A visual engineering monitoring method

Through visual engineering monitoring methods and physics engine simulation, combined with three-dimensional laser scanning and ant colony algorithm optimization, the problem of low control accuracy of traditional cluster towers is solved, the precise simulation of tower crane motion trajectory and optimization of layout scheme is realized, and the construction site is improved Safety and efficiency.

CN118643679BActive Publication Date: 2024-11-26CSCEC INT URBAN CONSTR CO LTD
View PDF 9 Cites 0 Cited by

Patent Information

Application Number
CN202411110051.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-14
Publication Date
2024-11-26
Estimated Expiration
2044-08-14

AI Technical Summary

Technical Problem

The traditional tower group control method relies on human operation, resulting in low control accuracy and insufficient efficiency, especially in large-scale construction projects, tower group coordination and control accuracy are difficult to achieve.

Method used

Using visual engineering monitoring methods, the tower crane motion trajectory is simulated by establishing a three-dimensional model and a physics engine, combining three-dimensional laser scanning technology to obtain construction site data, build a multi-objective optimization model, and use ant colony algorithm to optimize the tower crane layout scheme to improve control accuracy and efficiency.

Benefits of technology

The precise simulation and interference detection of the tower crane motion trajectory are realized, the tower crane layout plan is optimized, the safety and efficiency of the construction site are improved, blind spots and interference are reduced, and the overall performance of tower group operations is improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118643679B_ABST
    Figure CN118643679B_ABST
Patent Text Reader

Abstract

The present application discloses a visualized engineering monitoring method for tower crane group control, which relates to the field of tower crane group control, including: obtaining attribute parameters of the tower crane, using a parametric modeling method to establish a three-dimensional model of the tower crane; importing the established three-dimensional model of the tower crane into a physical engine, using a rigid body dynamics algorithm to simulate and solve the motion trajectory of the tower crane; obtaining a three-dimensional model of an obstacle, using a collision detection algorithm to determine the spatial interference between the motion trajectory of the tower crane and the three-dimensional model of the obstacle; obtaining point cloud data of the construction site; extracting the constraint conditions of the construction site through a point cloud segmentation and extraction algorithm; generating a tower crane group layout plan in a BIM software platform; establishing a multi-objective optimization function with the goal of minimizing the blind area of ​​the tower crane coverage and maximizing the site coverage rate; using an ant colony algorithm to obtain the optimal tower crane group layout plan that meets the multi-objective optimization function; in view of the low precision of tower crane group control in the prior art, the present application improves the tower crane group control precision.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of tower group control, and in particular to a visual engineering monitoring method. Background Art

[0002] In the field of construction engineering, especially in large-scale construction and infrastructure projects, the coordinated use of tower groups (multiple tower cranes) plays a vital role in improving construction efficiency and ensuring construction safety. As the main vertical transportation tool, tower cranes play a core role in material handling and equipment lifting at construction sites. However, with the continuous expansion and increasing complexity of construction projects, traditional tower crane operation and management methods face many challenges, especially in terms of tower group coordination and control accuracy. Therefore, the development of more advanced tower group control and monitoring technologies to improve construction efficiency and safety has become an important demand for the development of the industry.

[0003] Traditional tower crane operation relies on the experience of operators and the command of on-site managers. This method of relying on manual operation has obvious limitations in control accuracy and efficiency. The operation trajectory, operating range and mutual interference of tower cranes can only rely on on-site command and experience to avoid collisions and accidents, which can easily lead to problems such as large blind spots and inaccurate control. In addition, as the scale of the project expands, the efficiency of a single tower crane can no longer meet the construction needs, and the complexity and coordination difficulty of the tower crane system have also increased.

[0004] In the related art, for example, Chinese patent document CN115465784A provides a regional tower crane monitoring and control method and system based on project control, the method includes: obtaining a task progress list of multiple tower cranes in a project, the task progress list includes the material requirements, safety monitoring requirements, intelligent obstacle avoidance requirements, and priority scheduling requirements of each tower crane every day; according to various requirements, the transport vehicles are scheduled, each working mechanism of the tower crane is controlled in real time using real-time video, obstacle avoidance planning is performed between multiple tower cranes according to the spatial position relationship of multiple tower cranes, and the tasks of multiple tower cranes are performed according to the sorted overall task priority sorting list. However, the priority scheduling requirements in this scheme involve sorting the operating tasks of multiple tower cranes according to the importance of the tasks, which not only needs to consider the efficiency of a single tower base but also the entire project. This complex scheduling affects the accuracy of group tower control. Summary of the invention

[0005] In view of the problem of low control accuracy of tower groups existing in the prior art, the present application provides a visual engineering monitoring method, which improves the control accuracy of tower groups by establishing a three-dimensional model and constructing multi-objective optimization.

[0006] The purpose of this application is achieved through the following technical solutions.

[0007] This specification provides a visual engineering monitoring method, including: obtaining the attribute parameters of the tower crane, obtaining the geometric parameters such as the spatial position coordinates (x, y, z), tower height H and arm length L of the tower crane through digital measurement or BIM model extraction. Using the parametric modeling method, taking the attribute parameters of the tower crane as constraint variables, and quickly generating a three-dimensional solid model of the tower crane in the BIM software through parametric family file calling and instance modification. Importing the three-dimensional model of the tower crane into the physical engine, setting the material properties, gravity acceleration and other physical parameters of the tower crane. Using the rigid body dynamics algorithm based on the Newton-Euler equation, taking the position, velocity, angular velocity and the like of the tower crane as initial conditions, the motion trajectory of the tower crane is calculated by time step integration. Obtaining the three-dimensional geometric model of obstacles such as terrain and buildings in the construction site, constructing it through three-dimensional modeling software or reconstructing it through point cloud data. Importing the three-dimensional model of the obstacle into the physical engine, setting its material properties and collision layer. Using acceleration algorithms such as hierarchical bounding box trees, using the separating axis theorem to realize collision detection between the tower crane model and the obstacle model, it is determined whether the tower crane interferes with the obstacle during movement. The construction site is scanned at multiple sites and multiple perspectives using a 3D laser scanner to obtain high-precision point cloud data of the site. Through pre-processing such as point cloud registration and filtering, site elements such as terrain, roads, and buildings are extracted using segmentation algorithms such as regional growing and RANSAC to obtain constraints such as hardened area boundaries, terrain undulations, and building outlines. In the BIM software platform, the 3D model of the tower crane, obstacle model, and site constraints are imported into the same coordinate system. Through 3D scene roaming and graphic interaction, the initial position of the tower crane is set through human-computer interaction on the premise of meeting the site constraints, and a 3D layout plan for the tower group is generated. A multi-objective optimization model is established, with the optimization goals of minimizing the blind area of ​​the tower crane and maximizing the site coverage rate. The constraints include the tower crane's own parameters, site terrain, hardened area, building boundaries, and the safe distance between the tower cranes. The tower crane's motion trajectory, collision detection results, and the initial tower group layout plan are input into the optimization model, and the improved ant colony algorithm is used to solve it. With blind area coverage and coverage rate as optimization goals, and on-site constraints as search boundaries, the optimization is iterated continuously through the pheromone update mechanism and probability selection rules until the convergence conditions are met, and the optimal tower group layout coordinates are output as the final solution. The optimal tower group layout solution is imported into the 3D visualization engine, and the tower crane appearance texture, site environment map, etc. are added. The lighting material is set to generate a realistic 3D visualization model of the tower crane group. The tower crane operation process is simulated through roaming animation, combined with real-time hoisting parameter display, to intuitively present the overall operation of the tower crane group.

[0008] Among them, the physics engine is a computer software system that simulates the real physical world. It uses mathematical models and numerical calculation methods to approximate the motion behavior and interaction phenomena of objects under physical effects. The physics engine mainly includes two modules: rigid body dynamics and collision detection. It is used to calculate the motion trajectory, velocity, acceleration and other kinematic parameters of objects, as well as to detect collisions, contacts and spatial relationships between objects.

[0009] Furthermore, a three-dimensional model of a tower crane is established, including: defining the acquired tower crane attribute parameters as parameter variables, including the position coordinates (x, y, z) of the tower crane, the tower height H, the tower arm length L, the tower arm angle α, etc. of the tower crane. Parametric family files of different types of tower cranes are pre-created in the BIM software, such as boom type, inner climbing type, etc. The family file contains the main components of the tower crane, such as the tower body, tower arm, hook, etc., and the size of each component is defined as a parameter variable. A geometric constraint relationship between parameter variables is established in the family file, such as the tower body height is equal to the tower height H, the tower arm length is equal to L, the tower arm angle corresponds to α, etc. Flexible adjustment of the family file is achieved through parameter association. The acquired tower crane attribute parameters are assigned to the parameter variables of the family file, and a three-dimensional solid model of the tower crane with actual size is automatically generated through the parametric modeling engine.

