A Fast and Accurate Path Planning Method, Device, Equipment and Medium for Tower Cranes

Abstract

A fast and accurate tower crane path planning method, device, equipment and medium provided by the present invention relate to the technical field of path planning. The present invention obtains relevant data of the tower crane, the object to be lifted and obstacles, and establishes a Cartesian coordinate system with the projection of the rotation center of the tower crane on the ground as the origin; then generates a dynamic buffer space for the object to be lifted in the Cartesian coordinate system, performs voxelization segmentation of all obstacles in the scene based on the OBB bounding box to generate cubic voxels, and generates inscribed cylinder voxels; maps the buffer space of the object to be lifted and each inscribed cylinder voxel of all obstacles in the Cartesian coordinate system into the configuration space to form axis-aligned bounding boxes; performs collision detection and path planning based on the dynamic axis-aligned bounding boxes in the configuration space to generate a collision-free lifting path to complete the path planning of the tower crane. This application can significantly improve the calculation efficiency and operability of tower crane path planning, reduce the lifting risk, and is applicable to dynamic and complex construction environments.

Description

Technical Field

[0001] The present invention relates to the technical field of tower crane path planning, and more specifically, to a fast and accurate tower crane path planning method, device, equipment and medium. Background Art

[0002] In the field of construction, tower cranes, as important lifting equipment, their path planning has a crucial impact on construction efficiency and safety. Due to strict safety regulations, the tower body and boom of a tower crane are usually not prone to collision risks. However, during the hoisting process, the lifted object is prone to collide with obstacles on the construction site due to its randomness and complexity. Therefore, precise path planning is required. Traditionally, tower crane operators and assistants rely on their own observations and experience to perform hoisting tasks. This method is not only cumbersome but also error-prone. In recent years, researchers have proposed various path planning methods, attempting to improve efficiency by optimizing path length or planning time. However, these methods often ignore the execution ability of crane operators, resulting in paths that are difficult to operate in practice and have potential safety hazards.

[0003] Existing path planning methods mainly include Rapidly-Exploring Random Trees (RRT) and its variants, genetic algorithms, particle swarm optimization, simulated annealing, and deterministic path planning algorithms such as Dijkstra and A*. Although these methods can optimize path length or time cost to a certain extent, there are still many deficiencies. First, many algorithms lack constraints in the search direction, resulting in paths that involve complex crane operations, increasing the operation difficulty and collision risk. Second, the limited availability and high cost of high-computing-performance devices on the construction site limit the practical application of evolutionary algorithms. At the same time, the low interpretability and repeatability of these algorithms also prevent their use in scenarios that require stable and precise operations. In addition, the paths planned by existing methods are mostly polyline paths with many nodes, resulting in frequent swinging of the hook, increasing the collision risk and energy consumption. At the same time, they are not compatible with the rotation-luffing motion characteristics of tower cranes, making it difficult for operators to operate according to the planned path. More importantly, existing methods do not fully consider the possible inertial swing of the lifted object during planning. Even if they do consider it, they usually adopt a conservative strategy, resulting in inaccurate path planning and long calculation time.

[0004] In dynamic scenarios, the performance of existing methods is particularly insufficient. Most studies only focus on path planning in static scenarios and ignore the impact of dynamic obstacles, resulting in paths that cannot adapt to the real-time changing construction environment. At the same time, the lack of accuracy in collision detection is also a major defect of existing technologies. Complex geometric calculations not only take a long time but may also result in the loss of the optimal path.

[0005] In view of this, the applicant specifically proposes this application. Summary of the Invention

[0006] The present invention aims to provide a method, device, equipment and medium for fast and accurate tower crane path planning, so as to solve the problems of complex path, low operability, high risk of hoisting inertial swing, low calculation efficiency and poor adaptability to dynamic scenarios in the existing methods.

[0007] To solve the above technical problems, the present invention is realized through the following technical solutions:

[0008] A method for fast and accurate tower crane path planning, including:

[0009] S1, establish a Cartesian coordinate system with the projection of the rotation center of the tower crane on the ground as the origin;

[0010] S2, generate a dynamic buffer space for the object to be lifted in the Cartesian coordinate system;

[0011] S3, perform voxelization segmentation of all obstacles in the scene based on the OBB bounding box to generate cubic voxels, and generate inscribed cylinder voxels inside the cubic voxels;

[0012] S4, map the dynamic buffer space of the object to be lifted and each inscribed cylinder voxel of all obstacles in the Cartesian coordinate system into the configuration space to form axis-aligned bounding boxes;

[0013] S5, perform collision detection and path planning based on dynamic axis-aligned bounding boxes in the configuration space to generate a collision-free hoisting path and complete the path planning of the tower crane.

