Method and device for positioning a gantry crane, automated gantry crane and storage medium

Abstract

The present application relates to the technical field of navigation positioning, and discloses a gantry crane positioning method and device, an automatic gantry crane and a storage medium, wherein the three-dimensional point cloud data of the refrigerated shelves on the left and right sides of the traveling channel of the gantry crane is acquired under a positioning scenario; the point cloud plane positions of the refrigerated shelves on the left and right sides corresponding to the side close to the traveling channel are calculated respectively based on the three-dimensional point cloud data of the refrigerated shelves on the left and right sides; and the positioning position of the gantry crane is determined based on the point cloud plane positions of the refrigerated shelves on the left and right sides corresponding to the side close to the traveling channel. In this way, the laying cost of infrastructure is reduced, the positioning accuracy of the gantry crane is increased, and the operation safety is improved.

Description

Technical Field

[0001] This invention relates to the field of navigation and positioning technology, specifically to a method, device, automated yard crane, and storage medium for positioning large vehicles on yard cranes. Background Technology

[0002] In automated refrigerated container yards, precise positioning of the yard crane is a prerequisite for the smooth implementation of core operational functions such as automatic bottom opening, automatic container stacking, collision avoidance, and automatic power connection / disconnection. Currently, magnetic nails and GPS technologies are mainly used for trolley positioning. However, current solutions generally suffer from problems in practical applications, including high infrastructure costs, susceptibility to environmental interference, and positioning errors. Summary of the Invention

[0003] This invention provides a method for locating large trolleys in a yard crane, in order to solve the problems of high infrastructure laying costs, signal susceptibility to environmental interference, and positioning errors of large trolleys in yard cranes.

[0004] In a first aspect, the present invention provides a method for positioning a trolley on a yard crane, applicable to a yard scenario with refrigerated racks. A movable trolley is mounted on the crossbeam of the yard crane, and at least one lidar is installed at the bottom of the trolley. The method includes: Acquire 3D point cloud data of the refrigerated racks on both sides of the travel channel of the large gantry crane in the positioning scenario; Based on the three-dimensional point cloud data of the left and right refrigerated shelves, the point cloud plane positions of the left and right refrigerated shelves corresponding to the side closest to the passageway are calculated respectively. Based on the point cloud plane positions of the left and right refrigerated racks corresponding to the sides near the travel channel, the positioning position of the yard crane trolley is determined.

[0005] In this embodiment of the method, on the one hand, by acquiring the three-dimensional point cloud data of the refrigerated racks on both sides of the travel channel, and utilizing the relatively fixed structural features of the refrigerated racks, the point cloud plane of the refrigerated racks is used as a positioning reference to provide high-precision positioning information. This helps the yard crane trolley to achieve accurate stopping and operation within the travel channel, reducing safety accidents such as collisions caused by positioning errors and improving operational safety. On the other hand, by acquiring the three-dimensional point cloud data of the refrigerated racks and calculating the positioning position of the yard crane trolley, the impact of external infrastructure failures or poor signal on the positioning of the yard crane trolley can be reduced, thereby lowering construction and maintenance costs.

[0006] In one optional implementation, based on the three-dimensional point cloud data of the left and right refrigerated shelves, the point cloud planar positions of the left and right refrigerated shelves corresponding to the side closest to the travel aisle are calculated, including: Preprocessing the 3D point cloud data yields the target feature point cloud corresponding to the cold storage rack columns on the left and right sides; Based on the target feature point cloud corresponding to the left refrigeration rack column, calculate the first point cloud plane position of the left refrigeration rack near the travel aisle. Based on the target feature point cloud corresponding to the right-side refrigeration rack column, calculate the second point cloud plane position of the right-side refrigeration rack near the travel aisle.

[0007] In this embodiment of the method, the planar position of the point cloud of the left and right refrigerated racks near the travel channel is calculated based on the target feature point cloud, which reduces the deviation in planar position calculation caused by data errors and can more accurately determine the actual position of the yard crane truck.

[0008] In one optional implementation, the three-dimensional point cloud data is preprocessed to obtain the target feature point cloud corresponding to the left and right refrigeration rack columns, including: The 3D point cloud data is filtered and denoised to obtain the first processed feature point cloud. The first processed feature point cloud is filtered based on a preset height threshold to obtain the second processed feature point cloud. Clustering and segmentation are performed on the second-processed feature point cloud to obtain the target feature point cloud corresponding to the cold storage rack columns on the left and right sides.

[0009] In this method embodiment, on the one hand, isolated noise points are removed by statistical filtering, eliminating the risk that noise points will form their own clusters or interfere with normal clustering during subsequent clustering; on the other hand, ground point clouds are removed by height filtering, which greatly reduces the amount of data involved in clustering calculation and improves the calculation efficiency. In addition, by preprocessing the collected three-dimensional point cloud data, the accuracy of locating the large vehicle of the yard bridge using the three-dimensional point cloud data can be improved.

[0010] In one optional implementation, the positioning position of the yard crane is determined based on the point cloud plane positions of the left and right refrigerated racks corresponding to the sides near the travel aisle, including: Determine the average coordinates of the first position coordinates corresponding to the first point cloud plane position and the second position coordinates corresponding to the second point cloud plane position; The position corresponding to the mean coordinate is determined as the positioning position of the yard crane trolley.

