Artificial intelligence-based tower crane trajectory planning method and device

Abstract

The application provides a tower crane trajectory planning method and device based on artificial intelligence, and belongs to the technical field of path planning. In the method, a multi-agent collaborative architecture based on a shared situation blackboard is introduced, efficient interaction between a trajectory planning agent and a time-space conflict negotiation agent is utilized, dynamic prediction and real-time resolution of time-space conflicts of a group tower system are realized, compared with a traditional static layout anti-collision mode, the scheme not only overcomes the limitation of relying on physical distance planning, significantly reduces the collision risk of the tower crane in a complex cross-operation area, but also greatly improves the lifting efficiency and scheduling flexibility of the group tower system through intelligent path re-planning and time occupancy locking mechanism under the premise of ensuring construction safety.

Description

Technical Field

[0001] This invention relates to the technical field of path planning, and in particular to a method and apparatus for tower crane trajectory planning based on artificial intelligence. Background Technology

[0002] Tower cranes, common pieces of machinery on construction sites, are primarily used to transport various building materials to higher floors and are an indispensable part of modern construction. When two or more tower cranes operate simultaneously on a construction site, these systems are collectively referred to as a tower crane cluster system. In many construction projects, due to limitations in the work area, it is sometimes necessary to deploy a large number of tower crane cluster systems to meet construction needs. Because each tower crane may need to transport materials in different directions and at different heights, the risk of collisions between tower cranes and between tower cranes and other on-site facilities is greatly increased. Therefore, effectively preventing collisions becomes a crucial task during the operation of a tower crane cluster system.

[0003] Traditionally, before installing a tower crane cluster system, technicians meticulously plan the installation location of each tower crane based on the specific conditions of the construction site, striving to avoid collisions while ensuring that the operating range of all tower cranes fully covers the entire construction work area. However, despite detailed pre-planning, the risk of collisions between tower crane cluster systems is not only affected by the installation location of the tower cranes but also by the dynamically changing conditions of the construction site. This means that even with thorough preparation, the possibility of collisions cannot be completely eliminated, resulting in limited effectiveness of existing collision avoidance measures. Therefore, exploring more efficient and flexible collision avoidance strategies for tower crane clusters is crucial. This requires us to consider not only static layout design but also to integrate advanced technologies such as real-time monitoring and intelligent scheduling to improve the overall safety and operational efficiency of the tower crane cluster system.

[0004] In summary, how to avoid collisions between tower cranes in a multi-tower system has become an urgent technical problem to be solved. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide a tower crane trajectory planning method and device based on artificial intelligence, so as to alleviate the technical problem that it is impossible to avoid collisions between tower cranes in a multi-tower system in the traditional technology.

[0006] In a first aspect, embodiments of the present invention provide an artificial intelligence-based tower crane trajectory planning method, implemented by a shared situation blackboard and a cluster of intelligent agents communicatively connected thereto. The cluster of intelligent agents includes at least a trajectory planning agent and a spatiotemporal conflict negotiation agent. The method includes: The trajectory planning agent acquires the hoisting task and construction site model, generates candidate hoisting paths and their estimated time occupancy, and publishes them to the shared situation blackboard; The spatiotemporal conflict negotiation agent reads the information in the shared situation blackboard, detects the spatiotemporal occupancy conflict between the candidate hoisting path and the other tower cranes' planned optimal hoisting paths published on the shared situation blackboard, generates a conflict resolution instruction, and writes it into the shared situation blackboard; The trajectory planning agent adjusts the candidate hoisting paths according to the conflict resolution instructions, and publishes the conflict-free optimal hoisting path to the shared situation blackboard to guide the tower crane to execute the hoisting path.

[0007] Furthermore, the intelligent agent cluster also includes environmentally perceptive intelligent agents; The environmental perception agent establishes the construction site model and obtains the real-time status of each tower crane. It publishes the construction site model and the real-time status to the shared situation blackboard, so that the trajectory planning agent can obtain the construction site model and the real-time status from the shared situation blackboard and generate candidate hoisting paths in combination with the hoisting task.

[0008] Furthermore, the environmental perception agent establishes a construction site model, including: A three-dimensional mesh model of the construction site in a cylindrical coordinate system is established, and a set of mesh nodes is generated. The angular coordinates are divided equally with the central axis of the tower crane as the vertical axis, and the angular coordinates are used as the first index, while the radius coordinates and height coordinates are used as the second index.

