5g group tower intelligent dispatching system and method based on digital twinning

Abstract

The application relates to the technical field of industrial intelligent collaborative control, and provides a 5G group tower intelligent transfer system and method based on digital twinning, which comprises the following steps: a time-space joint state field is constructed, physical accessibility, operation occupancy and safety rules are uniformly coded into discrete voxel states to form a dynamic feasibility domain constraint; a feasible path set is generated by a dispatching module based on the constraint, and the optimal hoisting path is screened through weighted or hierarchical decision-making; a simulation module simulates the operation of the optimal hoisting path in the feasible domain constraint, acquires a safety probability, and if the safety probability is less than a preset safety threshold, a time window is slid to retry until a safety release condition is met; and a control module executes the optimal hoisting path according to hoisting time planning. The application realizes safe, efficient and self-adaptive group tower collaborative operation.

Description

Technical Field

[0001] This invention relates to the field of industrial intelligent collaborative control technology, and more specifically, to a 5G-based intelligent dispatching system and method for multiple towers based on digital twins. Background Technology

[0002] With the continuous advancement of my country's new urbanization construction and the in-depth implementation of the intelligent construction strategy, the construction industry is accelerating its transformation towards digitalization, automation, and collaboration. In complex engineering projects such as large public buildings, super high-rise complexes, and industrial plants, highly intensive construction space, closely overlapping work processes, and dense deployment of machinery and equipment have become the norm. How to achieve the safe, efficient, and orderly collaborative operation of multiple heavy construction equipment under limited time and space resources has become a key challenge to improve the overall productivity and inherent safety level of construction sites.

[0003] In large-scale construction projects, tower cranes (referred to as "tower cranes") are the core vertical transportation equipment on the construction site. Their scheduling efficiency is directly related to the speed of material flow, the smoothness of process connection, and even the entire project's schedule and cost.

[0004] In multi-tower operation environments, the working envelopes of each tower crane overlap and hoisting tasks occur concurrently, making the workspace highly dynamic and uncertain. Current mainstream tower crane management methods mainly rely on centralized command scheduling, i.e., tasks are manually assigned by dispatchers. Although this method can avoid physical interference to a certain extent, it has obvious limitations. It is slow to respond to changes in the construction environment and cannot reflect the actual feasible work space in real time. In addition, in real construction site operation environments, traditional collision avoidance systems use passive response mechanisms, only braking when a collision is imminent. Each tower crane makes independent decisions, lacks global coordination, and frequent emergency stops and starts result in serious efficiency losses.

[0005] Furthermore, if the implementation is not strictly followed, it is difficult to independently assess the safety risks under the new conditions, and subsequent plans cannot be dynamically adjusted. As a result, operations can only be interrupted and manual intervention can be relied upon multiple times, which seriously affects the continuity of construction.

[0006] Therefore, there is an urgent need for a 5G-based intelligent tower dispatching system and method based on digital twins to solve the technical problem that existing technologies lack a dynamic and computable adaptive optimization mechanism for collaborative scheduling of tower operations. Summary of the Invention

[0007] In view of this, the present invention proposes a 5G group tower intelligent dispatching system and method based on digital twins, aiming to solve the technical problem that the collaborative scheduling of group tower operations in the prior art lacks a dynamic and computable adaptive optimization mechanism.

[0008] This invention discloses a 5G cluster intelligent dispatching system based on digital twins, comprising:

[0009] The digital twin module constructs a joint time-space state field, which encodes the physical accessibility, work occupancy and safety management rules of the construction site into discrete voxel states, and uses these as constraints for the feasibility domain.

[0010] The dispatch module obtains the dispatch task type and dispatch initiation and destination points. Based on the feasible domain constraints, it obtains a set of feasible paths and filters the optimal hoisting path. It assigns multi-dimensional weights to each path and constructs a path cost function through weighted fusion or hierarchical decision-making to obtain the optimal hoisting path.

[0011] The simulation module simulates the operation of the optimal hoisting path within the feasible domain constraints, obtains the safety probability, and releases the operation when the safety probability is greater than a preset safety threshold. If the safety probability is less than the preset safety threshold, the sliding task time window is used to simulate again until the operation is released. The release output includes the optimal hoisting path, hoisting feasible domain constraints, and hoisting time planning.

[0012] The control module plans and executes the optimal hoisting path based on the hoisting time.

[0013] Preferably, the digital twin module divides the construction site into discrete voxel grids in three-dimensional space and introduces a discrete time dimension to construct a time-space joint state field.

[0014] For each voxel At any moment Define its voxel state vector as :

[0015] ;

[0016] in, These are the index coordinates of a three-dimensional discrete spatial grid, used to uniquely identify the position of each voxel in the construction site; Indicates physical accessibility, i.e., whether the voxel is located within the current working envelope of at least one tower crane; This indicates whether the voxel is occupied by operating equipment or personnel / vehicle entities. The safety risk weight is calculated comprehensively from no-lift zones, height-restricted zones, hazardous areas, and time-based safety rules.

[0017] Define a voxel feasibility indicator function based on voxel state vectors. :

[0018] ;

[0019] in, It is the maximum permissible safety risk threshold; It is a voxel Job occupancy at time t; It is the physical reachability of voxel v at time t;

[0020] This generates a time-varying feasible lifting region with the following constraints:

[0021] ;

[0022] in, It is the feasible region of hoisting based on the spatiotemporal joint state field, that is, the set of all voxels that are feasible at time t.