[0014] Preferably, the dynamic buffer space is a cylinder surrounding the object to be lifted, defined as ;

[0015] Wherein, is the x value of the centroid coordinate of the cylinder of the dynamic buffer space; is the y value of the centroid coordinate of the cylinder of the dynamic buffer space; is the radius of the cylinder of the dynamic buffer space; is the centroid height value of the dynamic buffer space in the Cartesian coordinate system; represents the cylinder height value of the dynamic buffer space;

[0016] Let the length of the object to be lifted be , the width be , the height be , and the centroid coordinate of the object to be lifted be ;

[0017] Then, the maximum horizontal projection radius after the object to be lifted rotates by an angle around the vertical axis is:

[0018] ;

[0019] Among them, is the maximum horizontal projection radius of the object being lifted;

[0020] According to the current lifting speed and acceleration, the radius of the dynamic buffer space is adjusted in real time , and the formula is:

[0021] ;

[0022] Among them, is the radius of the dynamic buffer space at time t; g is the acceleration due to gravity; is the swing angle; is the static margin; is the lifting speed at time t;

[0023] The height value of the cylinder of the dynamic buffer space changes dynamically with the lifting and lowering of the hook, and the formula is:

[0024] ;

[0025] ;

[0026] Among them, is the current height value of the object being lifted; is the current height value of the tower crane boom; is the z value of the centroid coordinate of the object being lifted;

[0027] In the Cartesian coordinate system, the centroid coordinates of the dynamic buffer space are the same as those of the object being lifted in the x-axis and y-axis directions, that is = , = ; the centroid coordinate of the dynamic buffer space on the z-axis is , = .

[0028] Preferably, the process of generating cubic voxels by voxelizing all obstacles in the scene based on the OBB bounding box is specifically as follows:

[0029] Generate an OBB bounding box for all dynamic and static obstacles in the scene;

[0030] Set the side length of the cubic voxel to , and divide the OBB bounding box to generate a voxel network; among them, when the side length of the OBB bounding box is insufficient, it is rounded up;

[0031] Then set the voxel coordinates of each obstacle to ; wherein, and are horizontal plane indices; is the vertical plane index;

[0032] For each voxel , perform intersection detection between the voxel and the triangular mesh, and retain the voxels that intersect with the triangular mesh of the obstacle, and delete the non-intersecting voxels.

[0033] Preferably, after generating an inscribed cylinder voxel inside the cube voxel, merge the voxels with the same horizontal index and adjacent vertical indices in the inscribed cylinder voxel to obtain a merged large cylinder; the voxel radius of the merged large cylinder remains unchanged, and the height is the sum of the heights of the original multiple merged voxels.

[0034] Preferably, the three axes of the configuration space are the luffing s of the tower crane, the slewing angle and the lifting height h;

[0035] When the dynamic buffer space of the lifted object and each inscribed cylinder voxel of the dynamic and static obstacles are mapped to the configuration space, they are respectively mapped in the luffing s, slewing angle and the lifting height h of the tower crane, specifically:

[0036] The mapping expression of the dynamic buffer space of the lifted object in the luffing s direction is:

[0037] ;

[0038] Wherein, and respectively represent the minimum and maximum luffing values in the luffing s direction; is the x value of the centroid coordinate of the cylinder of the dynamic buffer space; is the y value of the centroid coordinate of the cylinder of the dynamic buffer space; is the radius of the cylinder of the dynamic buffer space;

[0039] The mapping expression of the dynamic buffer space of the lifted object in the slewing angle direction is:

[0040] ;

[0041] Wherein, and respectively represent the minimum and maximum values of the dynamic buffer space of the lifted object in the slewing angle direction;

[0042] The mapping expression of the dynamic buffer space of the lifted object in the height h direction is:

[0043] ;

[0044] Among them, and respectively represent the minimum and maximum values of the dynamic buffer space of the suspended load in the lifting height h direction; represents the cylinder height value of the dynamic buffer space; is the centroid height value of the dynamic buffer space in the Cartesian coordinate system;

[0045] For each inscribed cylinder of the dynamic obstacle and the static obstacle, the mapping method in the luffing s, slewing angle and the lifting height h direction is the same as that of the dynamic buffer space of the suspended load;

[0046] The cylinder mapped into the configuration space is an axis-aligned bounding box, and the time dimension t is introduced to reflect the real-time dynamic changes of the configuration space.

[0047] Preferably, it further includes: considering the inertial swing of the suspended load in the configuration space, and predicting the risk of inertial swing by expanding the axis-aligned bounding box. Specifically:

[0048] Expand the axis-aligned bounding box in the current moving direction and the opposite direction of the suspended load at the same time;

[0049] Among them, the expansion formula of the dynamic buffer space in the luffing direction is:

[0050] ;

[0051] ;

[0052] ;

[0053] Among them, and respectively represent the minimum luffing value and the maximum luffing value after the expansion of the dynamic buffer space in the luffing s direction; is the expansion amount in the moving direction; is the expansion amount in the opposite direction; is the acceleration; is the reverse coefficient; is the current height value of the suspended load;

[0054] When the suspended load has inertial swing in the slewing direction, the expansion method of the axis-aligned bounding box is the same as that in the luffing direction, and the value of the reverse coefficient is adjusted. The reverse coefficient can be adjusted according to the height of the suspended load.

[0055] Preferably, collision detection and path planning based on dynamic axis-aligned bounding boxes are performed in the configuration space, specifically as follows:

[0056] Let the current state of the tower crane be ; where is the luffing of the tower crane; is the slewing angle of the tower crane; is the lifting height of the tower crane; t is the current time;

[0057] The object being lifted moves discretely along the , , axes with an adjustable step size, while updating the time t;

[0058] Set the cost function , and the calculation formula is:

[0059] ;

[0060] where is the weight of the path length (set to 0.6, for example); is the weight of the time cost (set to 0.3, for example); is the weight of the acceleration penalty (set to 0.1, for example); represents the time difference; , , respectively represent the luffing difference, slewing angle difference, and height difference of the tower crane under the time difference;

[0061] Initialize the lifting point as the starting point , and add it to the priority queue; where is the initial luffing; is the initial slewing angle; is the initial lifting height; The priority queue is used to store the states of the object being lifted to be explored and is sorted according to the value of the cost function so as to preferentially explore the state with the minimum cost;

[0062] Select the state with the minimum cost from the current priority queue to generate sub-states in three directions ;

[0063] Calculate the dynamic axis-aligned bounding box of each sub-state and detect whether there is a collision between the axis-aligned bounding box of the object being lifted and the axis-aligned bounding box of the obstacle;

[0064] If there is no collision, calculate the cost of the sub-state , and Add it to the priority queue; continue to select the state with the smallest cost from the current priority queue to generate sub-states in three directions for collision detection until reaching the end state or the queue is empty, and obtain the hoisting path of the tower crane planned.