[0011] In this embodiment of the method, by taking the average coordinates, the positioning instability caused by fluctuations in unilateral data can be reduced, thereby improving the positioning accuracy of the yard crane trolley.

[0012] In one optional implementation, when the target feature point cloud corresponding to the left or right refrigerated shelf column is less than a preset threshold number, based on the target feature point cloud corresponding to the left or right refrigerated shelf column, the first point cloud plane position or the second point cloud plane position corresponding to the left or right refrigerated shelf near the travel aisle is calculated, including: Joint registration is performed on multiple target feature point clouds, and the planar position of the first point cloud or the planar position of the second point cloud is determined by the preset geometric constraints of the cold storage rack.

[0013] In this embodiment of the method, when some pillars are obscured by goods, resulting in a reduction in the number of target feature point clouds, joint registration can integrate the point cloud information of other unobscured parts, thereby improving the accuracy and reliability of positioning.

[0014] In one alternative implementation, the method further includes: Based on location, the system controls the handling and release of refrigerated containers.

[0015] In this embodiment of the method, the target refrigerated container can be quickly reached based on the precise positioning location, reducing operation time and improving operation safety.

[0016] In one alternative implementation, the method further includes: In response to the received positioning signal from the refrigerated rack column, the yard crane truck is determined to have entered the positioning scene.

[0017] In this embodiment of the method, when the yard crane truck travels near the refrigerated rack and receives the positioning signal from the refrigerated rack column, it can accurately determine that the yard crane truck has entered a suitable scene for positioning. The scene truck is then controlled to use LiDAR to acquire three-dimensional point cloud data to locate the yard crane truck, further ensuring the accuracy and effectiveness of the positioning data.

[0018] Secondly, the present invention provides a positioning device for a yard crane trolley, wherein a movable trolley is mounted on the crossbeam of the yard crane, and at least one lidar is installed at the bottom of the trolley. The device includes: The acquisition module is used to acquire the three-dimensional point cloud data of the refrigerated racks on the left and right sides of the travel channel of the gantry crane in the positioning scenario; The calculation module is used to calculate the point cloud plane position of the left and right refrigerated shelves corresponding to the side closer to the passageway, based on the three-dimensional point cloud data of the left and right refrigerated shelves. The positioning module is used to determine the location of the yard crane truck based on the point cloud plane positions of the left and right refrigerated racks corresponding to the sides closest to the travel channel.

[0019] Thirdly, the present invention provides an automated yard crane, applicable to a yard scenario with refrigerated racks. A movable trolley is mounted on the crossbeam of the yard crane, and at least one lidar is installed at the bottom of the trolley. The yard crane also includes a controller, which includes a memory and a processor. The memory and the processor are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the yard crane trolley positioning method of the first aspect or any corresponding embodiment.

[0020] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to perform the gantry crane positioning method of the first aspect or any corresponding embodiment described above. Attached Figure Description

[0021] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0022] Figure 1 This is a schematic flowchart of the first method for positioning a yard crane trolley according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the second process of the positioning method for the yard crane trolley according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the regional structure of the refrigeration rack column according to an embodiment of the present invention; Figure 4 This is a structural block diagram of the yard crane trolley positioning device according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the hardware structure of the field bridge controller according to an embodiment of the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, 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.

[0024] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0025] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0026] In automated refrigerated container yards, precise positioning of the yard trolleys is a prerequisite for achieving functions such as automatic bottom opening, automatic container stacking, collision avoidance, and automatic power connection / disconnection. Currently, trolley positioning mainly relies on the following methods: (a) Magnetic nail positioning. Magnetic nails are buried at certain intervals on the ground of the storage yard. The position of the trolley is determined by detecting the magnetic nail signals through magnetic sensors installed on the yard bridge.

[0027] (ii) GPS / Differential GPS Positioning. Large vehicle positioning is achieved through satellite signals, with differential base stations providing correction data to improve accuracy.

[0028] (III) Calibration plate in conjunction with laser rangefinder. A reflective calibration plate is installed on the track of the main vehicle, and a laser rangefinder is installed on the bottom of the main vehicle. The position of the main vehicle is calculated by measuring the distance between the main vehicle and the calibration plate.

[0029] The above solutions each have the following problems: (I) Problems with magnetic nail positioning. ① The installation of magnetic nails requires large-scale excavation of the stockyard ground, resulting in high installation costs and long cycles, and once installed, they are difficult to adjust; ② Magnetic nails are prone to displacement or damage after being run over by heavy vehicles for a long time, making subsequent maintenance difficult; ③ Magnetic nail signals are easily interfered with by scattered metal debris in the stockyard, and are also easily affected by electromagnetic interference from surrounding electrical equipment; ④ Magnetic nail positioning is essentially discrete point positioning, and the calculation between two magnetic nails relies on encoders, which will produce cumulative errors.

[0030] (II) Problems with GPS / Differential GPS Positioning. The dense refrigerated racks and stacked containers in container yards create a "steel canyon" environment, which severely blocks and reflects satellite signals, making it difficult to consistently meet the centimeter-level positioning accuracy requirements; moreover, high-precision differential GPS equipment is expensive, and differential service fees need to be paid continuously.

