization processing on the point cloud data unified in the three-dimensional space coordinate system to obtain a normalization result.

[0135] A clustering processing module is configured to perform clustering on the normalization result by using a DBSCAN algorithm to obtain a clustering result.

[0136] The cabin boundary construction module is used for determining a category matching the cabin shape based on the clustering result, unifying the point cloud data corresponding to the category matching the cabin shape into a three-dimensional space coordinate system, and establishing the cabin boundary.

[0137] The collision judgment result is used for constructing a minimum bounding box according to the size of the roller and the platform where the roller is located, projecting the minimum bounding box and the cabin boundary onto the xy plane and the xz plane by using a projection method, setting a region of interest, and judging whether the projection of the minimum bounding box and the projection of the cabin boundary in the region of interest intersect or overlap, to obtain a judgment result. If the judgment result is yes, it is determined that there is a collision risk between the ship loader and the cabin, and collision prompt information is generated. If the judgment result is no, it is determined that there is no collision risk between the ship loader and the cabin.

[0138] Further, the present application further provides an electronic device, which comprises:

[0139] The memory is used for storing the computer program.

[0140] The processor is connected with the memory, and is used for calling and executing the computer program to implement the anti-collision method of the all-weather ship loader.

[0141] In addition, the computer program in the memory can be stored in a computer readable storage medium in the form of a software functional unit and sold or used as an independent product. Based on such understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The foregoing storage medium includes a U disk, a mobile hard disk, a read-only memory, a random access memory, a magnetic disk or an optical disk, and various media that can store program codes.

[0142] The embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same or similar parts of each embodiment can be referred to each other. For the system disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method part.

[0143] The principles and implementation manners of the present application are described by using specific examples in the present application, and the above examples are only used to help understand the method of the present application and its core idea; meanwhile, for the general technical personnel in the art, the specific implementation manners and application ranges will be changed according to the idea of the present application. In conclusion, the content of the present specification should not be understood as the limitation of the present application.

Claims

1. A collision avoidance method for an all-weather ship loader, characterized in that, include: The point cloud data of the ship loader in four directions during its movement is acquired, and the point cloud data is preprocessed. The point cloud data in four directions of the ship loader during its movement is collected by millimeter-wave radar; Acquire information for each data acquisition cycle during the movement of the ship loader and the pitch angle of the ship loader's boom; the information for each data acquisition cycle includes the coordinates of the ship loader's initial position and the installation position of the millimeter-wave radar; The information from each acquisition cycle and the pitch angle of the ship loader's boom determine the origin of the coordinate system. The seaward side of the ship loader's vertical direction is taken as the positive y-axis, and the seaward side of the ship loader's horizontal direction is taken as the positive x-axis, thus establishing a three-dimensional spatial coordinate system. According to the traveling direction of the ship loader, the preprocessed point cloud data is classified, and the classified point cloud data and the coordinate information of the ship loader are combined and unified in the three-dimensional spatial coordinate system; the traveling direction of the ship loader includes forward and backward travel and left and right travel. The point cloud data unified to the three-dimensional spatial coordinate system is normalized to obtain the normalized result; The normalization results are clustered using the DBSCAN algorithm to obtain the clustering results; Based on the clustering results, the categories that match the shape of the cabin are determined, the point cloud data corresponding to the categories that match the shape of the cabin are unified into a three-dimensional spatial coordinate system, and the cabin boundary is established. Construct a minimum bounding box based on the dimensions of the chute and the platform on which the chute is located. Use the projection method to project the minimum bounding box and the cabin boundary onto the xy plane and xz plane. Set the region of interest and determine whether the projection of the minimum bounding box and the projection of the cabin boundary in the region of interest intersect or overlap, and obtain the judgment result. If the judgment result is yes, then it is determined that there is a risk of collision between the ship loader and the ship's hold, and a collision warning message is generated; If the judgment result is negative, it is determined that there is no risk of collision between the ship loader and the ship's hold.

2. The anti-collision method for all-weather ship loaders according to claim 1, characterized in that, The origin of the three-dimensional spatial coordinate system is: (x0, y0*cos(β-90°), z0); where (x0, y0) are the coordinate information of the initial position of the ship loader, z0 is the installation position of the millimeter-wave radar, and β is the pitch angle of the ship loader's boom.

3. The anti-collision method for all-weather ship loaders according to claim 1, characterized in that, According to the traveling direction of the ship loader, the preprocessed point cloud data is classified. The classified point cloud data and the coordinate information of the ship loader are combined and unified in the three-dimensional spatial coordinate system. Specifically, this includes: According to the direction of travel of the ship loader, the preprocessed point cloud data is divided into front and rear point cloud data and left and right point cloud data; the front and rear point cloud data is acquired by the millimeter-wave radar on the left and right sides of the ship loader, and the left and right point cloud data is acquired by the millimeter-wave radar on the left and right sides of the ship loader. Map the front and rear point cloud data from the radar coordinate system to the YZ plane in the three-dimensional spatial coordinate system; The left and right side point cloud data are mapped from the radar coordinate system to the XZ plane in the three-dimensional spatial coordinate system.