[0065] The present invention also provides a fast and accurate tower crane path planning device, including:

[0066] A Cartesian coordinate system establishment unit, configured to obtain relevant data of the tower crane, the object to be lifted, and obstacles, and establish a Cartesian coordinate system with the projection of the rotation center of the tower crane on the ground as the origin;

[0067] A dynamic buffer space generation unit, configured to generate a dynamic buffer space for the object to be lifted in the Cartesian coordinate system;

[0068] A voxelization unit, configured to perform voxelization segmentation based on the OBB bounding box for all static obstacles in the scene to generate cubic voxels, and generate inscribed cylinder voxels inside the cubic voxels;

[0069] A configuration space mapping unit, configured to map the dynamic buffer space of the object to be lifted and each inscribed cylinder voxel of all obstacles in the Cartesian coordinate system to the configuration space to form axis-aligned bounding boxes;

[0070] A path planning unit, configured to perform collision detection and path planning based on dynamic axis-aligned bounding boxes in the configuration space, generate a collision-free hoisting path, and complete the path planning of the tower crane.

[0071] The present invention also provides a fast and accurate tower crane path planning device, including a processor and a memory, where the memory stores a computer program, and the computer program can be executed by the processor to implement a fast and accurate tower crane path planning method as described above.

[0072] The present invention also provides a computer-readable storage medium, where computer-readable instructions are stored on the computer-readable storage medium, and when the computer-readable instructions are executed by the processor of the device where the computer-readable storage medium is located, a fast and accurate tower crane path planning method as described above is implemented.

[0073] In summary, compared with the prior art, the present invention has the following beneficial effects:

[0074] Through voxelization and configuration space mapping, the present invention simplifies the collision detection process and significantly improves the calculation efficiency. The generation of the dynamic buffer space and the inertial swing expansion ensure the accuracy and safety of the path planning. The planned path conforms to the kinematic characteristics of the tower crane, which is convenient for the operator to execute, reduces the operation difficulty and risk.

[0075] In addition, the present invention introduces the time dimension, takes into account the influence of dynamic obstacles, and makes the path planning applicable to complex construction environments.

[0076] In particular, the number of redundant voxels is reduced through the voxelization and cylinder voxel merging process, and the computational complexity of collision detection is reduced. The method for generating the dynamic buffer space combines the real-time adjustment mechanism of the hoisting speed and acceleration, avoiding the conservativeness of the traditional static buffer space, thereby improving the flexibility of path planning. The configuration space mapping and inertial swing expansion process achieve the accurate modeling of the inertial swing risk by expanding the axis-aligned bounding box, providing a higher safety guarantee for path planning.

[0077] The path planning algorithm of the present invention ensures the efficiency and operability of path planning through the design of the priority queue and the cost function. The design of the cost function in the path planning algorithm comprehensively considers the path length, time cost, and acceleration penalty, ensuring the global optimality of the planned path.

[0078] In summary, through the voxelization process, dynamic buffer space generation, configuration space mapping, and the design of an efficient path planning algorithm, the present invention realizes the accuracy and efficiency of path planning, having significant technical advantages and application prospects. Brief Description of the Drawings

[0079] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention, and thus should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0080] Figure 1 Schematic diagram of a fast and accurate tower crane path planning method provided for Embodiment 1.

[0081] Figure 2 Schematic diagram of the voxelization and cylinder voxel merging process provided for Embodiment 1.

[0082] Figure 3 Schematic diagram of a fast and accurate tower crane path planning device provided for Embodiment 2.

[0083] The present invention will be further described in detail below in conjunction with the drawings and specific embodiments. Detailed Embodiments

[0084] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0085] Embodiment 1

[0086] Embodiment 1 of the present invention provides a fast and accurate tower crane path planning method, which can be implemented by a fast and accurate tower crane path planning device (hereinafter referred to as the path planning device), and particularly, is executed by one or more processors in the path planning device.

[0087] In this embodiment, the path planning device may be an electronic device equipped with a processor, and the processor has a computer program for the fast and accurate tower crane path planning method and the computer program can be executed, such as a computer, a smart phone, a smart tablet, a workstation, etc., which is not limited herein.

[0088] In this embodiment, the configuration space is an abstract space concept, and each point in it represents a specific configuration of a robot (such as a tower crane). Configuration refers to a set of independent variables required to completely describe the position and orientation of a robot in space. For example, for a robot with n degrees of freedom, the configuration space is an n-dimensional space, and each point in the space is determined by n coordinate values, and these n coordinate values respectively correspond to the positions or angles of the respective joints of the robot.

[0089] Voxelization: Divide the three-dimensional space into many small cubic units, namely voxels. Each voxel has specific attributes, such as position, color, density, etc. By assigning and managing the attributes of these voxels, the geometric shape, appearance characteristics, etc. of a three-dimensional object or scene are represented.