[0031] (III) Problems with the use of calibration plates in conjunction with laser ranging. Calibration plates are installed in the open environment on the side of the runway, and are easily deformed or damaged by passing vehicles, requiring frequent replacement and recalibration. At the same time, single-point laser ranging is significantly affected by weather factors such as rain, snow, and fog, and beam attenuation or scattering can lead to ranging failure or a significant decrease in accuracy.

[0032] In summary, current solutions all rely on manually deployed positioning facilities, which have problems such as high deployment and maintenance costs, easy degradation and failure of positioning references, poor environmental adaptability, and lack of multi-source redundancy verification.

[0033] According to an embodiment of the present invention, a method for positioning a yard crane trolley is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0034] This embodiment provides a method for positioning yard crane trolleys, which can be applied to yard scenarios with refrigerated racks. Figure 1 This is a schematic flowchart of the first method for positioning the trolley of the yard crane according to an embodiment of the present invention, as follows: Figure 1 As shown, the process includes the following steps: Step S101: Obtain the 3D point cloud data of the refrigerated racks on the left and right sides of the travel channel of the gantry crane in the positioning scene.

[0035] Here, the location scenario refers to the operating area of ​​the automated refrigerated container yard, that is, the entire site area where the yard crane trolley travels along the track and performs container loading and unloading operations.

[0036] The gantry crane is the traveling mechanism of a rail-mounted container gantry crane. It is responsible for moving between working bays along the longitudinal direction of the yard runway and serves as the mobile chassis for the entire crane.

[0037] The travel passage refers to the area between two ground rails when the gantry crane travels along them.

[0038] Refrigerated racks are metal support structures used in container yards for storing and supplying power to refrigerated containers. They are arranged in neat rows, with each row consisting of multiple vertical steel columns spaced at fixed intervals.

[0039] 3D point cloud data refers to data generated by a 3D lidar system that scans the surface of an object with a laser beam, receives the reflected signals, and measures the 3D spatial coordinates of each point. The collection of all points constitutes the point cloud. Point cloud data contains information about the geometric shape and spatial location of objects.

[0040] Here, a 3D lidar is installed at the bottom of the trolley frame of the yard crane, scanning downwards. When the trolley moves near the target working position, the lidar starts scanning, emitting laser beams towards the refrigerated rack areas on both sides and receiving reflected signals, acquiring 3D point cloud data of the refrigerated rack columns on both sides within the current field of view. The installation position of the lidar ensures that the scanning range can simultaneously cover the column areas on both sides.

[0041] Step S102: Based on the three-dimensional point cloud data of the left and right refrigerated shelves, calculate the point cloud plane position of the left and right refrigerated shelves corresponding to the side closer to the travel channel.

[0042] Here, the point cloud plane positions on the left and right sides can be calculated using any suitable method.

[0043] In some implementations, a spatial plane model can be constructed, and the spatial plane equation can be established using the three-dimensional coordinates of all three-dimensional point cloud data as input data. The least squares method can be used to solve for the plane parameters that minimize the sum of the squares of the distances from all points to the plane, thereby determining the position of the point cloud plane on the left and right sides respectively.

[0044] In some implementations, the Random Sample Consensus (RANSAC) algorithm can be used to randomly sample at least 3 points from the feature surface point cloud to fit a temporary plane, count the distances of all points to the plane, and mark the points with a distance less than a preset threshold as interior points; repeat the above random sampling process N times, select the plane model with the most interior points as the final plane, and output the point cloud plane positions on the left and right sides.

[0045] In some implementations, point sets of the left and right vertical edge lines of the pillar are extracted from the pillar point cloud, and spatial straight line equations of the two edge lines are fitted respectively. The inner plane is determined by the two parallel vertical edge lines, and the point cloud plane positions on the left and right sides are taken.

[0046] It should be noted that the refrigerated rack uprights are vertical steel columns, and their inner side facing the aisle is a vertical plane. This plane extends in the X direction (along the aisle) and in the Z direction (vertically). Its spatial position is uniquely determined by the Y coordinate value. Therefore, when obtaining the point cloud plane position, the Y coordinate value can be directly used as the point cloud plane position, or the Y coordinate of the point cloud plane position can be extracted in the subsequent calculation process. This method does not impose any restrictions on this.

[0047] Step S103: Based on the point cloud plane positions of the left and right refrigerated racks corresponding to the side near the travel channel, determine the positioning position of the yard crane trolley.

[0048] The positioning position of the yard crane trolley can be calculated using any suitable method.

[0049] In some implementations, the coordinates of the point cloud plane positions on the left and right sides can be substituted into the mean calculation formula to calculate the geometric center point on the left and right sides in the direction perpendicular to the channel, and the position of the geometric center point can be used as the positioning position of the yard bridge trolley.

[0050] In some implementations, the coordinates of the point cloud plane positions on the left and right sides can be input into a pre-established localization model to obtain the current driving scene. This localization model can be any suitable neural network model capable of achieving this function.

[0051] In this embodiment of the method, on the one hand, by acquiring the three-dimensional point cloud data of the refrigerated racks on both sides of the travel channel, and utilizing the relatively fixed structural features of the refrigerated racks, the point cloud plane of the refrigerated racks is used as a positioning reference to provide high-precision positioning information. This helps the yard crane trolley to achieve accurate stopping and operation within the travel channel, reducing safety accidents such as collisions caused by positioning errors and improving operational safety. On the other hand, by acquiring the three-dimensional point cloud data of the refrigerated racks and calculating the positioning position of the yard crane trolley, the impact of external infrastructure failures or poor signal on the positioning of the yard crane trolley can be reduced, thereby lowering construction and maintenance costs.