[0010] Furthermore, the simulation and solution of the motion trajectory of the tower crane includes: obtaining the kinematic parameters of the tower crane, including the position vector of the tower crane in the physical engine coordinate system , velocity vector , acceleration vector , angular velocity vector and the angular acceleration vector According to the Newton-Euler equation, the rigid body dynamics equation of the tower crane is established: , , where m is the mass of the tower crane, I is the moment of inertia of the tower crane, and are the resultant external force and moment of the crane. The implicit Euler algorithm is used for numerical solution. Assuming the time step is Δt, the current moment is t, and the next moment is t+Δt, the position, velocity and angular velocity of the crane satisfy: , , , where and They are The acceleration and angular acceleration at the moment. The motion trajectory parameters of the tower crane at each time step can be obtained , and .

[0011] Furthermore, judging the spatial interference between the motion trajectory of the tower crane and the three-dimensional model of the obstacle includes: obtaining the three-dimensional geometric data of the obstacles (such as terrain, buildings, etc.) in the construction site through three-dimensional laser scanning, oblique photogrammetry and other methods. Using the triangular patch algorithm, the three-dimensional point cloud data or surface model of the obstacle is converted into a triangular patch mesh model. The triangular patch method is a commonly used three-dimensional model representation method that approximates the surface of an object through a combination of plane triangles. The converted obstacle triangular patch model is imported into the physical engine, and scaled and translated according to the actual size and position. In the physical engine, based on the parameters of the tower crane motion trajectory, the spatial position and posture of the three-dimensional model of the tower crane at each time step are obtained through simulation calculation. For each time step, the hierarchical bounding box tree algorithm is used to first calculate the axis-aligned bounding box (AABB) of the three-dimensional model of the tower crane and the obstacle model, and by comparing whether the two bounding boxes intersect, the obvious non-intersection situation is quickly eliminated. If the two bounding boxes intersect, the triangle patch intersection detection algorithm is used to traverse all the triangle patches in the tower crane 3D model and the obstacle model. By judging whether the triangles are coplanar and whether there are shared vertices, the two models can be accurately detected to see whether there are intersecting patches. If there are intersecting triangle patches, it is determined that there is spatial interference between the tower crane's motion trajectory and the obstacle, and the tower crane's layout position needs to be adjusted or the motion path needs to be optimized. If there are no intersecting patches, it is considered that there is no interference between the tower crane and the obstacle. Repeat the traversal of all time steps on the tower crane's motion trajectory to achieve collision interference detection of the tower crane's full cycle motion process.

[0012] Furthermore, the established 3D model of the tower crane is imported into the physical engine, and the blind area of ​​the tower crane is obtained. The established 3D entity model of the tower crane is imported into the physical engine scene, and the material properties, collision layer and other physical parameters of the tower crane are set. According to the tower crane motion trajectory parameters obtained by simulation, the spatial position coordinates (x, y, z) of the tower crane hook at each time step are extracted to form the hook motion trajectory point set. For each point in the hook trajectory point set , taking this point as the starting point of the ray and the hoisting direction of the tower crane (usually vertically upward) as the ray direction, generate ray Ri. Use the ray casting algorithm to calculate the ray The intersection point with the 3D obstacle model imported into the physics engine. Read the triangles of the 3D obstacle model one by one, and for each triangle , determine the ray Whether it intersects with it. If the ray With triangles If they intersect, calculate the intersection point , and record the intersection to the starting point of the ray Distance For all rays Intersecting triangles, find the intersection point closest to the starting point of the ray , which is the first intersection point. Calculate the starting point of the ray The first intersection The distance between , and compare it with the preset line of sight threshold For comparison. , then the ray starting point is considered Intersection The area between is the sight-blocking area, and this area is added to the set S. , then the ray starting point is considered Intersection There is no line of sight blocking, so no recording is performed. Repeat all points in to get the set of all sight-blocked areas . All the sight-blocking areas in the set S are merged, and the overlapping areas are removed to obtain the blind area B covered by the tower crane during the entire lifting process. The blind area B is visualized together with the three-dimensional model of the tower crane and the obstacle model to provide an intuitive basis for the optimization of the tower crane layout plan. Different colors or transparencies can be used to mark the blind areas to facilitate analysis by construction management personnel.

[0013] Furthermore, the established three-dimensional model of the tower crane is imported into the physical engine, and the lifting efficiency of the tower crane is calculated, including: obtaining the following parameters from the product manual or measured data of the tower crane equipment: rated lifting weight Q, indicating the maximum lifting capacity of the tower crane at different amplitudes, in tons (t); maximum lifting height Hmax and minimum lifting height Hmin, indicating the vertical motion range of the tower crane hook, in meters (m); variable amplitude speed v, indicating the rotation speed of the tower crane at different amplitudes, in degrees / minutes (° / min) or meters / minutes (m / min). Discretization of motion trajectory: The calculated parameters of the tower crane motion trajectory are discretized using a numerical integration algorithm. Taking the Euler method as an example, the discretization steps are as follows: the time interval [0, T] of the motion trajectory is divided into n equally spaced time steps, with a step size of Δt=T / n; for the i-th time node ti (i=0, 1, 2,..., n), the position coordinates of the tower crane hook are calculated according to the kinematic equation : ,in, , and They are The velocity components of the tower crane hook in the x, y, and z directions at the moment. As the input for subsequent calculations. Calculate the hoisting height and amplitude: For each discretized motion trajectory coordinate point , using the tower crane dynamics equation and kinematics equation, calculate the corresponding lifting height and amplitude According to the hoisting height and amplitude of the tower crane, the spatial coverage volume V on the entire motion trajectory is calculated. Taking into account the equipment parameters of the tower crane and the actual operation coverage volume, the hoisting efficiency E of the tower crane is calculated using a weighted algorithm.

[0014] Furthermore, the tower crane dynamic equation and kinematic equation include: Establishment of the tower crane dynamic equation: The Lagrangian equation is used to establish the tower crane dynamic equation. The Lagrangian equation is a dynamic modeling method based on the energy principle, which is suitable for analyzing the motion characteristics of multi-degree-of-freedom systems. The establishment steps are as follows: a. Select generalized coordinates: According to the structural characteristics of the tower crane, select appropriate generalized coordinates , such as tower arm elevation angle, rotation angle, hook lifting distance, etc., where n is the number of degrees of freedom. Differentiate the generalized coordinate q with time to obtain the generalized velocity Solve the kinetic energy expression: Calculate the total kinetic energy of the system based on the mass, moment of inertia and velocity distribution of each tower crane member. Solve the potential energy expression: Calculate the total potential energy of the system based on the position height and spring stiffness of each tower crane member . Constructing the Lagrangian function: Defining the Lagrangian function ,in, represents generalized velocities, i.e., the first-order derivative of generalized coordinates with respect to time, which describes the motion state of the system; q represents generalized coordinates, which is used to describe the position or configuration of the system; Represents the system at a given position q and velocity The Lagrangian function value under .

[0015] Applying the Lagrange equations: For the i-th generalized coordinate , apply the following equation: ,

[0016] d / dt(∂L / ∂q̇i)-∂L / ∂qi=Qi, where represents the i-th generalized coordinate, which is used to describe the position or configuration of the system; represents the i-th generalized velocity, that is, the first-order derivative of the i-th generalized coordinate with respect to time; Represents the partial derivative Perform another time derivative; Qi is the non-conservative generalized force corresponding to the i-th generalized coordinate. Solve the differential equation of motion: combine the n equations obtained to form the dynamic differential equation group of the tower crane, and use the numerical integration algorithm to solve it to obtain the motion response of each tower crane member.