[0023] Preferably, the digital twin module defines the physical accessibility of the construction space. This includes: deploying tower crane operation status acquisition sensors at the construction site to obtain real-time data on the slewing angle, luffing length, hook height, and load mass of each tower crane; calculating its instantaneous working envelope based on the tower crane's geometric kinematic model; and determining if voxel v lies within the instantaneous working envelope of any tower crane. If it is 1, then if it is not within any working envelope. =0; where the instantaneous working envelope refers to the set of all spatial locations that can be reached during tower crane hoisting operations.

[0024] Preferably, the digital twin module determines the job occupancy. ,include:

[0025] The system uses PLC signals, motor current, or control commands to determine in real time whether each tower crane is in operation. If a tower crane stops operating for more than a preset de-vibration time threshold, its working envelope is no longer considered an obstacle. If a tower crane is in operation, all voxels covered by its current working envelope are marked as occupied.

[0026] Preferably, the digital twin module calculates security risk weights. ,include:

[0027] The This indicates the overall safety risk level of voxel v at time t; a larger value indicates greater danger and the object is unsuitable for lifting. By taking the maximum value, the overall risk level is high as long as any high-risk factor exists, including:

[0028] ;

[0029] in, Risk of restricted hoisting areas; The risk level in the height-restricted area changes over time t. The risks in the mobile zone originate from the real-time location of personnel, vehicles, and mobile devices, which change over time t. For construction plan control, it is issued by the manager and changes over time t;

[0030] Among them, the risk of the motor zone The locations of all entities are obtained from the positioning system. For each voxel, the distance to the nearest entity is calculated, mapped to a risk value using an exponential decay function, and the maximum value among all entities is taken as the risk of that voxel. This is the risk of the maneuver zone. It is calculated using the following formula:

[0031] ;

[0032] in, It is a moment The entire collection of hazardous entities includes people, vehicles, and mobile equipment; It is a physical entity Location, It is a voxel Center coordinates It is a preset influence radius parameter; It is entity weight. Among them, the entity weight of a person is The entity weight of the vehicle is The physical weight of large machinery is .

[0033] Preferably, the dispatch module obtains the dispatch task type and dispatch initiation point and destination, and generates a set of feasible paths, including:

[0034] First, based on the spatial coverage relationship and load constraints of the hoisting task, determine whether there is a clear hoisting mode definition item. If there is, directly enter the preset dispatch mode. The preset dispatch modes include single tower hoisting mode, relay hoisting mode and joint lifting mode.

[0035] If not, activate the reverse joint planning mechanism to generate and determine the joint lifting execution plan;

[0036] Different path construction and dispatch weight calculation strategies are adopted for different hoisting modes to avoid scheduling instability caused by mixing different coordination logics.

[0037] The dispatching point and destination can be located within the working envelope of the same tower crane at the same time, and the weight of the goods does not exceed the rated load of the tower crane when dispatching is assigned to the single tower crane transport mode.

[0038] The dispatching point and the destination cannot be covered by the same tower crane, but the segmented hoisting can be completed through at least one handover point and dispatched to the relay hoisting mode.

[0039] If the load of the cargo exceeds the load of a single tower crane, multiple tower cranes can share the same weight of cargo and move along the same trajectory under time synchronization conditions, and are assigned to the joint lifting mode;

[0040] When relay hoisting mode and joint lifting mode are not triggered, or when the single tower optimality is not met, or when there is no explicit relay definition, or when there is no explicit joint lifting definition, the path back-reasoning joint planning mode is entered. The shortest spatial path is generated based on the starting point and the ending point, and the shortest path is expanded into a path influence zone within a preset buffer range. The set of candidate tower cranes whose working envelope covers the path influence zone is filtered, the intersection of the feasible domains of the candidate tower cranes at each time is constructed, and the relay path with the fewest handovers is selected.

[0041] Preferably, the dispatch module, in reverse-engineering the joint planning pattern construction, includes:

[0042] Define the dispatch initiation point Ps and the destination point Pg in Within the feasible region for hoisting, a voxel mesh is used, retaining only voxels with an Av of 1 (i.e., physically reachable). Dijkstra's algorithm is used to find the spatial shortest path Pgeo, which is expressed by the following formula:

[0043] ;

[0044] Extend the shortest path Pgeo into a three-dimensional tubular region. Three-dimensional tubular region Expressed by the following formula:

[0045] ;

[0046] in, is the center coordinate of voxel v; k is the index variable used to traverse all spatial points in the path Pgeo; This represents the coordinates of the kth spatial point on the shortest path; dbuffer is the preset buffer radius, which is set according to the component size and safety clearance.

[0047] In a confined space In the middle, multi-agent path stitching is performed to obtain a three-dimensional tubular region. The intersection with the working envelope of each tower crane is used to connect the intersections of the involved tower cranes to generate feasible paths, where, for each tower crane dt, if the reachable voxels of two tower cranes are in If there are adjacent or overlapping tower cranes, connect them with edges, filter those that can cover the start and / or end points, enumerate all possible tower crane relay sequences, and generate feasible paths.

[0048] Preferably, the dispatch module obtains the optimal hoisting path by comprehensively forming a path cost function through a weighted or hierarchical decision-making method, including:

[0049] The dispatch module assigns multi-dimensional weights to the path for each mode. The multi-dimensional weights include the load adaptability of the tower crane involved in the path, the current task queue length, the energy consumption per unit time, the path length, and the path risk weight. The multi-dimensional weights are combined to form a path cost function through weighted or hierarchical decision-making. Based on the path cost function, the path with the minimum cost is selected from the set of feasible hoisting paths as the optimal hoisting path.