[0052] This embodiment provides a method for positioning the main trolley of a yard crane, which can be used in the aforementioned automated yard crane system. Figure 2 This is a schematic diagram of the second process of the yard crane trolley positioning method according to an embodiment of the present invention, as follows: Figure 2 As shown, the process includes the following steps: Step S201: Obtain the 3D point cloud data of the refrigerated racks on the left and right sides of the travel channel of the gantry crane in the positioning scene.

[0053] In some optional implementations, the above method further includes: Step a1: In response to the received positioning signal from the refrigerated rack column, determine that the yard crane has entered the positioning scene.

[0054] Here, the positioning signal refers to the position identification signal issued by the upper-level dispatching system or automatically triggered by the on-board sensors when the yard crane trolley moves to the vicinity of the target working position.

[0055] The positioning signal can be a rough position signal calculated by the dispatching system based on the vehicle encoder or odometer, or it can be a positioning signal detected by the sensor when the vehicle enters a specific area.

[0056] It should be noted that the positioning signal is used to determine that the yard crane trolley has entered the effective range where precise positioning can be performed.

[0057] In this embodiment of the method, when the yard crane truck travels near the refrigerated rack and receives the positioning signal from the refrigerated rack column, it can accurately determine that the yard crane truck has entered a suitable scene for positioning. The scene truck is then controlled to use LiDAR to acquire three-dimensional point cloud data to locate the yard crane truck, further ensuring the accuracy and effectiveness of the positioning data.

[0058] Step S202: Based on the three-dimensional point cloud data of the left and right refrigerated shelves, calculate the point cloud plane position of the left and right refrigerated shelves corresponding to the side closer to the travel channel.

[0059] Specifically, step S202 includes: Step S2021: Preprocess the three-dimensional point cloud data to obtain the target feature point cloud corresponding to the cold storage rack columns on the left and right sides.

[0060] In some optional implementations, step S2021 above includes: Step b1: Filter and denoise the 3D point cloud data to obtain the first processed feature point cloud. Step b2: Filter the first processed feature point cloud based on a preset height threshold to obtain the second processed feature point cloud. Step b3: Cluster and segment the second processed feature point cloud to obtain the target feature point clouds corresponding to the left and right refrigeration rack columns.

[0061] Here, filtering and denoising refers to the process of removing outliers from point clouds caused by environmental interference, sensor noise, or multipath reflections using specific algorithms. These noise points are isolated points whose spatial locations are significantly deviated from the surface of the real object. If they are not removed, the accuracy of subsequent processing will be affected.

[0062] Clustering segmentation refers to the algorithmic process of grouping spatially close points in a point cloud into the same cluster. Clustering algorithms determine the grouping based on the Euclidean distance between points; if the distance between two points is less than a preset threshold, they are grouped into the same cluster; otherwise, they are grouped into different clusters. Through clustering, the originally chaotic point cloud is segmented into multiple independent object point cloud clusters.

[0063] In one specific implementation, firstly, for each point in the point cloud, search for its K nearest neighbors (e.g., K=50) and calculate the average distance from that point to all its neighbors. Assuming that the average distance of all points follows a Gaussian distribution, calculate the global average distance μ and the standard deviation σ. If the average distance of a point is greater than μ+α×σ (α is a preset multiple, such as 1.0~2.0), then the point is determined to be an outlier and is removed, and the remaining points constitute the first processed feature point cloud.

[0064] Second, all points are judged based on their Z-coordinate (height value). A height threshold is set (e.g., 0.3 meters). Points with Z-coordinates less than or equal to the height threshold are judged as ground points and removed. Points with Z-coordinates greater than the height threshold are retained to form the second-process feature point cloud.

[0065] Third, a clustering algorithm based on Euclidean distance is used. For all points in the second-processed feature point cloud, a clustering distance threshold (e.g., 0.2~0.3 meters) is set, and the spatial distance between any two points is determined sequentially. If the distance between two points is less than the clustering distance threshold, they are grouped into the same cluster; otherwise, they belong to different clusters. After traversal, each cluster represents an independent object. Then, each cluster is further distinguished based on the geometric characteristics of the pillar (cluster height > 2 meters, width < 0.5 meters, verticality > 0.95). Clusters that meet the conditions are the pillar point cloud clusters, constituting the target feature point cloud.

[0066] In this method embodiment, on the one hand, isolated noise points are removed by statistical filtering, eliminating the risk that noise points will form their own clusters or interfere with normal clustering during subsequent clustering; on the other hand, ground point clouds are removed by height filtering, which greatly reduces the amount of data involved in clustering calculation and improves the calculation efficiency. In addition, by preprocessing the collected three-dimensional point cloud data, the accuracy of locating the large vehicle of the yard bridge using the three-dimensional point cloud data can be improved.

[0067] Step S2022: Based on the target feature point cloud corresponding to the left refrigerated shelf column, calculate the first point cloud plane position of the left refrigerated shelf corresponding to the side closer to the travel aisle. Step S2023: Based on the target feature point cloud corresponding to the right refrigerated shelf column, calculate the second point cloud plane position of the right refrigerated shelf corresponding to the side closer to the travel aisle.