[0017] Establishment of the kinematic equation of the tower crane: The kinematic equation of the tower crane is established using the Denavit-Hartenberg (DH) parameter method. The DH parameter method is a method for describing the kinematic relationship of an open-chain mechanism. By establishing a homogeneous transformation matrix between adjacent rods, the forward and inverse kinematics of the mechanism can be solved. Establishment of the coordinate system: According to the DH rule, corresponding coordinate systems are established on each moving rod of the tower crane, including the base coordinate system, tower body coordinate system, tower arm coordinate system, and hook coordinate system. Definition of DH parameters: For two adjacent coordinate systems i-1 and i, four DH parameters are defined: Indicates the shortest distance from the x-axis of the i-1th coordinate system to the z-axis of the i-th coordinate system; Indicates the angle of rotation of the x-axis of the i-1th coordinate system around the z-axis of the i-th coordinate system; The distance from the origin of the i-1th coordinate system to the common normal along the z-axis of the i-th coordinate system; The angle at which the x-axis of the ith coordinate system rotates around the z-axis of the ith coordinate system. Establish a transformation matrix: According to the DH parameters, establish a homogeneous transformation matrix between adjacent coordinate systems : ,in, Represents the homogeneous transformation matrix from the i-1th coordinate system to the i-th coordinate system; Indicates rotation around the z axis (Joint angle) angle, Indicates translation along the z-axis (connecting rod offset) distance, Indicates translation along the x-axis (connecting rod length) distance, Indicates rotation around the x-axis (Connecting rod torsion angle) angle. Solve the forward kinematics: Multiply the transformation matrices of each adjacent coordinate system to obtain the total transformation matrix from the base coordinate system to the hook coordinate system , according to the position and posture of the hook coordinate system, calculate the spatial position and posture of the hook relative to the base. Solve inverse kinematics: Given the target position and posture of the hook, infer the joint variables of each moving rod, and use analytical method, numerical method or optimization algorithm to solve the inverse kinematics equation to obtain the motion parameters of each rod of the tower crane.

[0018] Furthermore, the detailed technical solution for calculating the tower crane hoisting efficiency is as follows:

[0019] The formula adopts the form of weighted power function, which comprehensively considers multiple influencing factors such as the lifting capacity, lifting speed, lifting height and space coverage volume of the tower crane. Compared with the simple linear weighting method, this formula can more accurately evaluate the actual working performance of the tower crane. Among them, the weight coefficient The value of can be adjusted according to the actual project requirements and the characteristics of the tower crane model, reflecting the relative importance of different factors to the lifting efficiency. The introduction of can reflect the nonlinear relationship between various parameters and hoisting efficiency. When a certain parameter changes, its influence on efficiency will also change accordingly. At the same time, the ratio of the actual value of each parameter to the rated value is used in the formula, which eliminates the dimensional difference between different tower crane models and facilitates the horizontal comparison and optimization of the performance of different tower cranes.

[0020] Lifting height calculation formula: This formula takes into account the influence of the boom pitch angle β on the hoisting height during the tower crane luffing process, as well as the hanging height of the hook relative to the boom. Compared with the traditional simplified calculation method, this formula can more accurately describe the actual hoisting height of the tower crane. In practical applications, the tower crane's amplitude angle β and boom height can be obtained by measurement or calculation. , combined with the hook hanging height , solve the real-time hoisting height H of the tower crane. This formula can be used to guide the height control and obstacle avoidance planning of the tower crane, and improve the safety and accuracy of the hoisting operation.

[0021] Lifting range calculation formula: This formula is based on the geometric relationship of the tower crane, and takes the actual boom length L and the amplitude angle α of the tower crane as input to calculate the real-time lifting amplitude R of the tower crane. Compared with the setting of fixed amplitude value, this formula can dynamically reflect the amplitude state of the tower crane and provide a basis for optimizing the lifting range. In practical applications, the amplitude angle α of the tower crane can be collected in real time by the angle sensor, and the current lifting amplitude R of the tower crane can be calculated in combination with the boom length L. This information can be used to assist the anti-collision control and dynamic avoidance of the tower crane, and improve the collaborative operation efficiency of the tower crane group.

[0022] Hoisting speed calculation formula: , which comprehensively considers the influence of the tower crane's luffing motion and the hook lifting motion on the lifting speed. The product of the lifting amplitude R represents the horizontal movement speed of the hook; the lifting speed of the hook Characterizes the speed of the hook in the vertical direction. The vector sum of the two is the composite lifting speed v of the tower crane. The introduction of this formula can achieve refined control and optimization of the tower crane lifting speed. By adjusting the variable angular velocity and hook lifting speed , it can improve the hoisting efficiency and rhythm of the tower crane and reduce energy consumption and equipment loss while meeting the process requirements.

[0023] Spatial coverage volume calculation formula: , the formula uses the form of triple integral to solve the volume of the entire working space of the tower crane. The lower and upper limits of the integral correspond to the minimum and maximum lifting heights, rotation angles and lifting amplitudes of the tower crane, respectively, which fully describe the three-dimensional operating range of the tower crane. The spatial coverage volume V is introduced as a performance evaluation index of the tower crane, which can quantitatively analyze the operating capacity and utilization rate of the tower crane under actual working conditions. By comparing the coverage volumes of different tower crane layout schemes, the equipment configuration and construction organization of the tower crane can be optimized, the lifting blind spots and interference can be reduced, and the utilization rate of the site space can be improved. At the same time, substituting the coverage volume V into the lifting efficiency calculation formula can comprehensively evaluate the working performance of the tower crane, taking into account efficiency and coverage range, and provide a quantitative basis for tower crane selection and scheme optimization.

[0024] Furthermore, the optimal tower group layout scheme that satisfies the multi-objective optimization function is obtained. The multi-objective optimization function is established: Objective function 1: , objective function 2: The improved multi-objective optimization function takes into account the mean square error of the proportion of the blind area covered by the tower crane. and site coverage Two indicators. Among them, the mean square error of the blind area ratio It reflects the discrete degree of blind area of ​​each tower crane. The smaller the value, the closer the blind area of ​​each crane is to the others, and the more balanced the layout plan is. Site coverage It reflects the effective coverage of the construction site by the tower group layout plan. The larger the value, the wider the coverage of the layout scheme and the higher the utilization efficiency. Compared with the traditional single-objective optimization, the improved multi-objective function can more comprehensively evaluate the performance of the tower cluster layout scheme. While ensuring the coverage effect, it also takes into account the load balance between the tower cranes, avoiding the situation where the blind area of ​​individual tower cranes is too large or the utilization rate is too low, and improving the overall efficiency of the tower cluster operation.

[0025] Transformation of multi-objective optimization functions: , the weighted Tchebycheff method is used to transform the multi-objective optimization problem into a single-objective optimization problem. This method introduces the ideal point For reference, the weighting factor Balance the importance of different objectives and use the infinite norm to measure the deviation between the solution and the ideal point, so that the optimization solution is as close to the Pareto frontier as possible. At the same time, introduce the penalty function The constraints are processed. When the solution violates the constraints, the value of the penalty function will increase significantly, causing the objective function value to deteriorate, so that the algorithm automatically searches within the feasible domain. It is used to balance the weight between the objective function and the penalty term. Its value needs to be adjusted according to the characteristics of the problem and the solution requirements. Compared with the traditional weighted summation method, the weighted Tchebycheff method can better handle convex and non-convex Pareto fronts and generate a more evenly distributed non-inferior solution set. The introduction of the penalty function combines the constraint conditions with the objective function, simplifying the construction and solution process of the optimization model.

[0026] Solution of the optimal tower group layout scheme: Taking the tower crane 3D model, motion trajectory parameters, obstacle 3D model, spatial interference results and initial tower group layout scheme as input, and the constraints of the construction site as boundary parameters, the improved ant colony algorithm is used to solve the transformed single-objective optimization function. The ant colony algorithm is a heuristic optimization algorithm that simulates the foraging behavior of ant colonies, and has strong global search capabilities and robustness. According to the characteristics of the tower group layout problem, the classic ant colony algorithm can be improved from the following aspects: optimize the encoding method of decision variables, use real number encoding instead of discrete encoding, directly use the tower crane coordinates as decision variables, reduce encoding and decoding operations, and improve search efficiency; improve the pheromone update mechanism, introduce dynamic evaporation coefficients and local update strategies, accelerate algorithm convergence, and avoid premature and stagnation; combine the characteristics of the problem, design heuristic information and state transition rules, such as introducing factors such as tower crane spacing and height difference, to assist ants in selecting a better tower position combination; in the algorithm iteration process, call the tower crane configuration verification module and spatial interference check module in real time, perform feasibility checks and performance evaluations on the current solution, and dynamically adjust the search direction of the ants. By improving and adapting the ant colony algorithm, we can make full use of its advantages of parallel search and positive feedback mechanism to efficiently solve the optimization problem of tower group layout. Compared with traditional genetic algorithms and particle swarm algorithms, the improved ant colony algorithm can better handle multi-constraint and multi-objective combination optimization problems, and the solution efficiency and solution quality are more guaranteed.