[0050] Preferably, the simulation module simulates the operation of the optimal hoisting path within the feasible region constraints to obtain the safety probability. It can be calculated using the following formula:

[0051] ;

[0052] in After normalization, It is the start time; This is the optimal hoisting path; It is a voxel Risk weight at time t; It is the i-th voxel The time it takes for the crane to pass through is calculated using the following formula:

[0053] ;

[0054] in, These are the center coordinates of the k-th voxel; It is the distance from the k-th point to the (k+1)-th point; It is the average speed;

[0055] If the value is less than the preset safety threshold, the sliding task time window will be used to simulate again until clearance is granted, including:

[0056] Fixed optimal hoisting path Under the premise of adjusting the task start time Find a safe execution window such that the probability of safety is greater than a preset safety threshold.

[0057] This invention achieves a paradigm shift in multi-tower hoisting from static planning to dynamic, safe collaboration by constructing an automated control architecture based on digital twins, task dispatching, and path simulation. Specifically, the digital twin module encodes the physical accessibility of the construction site, the operational occupancy status, and safety management rules into a time-space joint discrete voxel state field, forming a high-dimensional dynamic feasible domain constraint. This significantly improves the completeness and timeliness of environmental perception. The dynamic feasible domain naturally covers the multi-tower overlapping operation area, automatically generating cross-tower hoisting paths and verifying the spatiotemporal safety of the entire link. It also enables relay hoisting of large components, breaking through the single-tower operation radius limitation and providing key technical support for the construction of super high-rise and large-span buildings. By automatically sliding the task time window and iteratively resimulating until a safe execution opportunity is found, the conservative strategy of rejecting orders upon encountering conflicts is abandoned. The innovative time window sliding + iterative simulation mechanism proactively identifies potential feasible execution periods while ensuring safety, effectively alleviating the problem of task backlog during peak periods. It requires no modification to existing tower crane hardware; simply connecting to common smart construction site data sources such as BIM models, UWB / RTK positioning, and AI recognition is sufficient to construct a voxel state field. The algorithm module can be deployed on edge servers, issuing commands via 5G low-latency networks, making it easy to implement and promote on existing construction sites. This not only significantly improves tower crane resource utilization and task release rates but also constructs a verifiable, traceable, and scalable intelligent hoisting safety base, laying a core technological foundation for future 5G+AI-driven unmanned smart construction sites.

[0058] On the other hand, the present invention also discloses a 5G cluster intelligent dispatching method based on digital twins, comprising:

[0059] Step S1: Construct a time-space joint state field, and encode the physical accessibility, work occupancy and safety management rules of the construction site into discrete voxel states, and use them as feasibility domain constraints.

[0060] Step S2: Obtain the dispatch task type and dispatch origin and destination. Based on the feasible domain constraints, obtain the set of feasible paths and filter the optimal hoisting path. Assign multi-dimensional weights to each path and construct the path cost function through weighted fusion or hierarchical decision-making to obtain the optimal hoisting path.

[0061] Step S3: Simulate the operation of the optimal hoisting path within the feasible region constraints to obtain the safety probability. If the safety probability is greater than a preset safety threshold, the operation is allowed. If it is less than the preset safety threshold, the task time window is slid and the simulation is repeated until the operation is allowed. The output of the operation includes the optimal hoisting path, the hoisting feasible region constraints, and the hoisting time plan.

[0062] Step S4: Execute the optimal hoisting path according to the hoisting time plan.

[0063] It is understandable that the aforementioned 5G cluster intelligent dispatching system and method based on digital twins has the same beneficial effects, and will not be elaborated further here. Attached Figure Description

[0064] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0065] Figure 1 A functional block diagram of a 5G cluster intelligent dispatching system based on digital twins provided in an embodiment of the present invention;

[0066] Figure 2 A digital twin architecture diagram of a 5G cluster intelligent dispatching system based on digital twins provided in an embodiment of the present invention;

[0067] Figure 3 A flowchart of a 5G cluster intelligent dispatching method based on digital twins provided in an embodiment of the present invention. Detailed Implementation

[0068] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0069] See Figure 1-2 As shown, this embodiment discloses a 5G cluster intelligent dispatching system based on digital twins, including:

[0070] The digital twin module constructs a joint time-space state field, which encodes the physical accessibility, work occupancy and safety management rules of the construction site into discrete voxel states, and uses these as constraints for the feasibility domain.

[0071] The dispatch module obtains the dispatch task type and dispatch initiation and destination points. Based on the feasible region constraint, it obtains a set of feasible paths and filters the optimal hoisting path. It assigns multi-dimensional weights to each path and constructs a path cost function through weighted fusion or hierarchical decision-making to obtain the optimal hoisting path.

[0072] The simulation module simulates the operation of the optimal hoisting path within the feasible region constraints, obtains the safety probability, and releases the operation when the safety probability is greater than the preset safety threshold. If the safety probability is less than the preset safety threshold, the sliding task time window is used to simulate again until the operation is released. The release output includes the optimal hoisting path, hoisting feasible region constraints, and hoisting time planning.

[0073] The control module plans and executes the optimal hoisting path based on the hoisting time.

[0074] Specifically, in this embodiment, the control module executes the optimal hoisting path according to the hoisting time plan. If the actual transfer time exceeds the transfer time limit, the real-time feasible domain constraint is obtained before the node is released, and the difference domain between it and the hoisting feasible domain constraint is compared. The collision probability of the difference domain is determined. If it is determined that adjustment is needed, all subsequent hoisting time plans are adjusted sequentially. When the number of consecutive adjustments exceeds a preset threshold, the hoisted item is dispatched to the isolation area, the task item is retained, and a global reassignment is triggered for the task. The collision probability and the safety probability are obtained by subtracting the safety probability from 1 using the same technique.