[0009] Furthermore, the trajectory planning agent generates candidate hoisting paths, including: A heuristic search algorithm is used to maintain a first list and a second list. The hoisting starting point is placed in the first list. If the first list is not empty, the node with the minimum total cost is taken as the current node and added to the second list. The adjacent reachable nodes of the current node are calculated to obtain the child nodes. After processing each child node accordingly, the nodes are repeatedly extracted until the hoisting endpoint is reached, and candidate hoisting paths are obtained by backtracking based on the parent node.

[0010] Furthermore, the calculation of the total cost introduces a vertical priority weight function, a regional suppression weight function, and a horizontal guidance weight function, and the weights of each weight function are dynamically adjusted according to environmental factors; The regional suppression weight function is calculated based on the horizontal distance from the current node to the hoisting start point, the total horizontal distance from the hoisting start point to the hoisting end point, and the preset reference height. The horizontal guidance weight function is calculated based on the coordinate relationship between the current node and the hoisting endpoint.

[0011] Furthermore, the conflict resolution instruction includes at least one of: discarding the conflict node, delaying the waiting for the conflict node, or replanning the local path.

[0012] Furthermore, the spatiotemporal conflict negotiation agent detects spatiotemporal occupancy conflicts, including: Obtain the estimated occupancy time of nodes in the candidate hoisting path, and the occupancy time of the node in other tower cranes' planned optimal hoisting paths; Calculate the absolute value of the difference between the estimated occupancy time and the occupancy time; When the absolute value is less than a preset time threshold, it is determined that there is a time-space occupancy conflict; The preset time threshold is dynamically determined based on at least one of wind speed, the swing amplitude of the suspended object, and the site congestion.

[0013] Furthermore, it also includes: When the trajectory planning agent publishes the optimal hoisting path to the shared situation blackboard, it locks the time occupancy commitment of each node in the optimal hoisting path as a constraint for other tower crane path planning.

[0014] Furthermore, the intelligent agent cluster also includes an execution guiding intelligent agent; The execution guidance agent obtains the optimal hoisting path from the shared situation blackboard, guides the current tower crane to perform hoisting according to the optimal hoisting path, and updates the execution progress to the shared situation blackboard. The intelligent agent cluster also includes an anomaly monitoring intelligent agent; The anomaly monitoring agent obtains the real-time status of the tower crane and the execution progress of the optimal hoisting path from the shared situation blackboard. When the execution deviation is detected to exceed a preset threshold, a replanning trigger event is issued to the shared situation blackboard. The trajectory planning agent and the spatiotemporal conflict negotiation agent respond to the replanning trigger event and replan the path.

[0015] Secondly, embodiments of the present invention also provide an artificial intelligence-based tower crane trajectory planning device, comprising: a shared situation blackboard, and a trajectory planning unit and a spatiotemporal conflict negotiation unit respectively communicatively connected to the shared situation blackboard; wherein, The trajectory planning unit is used to acquire the hoisting task and construction site model, generate candidate hoisting paths and their estimated time occupancy, and publish them to the shared situation blackboard; The spatiotemporal conflict negotiation unit is used to read the information in the shared situation blackboard, detect the spatiotemporal occupancy conflict between the candidate hoisting path and the other tower cranes' planned optimal hoisting paths published on the shared situation blackboard, generate a conflict resolution instruction, and write it into the shared situation blackboard. The trajectory planning unit is also used to adjust the candidate hoisting paths according to the conflict resolution instructions, and publish the conflict-free optimal hoisting path to the shared situation blackboard to guide the tower crane to execute.

