Double-box operation method, device and equipment based on adjacent box postures and medium

Abstract

The embodiment of the invention provides a double-box operation method, device and equipment based on adjacent box postures and a medium, and relates to the technical field of port logistics. The method comprises the following steps: acquiring pose information of at least one existing container adjacent to the position of a container to be placed, and taking the pose information as an adjacent container pose; according to the posture of the adjacent container, the placing posture of the to-be-placed container is adjusted; determining that the adjusted posture of the to-be-placed container and the adjusted posture of the adjacent container meet preset double-container grabbing conditions; and determining that the to-be-placed position of the to-be-placed container is located in a preset safety area, and executing container placing operation. According to the method, the problem that the container cannot be aligned with the adjacent container under the condition that it is guaranteed that the container meets the position requirement of the container in double-container operation in the related technology is effectively solved.

Description

Technical Field

[0001] This application relates to the field of port logistics technology, and in particular to a method, apparatus, equipment and medium for dual-container operation based on the attitude of adjacent containers. Background Technology

[0002] In automated container terminal yard operations, there is a demand for efficient "single-in, double-out" operations. This means that containers are transported to the port by truck and placed into the yard as single containers, while when loading onto ships, the yard cranes need to grab two adjacent containers at once for handling to improve efficiency. Therefore, precisely aligning a single container with adjacent containers in the yard during port entry creates the necessary conditions for efficient double-container grabbing operations, significantly improving the overall throughput efficiency of the terminal.

[0003] In related technologies, automated container placement is usually controlled by visual recognition of pre-drawn container corner lines on the yard floor. However, this method can only ensure that the container is placed in a safe position for a single time, which may cause a large positional deviation between the container placed later and the adjacent container, leaving hidden dangers for subsequent double container operations. However, if alignment with the adjacent container is given priority, the container may be caused to exceed the safe container corner line range, causing safety hazards. Summary of the Invention

[0004] This application provides a method, apparatus, equipment, and medium for dual-container operations based on the orientation of adjacent containers, in order to solve the problem in related technologies where dual-container operations cannot align with adjacent containers while ensuring that the containers meet their own positional requirements.

[0005] Firstly, this application provides a method for dual-box operations based on the attitude of adjacent boxes, the method comprising:

[0006] Obtain the pose information of at least one existing container adjacent to the location of the container to be placed, and use the pose information as the pose of the neighboring container.

[0007] Adjust the placement orientation of the container to be placed according to the orientation of the adjacent containers;

[0008] Determine that the adjusted orientation of the container to be placed and the orientation of the adjacent container meet the preset dual-container grasping conditions;

[0009] Once the intended placement location of the container is within the preset safe area, the container placement operation is executed.

[0010] In one embodiment of this disclosure, determining that the adjusted current container posture and the adjacent container posture meet the preset dual-container grasping conditions includes: determining the angle difference between the yaw angle of the container to be placed and the yaw angle of the adjacent container in the adjacent container posture, and calculating the position offset of the container to be placed and the adjacent container along a preset direction; if the angle difference is less than a first threshold and the position offset is less than a second threshold, determining that the container to be placed and the adjacent container meet the dual-container grasping conditions.

[0011] In one embodiment of this disclosure, before acquiring the pose information of at least one existing container adjacent to the location of the container to be placed, the method further includes: acquiring point cloud data of the neighborhood of the location of the container to be placed based on a ranging sensor installed on the hoisting equipment, wherein the ranging sensor is set at a relative angle to the bottom surface of the hoisting equipment, and the mirror of the ranging sensor faces downward perpendicular to the ground; determining that the number of point cloud data exceeds a third threshold; and determining that there is at least one existing container in the neighborhood of the location of the container to be placed.

[0012] In one embodiment of this disclosure, after determining that there is at least one existing container in the neighborhood of the location of the container to be placed, the method further includes: in response to the task information issued by the task management system, determining that the target neighboring container that needs to be aligned with the container to be placed is an existing container located in front of or behind the location to be placed; obtaining the pose information of at least one existing container adjacent to the location of the container to be placed, and using the pose information as the pose of the neighboring container, including: obtaining the pose information of the target neighboring container and using the pose information as the pose of the neighboring container.

[0013] In one embodiment of this disclosure, obtaining the pose information of the target neighboring container and using the pose information as the neighboring container's attitude includes: performing planar fitting on the point cloud data corresponding to the target neighboring container to obtain the planar point cloud of the upper surface of the target neighboring container; extracting the boundary features of the existing container based on the planar point cloud of the upper surface; and calculating the neighboring container's attitude based on the boundary features, wherein the neighboring container's attitude includes at least the neighboring container's yaw angle and the neighboring container's offset along the trolley's running direction.

[0014] In one embodiment of this disclosure, the boundary features of existing containers are extracted based on the upper surface planar point cloud, including: determining the maximum and minimum coordinate points of the upper surface planar point cloud in a preset direction; starting from the maximum and minimum coordinate points, iteratively searching for the lowest point of each local point cloud along the opposite direction to obtain a boundary point set; performing a straight line fitting on the boundary point set based on a random sampling consensus algorithm, and determining the yaw angle of adjacent containers based on the parameters of the fitted straight line.

[0015] In one embodiment of this disclosure, adjusting the placement posture of the container to be placed according to the posture of the adjacent container includes: generating a spreader adjustment command based on the yaw angle and offset of the adjacent container; outputting the spreader adjustment command and performing closed-loop control based on the posture information fed back in real time by the posture sensor on the spreader until the current placement posture of the container matches the posture of the adjacent container, wherein the spreader adjustment command is used to drive the spreader movement.

[0016] Secondly, embodiments of this disclosure provide a dual-box operation device based on adjacent box attitude, the dual-box operation device based on adjacent box attitude includes:

[0017] The acquisition module is used to acquire the pose information of at least one existing container adjacent to the position of the container to be placed, and to use the pose information as the pose of the neighboring container.