[0075] In some implementations of this application, the digital twin module divides the construction site into discrete voxel grids in three-dimensional space and introduces a discrete time dimension to construct a time-space joint state field.

[0076] For each voxel At any moment Define its voxel state vector as :

[0077] ;

[0078] in, These are the index coordinates of a three-dimensional discrete spatial grid, used to uniquely identify the position of each voxel in the construction site; Indicates physical accessibility, i.e., whether the voxel is located within the current working envelope of at least one tower crane; This indicates whether the voxel is occupied by operating equipment or personnel / vehicle entities. The safety risk weight is calculated comprehensively from no-lift zones, height-restricted zones, hazardous areas, and time-based safety rules.

[0079] Define a voxel feasibility indicator function based on voxel state vectors. :

[0080] ;

[0081] in, It is the maximum permissible safety risk threshold; It is a voxel Job occupancy at time t; It is the physical reachability of voxel v at time t;

[0082] This generates a time-varying feasible lifting region with the following constraints:

[0083] ;

[0084] in, It is the feasible region of hoisting based on the spatiotemporal joint state field, that is, the set of all voxels that are feasible at time t.

[0085] In some implementations of this application, the digital twin module defines the physical accessibility of the construction space. This includes: deploying tower crane operation status acquisition sensors at the construction site to obtain real-time data on the slewing angle, luffing length, hook height, and load mass of each tower crane; calculating its instantaneous working envelope based on the tower crane's geometric kinematic model; and determining if voxel v lies within the instantaneous working envelope of any tower crane. If it is 1, then if it is not within any working envelope. =0; where, the instantaneous working envelope refers to the set of all spatial locations that can be reached during tower crane hoisting operations.

[0086] Specifically, the spatial position of the hook is represented as a function of the slewing angle, luffing radius, and height. Combined with the tower crane's rated lifting capacity curve and safety margin constraints, the instantaneous working envelope of each tower crane at the current moment is calculated in real time. The tower crane's da at time... Instantaneous working envelope It is calculated using the following formula:

[0087] ;

[0088] in, It is the spatial position vector of the hook. It's a turning angle. It represents the current allowable turning angle range; u is the amplitude. This is the current allowable range of amplitude. It's about height. This is the current allowed altitude range; It is a curve showing the rated lifting capacity as the amplitude changes.

[0089] In some implementations of this application, the digital twin module determines job occupancy. ,include:

[0090] The system uses PLC signals, motor current, or control commands to determine in real time whether each tower crane is in operation. If a tower crane stops operating for more than a preset de-vibration time threshold, its working envelope is no longer considered an obstacle. If a tower crane is in operation, all voxels covered by its current working envelope are marked as occupied.

[0091] Specifically, the jitter removal time threshold can be set between 5 and 15 seconds; since the signal may have instantaneous jitter, such as brief pauses or fine adjustments, time filtering is introduced. The jitter removal example algorithm in this embodiment includes:

[0092] class CraneStateMonitor:

[0093] def __init__(self, crane_id, debounce_time=10.0): # Default is 10 seconds

[0094] self.crane_id = crane_id

[0095] self.debounce_time = debounce_time

[0096] self.last_active_time = current_time()

[0097] self.is_running = False def update(self, signal_active:bool):

[0098] if signal_active:

[0099] self.last_active_time = current_time()

[0100] self.is_running = True

[0101] else:

[0102] idle_duration = current_time() - self.last_active_time

[0103] if idle_duration >= self.debounce_time:

[0104] self.is_running = False

[0105] # Otherwise, maintain the original state (anti-shake).

[0106] In some implementations of this application, the digital twin module calculates security risk weights. ,include:

[0107] This indicates the overall safety risk level of voxel v at time t; a higher value indicates greater danger and the material is unsuitable for lifting. By taking the maximum value, the overall risk level is high as long as any high-risk factor exists, including:

[0108] ;

[0109] in, Risk of restricted hoisting areas; The risk level in the height-restricted area changes over time t. The risks in the mobile zone originate from the real-time location of personnel, vehicles, and mobile devices, which change over time t. For construction plan control, it is issued by the manager and changes over time t;

[0110] Among them, the risk of the motor zone The locations of all entities are obtained from the positioning system. For each voxel, the distance to the nearest entity is calculated, mapped to a risk value using an exponential decay function, and the maximum value among all entities is taken as the risk of that voxel. This is the risk of the maneuver zone. It is calculated using the following formula:

[0111] ;

[0112] in, It is a moment The entire collection of hazardous entities includes people, vehicles, and mobile equipment; It is a physical entity Location, It is a voxel Center coordinates It is a preset influence radius parameter; It is entity weight. Among them, the entity weight of a person is The entity weight of the vehicle is The physical weight of large machinery is .

[0113] Specifically, the risks of no-hogging zones The source is the permanently prohibited lifting areas marked in the BIM model, including but not limited to high-voltage power line corridors, and the logic is as follows:

[0114] ;

[0115] in, It involves extracting the no-lift zone polyhedron from BIM or CAD drawings, voxelizing it, and generating a set. Query voxels Is it in this set?

[0116] Risk of height restriction zone Temporary construction restrictions include, but are not limited to, nighttime height restrictions and weather-related height restrictions; Construction schedule control, including but not limited to the exclusion of hazardous areas during high-altitude welding periods.