[0016] In this embodiment of the invention, an artificial intelligence-based tower crane trajectory planning method is provided, implemented by a shared situation blackboard and a cluster of intelligent agents connected to it. The cluster of intelligent agents includes at least a trajectory planning intelligent agent and a spatiotemporal conflict negotiation intelligent agent. The method includes: the trajectory planning intelligent agent acquiring the hoisting task and construction site model, generating candidate hoisting paths and their estimated time occupancy, and publishing them to the shared situation blackboard; the spatiotemporal conflict negotiation intelligent agent reading the information in the shared situation blackboard, detecting spatiotemporal conflicts between the candidate hoisting paths and other tower cranes' planned optimal hoisting paths published on the shared situation blackboard, generating conflict resolution instructions, and writing them to the shared situation blackboard; the trajectory planning intelligent agent adjusting the candidate hoisting paths according to the conflict resolution instructions, and publishing the conflict-free optimal hoisting paths to the shared situation blackboard to guide the tower cranes to execute the hoisting paths. As described above, the tower crane trajectory planning method based on artificial intelligence of the present invention introduces a multi-agent collaborative architecture based on a shared situation blackboard, and utilizes the efficient interaction between the trajectory planning agent and the spatiotemporal conflict negotiation agent to realize the dynamic prediction and real-time resolution of spatiotemporal conflicts in the tower crane group system. Compared with the traditional static layout anti-collision method, this solution not only overcomes the limitations of relying solely on physical spacing planning and significantly reduces the collision risk of tower cranes in complex cross-operation areas, but also greatly improves the hoisting efficiency and scheduling flexibility of the tower crane group system under the premise of ensuring construction safety through intelligent path replanning and time occupancy locking mechanisms. Attached Figure Description

[0017] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 A flowchart illustrating an artificial intelligence-based tower crane trajectory planning method provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0019] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] In traditional technologies, it is impossible to avoid collisions between tower cranes in a multi-tower system.

[0021] Based on this, the tower crane trajectory planning method based on artificial intelligence of the present invention introduces a multi-agent collaborative architecture based on a shared situation blackboard, and utilizes the efficient interaction between the trajectory planning agent and the spatiotemporal conflict negotiation agent to realize the dynamic prediction and real-time resolution of spatiotemporal conflicts in the tower crane group system. Compared with the traditional static layout anti-collision method, this solution not only overcomes the limitations of relying solely on physical spacing planning and significantly reduces the collision risk of tower cranes in complex cross-operation areas, but also greatly improves the hoisting efficiency and scheduling flexibility of the tower crane group system under the premise of ensuring construction safety through intelligent path replanning and time occupancy locking mechanism.

[0022] To facilitate understanding of this embodiment, a detailed description of an artificial intelligence-based tower crane trajectory planning method disclosed in this embodiment of the invention will be provided first.

[0023] Example 1: According to an embodiment of the present invention, an embodiment of a tower crane trajectory planning method based on artificial intelligence is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0024] Figure 1 This is a flowchart of an artificial intelligence-based tower crane trajectory planning method according to an embodiment of the present invention, such as... Figure 1 As shown, the method includes the following steps: Step S102: The trajectory planning agent acquires the hoisting task and construction site model, generates candidate hoisting paths and their estimated time occupancy, and publishes them to the shared situation blackboard; The method of this invention is based on a multi-agent cooperative architecture with a shared situation blackboard, achieving collision-free trajectory planning for a cluster tower system through cooperation and communication among agents. This method is implemented by a shared situation blackboard and a cluster of agents communicating with it.

[0025] In this step, the trajectory planning agent, as the core unit for path generation, first needs to obtain basic planning inputs. The "lifting task" typically includes the coordinates of the starting point (e.g., material storage area) and the ending point (e.g., a designated location on the working floor). The "construction site model" is a digital 3D environment containing buildings, scaffolding, and other fixed obstacles.

[0026] After obtaining the above information, the trajectory planning agent initiates a path search algorithm (such as the A* algorithm or its variants) to search for feasible paths from the starting point to the destination in three-dimensional space. During this process, the agent not only calculates spatial coordinates but also needs to combine the tower crane's kinematic model (such as hoisting, luffing, and slewing speeds) to estimate the "estimated time occupancy" of each node on the path, i.e., the estimated time point or time period for the tower crane to reach that node. After completing the calculation, the generated "candidate hoisting paths," along with their time occupancy information, are uniformly published to the "shared situation blackboard."

[0027] The trajectory planning agent can obtain hoisting tasks by subscribing to the message queue of the task scheduling system; the construction site model can be obtained by converting the BIM model through LiDAR scanning, and can be implemented by the environmental perception agent. The estimated time occupancy is calculated by dividing the Euclidean distance or arc length between path nodes by the rated average speed of the tower crane under that action (lifting / luffing / slewing), and then summing the results to obtain the cumulative time.

[0028] By introducing the time dimension in the early stages of path planning, a foundation is laid for subsequent spatiotemporal conflict detection, making the path not only spatial obstacle avoidance, but also orderly passage in time.

[0029] In step S104, the spatiotemporal conflict negotiation agent reads the information in the shared situation blackboard, detects spatiotemporal occupancy conflicts between the candidate hoisting path and the other tower cranes' planned optimal hoisting paths published on the shared situation blackboard, generates conflict resolution instructions, and writes them into the shared situation blackboard.