[0068] Here, the target feature point cloud corresponding to the left refrigerated rack column refers to the set of point cloud data of all refrigerated rack columns located on the left side of the vehicle travel channel, including all point clouds of one or more columns on the left.

[0069] The target feature point cloud corresponding to the right-side refrigerated rack column refers to the set of point cloud data of all refrigerated rack columns located on the right side of the vehicle travel channel, including all point clouds of one or more columns on the right side.

[0070] The first point cloud plane position of the left refrigerated rack corresponding to the side closest to the travel channel refers to the spatial position coordinate value of the feature surface of each left column facing the inside of the channel (i.e., the right side of the left column) in the vertical channel direction (Y direction). This position value is obtained by fitting or calculating all point clouds on the feature surface and represents the lateral position of the left boundary of the channel.

[0071] The second point cloud plane position of the right-side refrigerated rack corresponding to the side closest to the passageway: refers to the spatial coordinate value of the feature surface of each right-side column facing the inside of the passageway (i.e., the left side of the right-side column) in the direction perpendicular to the passageway (Y direction). This position value represents the lateral position of the right-side boundary of the passageway.

[0072] In one optional implementation, the positioning position of the yard crane is determined based on the point cloud plane positions of the left and right refrigerated racks corresponding to the sides near the travel aisle, including: Step c1: Determine the average coordinates of the first position coordinates corresponding to the first point cloud plane position and the second position coordinates corresponding to the second point cloud plane position; Step c2: Determine the position corresponding to the average coordinates as the positioning position of the yard crane trolley.

[0073] In one specific implementation, the feature surface point cloud of the first left-side pillar facing the inner side of the passage (the part of points with the largest Y-coordinate on the right side of the pillar, approximately 2000 points) is extracted. The average Y-coordinate of these points is taken to obtain Y = +5.01 meters. Similarly, the Y-coordinate of the inner side of the second pillar is calculated to be +5.00 meters, and the Y-coordinate of the inner side of the third pillar is +4.99 meters. The positions of the inner sides of the three left-side pillars are +5.01 meters, +5.00 meters, and +4.99 meters, respectively. The average of these three values ​​is taken as the left-side passage boundary position C_left = +5.00 meters.

[0074] Extract the feature surface point cloud of the first pillar on the right facing the inside of the passage (the part of points with the smallest Y-coordinate on the left side of the pillar, approximately 2000 points). Average the Y-coordinates of these points to obtain Y = -4.98 meters. Similarly, calculate Y = -5.01 meters for the inside of the second pillar and -4.99 meters for the inside of the third pillar. The positions of the inside surfaces of the three pillars on the right are -4.98 meters, -5.01 meters, and -4.99 meters respectively. Take the average of these three values ​​as the boundary position of the right passage, C_right = -4.99 meters.

[0075] Now, substitute the positions of the first and second point cloud planes into the mean calculation formula to determine the mean coordinates. The mean calculation formula is as follows: C_mid = (C_left + C_right) / 2 Where C_mid is the mean coordinate, C_left is the position of the first point in the cloud plane, and C_right is the position of the second point in the cloud plane.

[0076] Here, substituting the cloud plane position of the first point C_left = +5.00 meters and the cloud plane position of the second point C_right = -4.99 meters into the mean calculation formula, the mean coordinate C_mid is obtained as +0.005 meters.

[0077] At this moment, the current position of the gantry crane is 0.001 meters to the left of the theoretical centerline of the passage.

[0078] In this embodiment of the method, the planar position of the point cloud of the left and right refrigerated racks near the travel channel is calculated based on the target feature point cloud, which reduces the deviation in planar position calculation caused by data errors and can more accurately determine the actual position of the yard crane truck.

[0079] In some optional implementations, when the target feature point cloud corresponding to the left or right refrigerated shelf column is less than a preset threshold number, step S2022 or step S2023 includes: Step d1 involves jointly registering multiple target feature point clouds and determining the planar position of the first point cloud or the planar position of the second point cloud using the preset geometric constraints of the refrigerated rack.

[0080] Here, the preset quantity threshold refers to the minimum number of refrigerated rack columns required for effective positioning of the yard crane trolley. For example, the preset threshold is "at least 2 columns on one side" or "at least 4 columns in total on both sides". When the actual number of detected columns is lower than this threshold, the reliability of single-side positioning is insufficient, and a multi-point joint registration strategy needs to be activated to perform joint calculation using multi-source information.

[0081] The preset geometric constraints of the refrigerated racks refer to the known and fixed geometric relationships between the rack columns, which mainly include three types: ① the spacing between adjacent columns is fixed (e.g., all are 3 meters); ② the columns on the same side are collinear along the direction of the runway; ③ the columns on the left and right sides are symmetrically distributed about the center line of the passage.

[0082] In some implementations, when the number of pillars detected on the left or right side is less than a preset threshold (e.g., less than 2 pillars on one side), the system uses all available pillar point clouds within the current field of view as input data for joint registration. For each pillar point cloud, its feature surface point cloud facing the inner side of the channel is extracted independently, and the position coordinates of the surface in the Y direction are calculated (using the centroid averaging method or least squares fitting method in the prior art), obtaining the single-point localization result of each pillar as the initial value.