[0018] The adjustment module is used to adjust the placement posture of the container to be placed based on the posture of the adjacent container;

[0019] The judgment module is used to determine whether the adjusted posture of the container to be placed and the posture of the adjacent container meet the preset double container grasping conditions.

[0020] The processing module is used to determine that the intended placement location of the container to be placed is within a preset safe area and to perform the container placement operation.

[0021] Optionally, the judgment module is specifically used to determine the angle difference between the yaw angle of the container to be placed and the yaw angle of the adjacent container in the attitude of the adjacent container, and to calculate the position offset of the container to be placed and the adjacent container along a preset direction; if the angle difference is less than a first threshold and the position offset is less than a second threshold, it is determined that the container to be placed and the adjacent container meet the dual container grabbing conditions.

[0022] Optionally, the acquisition module is further configured to, before acquiring the pose information of at least one existing container adjacent to the location of the container to be placed, acquire point cloud data of the neighborhood of the location of the container to be placed based on a ranging sensor installed on the hoisting equipment, wherein the ranging sensor is set at a relative angle to the bottom surface of the hoisting equipment, and the mirror of the ranging sensor faces downward perpendicular to the ground; determine that the number of point cloud data exceeds a third threshold; and determine that there is at least one existing container in the neighborhood of the location of the container to be placed.

[0023] Optionally, the acquisition module is further configured to, after determining that there is at least one existing container in the neighborhood of the location of the container to be placed, in response to the task information issued by the task management system, determine that the target neighboring container that needs to be aligned with the container to be placed is an existing container located in front of or behind the location to be placed; and acquire the pose information of the target neighboring container as the pose of the neighboring container.

[0024] Optionally, the acquisition module is specifically used to perform planar fitting on the point cloud data corresponding to the target neighboring container to obtain the planar point cloud of the upper surface of the target neighboring container; extract the boundary features of the existing container based on the planar point cloud of the upper surface; and calculate the attitude of the neighboring container based on the boundary features, wherein the attitude of the neighboring container includes at least the yaw angle of the neighboring container and the offset of the neighboring container along the running direction of the trolley.

[0025] Optionally, the acquisition module is specifically used to: determine the maximum and minimum coordinate points of the upper surface planar point cloud in a preset direction; starting from the maximum and minimum coordinate points, iteratively search for the lowest point of each local point cloud along the opposite direction to obtain the boundary point set; perform straight line fitting on the boundary point set based on the random sampling consensus algorithm, and determine the yaw angle of the adjacent box based on the parameters of the fitted straight line.

[0026] Optionally, the adjustment module is specifically used to generate spreader adjustment commands based on the yaw angle and offset of the adjacent container; output the spreader adjustment commands; and perform closed-loop control based on the attitude information fed back in real time by the attitude sensors on the spreader until the placement attitude of the container to be placed matches the attitude of the adjacent container. The spreader adjustment commands are used to drive the spreader movement.

[0027] Thirdly, embodiments of this application provide a control device, including: a memory and a processor;

[0028] The memory stores instructions that the computer executes;

[0029] The processor executes computer execution instructions stored in memory, causing the processor to perform a two-box operation method based on adjacent box orientation as described in the first aspect of this disclosure.

[0030] Fourthly, embodiments of this disclosure also provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the dual-box operation method based on adjacent box attitude as described in the first aspect of this disclosure.

[0031] Fifthly, embodiments of this disclosure also provide a computer program product comprising computer execution instructions, which, when executed by a processor, are used to implement the dual-box operation method based on adjacent box attitude as described in the first aspect of this disclosure.

[0032] The dual-container operation method, apparatus, equipment, and medium based on adjacent container posture provided in this disclosure first acquire the posture of the adjacent container and adjust the placement posture of the container to be placed accordingly. After confirming that the dual-container grabbing coordination conditions are met, the method then verifies whether the intended placement position is within a preset safety area before executing the container placement operation. This integrates the two mutually constraining objectives of precise alignment with adjacent containers and ensuring container safety—which are traditionally linked under a single control logic—into a single processing flow. It effectively solves the problem that existing automated container placement technologies cannot proactively pre-align for subsequent efficient dual-container grabbing operations while ensuring operational safety. This represents a significant technological advancement in improving the continuity and overall efficiency of automated terminal yard operations. Attached Figure Description

[0033] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0034] Figure 1 This diagram illustrates an application scenario of the dual-box operation method, apparatus, equipment, and medium based on adjacent box posture provided in this disclosure.

[0035] Figure 2 A flowchart illustrating a dual-box operation method based on adjacent box attitude provided in one embodiment of this disclosure;

[0036] Figure 3 A flowchart of a dual-box operation method based on adjacent box attitude provided in yet another embodiment of this disclosure;

[0037] Figure 4 A schematic diagram of the structure of a dual-box operation device based on the attitude of adjacent boxes, provided in yet another embodiment of this disclosure;

[0038] Figure 5 This is a schematic diagram of the structure of a control device provided in yet another embodiment of this disclosure.

[0039] The accompanying drawings have illustrated specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to specific embodiments. Detailed Implementation

[0040] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0041] In automated container terminal operations, to quickly grab containers, improve outbound efficiency, and achieve a highly efficient "single-in, double-out" logistics model, preparations must be made during the port entry phase for the simultaneous grabbing of two containers during subsequent outbound operations. The core challenge lies in the fact that placing a single container requires simultaneously meeting two mutually restrictive conditions: first, it must maintain precise alignment with adjacent containers (when grabbing two containers simultaneously, the spreader typically uses a uniform posture to grab both containers at the same time, rather than selectively grabbing different containers with different postures, as this would involve the same amount of data processing as grabbing individual containers sequentially), ensuring that the spreader can successfully grab both containers simultaneously; second, it must ensure that the container is completely within the designated safety area on the ground to guarantee stacking stability and operational safety. These two conditions rely on different and often conflicting spatial reference systems, making the simultaneous fulfillment of these conditions a critical bottleneck.