[0030] This step is the core of resolving collisions among multiple tower cranes. The "spatiotemporal conflict negotiation agent" acts as a "traffic policeman." It monitors the shared situation blackboard in real time, reading the "candidate hoisting paths" submitted by the current tower crane and the "optimal hoisting paths" already locked by other tower cranes.

[0031] During conflict detection, each spatial node on the candidate hoisting path is traversed to check if the node is occupied by the optimal hoisting path of another tower crane. If the "time difference" between two tower cranes occupying the same node is less than a safety threshold (i.e., they arrive at the point almost simultaneously), it is determined to be a "spatiotemporal occupancy conflict". Once a conflict is detected, the negotiating AI will generate a "conflict resolution instruction", such as "wait", "slow down", or "detour".

[0032] Conflict detection can employ time-of-flight constraint algorithms. For example, calculate the arrival time T1 of the current tower crane at node N and the arrival times T2 of other tower cranes at node N. If |T1-T2| < ΔT (e.g., ΔT = 2 seconds), a collision is considered to have occurred. Resolution instructions can be implemented by modifying the cost weights of path nodes or by directly inserting waiting timestamps.

[0033] Decentralized distributed negotiation is achieved, and information decoupling is achieved through a shared blackboard, which enables the path planning of multiple tower cranes to be independent and efficient, effectively avoiding physical collisions.

[0034] In step S106, the trajectory planning agent adjusts the candidate hoisting paths according to the conflict resolution instructions and publishes the optimal hoisting path without conflict to the shared situation blackboard to guide the tower crane to execute the hoisting path.

[0035] After publishing candidate hoisting paths, the trajectory planning agent continuously monitors the shared situation board for feedback on those paths. Upon receiving a "conflict resolution instruction," the agent treats it as a new constraint and readjusts its path search strategy. For example, if the instruction is "discard conflicting nodes," the node is marked as unreachable in the search graph; if it's "delay," the timeline is adjusted. After adjustment, the trajectory planning agent generates a new, conflict-free "optimal hoisting path," republishes it on the shared situation board, and locks it for the execution layer to use.

[0036] Path adjustments can employ an incremental replanning strategy, correcting only the conflicting localities rather than recalculating globally, thus improving response speed. The publication of the optimal hoisting path is typically accompanied by a "locking" mechanism to prevent other agents from modifying the path during this period.

[0037] This forms a closed-loop control system of "planning-detection-correction," ensuring that the final output path is safe and executable in complex and dynamic construction environments.

[0038] In the artificial intelligence-based tower crane trajectory planning method of this invention, a multi-agent collaborative architecture based on a shared situation blackboard is introduced. By utilizing the efficient interaction between the trajectory planning agent and the spatiotemporal conflict negotiation agent, dynamic prediction and real-time resolution of spatiotemporal conflicts in a group of tower cranes are achieved. Compared with the traditional static layout anti-collision method, this scheme not only overcomes the limitations of relying solely on physical spacing planning and significantly reduces the collision risk of tower cranes in complex cross-operation areas, but also greatly improves the hoisting efficiency and scheduling flexibility of the group of tower cranes under the premise of ensuring construction safety through intelligent path replanning and time occupancy locking mechanisms.

[0039] Based on the above, this embodiment further details the composition of the intelligent agent cluster and the environmental perception process. The intelligent agent cluster also includes environmental perception agents.

[0040] The environmental perception agent establishes a construction site model and obtains the real-time status of each tower crane, and publishes the construction site model and real-time status to the shared situation blackboard.

[0041] The environmental perception agent is responsible for building and maintaining the "digital foundation" of the entire system. It collects data from the construction site in real time through sensors (such as cameras, lidar, and encoders) and dynamically updates the "construction site model," for example, identifying newly arrived piles of rebar or moving cranes. At the same time, it also obtains the real-time status of each tower crane (such as the current hook position, slewing angle, and operating speed).

[0042] The trajectory planning agent obtains the construction site model and real-time status from the shared situation blackboard, and generates candidate hoisting paths in combination with the hoisting task.

[0043] The trajectory planning agent no longer passively receives static models, but instead obtains the latest dynamic information published by the environment-aware agent from the shared situational blackboard. This ensures that the environmental data used for path planning is real-time and accurate, rather than outdated drawing data.