[0117] In some implementations of this application, the dispatch module obtains the dispatch task type, dispatch origin, and destination, and generates a set of feasible paths, including:

[0118] First, based on the spatial coverage relationship and load constraints of the hoisting task, determine whether there is a clear hoisting mode definition item. If there is, directly enter the preset dispatch mode. The preset dispatch modes include single tower hoisting mode, relay hoisting mode and joint lifting mode.

[0119] If not, activate the reverse joint planning mechanism to generate and determine the joint lifting execution plan;

[0120] Different path construction and dispatch weight calculation strategies are adopted for different hoisting modes to avoid scheduling instability caused by mixing different coordination logics.

[0121] The dispatching point and destination can be located within the working envelope of the same tower crane at the same time, and the weight of the goods does not exceed the rated load of the tower crane when dispatching is assigned to the single tower crane transport mode.

[0122] The dispatching point and the destination cannot be covered by the same tower crane, but the segmented hoisting can be completed through at least one handover point and dispatched to the relay hoisting mode.

[0123] If the load of the cargo exceeds the load of a single tower crane, multiple tower cranes can share the same weight of cargo and move along the same trajectory under time synchronization conditions, and are assigned to the joint lifting mode;

[0124] When relay hoisting mode and joint lifting mode are not triggered, or when the single tower optimality is not met, or when there is no explicit relay definition, or when there is no explicit joint lifting definition, the path back-reasoning joint planning mode is entered. The shortest spatial path is generated based on the starting point and the ending point, and the shortest path is expanded into the path influence zone within the preset buffer range. The candidate tower crane set that covers the path influence zone is screened, the feasible domain intersection of the candidate tower cranes at each time is constructed, and the relay path with the fewest handover times is selected.

[0125] In some implementations of this application, the dispatch module and the construction of the reverse joint programming pattern include:

[0126] Define the dispatch initiation point Ps and the destination point Pg in Within the feasible region for hoisting, a voxel mesh is used, retaining only voxels with an Av of 1 (i.e., physically reachable). Dijkstra's algorithm is used to find the spatial shortest path Pgeo, which is expressed by the following formula:

[0127] ;

[0128] Extend the shortest path Pgeo into a three-dimensional tubular region. Three-dimensional tubular region Expressed by the following formula:

[0129] ;

[0130] in, is the center coordinate of voxel v; k is the index variable used to traverse all spatial points in the path Pgeo; This represents the coordinates of the kth spatial point on the shortest path; dbuffer is the preset buffer radius, which is set according to the component size and safety clearance.

[0131] In a confined space In the middle, multi-agent path stitching is performed to obtain a three-dimensional tubular region. The intersection with the working envelope of each tower crane is used to connect the intersections of the involved tower cranes to generate feasible paths, where for each tower crane da, if the reachable voxels of the two tower cranes are in If there are adjacent or overlapping tower cranes, connect them with edges, filter those that can cover the start and / or end points, enumerate all possible tower crane relay sequences, and generate feasible paths.

[0132] Specifically, there are multiple ways to solve this problem in the existing technology. This embodiment provides an example of one solution:

[0133] Get tower crane collection

[0134] Each tower crane The working envelope, the union of times within the task period, is expressed by the following formula:

[0135] ;

[0136] Define each tower crane in Reachable subfields in:

[0137]

[0138] like Then the tower crane Not participating in collaboration.

[0139] Defining the interoperability relationships between tower cranes includes: introducing the volume cable adjacency operator. : Represents voxels The 26 neighborhoods, including itself; define the tower crane. and It is permissible to hand over if and only if:

[0140] ;

[0141] in This is the intersection of morphological dilation, but it can be simplified to the following in discrete voxels:

[0142] ;

[0143] in, ;

[0144] Building a tower crane collaboration diagram ;

[0145] Then the node set ;

[0146] Edge set ;

[0147] Define the start / end point coverage of the tower crane, including:

[0148] ;

[0149] Define a valid tower crane sequence, including:

[0150] A tower crane sequence It is legal if and only if:

[0151] ;

[0152] The group of tower cranes that can be transferred must satisfy the following:

[0153] ;

[0154] Let all valid sequences be . ;

[0155] For a given sequence Define its set of collaborative paths as:

[0156] ;

[0157] in ,and The handover point; Indicates in the region From the inside arrive Given the set of all connected paths, the path concatenation operation results in the following set of feasible paths:

[0158] ;

[0159] In some implementations of this application, the dispatch module synthesizes a path cost function through weighted or hierarchical decision-making to obtain the optimal hoisting path, including:

[0160] The dispatch module assigns multi-dimensional weights to the path for each mode. The multi-dimensional weights include the load adaptability of the tower crane involved in the path, the current task queue length, the energy consumption per unit time, as well as the path length and path risk weight. The multi-dimensional weights are combined to form a path cost function through weighted or hierarchical decision-making. Based on the path cost function, the path with the minimum cost is selected from the set of feasible hoisting paths as the optimal hoisting path.

[0161] Specifically, the following weights are calculated for each candidate path, including:

[0162] Load adaptability weight The load adaptability coefficient is based on the weight of the goods and the rated lifting capacity of the tower crane at the maximum radius required for loading and unloading. The closer the value is to 1, the higher the utilization rate of the tower crane and the better the efficiency. However, it also means that the safety margin is smaller and the risk is relatively higher.

[0163] Task queue length weight The system counts the number of pending tasks on the tower cranes involved in executing the path. A shorter queue indicates longer idle time for the tower cranes, faster response times, and higher scheduling priority. The smaller the value, the better; this is a key indicator for measuring scheduling fairness and response speed.

[0164] Energy consumption weight per unit time The energy consumption per unit time during the path is estimated based on the motor power model, or it can be obtained through historical data calibration or simulation modeling. This weight is used to quantify operating costs and support green construction goals. Lower energy consumption is more conducive to energy conservation, emission reduction, and long-term operational optimization.