[0042] In related technologies, automated container placement technology mainly relies on identifying pre-drawn fixed container corner lines on the ground as the sole control reference. However, this method can only solve the absolute positional safety problem of the containers in one direction, and cannot perceive or adapt to the relative posture of adjacent containers. Furthermore, in order to ensure strict alignment with the fixed container corner lines, it is inherently unable to handle the contradictory requirement of aligning with the posture of adjacent containers. This either sacrifices alignment accuracy, affecting subsequent efficiency, or risks exceeding safety boundaries, leading to potential hazards. As a result, existing technologies lack a decision-making framework and control mechanism that integrates these two heterogeneous spatial constraints in a single operation, making it difficult to guarantee the reliability and stability of automated dual-container operations.

[0043] The method, apparatus, equipment, and medium for dual-container operations based on adjacent container attitude provided in this application adjust the current container to achieve alignment preparation by adjusting the attitude of the adjacent container, and then strictly verify whether it is in the safe zone. This transforms the originally mutually restrictive alignment requirements and safety constraints into a solvable problem of step-by-step decision-making and sequential satisfaction. Furthermore, by introducing the attitude of the adjacent container as the main adjustment benchmark and combining it with subsequent safety verification, a collaborative control mechanism is achieved to simultaneously handle heterogeneous spatial constraints.

[0044] Figure 1 This is a schematic diagram illustrating the application scenario of the dual-box operation method, apparatus, equipment, and medium based on adjacent box attitude provided in this application, such as... Figure 1 As shown, in the dual-container operation based on the attitude of adjacent containers, the controller 100 monitors the position and attitude of the container to be placed 110 and the adjacent container 120 respectively, and ensures that the container to be placed 110 can be conveniently grasped by the subsequent dual-container operation when it is placed.

[0045] It should be noted that, Figure 1The scenario shown includes a controller, a container to be placed, and adjacent containers. Only one or a specific number of these are used as examples for illustration, but this disclosure is not limited to this. That is to say, the number of controllers, containers to be placed, and adjacent containers can be arbitrary.

[0046] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0047] Figure 2 This is a flowchart illustrating the dual-box operation method based on adjacent box attitude provided in an embodiment of this application. The following is a summary of the process. Figure 2 The main process of the dual-box operation method based on the attitude of adjacent boxes is explained:

[0048] S201. Obtain the pose information of at least one existing container adjacent to the position of the container to be placed, and use the pose information as the pose of the neighboring container.

[0049] Specifically, in this embodiment of the disclosure, the executing entity is a controller used to control container lifting in an automated yard crane. This controller is typically an industrial computer or a high-performance PLC, integrating logic processing, motion control, and communication interfaces. It is responsible for receiving instructions, processing sensing data, and outputting control signals to drive the spreader to complete precise gripping and placement operations.

[0050] Unlike container placement control in related technologies, in this embodiment of the disclosure, the reference provided for the current container placement operation is a dynamically acquired spatial reference reference, rather than relying on fixed ground markings.

[0051] Therefore, the pose information here includes at least the orientation (yaw angle) and position (offset relative to the predetermined coordinate system) of the adjacent container on the horizontal plane to describe the actual state of the adjacent container in the yard (the adjacent container can be a container that is still being placed or a container that has already been placed, but the state controller of the adjacent container cannot control it in the current process to ensure the relative independence of the operation of different containers and avoid the safety risks caused by over-coupling).

[0052] In practice, the controller needs to actively perceive physical objects in the working environment.

[0053] In this embodiment, the controller can acquire data through an environmental sensing system integrated on the hoisting equipment (such as a crane trolley).

[0054] For example, two lidar sensors can be placed diagonally at the bottom of the vehicle to scan the area below in real time.

[0055] When the hoisted container arrives above the target bay (a bay is a standard unit of space in a ship or yard that is divided according to rules, and its function is to accurately locate and schedule operations), the controller will collect raw point cloud data of the area near the placement location through a sensing system (such as the lidar sensor mentioned above).

[0056] The controller will perform coordinate transformation on the collected data, transferring it from the sensor coordinate system to the gantry or trolley coordinate system.

[0057] Subsequently, instead of directly searching for containers, the controller determines whether there are any valid existing containers by analyzing the point cloud density and distribution characteristics within a specific region of interest.

[0058] For example, if the number of point clouds in a region is significantly higher than the ground background threshold and exhibits regular cuboid upper features, it can be preliminarily identified as the top surface of a container.

[0059] Once its existence is confirmed, the controller locks the target as a neighboring box and initiates the pose calculation process.

[0060] In practical applications, the controller can also fuse multi-frame scan data or combine short-term historical data to filter dynamic interference (such as temporary obstacles on the ground) to ensure that the identified neighboring containers are stably stacked containers.

[0061] In some embodiments, in actual terminal applications, this step needs to adapt to various complex yard scenarios, such as the reflection interference caused by snow or water accumulation on the surface of adjacent containers, or the difference in laser reflection intensity caused by the different shades of container colors. Therefore, the point cloud analysis of the controller can also be combined with preprocessing algorithms (such as filtering and noise reduction) to improve environmental adaptability.

[0062] S202. Adjust the placement posture of the container to be placed according to the posture of the adjacent container.

[0063] Specifically, in this step, the controller's control reference is switched from the simple corner position of the container to the relative and dynamic position of the adjacent container, so that the final placement posture of the currently hoisted container (i.e. the container to be placed) matches the posture of the adjacent container, creating conditions for close side-by-side arrangement in physical space, which facilitates subsequent double container grabbing.

[0064] The controller can compare the attitude of the adjacent container obtained in the first step with the attitude of the current spreader and the container being lifted, which is fed back in real time by attitude sensors (such as tilt sensors and encoders) on the spreader, and generate control deviation.

[0065] Then, the drive command is calculated by combining the motion control algorithm integrated inside the controller (such as PID control, model predictive control, etc., which are not limited here).