[0044] Environmental perception agents can use SLAM (Simultaneous Localization and Mapping) technology to build 3D point cloud maps in real time. After acquiring real-time point cloud data, trajectory planning agents can rasterize it for obstacle detection using the A* algorithm.

[0045] It enables rapid response to dynamically changing construction environments, improves the safety and adaptability of path planning, and avoids planning failures caused by environmental changes (such as temporary obstacles).

[0046] Based on the above, this embodiment describes in detail the method for establishing a construction site model. The establishment of the construction site model by the environmental perception agent includes specific coordinate system processing.

[0047] Establish a three-dimensional mesh model of the construction site in cylindrical coordinate system and generate a set of mesh nodes.

[0048] Considering the motion characteristics of tower cranes (mainly rotation and luffing), planning paths in a Cartesian coordinate system often involves large computational loads and does not conform to the intuitive understanding of tower crane motion. Therefore, this embodiment uses a cylindrical coordinate system for modeling. The central axis of the tower crane body is used as the vertical axis (Z-axis) of the cylindrical coordinate system, and the angular coordinate (θ) of the horizontal plane is equally divided.

[0049] Use angular coordinates as the first index, and radius and height coordinates as the second index.

[0050] To efficiently store and retrieve grid nodes, this embodiment designs a two-layer index structure. The first layer of the index is based on "angular coordinates," dividing the 360-degree circumference into several sectors; the second layer of the index is based on "radius coordinates" (distance from the tower body) and "height coordinates" (height of the hook). This structure allows for rapid location of nodes within a specific angle and radius range during path searching, greatly improving search efficiency.

[0051] In practice, the angular coordinate θ can be divided into equal parts with a step size of 1 degree or 5 degrees, generating an integer index from 0 to 359. A two-dimensional array or hash table is constructed in memory, with keys being (angle, radius, height) and values ​​being the state of the grid node (free / obstacle / occupied).

[0052] By utilizing a cylindrical coordinate system to fit the kinematic characteristics of tower cranes and combining it with a dual-index mechanism, the spatial complexity of the path planning algorithm is significantly reduced, and the path search speed is accelerated, making it particularly suitable for tower crane operation scenarios with rotational characteristics.

[0053] Based on the above, this embodiment details the specific algorithm flow for the trajectory planning agent to generate candidate hoisting paths. This process employs a heuristic search algorithm.

[0054] A heuristic search algorithm is used to maintain a first list and a second list.

[0055] The trajectory planning agent initiates an A* or other heuristic search algorithm. In this process, the "first list" typically refers to the Open List, which stores nodes to be examined; the "second list" typically refers to the Closed List, which stores nodes that have already been examined.

[0056] Place the hoisting starting point into the first list. If the first list is not empty, take the node with the smallest total cost as the current node and add it to the second list.

[0057] The algorithm begins by adding the "lifting start point" to the Open List. In each iteration, the node with the smallest "total cost" (f value, i.e., the estimated total cost from the start point through the current point to the end point) is taken from the Open List as the "current node" and moved to the Close List, indicating that the node has been processed.

[0058] Calculate the adjacent reachable nodes of the current node to obtain child nodes; after processing each child node accordingly, repeat the process of extracting nodes until the hoisting destination is reached.

[0059] For the current node, calculate all accessible and unobstructed "adjacent nodes" in its surroundings (e.g., up, down, left, right, front, back, etc.) and designate them as "child nodes". For each child node, process it according to standard A* logic (e.g., update the parent node, update the cost, etc.) and add it to the Open List. This process is repeated until the node removed from the Open List is the "lifting destination".

[0060] Candidate hoisting paths are obtained by backtracking from the parent node.

[0061] Once the search reaches the hoisting endpoint, the algorithm uses a "backtracking" mechanism to start from the endpoint and trace back to the starting point along the "parent node" pointers recorded at each node, thereby generating the final candidate path.

[0062] The total cost f is typically composed of the g value (the actual cost from the starting point to the current node) and the h value (the heuristically estimated cost from the current node to the end point). The calculation of adjacent nodes needs to consider the crane's motion constraints, such as the inability to pass through walls or drag heavy objects horizontally on the ground.

[0063] The algorithm uses a classic graph search algorithm to find the optimal path in a discretized grid model, ensuring the global optimality or suboptimalness of the path. The algorithm is mature, stable, and easy to implement in engineering.