[0090] Buffer Space: In normal collision detection, it is determined whether a collision occurs based on the exact geometric shape and position of an object. The setting of a buffer space is to demarcate a space range slightly larger than the actual geometric size around the object. When it is detected that another object enters this buffer space, it is considered that there may be a potential collision risk, and relevant processing or warnings are carried out in advance, rather than waiting until the object actually touches the actual boundary to react.

[0091] Path Planning: A sequence of points or curves connecting a starting position and an ending position is called a path, and the strategy for constructing the path is called path planning.

[0092] As Figure 1 shown, a fast and accurate tower crane path planning method includes steps S1 to S5.

[0093] S1. Taking the projection of the rotation center of the tower crane on the ground as the origin, a Cartesian coordinate system is established.

[0094] In the information acquisition stage of this embodiment, information such as the lifting point of the tower crane, the target end point, the position of the tower crane, the maximum radius and the minimum radius of the tower crane is obtained; information such as the position of the object to be lifted and the size of the object to be lifted is obtained; information such as the position and size of the obstacle is obtained; so as to carry out subsequent path planning and collision detection.

[0095] In an actual construction scenario, in order to achieve efficient tower crane path planning, it is first necessary to establish a Cartesian coordinate system. In this coordinate system, a dynamic buffer space for the object to be lifted is generated.

[0096] In this embodiment, taking the projection of the rotation center of the tower crane on the ground as the origin , a Cartesian coordinate system is established , where:

[0097] Axis: The initial direction of the tower crane boom;

[0098] Axis: The horizontal direction perpendicular to the Ox axis;

[0099] Axis: Perpendicular to the ground, with the upward direction being the positive direction.

[0100] S2. Generating a dynamic buffer space for the object to be lifted in the Cartesian coordinate system.

[0101] Specifically, the dynamic buffer space is a cylinder surrounding the object to be lifted, defined as . We need to generate a dynamic cylindrical buffer space for the lifted object. This cylinder needs to completely surround the lifted object regardless of how the lifted object rotates around the vertical axis (i.e., the z-axis). Calculate based on the geometric dimensions and motion characteristics of the lifted object, and update the buffer space in real time according to the motion of the tower crane (luffing, slewing angle, lifting height).

[0102] Among them, is the x value of the centroid coordinate of the cylinder of the dynamic buffer space; is the y value of the centroid coordinate of the cylinder of the dynamic buffer space; is the radius of the cylinder of the dynamic buffer space; is the centroid height value of the dynamic buffer space in the Cartesian coordinate system; represents the height and length value of the cylinder of the dynamic buffer space.

[0103] Let the length of the lifted object be , the width be , and the height be . The centroid coordinate of the lifted object is ;

[0104] Then, the maximum horizontal projection radius after the lifted object rotates by an angle around the vertical axis is:

[0105] ;

[0106] Among them, is the maximum horizontal projection radius of the lifted object;

[0107] According to the current lifting speed and acceleration, adjust the radius of the dynamic buffer space in real time. The formula is:

[0108] ;

[0109] Among them, is the radius of the dynamic buffer space at time t; g is the acceleration due to gravity; is the swing angle (which can be collected in real time by the lifting acceleration sensor); is the static margin;

[0110] The height and length value of the cylinder of the dynamic buffer space changes dynamically with the lifting and lowering of the hook. The formula is:

[0111] ;

[0112] ;

[0113] Among them, is the current height value of the lifted object; is the current height value of the tower crane boom; is the z value of the centroid coordinate of the lifted object.

[0114] In the Cartesian coordinate system, the centroid coordinates of the dynamic buffer space are the same as those of the lifted object in the x-axis and y-axis directions, that is = , = ; the centroid coordinate of the dynamic buffer space on the z-axis is , = .

[0115] By dynamically adjusting the buffer space, the conservatism of the traditional static buffer space is avoided, and at the same time, the adaptability of path planning is improved.

[0116] S3. Voxelize all obstacles in the scene to generate cubic voxels based on the OBB (Oriented Bounding Box) bounding box, and generate inscribed cylinder voxels inside the cubic voxels.

[0117] After generating the dynamic buffer space of the lifted object, in order to accurately plan the optimal feasible collision-free lifting route, it is necessary to perform voxelization segmentation on the dynamic and static obstacles in the construction scene according to the triangular mesh of the objects in the BIM model. Specifically:

[0118] Generate an OBB (Oriented Bounding Box) bounding box for all dynamic and static obstacles in the scene to ensure the accuracy of collision detection and not be affected by the turning of dynamic obstacles. The OBB bounding box is an aligned bounding box with the object, which can better adapt to the shape and direction of the object.

[0119] Based on the OBB bounding box, the object is segmented into cubic voxels with a preset side length, that is, voxelization. For the insufficient part of the bounding box side length, the ceiling method is used to ensure complete coverage of the object.

[0120] Let the side length of the cubic voxel be , ensure that is small enough to accurately represent the shape of the object, but not too small to avoid excessive computational overhead. Then perform cubic voxel segmentation on the OBB bounding box to generate a voxel network.

[0121] For example, if the side length of the bounding box is only enough to divide 8 voxels, but there are still some edges left, then it is finally divided into 9 voxels.

[0122] In the voxel network, let the cylinder voxel coordinates of each obstacle be ; among them, , is the horizontal plane index; is the vertical plane index; voxel coordinates include the x value of the centroid coordinates of the cylinder , y value , the radius of the cylinder , the height value of the cylinder and the centroid height value of the cylinder in the Cartesian coordinate system .