[0066] For example, if the attitude of the adjacent container shows a slight clockwise deflection angle, and the current spreader is oriented positively, the controller will generate a command to make a counterclockwise fine-tuning adjustment of the spreader's rotation mechanism to eliminate the angular deviation. At the same time, in the horizontal plane, the controller will also control the traveling mechanism of the trolley or carriage to make fine movements to align with the centerline of the adjacent container in the track direction.

[0067] Thus, the entire adjustment process forms a dynamic closed-loop control of output and feedback until the current estimated placement posture of the container and the posture of the target adjacent container reach a preset state.

[0068] S203. Determine that the adjusted orientation of the container to be placed and the orientation of the adjacent container meet the preset dual-container grabbing conditions.

[0069] Specifically, based on the previous adjustment, the controller does not directly and unconditionally execute the container placement, but must verify whether the spatial relationship between the two containers (the container to be placed and the adjacent container) after adjustment meets the "simultaneous grabbing" requirement of the subsequent process, that is, the dual container grabbing condition.

[0070] Dual-box gripping conditions are typically determined based on the tolerance range allowed by the mechanical structure of the dual-box spreader (such as a dual 20-foot spreader).

[0071] For example, the condition for grabbing two containers could be that the longitudinal centerlines of the two containers on the horizontal plane should be as parallel and close as possible.

[0072] The controller compares the real-time orientation of the container to be placed with the orientation of the neighboring container determined in the previous steps, and calculates the differences between the two in key dimensions.

[0073] Comparison dimensions typically include: the difference in yaw angles between the two (measuring parallelism), and the offset of their center positions in a direction perpendicular to the quayline (usually the track direction) (measuring proximity).

[0074] The controller has preset thresholds for these differences.

[0075] In some embodiments, the controller can be configured to dynamically load the corresponding allowable thresholds from a database or configuration file based on parameters such as the container size (e.g., 20 feet or 40 feet), the specific model of the spreader, and the current weather conditions (e.g., the threshold needs to be tightened in windy weather).

[0076] The controller determines that the dual-box grasping condition is met only when all calculated difference values ​​are less than their corresponding allowable thresholds, in order to ensure the reliability of automated operations.

[0077] S204. Determine that the intended placement location of the container to be placed is within the preset safe area, and perform the container placement operation.

[0078] Specifically, to prevent containers from being unintentionally placed outside the safety zone defined by the corner lines in the pursuit of perfect alignment with neighboring containers, which could lead to containers pressing against the lines, exceeding the load-bearing area, or being too close to other facilities and causing safety hazards, the controller will introduce an additional layer of safety constraints to ensure the safety of automated operations.

[0079] Specifically, before the final container placement, the controller uses a vision system (such as an industrial camera that shoots downwards) mounted on the lifting device to capture an image of the yard floor directly below it, and uses image recognition algorithms (such as target detection based on feature matching or deep learning) to identify the pre-painted corner markings of the containers in the image.

[0080] At this point, the controller can project the real-time precise outline of the current spreader (and the container to be placed) onto the image based on the calibration parameters of the visual model, or directly calculate the pixel coordinates of the four bottom corners of the container in the image and compare and analyze them with the identified corner line boundaries.

[0081] In practice, the controller can calculate the overlap between the container projection area and the standard container corner line area, or directly determine whether all bottom corner points are located inside the container corner line boundary.

[0082] The controller's final box-laying action is only triggered when the safety constraint is also met.

[0083] At this point, the controller will output the corresponding control command: first, stop all horizontal and rotational adjustment mechanisms, then control the lifting mechanism of the spreader to descend at a constant speed until the container lands stably, and finally open the turnlock to complete the container unloading operation.

[0084] In some embodiments, if the security check fails, the controller will not issue a descent command, but will instead enter an exception handling process, such as issuing an alarm, keeping the device hovering and reporting its status, and waiting for manual intervention.

[0085] Therefore, while ensuring efficiency improvement, we can guarantee the safety and reliability of automated dual-box operations without sacrificing basic safety.

[0086] The dual-container operation method based on adjacent container posture provided in this application first acquires the posture of the adjacent container and adjusts the placement posture of the container to be placed accordingly. After confirming that the conditions for dual-container grabbing coordination are met, it then verifies whether the intended placement position is within a preset safe area before executing the container placement operation. This integrates the two mutually constraining goals of precise alignment with adjacent containers and ensuring container safety—which are traditionally linked under a single control logic—into a single processing flow. It effectively solves the problem that existing automated container placement technologies cannot proactively pre-align for subsequent efficient dual-container grabbing operations while ensuring operational safety. This represents a significant technological advancement in improving the continuity and overall efficiency of automated terminal yard operations.

[0087] Figure 3 Another embodiment of this disclosure provides a dual-box operation method based on adjacent box attitude, in Figure 2 Based on the illustrated embodiment, the following is combined with Figure 3 The implementation process of the dual-box operation method based on the attitude of adjacent boxes is described in detail, and it includes the following steps:

[0088] S301. Based on the ranging sensor installed on the hoisting equipment, obtain point cloud data of the neighborhood of the location of the container to be placed.

[0089] The distance measuring sensor is positioned at a relative angle to the bottom surface of the hoisting equipment, and the mirror of the distance measuring sensor faces downwards perpendicular to the ground.

[0090] Specifically, the controller acquires raw environmental data through a specific sensor layout (such as a ranging sensor).

[0091] In this embodiment, the ranging sensor is preferably a lidar (LiDAR), and the two devices are specifically installed at opposite corners below the crane trolley, forming a diagonal layout.

[0092] Therefore, regardless of the orientation of the hoisting equipment (i.e., the trolley), the two sensors can cover a large diagonal area, increasing the viewing angle and reliability of scanning adjacent boxes, and avoiding blind spots that may occur due to box obstruction by a single sensor.

[0093] The sensor mirror is mounted vertically downwards to ensure that its scanning plane is basically parallel to the ground, enabling it to efficiently acquire vertical distance information of the surface of the storage yard below (including the ground and the top of the containers), thereby generating a three-dimensional point cloud.