[0064] Based on the above, this embodiment optimizes the cost evaluation function in the path search process. Multiple weighting functions are introduced to calculate the total cost.

[0065] The calculation of total cost incorporates a vertical priority weight function, a regional suppression weight function, and a horizontal guidance weight function, and the weights of each weight function are dynamically adjusted according to environmental factors.

[0066] To ensure that the planned path better conforms to actual construction safety regulations, this embodiment introduces three specific weight functions into the heuristic function: Vertical priority function: encourages tower cranes to rise vertically first to reach a safe height as quickly as possible and avoid collisions between suspended loads and ground personnel or equipment at low altitudes.

[0067] Region inhibition weight function: penalizes path nodes that are close to obstacles or dangerous areas, forcing paths away from these areas.

[0068] Horizontal guidance weight function: guides the tower crane to move horizontally at high altitudes towards the target point.

[0069] Furthermore, the weights of these functions are not fixed, but dynamically adjusted based on "environmental factors" such as wind speed, whether construction is being carried out at night, and site congestion. For example, when it is windy, the weight of vertical priority is increased to reduce the time that suspended objects remain in the air.

[0070] The regional suppression weight function is calculated based on the horizontal distance from the current node to the hoisting start point, the total horizontal distance from the hoisting start point to the hoisting end point, and the preset reference height.

[0071] The specific calculation logic of this function takes into account the progress and height of the path. For example, in the early stages of lifting (near the starting point), if the height has not reached the reference height, a larger penalty value is applied.

[0072] The horizontal guidance weight function is calculated based on the coordinate relationship between the current node and the hoisting endpoint.

[0073] This function is typically based on the straight-line distance or angular deviation between the current node and the endpoint. The smaller the deviation, the lower the cost, thus guiding the path to converge toward the endpoint.

[0074] The vertical priority function can be designed such that when the node height is below the safe height, the cost increases sharply with the horizontal movement distance. The area inhibition weight function can be based on an inverse proportional function of distance to obstacles. Dynamic weight adjustment can be achieved through a fuzzy logic controller, taking environmental parameters as input and outputting weight coefficients.

[0075] By using a customized cost function, the automatically generated path not only avoids static obstacles but also conforms to safe operating procedures such as "fast take-off and slow landing" and "high-altitude horizontal movement," thus improving the practicality and safety of the planned path.

[0076] Based on the above, this embodiment specifically illustrates the conflict resolution strategy. Conflict resolution instructions include various methods.

[0077] Conflict resolution instructions include at least one of the following: discarding the conflict node, delaying the waiting for the conflict node, or replanning the local path.

[0078] When a spatiotemporal conflict is detected, the system does not simply stop operating, but provides multiple solutions: Discard conflicting nodes: Mark the point in the path search graph as "unreachable" to force the algorithm to find an alternative path.

[0079] Delayed waiting: Instructs the current tower crane to wait for a certain period of time at a safe point (such as stopping in the air or returning to the lifting point) to avoid time conflicts with other tower cranes.

[0080] Replan local paths: Without changing the global path, only fine-tune the local road segments where conflicts occur to reduce computational overhead.

[0081] The spatiotemporal conflict negotiation agent can choose a strategy based on the severity of the conflict. If it is a slight temporal overlap, it can choose "delay"; if it is a spatial deadlock, it can choose "discard node" to trigger a detour.

[0082] It provides a flexible conflict resolution mechanism, enabling the system to select the optimal solution when faced with conflicts of varying degrees, thus balancing efficiency and security.

[0083] Based on the above, this embodiment describes in detail the logic for determining spatiotemporal conflicts. This logic is based on the absolute value of the time difference.

[0084] Obtain the estimated occupancy time of nodes in the candidate hoisting path, as well as the occupancy time of the node in other tower cranes' planned optimal hoisting paths.

[0085] For each path node to be evaluated, the system needs to obtain two sets of time data: the time window in which the current tower crane plans to pass through the node, and the time window in which other tower cranes have committed to passing through the node.

[0086] Calculate the absolute value of the difference between the estimated time occupied and the time already occupied; when the absolute value is less than the preset time threshold, it is determined that there is a time and space occupancy conflict.

[0087] The system calculates the absolute value of the time difference between the two cranes. If this difference is less than a "preset time threshold", it means that the two tower cranes arrive at the point almost simultaneously, posing a risk of collision, and is therefore determined to be a conflict.