[0123] For each voxel , consider the voxel as a small cube, perform the intersection detection between the voxel and the triangular mesh (i.e., check whether the vertices or edges of the triangle intersect with the cube), and retain the voxels that intersect with the triangular mesh of the obstacle, and delete the non-intersecting voxels.

[0124] In order to implement the method for establishing the inertial buffer space and simplifying the collision detection in the subsequent method, an inscribed cylinder voxel is generated inside the cubic voxel, and the voxel index remains unchanged.

[0125] Then, merge the voxels with the same horizontal index and adjacent vertical indices in the inscribed cylinder voxels to obtain a merged large cylinder; the voxel radius of the merged large cylinder remains unchanged, and the height is the sum of the heights of the original multiple merged voxels.

[0126] As Figure 2 shown, merge the voxels with the same horizontal index and adjacent vertical indices , , . The voxel radius of the merged large cylinder is the same as the original / / , but the height is the sum of the merged voxels.

[0127] S4. Map the dynamic buffer space of the object being lifted and each inscribed cylinder voxel of all obstacles in the Cartesian coordinate system into the configuration space to form an axis-aligned bounding box.

[0128] In order to better establish the inertial dynamic buffer space and simplify the collision detection process and enhance the executability of the planned path, a configuration space is designed. Specifically, the configuration space represents the state space of the tower crane, and its three axes are the luffing s of the tower crane, the slewing angle and the lifting height h.

[0129] When the dynamic buffer space of the object being lifted and each inscribed cylinder voxel of the dynamic and static obstacles are mapped into the configuration space, they are respectively mapped at the luffing s of the tower crane, the slewing angle and the lifting height h. Specifically:

[0130] The mapping expression of the dynamic buffer space of the lifted object in the luffing s direction is:

[0131] ;

[0132] Where, , respectively represent the minimum luffing value and the maximum luffing value of the dynamic buffer space of the lifted object in the luffing s direction; is the x value of the centroid coordinate of the cylinder of the dynamic buffer space; is the y value of the centroid coordinate of the cylinder of the dynamic buffer space; is the radius of the cylinder of the dynamic buffer space;

[0133] The mapping expression of the dynamic buffer space of the lifted object in the slewing angle direction is:

[0134] ;

[0135] Where, , respectively represent the minimum value and the maximum value of the dynamic buffer space of the lifted object in the slewing angle direction;

[0136] The mapping expression of the dynamic buffer space of the lifted object in the height h direction is:

[0137] ;

[0138] Where, , respectively represent the minimum value and the maximum value of the dynamic buffer space of the lifted object in the lifting height h direction; represents the height value of the cylinder of the dynamic buffer space; is the centroid height value of the dynamic buffer space in the Cartesian coordinate system.

[0139] The mapping calculation method of each inscribed cylinder of the dynamic obstacle and the static obstacle in the luffing s, slewing angle and the lifting height h direction is the same as that of the dynamic buffer space of the lifted object, which will not be elaborated here.

[0140] The cylinder mapped into the configuration space is an axis-aligned bounding box, and the time dimension t is introduced to reflect the real-time dynamic change of the configuration space. The axis-aligned bounding box is used to simplify the collision detection.

[0141] Considering the inertial swing of the lifted object, the modeling of the inertial swing space in the configuration space is simpler. The risk of inertial swing is predicted by expanding the axis-aligned bounding box. The specific expansion method is:

[0142] Expand the axis-aligned bounding box in both the current moving direction and the opposite direction of the object being lifted;

[0143] Among them, the expansion formula of the dynamic buffer space in the luffing direction is:

[0144] ;

[0145] ;

[0146] ;

[0147] Among them, , respectively represent the minimum luffing value and the maximum luffing value after the expansion of the dynamic buffer space in the luffing s direction; is the expansion amount in the moving direction; is the expansion amount in the opposite direction; is the acceleration; is the reverse coefficient (which can be calibrated according to historical hoisting data); is the current height value of the object being lifted.

[0148] When the object being lifted has an inertial swing in the slewing direction, the expansion method of the axis-aligned bounding box is the same as that in the luffing direction, except for the value of the reverse coefficient .

[0149] So far, the path planning problem that originally needed to perform complex geometric collision detection in the Cartesian coordinate system is transformed into the collision detection of a simple dynamic axis-aligned bounding box in the configuration space, and the dynamic buffer space of inertial swing can be better constructed, and the planned path is more efficient and highly executable.

[0150] S5. Perform collision detection and path planning based on the dynamic axis-aligned bounding box in the configuration space to generate a collision-free hoisting path and complete the path planning of the tower crane.

[0151] Specifically, let the current state of the tower crane be ; Among them, is the luffing of the tower crane; is the slewing angle of the tower crane; is the lifting height of the tower crane; t is the current time;

[0152] The object being lifted moves discretely with adjustable step sizes along the , , axes, and the time t is updated simultaneously;

[0153] Set the cost function , and the calculation formula is:

[0154] ;

[0155] Among them, is the weight of the path length; is the weight of the time cost; is the weight of the acceleration penalty; represents the time difference; 、 、 respectively represent the luffing difference, slewing angle difference, and height difference of the tower crane under the time difference;

[0156] Initialize the lifting point as the starting point , and add it to the priority queue; among them, is the initial luffing; is the initial slewing angle; is the initial lifting height; the priority queue is used to store the states of the suspended object to be explored and sorted according to the value of the cost function so as to preferentially explore the state with the minimum cost;

[0157] Select the state with the minimum cost from the current priority queue to generate sub-states in three directions ;

[0158] Calculate the dynamic axis-aligned bounding box of each sub-state and detect whether there is a collision between the axis-aligned bounding box of the suspended object and the axis-aligned bounding box of the obstacle;

[0159] If there is no collision, calculate the cost of the sub-state , and add it to the priority queue;

[0160] Continue to select the state with the minimum cost from the current priority queue to generate sub-states in three directions for collision detection until the target end state is reached or the queue is empty, and obtain the hoisting path of the planned tower crane.