[0027] Specifically, minF consists of two parts: the weighted Tchebycheff term and the constraint violation measure. is the weighted Tchebycheff term: and are the two objective functions in the original problem, and are their respective ideal values, that is, under the constraints, The minimum and The maximum value of and is the weight coefficient of the two objectives, which is used to balance the importance of the two objectives. The calculation of this item is to first calculate the difference between the two objective function values ​​and their ideal values. and , then multiply by their respective weight coefficients, and finally take the maximum of the two weighted differences. This method can balance multiple objectives to a certain extent and transform it into a single-objective optimization problem. However, there may be multiple non-inferior solutions (Pareto optimal solutions). Part II is a constraint violation measure: is the i-th inequality constraint, with a total of m constraints; When , the constraint is violated and the value is ; represents the sum of all constraint violations, reflecting the overall degree of violation of a solution to the constraints; finally, the constraint violation is squared, that is, , which means further amplifying the penalty for constraint violation, and then summing ; is the penalty factor, which is multiplied by the square sum of constraint violations to make it equal to the magnitude of the objective function, thereby balancing the contribution of the two to the entire function. In summary, the role of the constraint violation measure is to first check whether each constraint is violated for a solution x, and if so, take the violation amount , otherwise it is 0; then the sum of squares of all violations is calculated and finally multiplied by the penalty factor As a penalty value for constraint violation. This method adds the sum of squares of constraint violations to the objective function, reflecting the guiding role of avoiding constraint violations during the optimization process. At the same time, the square operation further amplifies the penalty for violation. Therefore, in the process of minimizing the objective function F, the algorithm will search in the direction of optimizing the objective function value and reducing constraint violations, and finally obtain a balanced solution of multiple objectives under the constraints.

[0028] Furthermore, the ant colony algorithm is used to solve the multi-objective optimization function. N ants are randomly generated as the initial solution. Each ant represents a tower layout plan. The position coordinates of the ants are Converted to the position coordinates of the i-th crane. This real number encoding method directly uses the crane coordinates as decision variables, avoiding the encoding and decoding operations caused by binary or discrete encoding, simplifying the algorithm process, and improving search efficiency. Calculate the objective function value of each ant and , that is, the mean square error of the percentage of the blind area covered by the crane and the site coverage rate. Substituting the position coordinates of the ants into the objective function can quickly evaluate the quality of the current solution and provide a basis for subsequent pheromone updates and solution selection. Update of the pheromone concentration matrix: , , the ranking factor method is used to update the pheromone concentration matrix, and the pheromone volatility coefficient ρ is introduced to simulate the process of pheromone decay over time to prevent the algorithm from converging to the local optimum too early. At the same time, according to the ranking factor ant's objective function value Calculating pheromone increment , only the top-ranked high-quality ants are allowed to release pheromones, which speeds up the convergence of the algorithm. Compared with the classic ant colony algorithm, the ranking factor method can better balance the exploration and utilization capabilities of the algorithm, retaining the diversity of solutions while highlighting the guiding role of high-quality solutions, thus improving the search efficiency of the algorithm and the quality of solutions.

[0029] Calculation of state transition probability: , according to the state transition rule, calculate the probability that the kth ant will transfer from the crane position i to the position j. Among them, the pheromone concentration Reflects historical search experience, expectation Reflects the inspiration information of the problem itself, distance As a weakening factor to prevent excessive transfer. By adjusting the relative importance of the three, the effective combination of empirical information and heuristic information is achieved. Compared with the classic ant colony algorithm, the improved transition probability formula introduces heuristic information related to the problem characteristics. and weakening factors , It depends on the improvement of site coverage after the transfer, Δf_2, which reflects the contribution of the transfer behavior to the optimization goal. This inhibits the long-distance transfer of ants, makes the search more focused on promising areas, avoids invalid searches, and improves the convergence speed of the algorithm and the quality of the solution. Represents edge The pheromone concentration on the edge reflects the historical accumulation of the edge selected by the ants. The higher the concentration, the more times the edge has been selected in the previous search, which usually means that the path where the edge is located has better quality. i and s represent the starting and ending nodes of the edge respectively. Represents edge Heuristic information on With edge The heuristic value is proportional to the inverse of the distance or cost, that is, the shorter the distance or the lower the cost, the greater the heuristic value and the stronger the attraction to the ants. Represents edge Distance cost

[0030] According to the state transition probability matrix, the roulette method is used to select the next position of the ant and update the ant's solution. The roulette method is a probability-based selection strategy. The position with a larger transition probability is more likely to be selected. It not only ensures the randomness of the search, but also takes into account the inheritance opportunity of high-quality positions. It is a classic selection method of the ant colony algorithm. Determine whether the maximum number of iterations is met or the change in the solution is less than the preset threshold. If so, output the current optimal solution as the optimal tower layout plan, otherwise return to continue iteration. The maximum number of iterations controls the global search time of the algorithm, and the change in the solution reflects the degree of convergence of the algorithm. The combination of the two can terminate the algorithm in time to avoid unnecessary computational overhead.

[0031] Compared with the prior art, the advantages of this application are:

[0032] The parametric modeling method is used to efficiently establish the three-dimensional model of the tower crane, and the physical engine is used to simulate the kinematic behavior of the tower crane to obtain the precise motion trajectory of the tower crane, providing data support for tower crane interference detection and coverage range analysis.

[0033] Three-dimensional laser scanning technology was introduced to obtain point cloud data of the construction site. Through point cloud segmentation and extraction, the constraint boundaries such as site terrain, hardened areas and existing buildings were obtained, which improved the adaptability and feasibility of the tower crane layout plan.

[0034] The tower crane 3D model and site constraints are integrated into the BIM platform, and the visualization of the tower crane layout plan is achieved through 3D interactive settings, providing an intuitive and reliable initial solution for plan optimization.

[0035] A multi-objective optimization model was constructed to minimize the blind area of ​​the tower crane and maximize the site coverage rate. The ant colony algorithm was used for optimization and solution to obtain the optimal layout plan that met the on-site constraints, thereby improving the scientificity and economy of the layout plan.

[0036] The tower crane lifting efficiency is calculated based on the tower crane dynamics and kinematics equations, which enables a quantitative evaluation of the tower crane performance and provides a basis for the formulation of the tower group plan. The optimal tower group layout plan is generated through 3D visualization engine rendering, which intuitively displays the tower crane operation status, making it easier for on-site managers to control the overall operation of the tower crane group.

[0037] The integration of physical engine simulation, multi-objective optimization and 3D visualization technologies is applied to tower crane layout management, which realizes the transformation of tower crane group control from empirical decision-making to quantitative analysis, and improves the safety, reliability and refined management and control level of tower crane group operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] This specification will be further described in the form of exemplary embodiments, which will be described in detail by the accompanying drawings. These embodiments are not restrictive, and in these embodiments, the same number represents the same structure, wherein:

[0039] Figure 1 is an exemplary flow chart of a visualized engineering monitoring method according to some embodiments of this specification;

[0040] Figure 2 is an exemplary flow chart of obtaining a motion trajectory according to some embodiments of this specification;

[0041] Figure 3 is an exemplary flow chart of obtaining spatial interference results according to some embodiments of this specification;

[0042] Figure 4 It is an exemplary flowchart for obtaining the tower crane coverage blind area and lifting efficiency according to some embodiments of this specification. DETAILED DESCRIPTION

[0043] The method and system provided in the embodiments of this specification are described in detail below with reference to the accompanying drawings.

[0044] Figure 1 This is an exemplary flow chart of a visualized engineering monitoring method shown in some embodiments of the present specification, a visualized engineering monitoring method, used for tower crane group control, including: obtaining attribute parameters of the tower crane, the attribute parameters including the position, height and arm length of the tower crane; using the obtained attribute parameters, adopting a parametric modeling method to establish a three-dimensional model of the tower crane; importing the established three-dimensional tower crane model into a physical engine, using a rigid body dynamics algorithm, inputting the kinematic parameters of the three-dimensional tower crane model into the rigid body dynamics equation, and simulating and solving the motion trajectory of the tower crane; obtaining a three-dimensional model of an obstacle, importing the three-dimensional obstacle model into the physical engine, using a collision detection algorithm, judging the spatial interference between the motion trajectory of the tower crane and the three-dimensional model of the obstacle; using a three-dimensional laser scanner to obtain point cloud data of the construction site ; Based on the acquired point cloud data, the constraints of the construction site are extracted through the point cloud segmentation and extraction algorithm. The constraints include the site terrain, hardened area and existing building constraint boundaries; in the BIM software platform, the three-dimensional model of the tower crane and the constraints are used as input, and the position parameters of the tower group are interactively set in the three-dimensional scene to generate a tower group layout plan; a multi-objective optimization function with the goals of minimizing the blind area of ​​the tower crane and maximizing the site coverage rate is established; the obtained tower crane motion trajectory, spatial interference results and tower group layout plan are used as input, and the constraints are used as boundary parameters. The ant colony algorithm is used to obtain the optimal tower group layout plan that meets the multi-objective optimization function; the obtained optimal tower group layout plan is rendered in a three-dimensional visualization engine to generate a three-dimensional visualization model of the optimal tower group layout plan.