[0165] Path length weight The total length of the path in voxel space is calculated, which is the sum of the number of voxels traversed along the path. A shorter path results in less movement time, higher work efficiency, and a reduced probability of accidents. Therefore, Smaller is better, which is one of the core objectives of traditional path planning.

[0166] Path risk weight The risk of all voxels along the path is integrated and summed, and the results are obtained from the real-time voxel state field of the digital twin system. This reflects factors such as the collision probability, personnel density, and obstacle distance of the voxel at the current moment. This weight is directly related to safety. The lower the value, the safer the path.

[0167] After all five weights are normalized within the same mode, the path cost function is formed by weighted fusion or hierarchical decision-making. This method is existing technology and will not be described in detail in this embodiment.

[0168] In some embodiments of this application, a simulation module simulates the operation of the optimal hoisting path within the feasible region constraints to obtain the safety probability. It can be calculated using the following formula:

[0169] ;

[0170] in, It is the start time; This is the optimal hoisting path; It is a voxel Risk weight at time t Normalized; It is the i-th voxel The time it takes for the crane to pass through is calculated using the following formula:

[0171] ;

[0172] in, These are the center coordinates of the k-th voxel; It is the distance from the k-th point to the (k+1)-th point; It is the average speed;

[0173] If the value is less than the preset safety threshold, the sliding task time window will be used to simulate again until clearance is granted, including:

[0174] Fixed optimal hoisting path Under the premise of adjusting the task start time Find a safe execution window such that the probability of safety is greater than a preset safety threshold.

[0175] Understandably, i is the index of the target voxel, k is the index of the path segment, and the summation from k=0 to i−1 is to accumulate the time of the first i segments, because the first i segments must be completed before the i-th voxel can be reached.

[0176] In summary, this embodiment achieves a fundamental shift from static planning to dynamic, safe, and collaborative operation in tower crane operations by constructing an automatic control architecture that integrates digital twin modeling, intelligent task assignment, and dynamic path simulation.

[0177] The architecture of this embodiment fully relies on the smart construction site infrastructure. No hardware modification is required for the tower crane itself. By simply accessing the BIM model, UWB or RTK high-precision positioning system, AI recognition results, PLC operation signals and environmental perception data, a time-space joint discrete voxel state field can be constructed in real time on the edge computing node.

[0178] This voxel state field divides the entire construction site into three-dimensional voxel units with a fixed spatial resolution, and assigns multi-dimensional state information to each voxel at each moment, including whether it is within the geometric reach of any tower crane, whether it is occupied by operating equipment or personnel, whether it is located in a restricted space such as a high-voltage line protection zone or an adjacent danger zone, and a dynamic risk score calculated based on motion prediction. The resulting high-dimensional feasible domain constraint not only fully expresses the accessibility of physical space, but also accurately characterizes the safety window in the time dimension, significantly improving the completeness and timeliness of environmental perception, and providing a highly reliable decision-making basis for subsequent path planning.

[0179] Thanks to the unified encoding capability of this state field for the working envelopes of multiple tower cranes, the system can automatically identify the spatial intersection areas between tower cranes and generate cross-tower crane transport paths based on this. For example, when the lifting point and landing point of a component are located within the coverage areas of different tower cranes, the system can plan a relay scheme consisting of multiple sub-paths: the first tower crane lifts the component to the handover area, the second tower crane takes over the operation within a safe time window, and finally delivers it to the target location. In practical engineering applications, due to limitations in lifting efficiency, synchronous control difficulty, and safety redundancy requirements, the number of relay segments is usually extremely limited. Most relay lifting tasks can be completed within 1 to 3 segments, and schemes with more than 3 segments are rarely used due to excessive handovers and significantly increased coordination complexity. Therefore, this embodiment defaults to limiting the maximum number of relay segments to 3 or more during the path generation stage, covering all typical working conditions.

[0180] After the entire link is generated, it undergoes end-to-end spatiotemporal security verification to ensure that each path segment is executed within the dynamic feasible domain of its respective tower crane, and that there are no collisions or conflicts during the handover process. This mechanism effectively overcomes the limitations of the operating radius and load of a single tower crane, providing practical technical support for complex engineering scenarios such as the construction of the core tube of super high-rise buildings and the installation of large-span steel structures.

[0181] At the task scheduling level, this embodiment abandons the traditional conservative strategy and innovatively adopts a dynamic scheduling mechanism that combines time window sliding and iterative re-simulation. When the initially planned path cannot meet the safety threshold due to spatiotemporal conflicts with other tasks, the task is not directly rejected. Instead, the execution window of the task is automatically slid within a preset time elasticity range (e.g., ±30 seconds), and a full path simulation is re-performed based on the updated voxel state field. This process can be iteratively executed until the first feasible execution opportunity that meets the safety constraints is found. Once verified, a complete release instruction package is immediately generated. This mechanism, while ensuring inherent safety, can effectively explore potential feasible time periods, significantly alleviate the task backlog problem during peak construction periods, and improve overall scheduling efficiency.

[0182] The entire algorithm module can be deployed on field edge servers or 5G MEC nodes, and sends commands to the tower crane control system through low-latency communication networks (such as 5G uRLLC) to ensure the real-time performance and reliability of the control closed loop.

[0183] It supports interfaceing with mainstream tower crane brands via standard industrial protocols (such as Modbus TCP and CANopen), reading operational status and issuing path guidance signals, demonstrating excellent compatibility and engineering feasibility. Simultaneously, all decision-making processes automatically generate traceable digital credentials, including voxel occupancy records, safety probability logs, and time planning snapshots, providing highly reliable evidence for accident retrospective analysis, liability determination, and insurance claims.