4. The anti-collision method for all-weather ship loaders according to claim 3, characterized in that, Based on the clustering results, categories matching the shape of the ship's cabin are determined, specifically including: After obtaining the clustering results, the average y-value and average x-value of each category in the front and rear point cloud data are determined. The category corresponding to each side of the cabin is determined based on the average y-value and the average x-value; The point cloud data are labeled with corresponding tags based on the installation locations of the millimeter-wave radars in the four directions and then unified in the same coordinate system. After clustering the point cloud data along the y-axis and the point cloud data along the x-axis in the same coordinate system using the DBSCAN algorithm, the category that matches the shape of the ship cabin is selected again.

5. The anti-collision method for all-weather ship loaders according to claim 1, characterized in that, The preprocessing includes pass-through filtering and standard deviation filtering.

6. The anti-collision method for all-weather ship loaders according to claim 1, characterized in that, Preprocessing the point cloud data includes: Define a window, and within the window, determine the mean of the X value and the mean of the Y value in the point cloud data, and determine the standard deviation corresponding to the X value and the standard deviation corresponding to the Y value; For each point in the point cloud data, determine the difference between the X value and the mean of the X values, and the difference between the Y value and the mean of the Y values; Determine whether the difference between the X value and the mean of the X values ​​is less than 1 times the standard deviation corresponding to the X value, and determine whether the difference between the Y value and the mean of the Y values ​​is less than 1 times the standard deviation corresponding to the Y value; If the difference between the X value and the mean of the X values ​​is less than 1 times the standard deviation corresponding to the X value, and the difference between the Y value and the mean of the Y values ​​is less than 1 times the standard deviation corresponding to the Y value, then retain that point; If the difference between the X value and the mean of the X values ​​is greater than or equal to 1 times the standard deviation corresponding to the X value, or the difference between the Y value and the mean of the Y values ​​is greater than or equal to 1 times the standard deviation corresponding to the Y value, then delete that point.

7. A collision avoidance system for an all-weather ship loader, characterized in that, The system is used to implement the collision avoidance method for all-weather ship loaders as described in any one of claims 1-6; the system includes: The point cloud data acquisition module is used to acquire point cloud data in four directions during the movement of the ship loader and to preprocess the point cloud data; the point cloud data in four directions during the movement of the ship loader is acquired by millimeter-wave radar. The information pitch angle acquisition module is used to acquire information and the pitch angle of the boom of the ship loader during each acquisition cycle during the movement of the ship loader; the information in each acquisition cycle includes the coordinate information of the initial position of the ship loader and the installation position of the millimeter-wave radar. The spatial coordinate system construction module is used to determine the coordinate origin based on the information from each acquisition cycle and the pitch angle of the ship loader's boom. The vertical seaward side of the ship loader is taken as the positive y-axis, and the horizontal seaward side of the ship loader is taken as the positive x-axis, thus establishing a three-dimensional spatial coordinate system. The point cloud data mapping module is used to classify the preprocessed point cloud data according to the traveling direction of the ship loader, and combine the classified point cloud data with the coordinate information of the ship loader and unify them in the three-dimensional spatial coordinate system; the traveling direction of the ship loader includes forward and backward travel and left and right travel. The point cloud data normalization module is used to normalize point cloud data unified to the three-dimensional spatial coordinate system to obtain normalization results. The clustering processing module is used to cluster the normalization results using the DBSCAN algorithm to obtain clustering results. The cabin boundary construction module is used to determine the category that matches the cabin shape based on the clustering results, unify the point cloud data corresponding to the category that matches the cabin shape into a three-dimensional spatial coordinate system, and establish the cabin boundary. The collision assessment result is used to construct a minimum bounding box based on the dimensions of the chute and the platform on which the chute is located. The minimum bounding box and the ship's cabin boundary are projected onto the xy and xz planes using a projection method. A region of interest is set, and it is determined whether the projection of the minimum bounding box in the region of interest intersects or overlaps with the projection of the ship's cabin boundary to obtain the assessment result. If the assessment result is yes, it is determined that there is a collision risk between the ship loader and the ship's cabin, and a collision warning message is generated. If the assessment result is no, it is determined that there is no collision risk between the ship loader and the ship's cabin.

8. An electronic device, characterized in that, Its features include: Memory, used to store computer programs; A processor, connected to the memory, is configured to retrieve and execute the computer program to implement the collision avoidance method for all-weather ship loaders as described in any one of claims 1-6.

Images