[0083] The design parameters of the refrigerated racks for the storage yard area are retrieved from the database to obtain the following geometric constraints: the spacing D between adjacent columns on the same side (e.g., 3.00 meters, guaranteed by the construction precision of the steel structure, with an error typically within ±2mm); the theoretical alignment of each column along the runway direction; and the symmetry of the columns on the left and right sides about the centerline of the passageway. These constraints were determined when the storage yard was built and remain unchanged throughout its entire service life.

[0084] An optimization objective function is constructed based on the deviation between the initial value of the single-point positioning of each column and the preset geometric constraints. The optimization objective of the objective function is to minimize the weighted sum of the following two types of errors: ① the deviation between the measured position and the theoretical position of each column (point cloud registration error term); ② the deviation between the relative position of each column and the preset geometric constraints (spacing constraint error term, collinearity constraint error term, and symmetry constraint error term).

[0085] When the number of columns on one side is insufficient, the effective columns on the other side and the constraints across both sides (such as left-right symmetry constraints and spacing consistency constraints) will play a crucial compensatory role. For example, when there is only one column on the left side, although its absolute position cannot be verified by other columns on the same side, it can still be effectively constrained to the correct position near the right position through the symmetry constraints and spacing consistency constraints with the two columns on the right side.

[0086] A nonlinear least squares optimization algorithm (such as the Levenberg-Marquardt algorithm) is used for iterative solution. In each iteration, the algorithm fine-tunes the position estimates of each column, causing the objective function value to gradually decrease. The algorithm considers a convergence threshold (e.g., 1 × 10⁻⁶) to be reached when the change in the objective function value between two consecutive iterations is less than a preset convergence threshold. -6 The iteration stops when the maximum number of iterations is reached, and the optimized column position estimate is output as the final result of the first point cloud plane position or the second point cloud plane position.

[0087] For example, in a certain operation, the second column on the right was obscured by stacked containers. The system only detected two columns on the right (Y=-4.98, -4.99), while three columns on the left (Y=+5.01, +5.00, +4.99) were still detected. The system determined that the number of columns on the right was two, which is lower than the preset threshold (three columns on one side), and initiated the multi-point joint registration process. The yard design drawings were reviewed to confirm that the column spacing in the area was 3.00 meters and symmetrical. The feature point clouds of five targets—three columns on the left and two on the right—were input into the joint registration algorithm. The single-point results of the three columns on the left (+5.01, +5.00, +4.99) largely matched the known spacing constraint of 3.00 meters, indicating high confidence. The single-point results of the two columns on the right (-4.98, -4.99) required symmetry (Y_left + Y_right ≈ 0) with the second and third columns on the left, respectively. Specifically, +5.00 on the left should correspond to -5.00 on the right; the measured deviation of -4.98 was +0.02 meters, and +4.99 on the left should correspond to -4.99 on the right; the measured deviation of -4.99 was 0.00 meters. Simultaneously, the adjacent column spacing constraint (both 3.00 meters) was used to verify the relative positions of the two columns on the right.

[0088] The symmetry and spacing constraints in the objective function play a role in the iterative solution process. The spacing between the three columns on the left is consistent, and the spacing between two adjacent columns on the right is |(-4.98)-(-4.99)|=0.01 meters, which matches the theoretical spacing of 3.00 meters. Through iterative solution, the single-point results of each column gradually approach the optimal solution that satisfies all constraints. Finally, the cloud plane position of the second point on the right is -4.995 meters, and the cloud plane position of the first point on the left is +5 meters. Substituting into the formula: mean position = (+5.00 + (-4.995)) / 2 = +0.0025 meters, the position of the vehicle is about 0.0025 meters to the left (north), which is still within the centimeter-level accuracy range, and the positioning is effective.

[0089] In this embodiment of the method, when some pillars are obscured by goods, resulting in a reduction in the number of target feature point clouds, joint registration can integrate the point cloud information of other unobscured parts, thereby improving the accuracy and reliability of positioning.

[0090] Step S203: Based on the point cloud plane positions of the left and right refrigerated racks corresponding to the sides closest to the travel aisle, determine the positioning position of the yard crane trolley. For details, please refer to [link to details]. Figure 1 Step S103 of the illustrated embodiment will not be described again here.

[0091] In some optional implementations, the above method further includes: Step e1: Control the grabbing and releasing of refrigerated containers based on the positioning location.

[0092] Here, the refrigerated container is controlled by ensuring that the positioning of the yard crane is within the allowable deviation threshold.

[0093] The control of grabbing and placing refrigerated containers can be performed through the following steps. After confirming that the positioning of the trolley is within the allowable deviation threshold, the control command "Trolley is accurately in position, stacking operation can be performed" is issued, and the stacking command sequence is issued simultaneously: First, the trolley moves laterally to directly above the target container position, the spreader descends to the top of the container on the ground, and the twistlock mechanism rotates and locks the container; Second, the lifting mechanism lifts the container to a safe height of about 0.5 meters above the existing stacked containers; Third, the trolley and spreader work together to fine-tune the horizontal position of the spreader, aligning the container being lifted with the four corner pieces of the existing stacked containers (this alignment process depends on the precise position feedback of the trolley and spreader); Fourth, the lifting mechanism slowly lowers the container to ensure it falls smoothly, and the four corner pieces accurately embed into the corner pieces holes of the container below (if the positioning deviation of the trolley exceeds the limit, the corner pieces cannot be aligned, which may cause the container to tip over or be damaged); Fifth, the twistlock mechanism unlocks and releases the container, and the lifting mechanism lifts it away unloaded; Sixth, after confirming that the container position is stable, the stacking operation is completed.