[0184] On the other hand, see Figure 3 As shown, this embodiment also discloses a 5G cluster intelligent dispatching method based on digital twins, including:

[0185] Step S1: Construct a time-space joint state field, and encode the physical accessibility, work occupancy and safety management rules of the construction site into discrete voxel states, and use them as feasibility domain constraints.

[0186] Step S2: Obtain the dispatch task type and dispatch origin and destination. Based on the feasible region constraint, obtain the set of feasible paths and filter the optimal hoisting path. Assign multi-dimensional weights to each path and construct the path cost function through weighted fusion or hierarchical decision-making to obtain the optimal hoisting path.

[0187] Step S3: Simulate the operation of the optimal hoisting path within the feasible region constraints, obtain the safety probability, and release the operation when the safety probability is greater than the preset safety threshold. If it is less than the preset safety threshold, slide the task time window and simulate again until release. The release output includes the optimal hoisting path, hoisting feasible region constraints, and hoisting time planning.

[0188] Step S4: Execute the optimal hoisting path according to the hoisting time plan.

[0189] It is understandable that the aforementioned 5G cluster intelligent dispatching system and method based on digital twins has the same beneficial effects, and will not be elaborated further here.

[0190] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0191] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0192] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0193] These computer program instructions can also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0194] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A 5G-based intelligent dispatching system for multiple towers based on digital twins, characterized in that, include: The digital twin module constructs a joint time-space state field, which encodes the physical accessibility, work occupancy and safety management rules of the construction site into discrete voxel states, and uses these as constraints for the feasibility domain. The dispatch module obtains the dispatch task type and dispatch initiation and destination points. Based on the feasible domain constraints, it obtains a set of feasible paths and filters the optimal hoisting path. It assigns multi-dimensional weights to each path and constructs a path cost function through weighted fusion or hierarchical decision-making to obtain the optimal hoisting path. The simulation module simulates the operation of the optimal hoisting path within the feasible domain constraints, obtains the safety probability, and releases the operation when the safety probability is greater than a preset safety threshold. If the safety probability is less than the preset safety threshold, the sliding task time window is used to simulate again until the operation is released. The release output includes the optimal hoisting path, hoisting feasible domain constraints, and hoisting time planning. The control module plans and executes the optimal hoisting path based on the hoisting time.

2. The 5G cluster intelligent dispatching system based on digital twins according to claim 1, characterized in that, The digital twin module divides the construction site into discrete voxel grids in three-dimensional space and introduces a discrete time dimension to construct a time-space joint state field. For each voxel At any moment Define its voxel state vector as : ； in, These are the index coordinates of a three-dimensional discrete spatial grid, used to uniquely identify the position of each voxel in the construction site; Indicates physical accessibility, i.e., whether the voxel is located within the current working envelope of at least one tower crane; This indicates whether the voxel is occupied by operating equipment or personnel / vehicle entities. The safety risk weight is calculated comprehensively from no-lift zones, height-restricted zones, hazardous areas, and time-based safety rules. Define a voxel feasibility indicator function based on voxel state vectors. : ； in, It is the maximum permissible safety risk threshold; It is a voxel Job occupancy at time t; It is the physical reachability of voxel v at time t; This generates a time-varying feasible lifting region with the following constraints: ； in, It is the feasible region of hoisting based on the spatiotemporal joint state field, that is, the set of all voxels that are feasible at time t.

3. The 5G cluster intelligent dispatching system based on digital twins according to claim 2, characterized in that, The digital twin module defines the physical accessibility of the construction space. This includes: deploying tower crane operation status acquisition sensors at the construction site to obtain real-time data on the slewing angle, luffing length, hook height, and load mass of each tower crane; calculating its instantaneous working envelope based on the tower crane's geometric kinematic model; and determining if voxel v lies within the instantaneous working envelope of any tower crane. If it is 1, then if it is not within any working envelope. =0; where the instantaneous working envelope refers to the set of all spatial locations that can be reached during tower crane hoisting operations.

4. The 5G cluster intelligent dispatching system based on digital twins according to claim 3, characterized in that, The digital twin module determines job occupancy. ,include: The system uses PLC signals, motor current, or control commands to determine in real time whether each tower crane is in operation. If a tower crane stops operating for more than a preset de-vibration time threshold, its working envelope is no longer considered an obstacle. If a tower crane is in operation, all voxels covered by its current working envelope are marked as occupied.

5. The 5G cluster intelligent dispatching system based on digital twins according to claim 4, characterized in that, The digital twin module calculates security risk weights. ,include: The This indicates the overall safety risk level of voxel v at time t; a larger value indicates greater danger and the object is unsuitable for lifting. By taking the maximum value, the overall risk level is high as long as any high-risk factor exists, including: ； in, Risk of restricted hoisting areas; The risk level in the height-restricted area changes over time t. The risks in the mobile zone originate from the real-time location of personnel, vehicles, and mobile devices, which change over time t. For construction plan control, it is issued by the manager and changes over time t; Among them, the risk of the motor zone The locations of all entities are obtained from the positioning system. For each voxel, the distance to the nearest entity is calculated, mapped to a risk value using an exponential decay function, and the maximum value among all entities is taken as the risk of that voxel. This is the risk of the maneuver zone. It is calculated using the following formula: ； in, It is a moment The entire collection of hazardous entities includes people, vehicles, and mobile equipment; It is a physical entity Location, It is a voxel Center coordinates It is a preset influence radius parameter; It is entity weight. .