[0094] When the controller receives a work instruction and moves the hoisting equipment above the target location, it triggers the two lidars to perform synchronous scanning.

[0095] In the acquired raw point cloud data, each point contains three-dimensional coordinates (x, y, z) in the sensor's own coordinate system and possible reflection intensity information.

[0096] The controller needs to unify these data into a fixed reference coordinate system (such as the vehicle coordinate system or geographic coordinate system) through the calibrated transformation relationship (including the rotation matrix R and the translation vector T), laying the foundation for subsequent regional analysis and calculation. The formula can be expressed as:

[0097] [x1, y1, z1] T = [RT] [x, y, z] T ,

[0098] This formula is a classic formula in coordinate transformation, so it will not be explained further here.

[0099] In some embodiments, the controller can also dynamically weight and fuse point clouds from multiple consecutive scans to filter out instantaneous interference points such as birds and moving people, ensuring data stability.

[0100] S302, It is determined that the number of point cloud data exceeds the third threshold.

[0101] Specifically, after completing the coordinate transformation, the controller does not directly process the entire point cloud, but instead delineates a region of interest (ROI) in the reference coordinate system based on the placement position specified by the task.

[0102] The ROI is typically a three-dimensional cubic space that covers the target container location and the ground adjacent to the front and rear container locations.

[0103] The controller will count the number of all point clouds falling within the ROI.

[0104] The third threshold is a value set based on actual testing. The basis for this setting is that, given a fixed sensor accuracy, installation height, and scanning angle, the number of point clouds generated within an ROI from an open ground (with only ground echoes) is a relatively stable and low value. However, when there are tall objects such as containers within the ROI, their top and sides will generate a large number of additional point clouds, causing the total number to surge.

[0105] For example, the third threshold can be set to 1.5 to 2 times the number of ground base point clouds (i.e., the lower value mentioned above).

[0106] If the number of point clouds counted by the controller exceeds the third threshold, it can be preliminarily determined that there are non-ground objects in the ROI, which are very likely containers.

[0107] Therefore, by conducting preliminary screening based on the number of point clouds, most operational scenarios without neighboring boxes can be quickly eliminated, reducing unnecessary complex calculations.

[0108] S303. Determine that there is at least one existing container in the neighborhood of the location of the container to be placed.

[0109] Specifically, based on the initial screening, when the number of point clouds exceeds the threshold, the controller needs to make a more precise judgment on whether the object is a valid "existing container".

[0110] At this point, the controller can perform clustering analysis on the point cloud within the ROI, for example, by using a clustering algorithm based on Euclidean distance (such as DBSCAN).

[0111] In practice, the controller uses a clustering algorithm to group spatially adjacent points into a class and calculates the geometric features of each point cloud cluster, such as volume, approximate size (length, width, height) and shape.

[0112] A standard container top dot cloud cluster should roughly present a rectangular plane, and its dimensions should be similar to the width and length of a standard-sized container (such as a 20-foot or 40-foot container).

[0113] The controller compares the size characteristics of the point cloud cluster with the standard size library of containers (allowing a certain error tolerance) to ultimately confirm whether the point cloud cluster corresponds to an "existing container".

[0114] If its existence is confirmed, the controller locks the point cloud cluster and marks it as a candidate neighbor box that requires further analysis.

[0115] Therefore, the initial screening based on the number of point clouds is upgraded to a further identification of objects that match the characteristics of containers.

[0116] S304. In response to the task information received from the task management system, determine that the target adjacent container that needs to be aligned with the container to be placed is an existing container located in front of or behind the placement position.

[0117] Specifically, in an automated terminal, the controller receives work instructions (i.e., task information) from the higher-level Task Management System (TOS).

[0118] The task information includes not only the target location coordinates, but also heap storage plan information.

[0119] For example, in order to achieve "single entry and double exit", TOS has planned to pair this single container with an existing container in the yard when arranging this single container collection task, so that it can be picked up at the same time later.

[0120] Therefore, the task information issued by TOS usually includes instructions such as "align with the container in front" or "align with the container behind".

[0121] At this point, the controller can determine the specific location (front or back) of the target adjacent box that needs to be aligned with in this operation by parsing the task information.

[0122] By combining one or more candidate neighbor boxes identified in step S303 and their positions in the coordinate system, the controller can uniquely determine the target neighbor box whose attitude needs to be calculated.

[0123] For example, if the task information requires "aligning with the container behind", and a container point cloud cluster is identified behind the target container position through step S303, the container behind will be identified as the target neighbor container regardless of whether a point cloud cluster corresponding to the container is identified in front.

[0124] This ensures the purposefulness and accuracy of the adjustment movements.

[0125] In some embodiments, if a container is not identified in the direction corresponding to the task information, the subsequent process can be terminated and an alarm message can be generated to indicate the existence of an execution anomaly.

[0126] S305. Obtain the pose information of the target neighboring box and use the pose information as the pose of the neighboring box.

[0127] Specifically, in this step, the controller will calculate the precise attitude parameters of the target neighbor box used for alignment control from the point cloud data of the target neighbor box obtained in the previous steps.

[0128] Furthermore, the method for determining the attitude of adjacent boxes specifically includes the following steps:

[0129] Step A1: Perform planar fitting on the point cloud data corresponding to the target neighbor box to obtain the planar point cloud of the upper surface of the target neighbor box.

[0130] Specifically, the controller first extracts the point cloud clusters that belong to the target neighboring boxes.

[0131] Since the top surface of the container is flat on a large scale, the controller can use a plane fitting algorithm (such as least squares or RANSAC) to fit these point clouds and obtain an optimal plane equation, for example:

[0132] Ax + By + Cz + D = 0.

[0133] This equation is a standard plane equation, so it will not be explained further here.

[0134] The fitted plane represents the average orientation of the container's top surface.

[0135] This allows us to filter out outliers caused by unevenness, markings, or a few small debris on the top of the box, resulting in a point cloud of the standard upper surface plane for subsequent accurate calculations.