[0094] In this embodiment of the method, the target refrigerated container can be quickly reached based on the precise positioning location, reducing operation time and improving operation safety.

[0095] In some implementation methods, the specific implementation of the yard crane trolley positioning method is as follows: I. LiDAR deployment and calibration.

[0096] One or more 3D LiDAR sensors are installed at the bottom of the trolley frame of the yard crane, each facing the refrigerated container stacking area below, to collect 3D point cloud data of the refrigerated racks and lifting devices on both sides. For example... Figure 3 As shown, Figure 3 This is a schematic diagram of the regional structure of the refrigerated rack columns. The installation position of the lidar ensures that the scanning range covers the column areas of the refrigerated racks on both the left and right sides, as well as the edge outline of the hangers.

[0097] II. Large vehicle positioning based on real-time cold storage rack point cloud registration.

[0098] 1. When the large vehicle (equivalent to the above-mentioned yard crane large vehicle) runs to the vicinity of the current refrigerated rack's working position, the point cloud data of the refrigerated rack columns in the current field of view is collected in real time by 3D LiDAR.

[0099] 2. Preprocess the real-time point cloud, including: statistical filtering to remove outlier noise points, elevation-based ground point separation, and column point cloud cluster segmentation based on clustering algorithms.

[0100] 3. Calculate the point cloud plane positions C_left and C_right on the inner sides of the refrigerated racks in real time to accurately position the trolley: C_mid=(C_left+C_right) / 2.

[0101] 4. When the field of view contains multiple columns, multi-point joint registration is adopted. Through multi-constraint optimization, the error of single-point registration is effectively eliminated, and the positioning accuracy of the large vehicle on both sides is centimeter-level.

[0102] This embodiment also provides a trolley positioning device for a yard crane, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, hardware implementations, or a combination of software and hardware, are also possible and contemplated.

[0103] This embodiment provides a positioning device for a yard crane trolley, such as Figure 4 As shown, it includes: The acquisition module 401 is used to acquire the three-dimensional point cloud data of the refrigerated racks on the left and right sides of the travel channel of the gantry crane in the positioning scene.

[0104] The calculation module 402 is used to calculate the point cloud plane position of the left and right refrigerated shelves on the side closest to the travel channel, based on the three-dimensional point cloud data of the left and right refrigerated shelves.

[0105] The determination module 403 is used to determine the positioning position of the yard crane truck based on the point cloud plane position of the left and right refrigerated racks corresponding to the side near the travel channel.

[0106] In some alternative implementations, the computing module 402 includes: The preprocessing unit is used to preprocess the three-dimensional point cloud data to obtain the target feature point cloud corresponding to the cold storage rack columns on the left and right sides.

[0107] The first calculation unit is used to calculate the first point cloud plane position of the left refrigerated shelf corresponding to the side close to the travel channel, based on the target feature point cloud corresponding to the left refrigerated shelf column.

[0108] The second calculation unit is used to calculate the second point cloud plane position of the right refrigerated shelf corresponding to the side close to the travel passage based on the target feature point cloud corresponding to the right refrigerated shelf column.

[0109] In some optional implementations, the preprocessing unit includes: The filtering and denoising subunit is used to filter and denoise the 3D point cloud data to obtain the first processed feature point cloud.

[0110] The filtering subunit is used to filter the first processed feature point cloud based on a preset height threshold to obtain the second processed feature point cloud.

[0111] The clustering segmentation subunit is used to cluster and segment the second processing feature point cloud to obtain the target feature point cloud corresponding to the cold storage rack columns on the left and right sides.

[0112] In some alternative implementations, the determining module 403 includes: The first determining unit is used to determine the average coordinates of the first position coordinates corresponding to the first point cloud plane position and the second position coordinates corresponding to the second point cloud plane position.

[0113] The second determining unit is used to determine the position corresponding to the mean coordinate as the positioning position of the yard crane trolley.

[0114] In some optional implementations, when the target feature point cloud corresponding to the left or right refrigerated shelf column is less than a preset threshold number, the first calculation unit or the second calculation unit includes: The joint registration subunit is used to jointly register multiple target feature point clouds and determine the first point cloud planar position or the second point cloud planar position through the preset geometric constraints of the cold storage rack.

[0115] In some alternative embodiments, the apparatus further includes: Based on location, the system controls the handling and release of refrigerated containers.

[0116] In some alternative embodiments, the apparatus further includes: In response to the received positioning signal from the refrigerated rack column, the yard crane truck is determined to have entered the positioning scene.

[0117] The trolley positioning device for yard cranes provided in this embodiment of the invention can execute the trolley positioning method for yard cranes provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0118] The yard bridge has a movable trolley mounted on its crossbeams, and at least one lidar sensor is installed on the bottom of the trolley. The yard bridge also includes a controller. Figure 5This is a schematic diagram of the structure of a controller for an automated field bridge provided in an embodiment of the present invention.