[0161] The queue being empty means that the target cannot be reached.

[0162] In this step, the framework of the A* algorithm can also be used for path planning.

[0163] The method of the present invention combines a dynamic buffer space, OBB bounding box voxelization, configuration space mapping, and collision detection based on an axis-aligned bounding box to ensure the accuracy and efficiency of path planning.

[0164] In an actual application scenario, assuming that there are multiple dynamic obstacles and static obstacles at the construction site, the tower crane needs to transport the suspended object safely from the starting position to the target position First, generate the dynamic buffer space for the lifted object according to the above steps, and perform voxelization segmentation and cylinder voxel merging on the obstacles in the scene. Subsequently, map all relevant objects into the configuration space to form a dynamic axis-aligned bounding box. Anticipate the risk of inertial swing by expanding the axis-aligned bounding box, and use the design of the priority queue and cost function during the path planning process to ensure that the planned path not only conforms to the kinematic characteristics of the tower crane but also can avoid all obstacles. The finally generated path not only reduces the collision risk and energy consumption caused by the frequent swing of the hook but also significantly improves the lifting efficiency and safety.

[0165] The technical solution of the present invention provides clear technical guidance through clear mathematical formulas and specific implementation steps. For example, the generation formula of the dynamic buffer space and the inertial swing expansion formula are both derived based on physical models, with high scientificity and practicality. The cost function design in the path planning algorithm comprehensively considers the path length, time cost, and acceleration penalty to ensure the global optimality of the planned path. In addition, the voxelization and cylinder voxel merging process reduces the computational complexity of collision detection by reducing the number of redundant voxels. The generation method of the dynamic buffer space combines the real-time adjustment mechanism of the lifting speed and acceleration, avoiding the conservativeness of the traditional static buffer space, thereby improving the flexibility of path planning. The configuration space mapping and inertial swing expansion process achieve accurate modeling of the inertial swing risk by expanding the axis-aligned bounding box, providing higher safety guarantees for path planning.

[0166] In summary, the present invention solves various problems in the prior art through voxelization processing, dynamic buffer space generation, configuration space mapping, and the design of an efficient path planning algorithm, and has significant technical advantages and application prospects.

[0167] Embodiment 2

[0168] As Figure 3 shown, the second embodiment of the present invention also provides a fast and accurate tower crane path planning device, including:

[0169] A Cartesian coordinate system establishment unit, configured to obtain relevant data of the tower crane, the lifted object, and the obstacles, and establish a Cartesian coordinate system with the projection of the rotation center of the tower crane on the ground as the origin;

[0170] A dynamic buffer space generation unit, configured to generate a dynamic buffer space for the lifted object in the Cartesian coordinate system;

[0171] A voxelization unit, configured to perform voxelization segmentation based on the OBB bounding box for all obstacles in the scene to generate cubic voxels, and generate inscribed cylinder voxels inside the cubic voxels;

[0172] A configuration space mapping unit, configured to map the dynamic buffer space of the suspended object and each inscribed cylinder voxel of all obstacles in the Cartesian coordinate system into the configuration space to form axis-aligned bounding boxes;

[0173] A path planning unit, configured to perform collision detection and path planning based on the dynamic axis-aligned bounding boxes in the configuration space, generate a collision-free hoisting path, and complete the path planning of the tower crane.

[0174] Embodiment III

[0175] The third embodiment of the present invention further provides a fast and accurate tower crane path planning device, which includes a memory and a processor. The memory stores a computer program, and the computer program can be executed by the processor to implement the fast and accurate tower crane path planning method as described above.

[0176] Embodiment IV

[0177] The fourth embodiment of the present invention further provides a computer-readable storage medium. Computer-readable instructions are stored on the computer-readable storage medium. When the computer-readable instructions are executed by the processor of the device where the computer-readable storage medium is located, the fast and accurate tower crane path planning method as described above is implemented.

[0178] In several embodiments provided by the embodiments of the present invention, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device and method embodiments described above are merely illustrative. For example, the flowcharts in the drawings show the possible architectures, functions, and operations of devices, methods, and computer program products according to multiple embodiments of the present invention. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and the module, program segment, or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0179] In addition, the functional modules in each embodiment of the present invention may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

[0180] When the above-mentioned functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, an electronic device, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs. It should be noted that in this article, the term "including", "comprising", or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article, or device including a series of elements not only includes those elements but also includes other elements not explicitly listed, or also includes elements inherent to such a process, method, article, or device. Without further limitations, an element defined by the statement "including a..." does not exclude the existence of additional identical elements in the process, method, article, or device including the said element.

[0181] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The singular forms "a", "the", and "said" used in the embodiments of the present invention and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0182] It should be understood that the term "and / or" used herein is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this article generally represents an "or" relationship between the associated objects before and after.