[0045] For group projects, the layout of the tower group is a key and difficult point in deployment. A certain project applied this application, simulated the operation and working conditions of the tower group through parametric simulation, determined the boundary conditions according to the on-site construction factors, and optimized the layout of the tower group using parametric design. It can achieve the effect of optimizing the plane layout and coverage of the tower crane and reducing blind spots. By simulating the lifting of the tower crane, the reasonable height of each lifting was determined, reducing the impact of the tower group and ensuring the efficient operation of the tower group.

[0046] Data collection: Use total stations, GPS and other measuring equipment to obtain the spatial position coordinates (x, y, z), tower height H and arm length L and other attribute parameters of the tower crane, use 3D laser scanners to obtain high-precision point cloud data of the construction site, and perform 3D modeling of obstacles such as terrain and buildings to obtain their geometric model data. 3D modeling: Input the collected tower crane attribute parameters into the BIM software, call the tower crane parametric family file, and generate a 3D solid model of the tower crane through parametric modeling methods. At the same time, import the 3D geometric model of the obstacle into the BIM software and integrate it with the tower crane model in the same coordinate system. Constraint extraction: Input the construction site point cloud data into the point cloud processing software, and use point cloud segmentation algorithms to extract site elements such as terrain undulations, road hardening areas, and building outlines through pre-processing such as point cloud registration and filtering to obtain site constraint data.

[0047] Specifically, a certain project used total stations, GPS and other measuring equipment to obtain the spatial position coordinates and attribute parameters of four tower cranes, among which the coordinates of tower crane TC1 are (20, 30, 0), tower height H=60m, front arm length L1=50m, and rear arm length L2=20m; the coordinates of tower crane TC2 are (50, 60, 0), tower height H=50m, front arm length L1=45m, and rear arm length L2=18m; the coordinates of tower crane TC3 are (80, 35, 0), tower height H=55m, front arm length L1=48m, and rear arm length L2=22m; the coordinates of tower crane TC4 are (40, 80, 0), tower height H=65m, front arm length L1=55m, and rear arm length L2=25m. At the same time, a 3D laser scanner was used to obtain high-precision point cloud data within the range of 500m×500m at the construction site, with a point cloud density of 100 points per square meter. The tower crane attribute parameters are input into the Revit software, and the pre-created boom tower crane family file is called to generate the 3D solid models of the four tower cranes. In the family file, geometric parameters such as tower height H, front arm length L1, and rear arm length L2 have been associated with the corresponding family parameters, and the constraint conditions of the tower arm angle α are preset. Therefore, after modifying the family parameter values ​​of the tower crane instance, the parametric modeling engine can automatically update the size and position of the tower crane components and generate a 3D solid model that conforms to the actual working conditions. The construction site point cloud data is imported into the Cyclone point cloud processing software. After pre-processing such as point cloud registration and filtering, a high-precision point cloud model in a unified coordinate system is obtained. In order to extract the site constraints, the point cloud model needs to be segmented and identified.

[0048] Extraction of terrain undulation: The ground point cloud is extracted using a curvature-based point cloud segmentation algorithm. The algorithm assumes that the ground is a smooth surface. By calculating the curvature value of each point, points with a curvature value less than the threshold are classified as ground points, and non-ground points with obvious height differences are eliminated to obtain a subset of ground point clouds that represent terrain undulations. After segmentation, a total of 12,857,962 points of the ground point cloud of the construction site are obtained, accounting for 65.8% of the total number of points. Extraction of road hardening area: Based on the ground point cloud, the road hardening area is further segmented using the regional growing algorithm. Starting from the seed point, the algorithm searches for neighboring points with normal angles and distances that meet the conditions, and continuously expands the regional boundaries until the growth stops. After segmentation, a total of 3,521,408 points of the point cloud of the road hardening area of ​​the construction site are obtained, accounting for 27.4% of the ground point cloud, corresponding to a road area of ​​approximately 12,500 square meters. Extraction of building contours: The RANSAC algorithm is used to fit the plane features in the point cloud, and the key surfaces of the building such as the wall and roof are identified through the direction and position relationship of the plane. The outer contour of the building can be obtained by connecting the boundary points of the key surface to form a polygon. After identification, the outer contours of the five buildings on the construction site were obtained, among which the contour of Building 1# is a rectangle of 60m×80m, the contour of Building 2# is a square of 50m×50m, the contour of Building 3# is three broken lines with a total length of 120m, the contour of Building 4# is a semicircle with a diameter of 80m, and the contour of Building 5# is a rectangle of 30m×70m. The extracted site elements such as terrain, roads, and building contours are imported into the Revit model to obtain a three-dimensional constraint model of the construction site. Based on this model, the rationality and feasibility of the tower crane layout plan can be further analyzed.

[0049] Figure 2This is an exemplary flow chart for obtaining motion trajectories according to some embodiments of this specification, physical engine simulation: import the three-dimensional model of the tower crane and the obstacle model into the physical engine, input the physical parameters of the tower crane such as gravity acceleration and speed, and use the rigid body dynamics algorithm to simulate and calculate the motion trajectory data of the tower crane. Specifically, the BIM three-dimensional models of four tower cranes (TC1~TC4) and the constraint condition model of the construction site have been obtained. In order to further analyze the actual movement of the tower crane, the tower crane model needs to be imported into the physical engine for simulation. This embodiment uses Nvidia Flex software based on the PhysX physics engine as the simulation platform. Import the tower crane BIM model into the Flex software and set the physical parameters of the tower crane. According to the technical parameters of the tower crane, input the material, density, mass and other properties of each component. The tower body of the TC1 tower crane is made of Q345 steel, with a density of 7850kg / m³ and a total mass of 120t; the tower arm is made of Q235 steel, with a density of 7750kg / m³, a front arm mass of 18t, and a rear arm mass of 12t; the hook is made of alloy steel, with a density of 8500kg / m³ and a mass of 2.5t. At the same time, define the degrees of freedom and constraint types of the tower crane, such as the tower body can rotate around the Z axis and the tower arm can swing around the Y axis. Obtain the kinematic parameters of the tower crane. Using the tower crane attribute parameters in the BIM model, calculate the kinematic parameters of the tower crane such as position and speed in the initial state. Taking the TC1 tower crane as an example, the hook is located at (35, 40, 0), the hook height h=20m, the lifting weight m=10t, the hook rope speed v=20m / min, the tower arm angle α=60°, and the rotation angular velocity ω=2.5r / min. According to the kinematic formula, the velocity parameter of the hook is calculated to be (0, 0.33, 0) m / s, the acceleration parameter of the hook is (0, 0, -9.8) m / s², the angular velocity parameter of the tower arm is (0, 0.105, 0) rad / s, and the angular acceleration parameter of the tower arm is (0, 0.042, 0) rad / s². Input the kinematic parameters into the rigid body dynamics equation. In the physics engine, a rigid body dynamics model based on the Newton-Euler equation is used to establish a group of differential equations of motion for the tower crane. Taking the TC1 tower crane as an example, the tower body is a homogeneous cylinder with a moment of inertia J=1 / 2MR², where M is the mass of the tower body and R is the radius of the tower body. The tower arm is simplified to a slender rod with a moment of inertia J=1 / 3mL², where m is the mass of the tower arm and L is the length of the tower arm. The hook is regarded as a point mass and the moment of inertia is ignored. According to the force analysis of the tower crane, the dynamic equations of the tower body, tower arm, and hook are listed as follows: Tower body: (Rotation equation), tower arm: (Swing equation), hook: T=mg+ma (lift equation), (Cart equation), (Trolley equation), where β is the tower body rotation angle, β' is the angular velocity, and β'' is the angular acceleration; α is the tower arm elevation angle, α' is the angular velocity, and α'' is the angular acceleration; is the wind load moment, c is the damping coefficient; m is the weight, g is the acceleration of gravity, v is the hook speed, a is the hook acceleration; T is the hook tension, For the driving force of the vehicle, is the driving force of the trolley, and θ is the hook deflection angle.