[0136] Step A2: Based on the planar point cloud of the upper surface, extract the boundary features of the existing containers.

[0137] Specifically, determining the plane alone is insufficient to determine the container's orientation and precise boundaries. Therefore, in actual calculations, it is necessary to further extract the linear features of the container's top edge, such as the equation of the boundary line and the yaw angle of adjacent containers.

[0138] Step A3: Calculate the pose of the neighboring boxes based on the boundary features.

[0139] The attitude of the adjacent container includes at least the yaw angle of the adjacent container and the offset of the adjacent container along the direction of the trolley's movement.

[0140] Specifically, through the calculations in the aforementioned steps, two quantified core parameters can be obtained to represent the attitude of the neighboring container: the neighboring container yaw angle (yaw) and the neighboring container offset (y) along the direction of the trolley's movement.

[0141] This accurately describes the orientation and position of the adjacent boxes on the horizontal plane.

[0142] S306. Generate a spreader adjustment command based on the yaw angle and offset of the adjacent container.

[0143] Specifically, based on the calculated yaw angle of the adjacent container (which can be denoted as yaw) neighbor ) and the offset y of the adjacent box (which can be denoted as y neighbor The value of ) can be used by the controller as the target setpoint for the control system.

[0144] At the same time, the controller reads the real-time attitude of the container to be placed, fed back by the spreader encoder, tilt sensor, etc., and measures it using the current spreader yaw angle. current and current position y current Two quantities are used to represent this.

[0145] Then, the controller calculates the corresponding yaw angle deviation Δyaw and position deviation Δy, and the calculation formula can be expressed as:

[0146] Δyaw = yaw neighbor - yaw current Δy = y neighbor - y current .

[0147] Based on these deviations, the controller uses built-in control algorithms (such as proportional-integral controller algorithm, i.e. PID algorithm) to calculate specific drive commands.

[0148] For example, a positive Δyaw might require controlling the spreader's rotary motor to rotate clockwise, with the command size being proportional to Δyaw; a positive Δy might require controlling the trolley to move a certain distance in a specific direction.

[0149] At this point, the generated instructions are signals (such as pulses, speed, or position commands) that the underlying drive mechanism (rotary motor, trolley travel motor) can directly execute.

[0150] S307 outputs spreader adjustment commands and performs closed-loop control based on the attitude information fed back in real time by the attitude sensors on the spreader until the placement attitude of the container to be placed matches that of the adjacent container.

[0151] Among them, the spreader adjustment command is used to drive the spreader movement.

[0152] Specifically, the controller will send the adjustment command generated in step S306 to the corresponding drive actuator.

[0153] Meanwhile, the attitude sensors on the spreader (such as gyroscopes and encoders) will continuously feed the latest attitude data back to the controller.

[0154] The controller will continuously repeat the closed-loop feedback control process of "reading feedback → calculating new deviation → generating new instructions → output".

[0155] When real-time feedback yaw current and y current With the target value yaw neighbor and y neighbor When the differences are all less than a certain minimum threshold (i.e., entering a matching state), the controller considers the adjustment to be in place, stops outputting adjustment commands, and the system enters a steady-state waiting state.

[0156] This ensures the accuracy and dynamic stability of the adjustment.

[0157] S308. Determine the angle difference between the yaw angle of the container to be placed and the yaw angle of the adjacent container in its attitude, and calculate the positional offset between the container to be placed and the adjacent container along the preset direction.

[0158] Specifically, after the adjustments are completed, the controller needs to further verify the coordination conditions.

[0159] At this time, the yaw angle of the container to be placed is... box (i.e., yaw at this time) current (To distinguish it from the calculation formula in step S306) is the current stable yaw angle of the spreader (feedback from the sensor), and the yaw angle of the adjacent container. neighbor Given that the angle difference θ satisfies:

[0160] θ = |yaw box - yaw neighbor |,

[0161] The position offset is usually calculated with the direction perpendicular to the long side of the container as the preset direction, which is the usual trolley running direction (or the quay crane's main trolley direction).

[0162] The controller uses the coordinates of the current spreader center (or the estimated container center) in the reference coordinate system, and the offset y of the adjacent container center calculated by S305. neighbor (These coordinates need to be converted to absolute coordinates in the same coordinate system) Compare them and calculate the positional offset ΔY (y) of the two in that direction. box That is, y at this time current ):

[0163] ΔY = |y box - y neighbor |

[0164] S309. If the angle difference is less than the first threshold and the position offset is less than the second threshold, it is determined that the container to be placed and the adjacent container meet the double container grabbing conditions.

[0165] Specifically, the controller compares the calculated θ and ΔY with preset first and second thresholds.

[0166] For example, in the dual-box gripping condition, both the first threshold and the second threshold can be set according to the required accuracy. For instance, based on the mechanical design tolerance of the dual 20-foot spreaders and the engineering experience value of the alignment accuracy requirement of the automated lock head to the box hole, the first threshold and the second threshold can be set as: θ < 0.05°, ΔY < 0.05m.

[0167] Therefore, the specific setup method can be determined comprehensively based on the mechanical tolerance of the double-box lifting device, the fitting accuracy of the lock head to the box hole, and the safety operation specifications.

[0168] Only when both of these strict inequality conditions are met simultaneously will the controller logically determine that the dual-box grasping condition is satisfied.

[0169] This ensures the physical feasibility and safety of subsequent dual-box operations.

[0170] S310. Determine that the intended placement location of the container to be placed is within the preset safe area, and perform the container placement operation.

[0171] Specifically, this step is related to Figure 2 The corresponding steps in the illustrated embodiments are the same and will not be repeated here.

[0172] To facilitate understanding, the following provides a more detailed explanation of some of the calculation processes in the above steps:

[0173] The above steps involve the action of "extracting the boundary features of existing containers based on the planar point cloud of the upper surface," which is specifically implemented through the following steps:

[0174] Step B1: Determine the maximum and minimum coordinates of the upper surface planar point cloud in the preset direction.