[0119] The following is a detailed reference. Figure 5 The diagram illustrates a structural schematic suitable for implementing a controller in an embodiment of the present invention. The controller may include a processor (e.g., a central processing unit, a graphics processing unit, etc.) 501, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 502 or a program loaded from memory 508 into random access memory (RAM) 503. The RAM 503 also stores various programs and data required for controller operation. The processor 501, ROM 502, and RAM 503 are interconnected via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0120] Typically, the following devices can be connected to I / O interface 505: input devices 506 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 507 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 508 including, for example, magnetic tapes, hard disks, etc.; and communication devices 509. Communication device 509 allows the controller to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 5 A controller with various devices is shown, but it should be understood that it is not required to implement or have all of the devices shown, and may alternatively implement or have more or fewer devices.

[0121] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 509, or installed from a memory 508, or installed from a ROM 502. When the computer program is executed by the processor 501, it performs the functions defined in the gantry crane positioning method of the embodiments of the present invention.

[0122] Figure 5 The controller shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0123] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the trolley positioning method shown in the above embodiments is implemented.

[0124] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0125] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for positioning a large trolley in a yard crane, characterized in that, Applied to a storage yard scenario with refrigerated racks, a movable trolley is mounted on the crossbeam of the yard crane, and at least one lidar is installed on the bottom of the trolley. The method includes: Acquire the three-dimensional point cloud data of the refrigerated racks on both sides of the travel channel of the large crane in the positioning scenario; Based on the three-dimensional point cloud data of the left and right refrigerated shelves, the point cloud plane positions of the left and right refrigerated shelves corresponding to the side closest to the travel channel are calculated respectively. The positioning position of the yard crane is determined based on the point cloud plane positions of the left and right refrigerated racks corresponding to the sides close to the travel channel.

2. The method according to claim 1, characterized in that, The calculation of the point cloud plane positions of the left and right refrigerated shelves near the travel aisle, based on the three-dimensional point cloud data of the left and right refrigerated shelves, includes: The three-dimensional point cloud data is preprocessed to obtain the target feature point cloud corresponding to the cold storage rack columns on the left and right sides; Based on the target feature point cloud corresponding to the left refrigeration rack column, calculate the first point cloud plane position of the left refrigeration rack near the side of the travel channel. Based on the target feature point cloud corresponding to the right-side refrigeration rack column, calculate the second point cloud plane position of the right-side refrigeration rack near the side of the travel channel.

3. The method according to claim 2, characterized in that, The preprocessing of the three-dimensional point cloud data to obtain the target feature point clouds corresponding to the left and right refrigeration rack columns includes: The three-dimensional point cloud data is filtered and denoised to obtain the first processed feature point cloud; The first processed feature point cloud is filtered based on a preset height threshold to obtain the second processed feature point cloud. Clustering and segmentation are performed on the second processed feature point cloud to obtain the target feature point cloud corresponding to the cold storage rack columns on the left and right sides.

4. The method according to claim 2, characterized in that, The method of determining the positioning position of the yard crane trolley based on the point cloud plane positions of the left and right refrigerated racks corresponding to the sides near the travel channel includes: Determine the average coordinates of the first position coordinates corresponding to the first point cloud plane position and the second position coordinates corresponding to the second point cloud plane position; The position corresponding to the mean coordinates is determined as the positioning position of the yard crane trolley.

5. The method according to claim 2, characterized in that, When the target feature point cloud corresponding to the left or right refrigerated shelf column is less than a preset threshold, based on the target feature point cloud corresponding to the left or right refrigerated shelf column, calculate the first point cloud plane position or the second point cloud plane position corresponding to the left or right refrigerated shelf near the travel channel side, including: Joint registration is performed on multiple target feature point clouds, and the planar position of the first point cloud or the planar position of the second point cloud is determined by the preset geometric constraints of the cold storage rack.

6. The method according to claim 1, characterized in that, The method further includes: Based on the location, the refrigerated container is controlled to be picked up and placed.

7. The method according to any one of claims 1 to 6, characterized in that, The method further includes: In response to the received positioning signal of the refrigerated rack column, it is determined that the yard crane trolley has entered the positioning scene.

8. A positioning device for a yard crane trolley, characterized in that, A movable trolley is mounted on the crossbeam of the yard bridge, and at least one lidar is installed on the bottom of the trolley. The device includes: The acquisition module is used to acquire the three-dimensional point cloud data of the refrigerated racks on the left and right sides of the travel channel of the gantry crane in the positioning scenario; The calculation module is used to calculate the point cloud plane position of the left and right refrigerated shelves corresponding to the side closer to the travel channel, based on the three-dimensional point cloud data of the left and right refrigerated shelves. The determination module is used to determine the positioning position of the yard crane trolley based on the point cloud plane positions of the left and right refrigerated racks corresponding to the sides close to the travel channel.

9. An automated field bridge, characterized in that, Applied to storage yard scenarios with refrigerated racks, the yard crane has a movable trolley mounted on its crossbeams, and at least one lidar sensor is installed on the bottom of the trolley. The yard crane also includes a controller, which comprises: A memory and a processor are interconnected, the memory storing computer instructions, and the processor executing the computer instructions to perform the trolley positioning method of any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to perform the gantry crane positioning method according to any one of claims 1 to 7.

Images