[0050] The implicit Euler algorithm is used to solve the dynamic equations. The above kinematic parameters and dynamic equations are input into the Flex software, and the differential equations are numerically solved using the implicit Euler integrator to obtain the displacement, velocity, acceleration and other motion parameters of each component of the tower crane in the time series. The calculation formula is as follows: , , where v is velocity, x is displacement, a is acceleration, Δt is time step, subscript i represents the current moment, and i+1 represents the next moment. For the hook of TC1 tower crane, assuming the time step Δt=0.01s, according to the initial conditions and lifting equation, the motion parameters of the hook at the first three moments are iteratively calculated as follows: t=0s: , , , t=0.01s: , , t=0.02s: , , t=0.03s: , Through iterative calculation, the complete trajectory parameters of the tower crane hook in the entire motion cycle can be obtained. Using the visualization tools provided by Flex software, the three-dimensional motion process of the tower crane can be rendered in real time to generate realistic simulation animations.

[0051] Figure 3This is an exemplary flow chart for obtaining spatial interference results according to some embodiments of this specification. The collision layer data of the obstacle model is imported into the physical engine, and the collision detection algorithm is used to determine whether there is interference between the tower crane motion trajectory and the obstacle. Specifically, the motion simulation data of the four tower cranes TC1~TC4 and the three-dimensional model of the construction site obstacle constructed by point cloud data have been obtained. In order to check whether the tower crane collides with the obstacle during movement and analyze the blind spot of sight obstruction during the tower crane hoisting process, collision detection and ray intersection calculation are required in the physical engine. The obstacle model is imported into the physical engine. Taking the No. 1 building at the construction site as an example, its outer contour is a rectangle of 60m×80m. According to the building structure information inverted from the point cloud data, the geometric model of the building is meshed using the triangular patch algorithm. The maximum side length of the triangular patch is set to 1m, and the walls, floors, roofs and other components of the building are meshed to generate a surface mesh model consisting of 45,920 triangles. The vertex coordinates, normal vectors, texture coordinates and other geometric data of the mesh model are input into the physical engine, and the collision layer of the 1# building is generated at the specified position. Similarly, the three-dimensional models of buildings 2#~5# are meshed and the collision layer is set. Collision detection is performed in the physical engine. Taking the motion trajectory of the TC1 tower crane as an example, the motion path points of the hook are extracted according to the simulation data. The coordinates of the hook at t=10s are (38.5, 42.8, 15.2), and this point is used as the sampling point for collision detection. With this point as the center, a 0.5m×0.5m×0.5m bounding box is constructed as a simplified collision model of the tower crane. At the same time, an AABB bounding box tree is constructed for the triangular facet mesh model of the 1# building as a simplified collision model of the obstacle. The SAT separating axis theorem is used to determine whether the two bounding boxes intersect. If the bounding boxes intersect, the OBB tree algorithm is used to further detect whether there is a real collision between the hook bounding box and the building mesh facet. After detection, the hook does not collide with the 1# building at t=10s. Repeat the above detection process, perform frame-by-frame collision detection on all sampling points of the hook trajectory, and determine whether there is a collision risk in the entire motion process. Analyze the blind spot of the tower crane hoisting process. Taking the motion trajectory of the TC1 tower crane as an example, extract the motion path points of the hook according to the simulation data. Select the hook coordinates (42.3, 38.6, 22.5) at t=15s as the starting point of the ray, and establish a local coordinate system with this point as the origin. According to the hoisting direction and the line of sight direction, set the direction vector of the ray to (-0.342, -0.643, -0.685) to generate ray R1. Use the ray casting algorithm to calculate the intersection of ray R1 and the obstacles at the construction site. By traversing the obstacle bounding box tree, it is found that the ray intersects with the bounding box of building 1#. The Möller-Trumbore algorithm is further used to calculate the intersection of the ray and the triangle face of building 1#.After calculation, the ray intersects with the triangle patch numbered TRI_4382, and the intersection coordinates are (17.6, 22.1, 8.4), and the distance from the starting point of the ray is 28.7m. The distance threshold of line of sight obstruction is set to 15m, so it is judged that there is line of sight obstruction between the ray and the 1# building. Generate a blind area model for the tower crane hoisting process. Repeat the sampling and extraction of the motion trajectory points of the TC1 tower crane hook at a time interval of 5s to generate a series of rays. Execute the ray casting algorithm for each ray, calculate its intersection and obstruction distance with the obstacle, and mark the ray interval with an obstruction distance less than the threshold as a blind area. The starting point and intersection set of all blind area rays are exported to obtain a blind area mesh model consisting of 1,826 vertices and 3,573 triangles. Use the mesh visualization tool provided by the Flex software to render and display the line of sight obstruction blind area of ​​the TC1 tower crane. The blind area is in the shape of an irregular cone with a volume of approximately 3,520m³, and is mainly distributed on the southeast side of the 1# and 3# buildings. Improve and optimize the tower crane layout. Based on the results of collision detection and blind spot analysis, optimization suggestions are put forward for the layout of the TC1 tower crane: Shift the TC1 tower crane 8m in the positive direction of the X axis to keep it away from the 1# building to reduce the risk of collision during the hoisting process; at the same time, reduce the turning radius of the TC1 tower crane from 60m to 55m to avoid interference between the boom end and the 3# building; adjust the foundation elevation of the TC1 tower crane and raise it by 2m to reduce the line of sight between the tower crane and the 3# building. According to the optimized layout parameters, update the BIM model of the TC1 tower crane in Revit to generate an improved 3D layout drawing.

[0052] Figure 4 This is an exemplary flow chart for obtaining the tower crane coverage blind area and lifting efficiency according to some embodiments of this specification. In order to evaluate the actual operating efficiency of the tower crane group, it is necessary to further analyze the lifting height, amplitude, speed and other parameters of each tower crane during the movement, and calculate the spatial coverage volume and lifting efficiency index of the tower crane. Obtain the equipment parameters of the tower crane. Taking the TC1 tower crane as an example, according to the equipment selection data, the rated lifting capacity of the TC1 tower crane is , rated lifting height , rated luffing speed According to the construction organization design, the actual operating parameters of the TC1 tower crane are: boom length L = 55m, maximum amplitude angle , maximum angular velocity , rated lifting speed of hook . Input the above parameters into the physical engine as the initial attributes of the TC1 tower crane 3D model. Discretize the motion trajectory of the tower crane. According to the simulated motion data of the TC1 tower crane, sample the motion trajectory of the hook, select the time step Δt=0.5s, and obtain a series of discrete trajectory point coordinates. , i=1, 2, …, n. The coordinates of the hook at t=15s are (38.2, 56.7, 22.4). This point is used as the input condition for trajectory discretization and substituted into the subsequent lifting parameter calculation. Calculate the lifting height and amplitude of the tower crane. In the physics engine, based on the Lagrange equation and the DH parameter method, the dynamic equation and kinematic equation of the TC1 tower crane are established. Taking the coordinates of the hook at t=15s as an example, according to the kinematic equation, the height of the boom relative to the tower body at this time is calculated through the homogeneous transformation matrix. , the angle between the boom and the horizontal plane β=60.3°.

[0053] The hanging height of the hook relative to the boom By solving the kinetic equations, Substituting into the formula , the actual lifting height H=25.1m of the TC1 tower crane at this trajectory point is calculated. Similarly, according to the boom length L=55m and the amplitude angle α=63.8°, substitute the formula R=L*sinα to calculate the actual lifting range R=50.4m at this point. Calculate the lifting speed of the tower crane. According to the kinematic equation of the TC1 tower crane, the amplitude angular velocity ω=1.8r / min of the boom at t=15s is calculated. Substitute the known lifting range R=50.4m and the hook lifting speed , according to the formula , calculate the lifting speed v=53.6m / min at this point. Calculate the spatial coverage volume of the tower crane operation. Extract the height range of the hook trajectory point based on the simulated motion data of the TC1 tower crane . Set the integration step Δh=0.5m, Δθ=1°, Δr=0.2m, and use the numerical integration algorithm to discretize the hoisting space. Take the hook coordinates (r, θ, h) corresponding to a certain integration step as an example, and calculate the hoisting height angle φ=34.6° based on the kinematic equation. Substitute r=28.5m and φ=34.6° into the volume microelement , the volume of the microelement is calculated to be ΔV = 0.38m³. Traverse all the integral steps of the hoisting space, accumulate the volume of each microelement, and obtain the working space coverage volume of TC1 tower crane V = 385,940m³. Calculate the hoisting efficiency of the tower crane. Set the weight coefficient of TC1 tower crane efficiency evaluation to , , , , the power index is , , , Substitute the actual lifting height H=25.1m, actual lifting speed v=53.6m / min, space coverage volume V=385,940m³, and actual lifting weight G=8.5t in the equipment parameters obtained above into the lifting efficiency formula: The hoisting efficiency of TC1 tower crane at t=15s is E=1.118, indicating that the actual efficiency of the tower crane is slightly higher than the rated working condition.