[0183] Depending on the context, the word "if" as used herein can be interpreted as "when", "while", "in response to determining", or "in response to detecting". Similarly, depending on the context, the phrase "if determined" or "if detected (stated condition or event)" can be interpreted as "when determined", "in response to determining", "when detecting (stated condition or event)", or "in response to detecting (stated condition or event)".

[0184] The "first / second" mentioned in the embodiments is only used to distinguish similar objects and does not represent a specific order for the objects. It can be understood that the "first / second" can be interchanged in a specific order or sequence when permitted. It should be understood that the objects distinguished by the "first / second" can be interchanged under appropriate circumstances so that the embodiments described herein can be implemented in an order other than those illustrated or described herein.

[0185] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A fast and accurate tower crane path planning method, characterized in that, Including: S1. Obtain the relevant data of the tower crane, the lifted object and the obstacles, and establish a Cartesian coordinate system with the projection of the rotation center of the tower crane on the ground as the origin; S2. Generate a dynamic buffer space for the object being lifted in the Cartesian coordinate system; wherein, the dynamic buffer space is a cylinder surrounding the object being lifted and is defined as ; Among them, is the x value of the centroid coordinate of the cylinder in the dynamic buffer space; is the y value of the centroid coordinate of the cylinder in the dynamic buffer space; is the radius of the cylinder in the dynamic buffer space; is the centroid height value of the dynamic buffer space in the Cartesian coordinate system; represents the height and length value of the cylinder in the dynamic buffer space; Let the length of the object to be lifted be , the width be , and the height be . The centroid coordinates of the object to be lifted are ; Then, the maximum horizontal projection radius after the suspended object rotates by an angle about the vertical axis is: ​ ； Among them, is the maximum horizontal projection radius of the lifted object; According to the current hoisting speed and acceleration, adjust the radius of the dynamic buffer space in real time , and the formula is: ； Among them, is the radius of the dynamic buffer space at time t; g is the acceleration due to gravity; is the swing angle; is the static margin; is the hoisting speed at time t; The height-length value of the cylinder in the dynamic buffer space changes dynamically with the lifting and lowering of the hook. The formula is: ； ； Among them, is the current height value of the object being lifted; is the current height value of the tower crane jib; is the z-axis value of the centroid coordinates of the object being lifted; In the Cartesian coordinate system, the centroid coordinates of the dynamic buffer space are the same as those of the suspended object's centroid in the x-axis and y-axis directions, that is = , = ; the centroid coordinate of the dynamic buffer space on the z-axis is , = ; S3. Perform voxelization segmentation based on the OBB bounding box for all obstacles in the scene to generate cubic voxels, and generate inscribed cylinder voxels inside the cubic voxels; S4. Map the dynamic buffer space of the lifted object and each inscribed cylinder voxel of all obstacles in the Cartesian coordinate system into the configuration space to form axis-aligned bounding boxes; S5. Perform collision detection and path planning based on the dynamic axis-aligned bounding boxes in the configuration space to generate a collision-free lifting path and complete the path planning of the tower crane.

2. A method for rapid and accurate tower crane path planning according to claim 1, characterized in that The process of performing voxelization segmentation based on the OBB bounding box for all obstacles in the scene to generate cubic voxels is specifically: Generate an OBB bounding box for all dynamic and static obstacles in the scene; Set the side length of the cubic voxel to be , and divide the OBB bounding box to generate a voxel network; among them, when the side length of the OBB bounding box is insufficient, round up; Then set the voxel coordinates of each obstacle to be ; where , is the horizontal plane index; is the vertical plane index; For each voxel , perform voxel-triangle mesh intersection detection, and retain the voxels that intersect with the triangle mesh of the obstacle, and delete the non-intersecting voxels.

3. A rapid and accurate tower crane path planning method according to claim 1, characterized in that After generating the inscribed cylinder voxels inside the cubic voxels, merge the voxels with the same horizontal index and adjacent vertical indices in the inscribed cylinder voxels to obtain a merged large cylinder; the voxel radius of the merged large cylinder remains unchanged, and the height is the sum of the heights of the original multiple merged voxels.

4. A method for rapid and accurate tower crane path planning according to claim 1, characterized in that , the three axes of the configuration space are respectively the luffing s, slewing angle and lifting height h of the tower crane; When the dynamic buffer space of the lifted object and each inscribed cylinder voxel of the dynamic and static obstacles are mapped to the configuration space, they are respectively mapped to the luffing s, slewing angle and lifting height h of the tower crane. Specifically: The mapping expression of the dynamic buffer space of the lifted object in the luffing s direction is: ； Among them, and respectively represent the minimum and maximum variable amplitudes in the luffing s direction; is the x value of the centroid coordinate of the cylinder in the dynamic buffer space; is the y value of the centroid coordinate of the cylinder in the dynamic buffer space; is the radius of the cylinder in the dynamic buffer space; The mapping expression of the dynamic buffer space of the lifted object at the slewing angle direction is: ； Among them, and respectively represent the minimum and maximum values of the dynamic buffer space of the suspended object in the direction of the slewing angle ; The mapping expression of the dynamic buffer space of the lifted object in the height h direction is: ； Among them, , respectively represent the minimum and maximum values of the dynamic buffer space of the suspended object in the lifting height h direction; represents the cylinder height value of the dynamic buffer space; is the centroid height value of the dynamic buffer space in the Cartesian coordinate system; The mapping calculation method of each inscribed cylinder of the dynamic obstacle and the static obstacle in the boom amplitude s, slewing angle in the hoisting height h direction is the same as that of the dynamic buffer space of the load to be lifted; The cylinder mapped into the configuration space is an axis-aligned bounding box, and the time dimension t is introduced to reflect the real-time dynamic changes in the configuration space.