6. The 5G cluster intelligent dispatching system based on digital twins according to claim 5, characterized in that, The dispatch module obtains the dispatch task type, dispatch initiation point, and dispatch destination, and generates a set of feasible paths, including: First, based on the spatial coverage relationship and load constraints of the hoisting task, determine whether there is a clear hoisting mode definition item. If there is, directly enter the preset dispatch mode. The preset dispatch modes include single tower hoisting mode, relay hoisting mode and joint lifting mode. If not, activate the reverse joint planning mechanism to generate and determine the joint lifting execution plan; Different path construction and dispatch weight calculation strategies are adopted for different hoisting modes to avoid scheduling instability caused by mixing different coordination logics. The dispatching point and destination can be located within the working envelope of the same tower crane at the same time, and the weight of the goods does not exceed the rated load of the tower crane when dispatching is assigned to the single tower crane transport mode. The dispatching point and the destination cannot be covered by the same tower crane, but the segmented hoisting can be completed through at least one handover point and dispatched to the relay hoisting mode. If the load of the cargo exceeds the load of a single tower crane, multiple tower cranes can share the same weight of cargo and move along the same trajectory under time synchronization conditions, and are assigned to the joint lifting mode; When relay hoisting mode and joint lifting mode are not triggered, or when single-tower optimality is not met, or when there is no explicit relay definition, or when there is no explicit joint lifting definition, the path back-reasoning joint planning mode is entered. Based on the start and end points, the shortest spatial path is generated, and the shortest path is expanded into a path influence zone within a preset buffer range. This path influence zone is a three-dimensional tubular region. The set of candidate tower cranes whose working envelopes cover the influence zone of the path is filtered, the intersection of the feasible regions of the candidate tower cranes at each time is constructed, and the relay path with the fewest handovers is selected.

7. The 5G cluster intelligent dispatching system based on digital twins according to claim 6, characterized in that, The dispatch module, in reverse engineering the joint planning pattern, includes: Define the dispatch initiation point Ps and the destination point Pg in Within the feasible region for hoisting, a voxel mesh is used, retaining only voxels with an Av of 1 (i.e., physically reachable). Dijkstra's algorithm is used to find the spatial shortest path Pgeo, which is expressed by the following formula: ； Extend the shortest path Pgeo into a three-dimensional tubular region. Three-dimensional tubular region Expressed by the following formula: ； in, is the center coordinate of voxel v; k is the index variable used to traverse all spatial points in the path Pgeo; This represents the coordinates of the kth spatial point on the shortest path; dbuffer is the preset buffer radius, which is set according to the component size and safety clearance. In a confined space In the middle, multi-agent path stitching is performed to obtain a three-dimensional tubular region. The intersection with the working envelope of each tower crane is used to connect the intersections of the involved tower cranes to generate feasible paths, where, for each tower crane dt, if the reachable voxels of two tower cranes are in If there are adjacent or overlapping tower cranes, connect them with edges, filter those that can cover the start and / or end points, enumerate all possible tower crane relay sequences, and generate feasible paths.

8. The 5G cluster intelligent dispatching system based on digital twins according to claim 7, characterized in that, The dispatch module, through a weighted or hierarchical decision-making method, synthesizes a path cost function to obtain the optimal hoisting path, including: The dispatch module assigns multi-dimensional weights to the path for each mode. The multi-dimensional weights include the load adaptability of the tower crane involved in the path, the current task queue length, the energy consumption per unit time, the path length, and the path risk weight. The multi-dimensional weights are combined to form a path cost function through weighted or hierarchical decision-making. Based on the path cost function, the path with the minimum cost is selected from the set of feasible hoisting paths as the optimal hoisting path.

9. The 5G cluster intelligent dispatching system based on digital twins according to claim 8, characterized in that, The simulation module simulates the operation of the optimal hoisting path within the feasible region constraints to obtain the safety probability. It can be calculated using the following formula: ； in After normalization, It is the start time; This is the optimal hoisting path; It is a voxel Risk weight at time t; It is the i-th voxel The time it takes for the crane to pass through is calculated using the following formula: ； in, These are the center coordinates of the k-th voxel; It is the distance from the k-th point to the (k+1)-th point; It is the average speed; If the value is less than the preset safety threshold, the sliding task time window will be used to simulate again until clearance is granted, including: Fixed optimal hoisting path Under the premise of adjusting the task start time Find a safe execution window such that the probability of safety is greater than a preset safety threshold.

10. A 5G cluster intelligent dispatching method based on digital twins, used in the 5G cluster intelligent dispatching system based on digital twins as described in any one of claims 1-9, characterized in that, include: Step S1: Construct a time-space joint state field, and encode the physical accessibility, work occupancy and safety management rules of the construction site into discrete voxel states, and use them as feasibility domain constraints. Step S2: Obtain the dispatch task type and dispatch origin and destination. Based on the feasible domain constraints, obtain the set of feasible paths and filter the optimal hoisting path. Assign multi-dimensional weights to each path and construct the path cost function through weighted fusion or hierarchical decision-making to obtain the optimal hoisting path. Step S3: Simulate the operation of the optimal hoisting path within the feasible region constraints to obtain the safety probability. If the safety probability is greater than a preset safety threshold, the operation is allowed. If it is less than the preset safety threshold, the task time window is slid and the simulation is repeated until the operation is allowed. The output of the operation includes the optimal hoisting path, the hoisting feasible region constraints, and the hoisting time plan. Step S4: The control module plans and executes the optimal hoisting path according to the hoisting time.

Images