[0054] Initial plan generation: In the BIM software platform, the tower crane 3D model, obstacle model and site constraint data are used as input, and the initial position coordinates of the tower crane are set through 3D scene interaction to generate a 3D layout plan for the tower group.

[0055] Import the tower crane model. In the Revit equipment library, select the parametric family file of Zoomlion ZT7050 tower crane, create multiple instances of the same family in the site model, and obtain the 3D models of TC1~TC4. According to the equipment selection data, input the basic properties of the tower crane into the parametric model, including the tower height. , arm length , radius of gyration , maximum lifting weight m=12t, rated power P=132kW, etc. Import obstacle model. Take the BIM models of the 5 buildings and 2 temporary buildings that have been constructed as obstacles and integrate the models in Revit. The height, length, and width information of each building can be extracted from its attribute parameters. ,length ,width ; Temporary wall height ,length Use Revit's 3D measurement tools to obtain the minimum distance between obstacles, such as the minimum distance between buildings 1# and 2#. . Set site constraints. In Revit, draw the site boundary line and define the site range as a constraint area. Use the elevation point tool to measure the terrain height difference inside the site and convert it into contour constraints. Generate the three-dimensional terrain surface of the site through topology as a reference surface for tower crane layout. Add a 2m wide safety belt on both sides of the site road to prohibit the layout of tower cranes. Mark the 5m range adjacent to each building as a no-go zone to avoid interference between tower crane foundation construction and building construction. Interactively set the tower crane position. Open the three-dimensional scene view in Revit and set the initial position of the tower crane model through human-computer interaction by dragging the mouse. Arrange the TC1~TC3 tower cranes outside the 1#~3# buildings, and the TC4 tower crane between the 4# and 5# buildings, so as to initially achieve the coverage of each construction area by the tower crane. Fine-tune the orientation angle of each tower crane to avoid major obstacles in the direction of the turning radius. Preliminary evaluation of the feasibility of the tower crane layout plan through perspective switching and roaming inspection. The TC1 tower crane is initially arranged at point A, with absolute coordinates of (65.8, 37.2, 8.2), and the offset of the local coordinate system origin relative to the northwest corner of building 1# is (12.5, -8.6). Similarly, the initial coordinates of the TC2~TC4 tower cranes are: TC2: (126.3, 82.5, 9.4), offset relative to building 2# (18.2, 10.8); TC3: (152.7, 40.6, 7.8), offset relative to building 3# (15.6, -12.3); TC4: (75.2, 112.4, 8.6), offset relative to building 4# (-10.3, 17.5). The initial coordinates of each tower crane are saved as the input of the parametric family to generate the initial plan for the three-dimensional layout of the tower group. The interference check function of Revit is used to perform collision detection on the plan to check the geometric conflicts between the tower cranes and between the tower cranes and obstacles. The tower crane layout parameters are exported. Select the tower group model, export its geometric parameters and position coordinates into an EXCEL table, and obtain the following data (unit: m): tower crane number, tower height, arm length, turning radius, x-coordinate, y-coordinate, z-coordinate TC1, 47.5, 55.0, 57.0, 65.8, 37.2, 8.2; TC2, 47.5, 55.0, 57.0, 126.3, 82.5, 9.4; TC3, 47.5, 55.0, 57.0, 152.7, 40.6, 7.8; TC4, 47.5, 55.0, 57.0, 75.2, 112.4, 8.6. At the same time, export the key geometric parameters of the site model and obstacle model into a JSON file for subsequent spatial collision detection and coverage analysis.

[0056] Further optimize the tower crane layout to achieve the goal of minimizing the tower crane coverage blind area and maximizing the site coverage rate. Establish an optimization model. Define the optimization objectives as: It represents the mean square error of the percentage of blind area covered by the tower crane, Indicates the site coverage rate. Among them, the mean square error of the blind area ratio reflects the degree of dispersion of the blind area of ​​each tower crane relative to the total construction area. The smaller the mean square error, the more balanced the coverage of each tower crane; the larger the site coverage rate, the better the overall coverage effect of the tower group layout plan. According to the general construction plan, the total area of ​​the construction site is S=30000m². In the physical engine simulation analysis, the coverage blind area of ​​each tower crane is obtained as follows (unit: m²): TC1 blind area: ,TC2 blind area: , TC3 blind area: , TC4 blind area: Substitute into the formula , calculate the mean square error of the percentage of blind area covered by the tower crane: . Use the spatial analysis tools of BIM software to calculate the actual coverage area of ​​the initial tower group layout plan Substituting into the formula , calculate the site coverage: , thus obtaining the two optimization target values ​​of the initial layout scheme: , .

[0057] Transform the optimization function. Use the weighted Tchebycheff method to transform the multi-objective optimization problem into a single-objective optimization problem. Set the weight coefficient , , the penalty factor ρ = 10. According to empirical estimates, The ideal minimum , The ideal maximum value The site constraints include: the minimum distance between the tower crane foundation and the building is not less than 5m, the minimum distance between tower cranes is not less than 1.2 times the length of the tower crane arm, and the minimum distance between the tower crane and the site boundary is not less than 3m. These constraints are expressed as mathematical inequalities as boundary conditions of the optimization model. The distance constraint between TC1 and Building 1 is: ,in is the coordinate of TC1 tower crane, is the reference point coordinate of building 1#. Similarly, distance constraints between other cranes and buildings, between cranes, and between cranes and site boundaries can be established, for a total of m constraints. Substituting the above parameters into the weighted Tchebycheff formula, the objective function of the optimization model is obtained: ,in, Coordinates are arranged by crane The only certainty is that To optimize the decision variables.

[0058] Considering the nonlinear characteristics of the optimization objective and the complexity of the constraints, the ant colony algorithm is used to solve the above single-objective optimization function. The relevant parameters of the ant colony algorithm are set as follows: population size m = 50, maximum number of iterations , pheromone importance factor α=1, heuristic function importance factor β=5, pheromone volatility factor ρ=0.1, pheromone constant Q=100. Randomly generate m initial ant colony positions in the search space, each of which corresponds to a set of tower crane layout coordinates. Calculate the objective function value of each position, and record the position with the smallest function value as the global optimal solution. In each iteration, the ant colony moves in the solution space according to the state transition rule and updates the pheromone of the passed position. At the same time, the newly generated solution is repaired to satisfy the constraints. Calculate the objective function value of the new solution, and update the current optimal solution and the global optimal solution. After N_max iterations, the global optimal solution is output as the final tower crane layout coordinates. The optimal solution is: TC1: (68.2, 42.5); TC2: (122.8, 88.3); TC3: (148.5, 35.2); TC4: (80.6, 108.1). The above coordinates are used as the input of the tower crane parametric model, and the tower group layout plan is updated in the BIM platform. Calculate the blind area mean square error of the optimized solution , site coverage , which are better than the initial solution. In the BIM platform, the tower crane layout parameters, coverage blind spots, site coverage rate and other data before and after optimization are extracted and exported as an EXCEL table, and a three-dimensional layout rendering before and after optimization is attached to generate a tower crane group layout optimization report. The report shows that the mean square error of the blind spots of the optimized solution is reduced by 38.1%, the site coverage rate is increased by 3.0%, and the overall layout is more balanced and reasonable.

[0059] Visual rendering: Take the optimal tower group layout coordinates as input, render the tower crane group in real time in the 3D visualization engine, simulate the tower crane operation animation, and generate a high-precision 3D visualization model of the tower crane group. Monitoring and management: Connect the 3D visualization model with the construction management system, transmit the tower crane operation parameters in real time, and display the lifting height, lifting weight, wind speed and other data in the visualization interface to achieve visual monitoring and refined management of the tower crane group.