5. A fast and accurate tower crane path planning method according to claim 4, characterized in that It also includes: Considering the inertial swing of the lifted object in the configuration space, and predicting the risk of inertial swing by expanding the axis-aligned bounding box. Specifically: Expand the axis-aligned bounding box in both the current moving direction and the opposite direction of the lifted object; Among them, the expansion formula of the dynamic buffer space in the luffing direction is: ； ； ； Among them, and respectively represent the minimum variable amplitude value and the maximum variable amplitude value after the dynamic buffer space expands in the variable amplitude s direction; is the expansion amount in the moving direction; is the reverse expansion amount; is the acceleration; is the reverse coefficient; is the current height value of the suspended object; When the suspended object swings inertially in the slewing direction, the axis-aligned bounding box expansion method is the same as that in the luffing direction, and the value of the reverse coefficient is adjusted.

6. The rapid and accurate tower crane path planning method according to claim 1, wherein Performing collision detection and path planning based on the dynamic axis-aligned bounding boxes in the configuration space is specifically: Let the current state of the tower crane be ; where is the luffing of the tower crane; is the slewing angle of the tower crane; is the lifting height of the tower crane; t is the current time; The object being lifted moves discretely along the , , axis with an adjustable step size, while updating the time t; Set the cost function , and the calculation formula is as follows: ； Among them, is the weight of the path length; is the weight of the time cost; is the weight of the acceleration penalty; represents the time difference; , , respectively represent the luffing difference, slewing angle difference, and height difference of the tower crane under the time difference; Initialize the lifting point as the starting point , and add it to the priority queue; where is the initial luffing; is the initial slewing angle; is the initial lifting height; the priority queue is used to store the states of the suspended object to be explored and sorted according to the cost function f(S) to give priority to exploring the state with the minimum cost; Select the state with the minimum cost from the current priority queue Generate sub-states in three directions ; Calculate each sub-state of the dynamic axis-aligned bounding box, and detect whether a collision occurs between the axis-aligned bounding box of the object being lifted and that of the obstacle; If there is no collision, calculate the cost of the sub-state , and add it to the priority queue; Continue to select the state with the minimum cost from the current priority queue to generate sub-states in three directions for collision detection until reaching the end state or the queue is empty, and obtain the lifting path of the planned tower crane.

7. A fast and accurate tower crane path planning device, characterized in that Including: A Cartesian coordinate system establishment unit, which is used to obtain the relevant data of the tower crane, the lifted object and the obstacles, and establish a Cartesian coordinate system with the projection of the rotation center of the tower crane on the ground as the origin; A dynamic buffer space generation unit for generating a dynamic buffer space for the object being lifted in a Cartesian coordinate system; wherein, the dynamic buffer space is a cylinder surrounding the object being lifted, defined as ; Among them, is the x value of the cylindrical centroid coordinate of the dynamic buffer space; is the y value of the cylindrical centroid coordinate of the dynamic buffer space; is the radius of the cylinder of the dynamic buffer space; is the centroid height value of the dynamic buffer space in the Cartesian coordinate system; represents the cylinder height value of the dynamic buffer space; Let the length of the object to be lifted be , the width be , and the height be . Let the centroid coordinates of the object to be lifted be ; Then, the maximum horizontal projection radius after the suspended object rotates by an angle about the vertical axis is as follows: ； Among them, is the maximum horizontal projection radius of the object to be lifted; According to the current hoisting speed and acceleration, adjust the radius of the dynamic buffer space in real time , the formula is: ； Among them, is the radius of the dynamic buffer space at time t; g is the acceleration due to gravity; is the swing angle; is the static margin; is the hoisting speed at time t; The height-length value of the cylinder in the dynamic buffer space changes dynamically with the lifting and lowering of the hook. The formula is: ； ； Among them, is the current height value of the object to be lifted; is the current height value of the tower crane's jib; is the z-axis value of the centroid coordinates of the object to be lifted; In the Cartesian coordinate system, the centroid coordinates of the dynamic buffer space are the same as those of the suspended object's centroid in the x-axis and y-axis directions, that is = , = ; the centroid coordinate of the dynamic buffer space on the z-axis is , = ; A voxelization unit, which is used to perform voxelization segmentation based on the OBB bounding box for all obstacles in the scene to generate cubic voxels, and generate inscribed cylinder voxels inside the cubic voxels; A configuration space mapping unit, which is used to map the dynamic buffer space of the lifted object and each inscribed cylinder voxel of all obstacles in the Cartesian coordinate system into the configuration space to form axis-aligned bounding boxes; A path planning unit, which is used to perform collision detection and path planning based on the dynamic axis-aligned bounding boxes in the configuration space to generate a collision-free lifting path and complete the path planning of the tower crane.

8. A fast and accurate tower crane path planning device, characterized in that, It includes a processor and a memory, and a computer program is stored in the memory. The computer program can be executed by the processor to implement a fast and accurate tower crane path planning method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, Computer-readable instructions are stored on the computer-readable storage medium. When the computer-readable instructions are executed by the processor of the device where the computer-readable storage medium is located, a fast and accurate tower crane path planning method according to any one of claims 1-6 is implemented.

Images