[0088] The preset time threshold is dynamically determined based on at least one of wind speed, the swing amplitude of the suspended object, and the site congestion.

[0089] To avoid misjudgments or missed judgments caused by the swinging of the suspended load in harsh environments, this time threshold is dynamic. For example, when the wind speed is high, the swing amplitude of the suspended load is large, and the safety time interval needs to be set to be larger.

[0090] The formula for calculating the time threshold can be ΔT = ΔT base *(1+k1*wind speed+k2*swing amplitude). By introducing a dynamic threshold, the collision avoidance mechanism becomes more robust.

[0091] Conflict detection based on dynamic time windows is more accurate than simple spatial overlap detection, effectively preventing "time collisions," and the dynamic threshold adapts to complex and ever-changing field conditions.

[0092] Building upon the above, this embodiment introduces a resource locking mechanism. The method also includes a path locking step.

[0093] When the trajectory planning agent publishes the optimal hoisting path to the shared situation blackboard, it locks the time occupancy commitment of each node in the optimal hoisting path.

[0094] Once a path is identified as "optimal" and published, the trajectory planning agent (or the spatiotemporal conflict negotiation agent) requests the shared situation blackboard to "lock" the "time occupancy commitments" of all nodes on that path. This means that at these time points, these spatial locations have been reserved.

[0095] The locking mechanism can employ distributed locks or version control. When other tower cranes under planning read the shared situational information board, they will treat these locked spatiotemporal nodes as "rigid obstacles" and will no longer attempt to pass through them.

[0096] This prevents oscillations and deadlocks in path planning, ensuring that once a path is successfully planned, it can be executed, thus guaranteeing the stability and determinism of the system.

[0097] Building upon the above, this embodiment improves the system's execution monitoring and anomaly handling closed loop. The system adds an execution guidance and anomaly monitoring intelligent agent.

[0098] The execution guidance agent obtains the optimal hoisting path from the shared situation blackboard, guides the current tower crane to perform hoisting according to the optimal hoisting path, and updates the execution progress to the shared situation blackboard.

[0099] The "execution guidance agent" is responsible for translating the digitized path into control commands to drive the tower crane motor. Simultaneously, it provides real-time feedback on the tower crane's actual position and progress to a shared situational awareness board for monitoring by other agents.

[0100] The anomaly monitoring agent obtains the real-time status of the tower crane and the execution progress of the optimal hoisting path from the shared situation blackboard. When the execution deviation is detected to exceed the preset threshold, a replanning trigger event is issued to the shared situation blackboard.

[0101] The "anomaly monitoring agent" acts as a safety guardian. It compares the tower crane's "planned progress" with its "actual progress." If an anomaly occurs (such as the tower crane stopping to avoid workers, or the speed slowing down due to excessive wind), causing the deviation to exceed a threshold, it will immediately issue a "replanning trigger event."

[0102] The trajectory planning agent and the spatiotemporal conflict negotiation agent respond to the replanning trigger event and replan the path.

[0103] Once a replanning trigger event is received, the original planning agent and negotiation agent will interrupt their current state and restart the path planning process based on the latest real-time data to generate a new obstacle avoidance path.

[0104] Execution deviations can be location deviations (e.g., the actual location deviates from the path node by more than 1 meter) or time deviations (e.g., arriving 10 seconds later than planned). The replanning trigger event can be a specific message queue signal.

[0105] A complete closed loop of "perception-planning-execution-monitoring" has been constructed, which makes the system extremely robust and able to cope with various emergencies on the construction site, ensuring the safe completion of hoisting tasks.

[0106] Example 2: This invention also provides an artificial intelligence-based tower crane trajectory planning device, which is mainly used to execute the artificial intelligence-based tower crane trajectory planning method provided in Embodiment 1 of this invention. The following is a detailed description of the artificial intelligence-based tower crane trajectory planning device provided in this invention.

[0107] An artificial intelligence-based tower crane trajectory planning device according to an embodiment of the present invention includes: a shared situation blackboard, and a trajectory planning unit and a spatiotemporal conflict negotiation unit respectively communicatively connected to the shared situation blackboard; wherein, The trajectory planning unit is used to acquire the hoisting task and construction site model, generate candidate hoisting paths and their estimated time occupancy, and publish them to the shared situation blackboard; The spatiotemporal conflict negotiation unit is used to read information from the shared situation blackboard, detect spatiotemporal occupancy conflicts between candidate hoisting paths and other tower cranes' planned optimal hoisting paths published on the shared situation blackboard, generate conflict resolution instructions, and write them to the shared situation blackboard. The trajectory planning unit is also used to adjust candidate hoisting paths according to conflict resolution instructions and publish the optimal hoisting path without conflict to the shared situation blackboard to guide the tower crane to execute the hoisting.