Claims

1. A visual engineering monitoring method for tower crane group control, comprising: Get the attribute parameters of the tower crane, which include the position, height and arm length of the tower crane; Using the acquired attribute parameters and parametric modeling methods, a three-dimensional model of the tower crane is established; Import the established 3D tower crane model into the physical engine, use the rigid body dynamics algorithm, input the kinematic parameters of the 3D tower crane model into the rigid body dynamics equation, and simulate and solve the motion trajectory of the tower crane; Obtain the 3D model of the obstacle, import the 3D model of the obstacle into the physical engine, and use the collision detection algorithm to determine the spatial interference between the motion trajectory of the tower crane and the 3D model of the obstacle; Use a 3D laser scanner to obtain point cloud data of the construction site. Based on the acquired point cloud data, use the point cloud segmentation and extraction algorithm to extract the constraints of the construction site. The constraints include site topography, hardened areas, and existing building constraint boundaries. In the BIM software platform, the 3D model of the tower crane and the constraints are used as input, and the position parameters of the tower groups are interactively set in the 3D scene to generate a tower group layout plan; Establish a multi-objective optimization function with the goal of minimizing the blind area of ​​the tower crane and maximizing the site coverage rate; The obtained tower crane motion trajectory, spatial interference results and tower group layout plan are used as input, and the constraint conditions are used as boundary parameters. The ant colony algorithm is used to obtain the optimal tower group layout plan that meets the multi-objective optimization function. The obtained optimal tower group layout scheme is rendered in a 3D visualization engine to generate a 3D visualization model of the optimal tower group layout scheme; Establish a 3D model of the tower crane, including: Define the acquired attribute parameters of the tower crane as parameter variables; Using the geometric constraint relationship between parameter variables, a three-dimensional tower crane model is established through a preset tower crane type parametric family file; Simulate and solve the motion trajectory of the tower crane, including: Obtaining kinematic parameters of the tower crane; the kinematic parameters include the position, velocity, acceleration, angular velocity and angular acceleration of the tower crane; The obtained kinematic parameters are input into the rigid body dynamics equations based on the Newton-Euler equations; The implicit Euler algorithm is used to solve the rigid body dynamics equation to obtain the motion trajectory parameters of the tower crane 3D model in the physical engine. Determine the spatial interference between the motion trajectory of the tower crane and the three-dimensional model of the obstacle, including: Obtain the three-dimensional geometric data of obstacles at the construction site; convert the three-dimensional geometric data of obstacles into a triangular patch mesh model through a triangular patch algorithm; Import the converted triangular face mesh model into the physical engine as a three-dimensional obstacle model; In the physical engine, based on the tower crane motion trajectory parameters, the bounding box of the tower crane 3D model is intersected with the bounding box of the obstacle 3D model; If the two bounding boxes intersect, a triangle patch intersection detection algorithm is used to determine whether the triangle patches of the tower crane 3D model and the triangle patches of the obstacle 3D model intersect. If there are intersecting triangles, it is judged that there is spatial interference between the motion trajectory of the tower crane and the obstacle; Import the established tower crane 3D model into the physical engine, including: Get the equipment parameters of the tower crane, which include the lifting capacity, lifting height and luffing speed of the tower crane; Based on the motion trajectory parameters, the motion trajectory of the tower crane is discretized using a numerical integration algorithm. The coordinates of the discretized motion trajectory points are used as input, and the tower crane dynamic equation and kinematic equation are used to calculate the tower crane hoisting height H and amplitude R of each motion trajectory point; The hoisting height and amplitude of all motion trajectory points are integrated to obtain the spatial coverage volume of the tower crane operation; According to the equipment parameters and the spatial coverage volume of the tower crane operation, a weighted algorithm is used to calculate the tower crane's hoisting efficiency E; Calculate the lifting efficiency of the tower crane, including: ; in are the weight coefficients of tower crane lifting weight, hoisting speed, hoisting height and space coverage volume respectively; are the power exponents of each parameter respectively; are the rated values ​​of each parameter, G is the actual lifting capacity of the tower crane, v is the actual lifting speed of the tower crane, H is the actual lifting height of the tower crane, and V is the actual space coverage volume of the tower crane operation; The calculation formula for the tower crane hoisting height H is: ; in, It indicates the angle between the boom and the horizontal plane when the tower crane is luffing; Indicates the height of the boom relative to the tower body when the tower crane is luffing; Indicates the hanging height of the hook relative to the boom; The calculation formula for the tower crane lifting range R is: ; Wherein, L represents the actual arm length of the tower crane; α represents the luffing angle of the tower crane, that is, the angle between the boom and the tower body; The calculation formula of tower crane hoisting speed v is: ; in, It represents the angular velocity of the tower crane, R is the actual lifting amplitude, is the lifting speed of the hook; The calculation formula for the spatial coverage volume V of tower crane operation is: ; in, and are the minimum and maximum values ​​of the tower crane hoisting height, respectively; R is the actual hoisting amplitude; r is the hoisting radius; φ is the hoisting height angle, i.e. the angle between the line connecting the hook and the tower crane base and the horizontal plane; θ is the tower crane rotation angle; Obtain the optimal tower layout solution that meets the multi-objective optimization function, including: A multi-objective optimization function is established with the goal of minimizing the blind area of ​​the tower crane and maximizing the site coverage rate; the multi-objective optimization function includes: ; ; in, is the mean square error of the percentage of blind area covered by the tower crane, is the blind area covered by the i-th tower crane, n is the number of tower cranes, and S is the total construction area; is the site coverage rate, The effective coverage area of ​​the tower cluster layout plan; The weighted Tchebycheff method is used to convert the multi-objective optimization function into a single-objective optimization function: ; in, and They are and The ideal value of and is the corresponding weight coefficient, is the penalty factor, m is the number of constraints, is the i-th constraint condition, x is the decision variable, i.e., the crane position coordinate; The obtained tower crane 3D model, tower crane motion trajectory parameters, obstacle 3D model, spatial interference results and initial tower group layout plan are used as input, and the constraints of the construction site are used as boundary parameters. The improved ant colony algorithm is used to solve the single-objective optimization function to obtain the optimal tower group layout plan that meets the multi-objective optimization function. Ant colony algorithm is used to solve multi-objective optimization functions, including: Randomly generate N ants as the initial solution, each ant represents a tower layout plan, and the position coordinates of the ants are Convert to the position coordinates of the i-th tower crane; Calculate the objective function value of each ant and ; According to the objective function value, the ranking factor method is used to update the pheromone concentration matrix of the tth iteration: ; ; in, For the tth iteration, The pheromone concentration on is the pheromone volatility coefficient, R is the ranking factor number, and its value range is 10% to 20% of the number of ants; For the rth sorting ant on the edge The pheromone increment left on is the objective function value of the rth sorted ant; Indicates that the rth ant has passed the edge in the current iteration ; According to the state transition rule, calculate the probability that the kth ant transfers from the tower crane position i to the position j: ; in, represents the probability that the kth ant moves from node i to node j at time t; represents the edge at time t pheromone concentration on; Represents the heuristic information from node i to node j; are the importance parameters of pheromone, desirability and distance respectively; is the heuristic information, which indicates the expected degree of the ant moving from position i to position j; is the distance between the two cranes, which serves as a weakening factor to weaken the probability of long-distance transfer; Represents the set of neighboring nodes of the node where the kth ant is currently located; Represents edge pheromone concentration on; Represents edge Heuristic information on Represents edge Distance cost on According to the state transition probability matrix, the roulette method is used to select the next position of the ant and update the ant's solution; Determine whether the maximum number of iterations is met or the change in the solution is less than the preset threshold. If so, output the current optimal solution as the optimal tower group layout plan, otherwise return the calculated objective function value.

2. The visual engineering monitoring method according to claim 1, characterized in that: Import the established 3D tower crane model into the physical engine, including: According to the motion trajectory parameters of the tower crane, the motion trajectory points of the tower crane hook are obtained; The ray is generated by taking the motion trajectory point of the hook as the starting point of the ray and the hoisting direction of the tower crane as the ray direction; Using the ray casting algorithm, the intersection point between the generated ray and the three-dimensional model of the obstacle imported into the physical engine is calculated to obtain the first intersection point of the ray; Calculate the distance between the starting point of the ray and the first intersection point, and compare the calculated distance with a preset threshold; If the calculated distance is less than the preset threshold, the area between the ray starting point and the first intersection point is set as the sight blocking area during the lifting process; Repeat the above ray intersection calculation for all the motion trajectory points of the tower crane hook to obtain all the sight-blocking areas during the tower crane hoisting process; Merge all the sight-blocked areas to obtain the tower crane’s blind coverage area.

3. The visual engineering monitoring method according to claim 1, characterized in that: The tower crane dynamics equation and kinematics equation are used, including: The dynamic equation of the tower crane is established using the Lagrange equation; The kinematic equation of the tower crane is established using the Denavit-Hartenberg parameter method, and the relative motion relationship between the tower crane members is described by the homogeneous transformation matrix.

Citation Information

Patent Citations

  • Project control-based regional tower crane monitoring control method and system

    CN115465784A

  • Large prefabricated part hoisting safety control method based on BIM

    CN115271269A

  • Supervision method and device for construction tower crane, electronic equipment and storage medium

    CN116226977A

  • Tower crane collision early warning method based on multiple integrated systems

    CN116605772A

  • Tower crane control method and system, control terminal and storage medium

    CN116621040A