[0175] Specifically, as mentioned above, the preset direction is usually chosen to be parallel or perpendicular to the direction of the dock track (such as the direction of the trolley's movement).

[0176] The controller traverses the point cloud to find the two points with the largest and smallest coordinate values ​​in that direction. These two points are usually located roughly at the two ends of a long side of the container's top surface.

[0177] Step B2: Starting with the maximum and minimum coordinate points, iteratively search for the lowest point of each local point cloud along the opposite direction to obtain the boundary point set.

[0178] Specifically, when the lidar shines vertically downwards, the reflection point on the top surface of the container has a relatively high y-coordinate (assuming the y-axis is upwards), while the reflection point on the side of the container or at the edge where the top and side meet will have a suddenly decreasing y-coordinate. By searching for the point with the minimum y-value within a small local area (width d), the top-view outline of the container's top edge can be captured.

[0179] Based on this principle, the controller first sets a small search range d (e.g., 0.1 meters) around the maximum and minimum points in the direction perpendicular to that side (i.e., the width direction of the container). Within this range, it searches for the point with the smallest y-coordinate value (assuming the y-axis is vertically upward), i.e., the local minimum point. The local minimum point is usually near the edges of the top and side surfaces.

[0180] Then, starting from the maximum and minimum points, move towards each other along the edge with a step size d. Each time you move, search for the lowest point within the new vertical slice.

[0181] This iterative process eventually yields a set of boundary points that depict the outline of a long side of the container's top surface.

[0182] Therefore, it can effectively capture edge information of containers and has strong noise resistance.

[0183] Step B3: Fit a straight line to the boundary point set based on the random sampling consensus algorithm, and determine the yaw angle of the adjacent box based on the parameters of the fitted straight line.

[0184] Specifically, after obtaining the set of boundary points, the controller can use the RANSAC algorithm to fit a straight line to it.

[0185] RANSAC can overcome interference from individual erroneous boundary points (such as those from bulges on the top of the container) and robustly fit the optimal straight line. The equation of this straight line (e.g., y = kx + b) represents the direction of one of the long sides of the container.

[0186] The slope k or angle arctan(k) of a straight line directly reflects the deflection angle of the container relative to the coordinate axis. This angle is called the yaw angle of the adjacent container.

[0187] Meanwhile, the offset of the adjacent box in the directional direction can be calculated by using the average coordinates of the boundary point set in the vertical direction (such as the direction of the vehicle).

[0188] The dual-container operation method based on adjacent container posture provided in this disclosure acquires the positional information of adjacent containers and dynamically adjusts the posture of the container to be placed accordingly. Then, it rigorously verifies whether the container meets quantified dual-container grabbing conditions and is within a preset safety area before placing it. This integrates the dynamic coordination requirement of alignment with adjacent containers and the static safety constraint of being within the container corner line into a unified control process, effectively overcoming the contradiction of the difficulty in balancing these two aspects in existing technologies. While ensuring absolute safety in container stacking, it proactively creates the necessary conditions for subsequent efficient single-in, double-out dual-container grabbing operations, thereby significantly improving the continuity and overall operational efficiency of automated terminal yard operations.

[0189] Figure 4 This is a schematic diagram of a dual-box operation device based on adjacent box posture, provided as an embodiment of this disclosure. Figure 4 As shown, the dual-box operation device 400 based on adjacent box attitude includes:

[0190] The acquisition module 410 is used to acquire the pose information of at least one existing container adjacent to the position of the container to be placed, and to use the pose information as the pose of the neighboring container.

[0191] The adjustment module 420 is used to adjust the placement posture of the container to be placed according to the posture of the adjacent container;

[0192] The judgment module 430 is used to determine whether the adjusted posture of the container to be placed and the posture of the adjacent container meet the preset dual-container grasping conditions.

[0193] The processing module 440 is used to determine that the intended placement location of the container to be placed is within a preset safe area and to perform the container placement operation.

[0194] Optionally, the judgment module 430 is specifically used to determine the angle difference between the yaw angle of the container to be placed and the yaw angle of the adjacent container in the attitude of the adjacent container, and to calculate the position offset of the container to be placed and the adjacent container along the preset direction; if the angle difference is less than the first threshold and the position offset is less than the second threshold, it is determined that the container to be placed and the adjacent container meet the dual container grabbing conditions.

[0195] Optionally, the acquisition module 410 is further configured to, before acquiring the pose information of at least one existing container adjacent to the location of the container to be placed, acquire point cloud data of the neighborhood of the location of the container to be placed based on a ranging sensor installed on the hoisting equipment, wherein the ranging sensor is set at a relative angle to the bottom surface of the hoisting equipment, and the mirror of the ranging sensor faces downward perpendicular to the ground; determine that the number of point cloud data exceeds a third threshold; and determine that there is at least one existing container in the neighborhood of the location of the container to be placed.

[0196] Optionally, the acquisition module 410 is further configured to, after determining that there is at least one existing container in the neighborhood of the location of the container to be placed, in response to the task information issued by the task management system, determine that the target neighboring container that needs to be aligned with the container to be placed is an existing container located in front of or behind the location to be placed; and acquire the pose information of the target neighboring container as the pose of the neighboring container.

[0197] Optionally, the acquisition module 410 is specifically used to perform planar fitting on the point cloud data corresponding to the target neighboring container to obtain the planar point cloud of the upper surface of the target neighboring container; extract the boundary features of the existing container based on the planar point cloud of the upper surface; and calculate the attitude of the neighboring container based on the boundary features, wherein the attitude of the neighboring container includes at least the yaw angle of the neighboring container and the offset of the neighboring container along the running direction of the trolley.

[0198] Optionally, the acquisition module 410 is specifically used to: determine the maximum and minimum coordinate points of the upper surface planar point cloud in a preset direction; starting from the maximum and minimum coordinate points, iteratively search for the lowest point of each local point cloud along the opposite direction to obtain the boundary point set; perform straight line fitting on the boundary point set based on the random sampling consensus algorithm, and determine the yaw angle of the adjacent box based on the parameters of the fitted straight line.

[0199] Optionally, the adjustment module 420 is specifically used to generate a spreader adjustment command based on the yaw angle and offset of the adjacent container; output the spreader adjustment command; and perform closed-loop control based on the attitude information fed back in real time by the attitude sensor on the spreader until the placement attitude of the container to be placed matches the attitude of the adjacent container. The spreader adjustment command is used to drive the spreader movement.

[0200] In this embodiment, the dual-container operation device based on the adjacent container posture solves the problem in related technologies that it is impossible to align the container with the adjacent container while ensuring that the container meets its own position requirements during dual-container operation by combining various modules.

[0201] Figure 5 This is a schematic diagram of the structure of a control device provided in one embodiment of the present disclosure, as shown below. Figure 5 As shown, the control device 500 includes a memory 510 and a processor 520.

[0202] The memory 510 stores a computer program that can be executed by at least one processor 520. This computer program is executed by at least one processor 520 to enable the control device to implement the battery SOC estimation method provided in any of the above embodiments.

[0203] The memory 510 and the processor 520 can be connected via a bus 530.

[0204] The relevant explanations can be understood by referring to the corresponding descriptions and effects in the method embodiments, and will not be repeated here.

[0205] One embodiment of this disclosure provides a computer-readable storage medium having a computer program stored thereon, the computer program being executed by a processor to implement the dual-box operation method based on adjacent box attitude as provided in any of the above embodiments.

[0206] The computer-readable storage medium may be ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0207] One embodiment of this disclosure provides a computer program product comprising computer-executable instructions that, when executed by a processor, are used to implement the dual-box operation method based on adjacent box attitude as provided in any of the above embodiments.

[0208] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0209] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments. The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification.

[0210] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0211] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A method for dual-box operation based on adjacent box posture, characterized in that, Includes the following steps: Obtain the pose information of at least one existing container adjacent to the location of the container to be placed, and use the pose information as the pose of the neighboring container. Adjust the placement posture of the container to be placed according to the posture of the adjacent container; Determine that the adjusted orientation of the container to be placed and the orientation of the adjacent container meet the preset dual-container grasping conditions; Once the intended placement location of the container is within the preset safe area, the container placement operation is executed.

2. The method according to claim 1, characterized in that, The determination that the adjusted orientation of the container to be placed and the orientation of the adjacent container meet the preset dual-container grasping conditions includes: Determine the angle difference between the yaw angle of the container to be placed and the yaw angle of the adjacent container in the adjacent container's attitude, and calculate the positional offset of the container to be placed and the adjacent container along a preset direction; If the angle difference is less than the first threshold and the position offset is less than the second threshold, it is determined that the container to be placed and the adjacent container meet the dual-container grabbing condition.

3. The method according to claim 1, characterized in that, Before acquiring the pose information of at least one existing container adjacent to the location of the container to be placed, the method further includes: Based on the distance measuring sensor installed on the hoisting equipment, point cloud data of the neighborhood of the location of the container to be placed is obtained, wherein the distance measuring sensor is set at a relative angle to the bottom surface of the hoisting equipment, and the mirror of the distance measuring sensor is facing downwards perpendicular to the ground. It is determined that the amount of point cloud data exceeds a third threshold; It is determined that there is at least one existing container in the neighborhood of the location of the container to be placed.

4. The method according to claim 3, characterized in that, After determining that there is at least one existing container in the neighborhood of the location of the container to be placed, the method further includes: In response to the task information received from the task management system, it is determined that the target adjacent container that needs to be aligned with the container to be placed is an existing container located in front of or behind the placement position. The step of acquiring the pose information of at least one existing container adjacent to the location of the container to be placed, and using the pose information as the pose of the neighboring container, includes: Obtain the pose information of the target neighboring box and use the pose information as the pose of the neighboring box.

5. The method according to claim 4, characterized in that, The step of obtaining the pose information of the target neighboring box and using the pose information as the pose of the neighboring box includes: Plane fitting is performed on the point cloud data corresponding to the target neighbor box to obtain the planar point cloud of the upper surface of the target neighbor box; Based on the planar point cloud of the upper surface, the boundary features of the existing container are extracted; Based on the boundary features, the attitude of the neighboring container is calculated, wherein the attitude of the neighboring container includes at least the yaw angle of the neighboring container and the offset of the neighboring container along the direction of the trolley's movement.

6. The method according to claim 5, characterized in that, The step of extracting the boundary features of the existing container based on the planar point cloud of the upper surface includes: Determine the maximum and minimum coordinate points of the upper surface planar point cloud in a preset direction; Starting with the maximum and minimum coordinate points, iteratively search for the lowest point of each local point cloud along the opposite direction to obtain the boundary point set; The boundary point set is fitted with a straight line based on the random sampling consensus algorithm, and the yaw angle of the adjacent box is determined based on the parameters of the fitted straight line.

7. The method according to claim 5, characterized in that, The step of adjusting the placement posture of the container to be placed according to the posture of the adjacent container includes: Based on the yaw angle and offset of the adjacent container, a spreader adjustment command is generated; The spreader adjustment command is output, and closed-loop control is performed based on the attitude information fed back in real time by the attitude sensor on the spreader until the placement attitude of the container to be placed matches the attitude of the adjacent container. The spreader adjustment command is used to drive the spreader movement.

8. A dual-box operation device based on adjacent box posture, characterized in that, include: The acquisition module is used to acquire the pose information of at least one existing container adjacent to the position of the container to be placed, and to use the pose information as the pose of the neighboring container. An adjustment module is used to adjust the placement posture of the container to be placed according to the posture of the adjacent container; The judgment module is used to determine whether the adjusted posture of the container to be placed and the posture of the adjacent container meet the preset double container grasping conditions. The processing module is used to determine that the intended placement location of the container to be placed is within a preset safe area and to perform the container placement operation.

9. A control device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1 to 7.

Images