[0108] Optionally, an environmental perception agent is used to establish a construction site model and obtain the real-time status of each tower crane. The construction site model and real-time status are published to a shared situation blackboard, so that the trajectory planning unit can obtain the construction site model and real-time status from the shared situation blackboard and generate candidate hoisting paths in combination with the hoisting task.

[0109] Optionally, the environmental perception agent is also used to establish a three-dimensional mesh model of the construction site in a cylindrical coordinate system and generate a set of mesh nodes; wherein, the angular coordinates are divided equally with the central axis of the tower crane body as the vertical axis, and the angular coordinates are used as the first index, and the radius coordinates and height coordinates are used as the second index.

[0110] Optionally, the trajectory planning unit is also used to maintain a first list and a second list using a heuristic search algorithm; the hoisting starting point is placed in the first list, and if the first list is not empty, the node with the smallest total cost is taken as the current node and added to the second list, and the adjacent reachable nodes of the current node are calculated to obtain child nodes; after processing each child node accordingly, the nodes are repeatedly taken out until the hoisting endpoint is reached, and the candidate hoisting path is obtained by backtracking based on the parent node.

[0111] Optionally, the calculation of the total cost introduces a vertical priority weight function, a regional suppression weight function, and a horizontal guidance weight function, and the weights of each weight function are dynamically adjusted according to environmental factors; the regional suppression weight function is calculated based on the horizontal distance from the current node to the hoisting start point, the total horizontal distance from the hoisting start point to the hoisting end point, and the preset reference height; the horizontal guidance weight function is calculated based on the coordinate relationship between the current node and the hoisting end point.

[0112] Optionally, the conflict resolution instruction includes at least one of: discarding the conflict node, delaying the waiting for the conflict node, or replanning the local path.

[0113] Optionally, the spatiotemporal conflict negotiation unit is also used to obtain the estimated occupancy time of a node in a candidate hoisting path, as well as the occupancy time of that node in other tower cranes' planned optimal hoisting paths; calculate the absolute value of the difference between the estimated occupancy time and the occupancy time; and determine that there is a spatiotemporal occupancy conflict when the absolute value is less than a preset time threshold; wherein the preset time threshold is dynamically determined based on at least one of wind speed, the swing amplitude of the hoisted object, and site congestion.

[0114] Optionally, when publishing the optimal hoisting path to the shared situation blackboard, the time occupancy commitment of each node in the optimal hoisting path is locked as a constraint for other tower crane path planning.

[0115] Optionally, the execution guidance unit is used to obtain the optimal hoisting path from the shared situation blackboard, guide the current tower crane to perform hoisting according to the optimal hoisting path, and update the execution progress to the shared situation blackboard; the anomaly monitoring unit is used to obtain the real-time status of the tower crane and the execution progress of the optimal hoisting path from the shared situation blackboard, and when the execution deviation is detected to exceed a preset threshold, it issues a replanning trigger event to the shared situation blackboard; the trajectory planning unit and the spatiotemporal conflict negotiation unit are also used to respond to the replanning trigger event and replan the path.

[0116] The device provided in this embodiment of the invention has the same implementation principle and technical effect as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0117] like Figure 2 As shown in the embodiment of this application, an electronic device 600 includes a processor 601, a memory 602, and a bus. The memory 602 stores machine-readable instructions that can be executed by the processor 601. When the electronic device is running, the processor 601 communicates with the memory 602 via the bus, and the processor 601 executes the machine-readable instructions to perform the steps of the artificial intelligence-based tower crane trajectory planning method described above.

[0118] Specifically, the memory 602 and processor 601 mentioned above can be general-purpose memory and processor, without any specific limitations. When the processor 601 runs the computer program stored in the memory 602, it can execute the above-mentioned tower crane trajectory planning method based on artificial intelligence.

[0119] The processor 601 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of the processor 601 or by instructions in software form. The processor 601 may be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software module can reside in a mature storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory 602, and processor 601 reads the information from memory 602 and, in conjunction with its hardware, completes the steps of the above method.

[0120] Corresponding to the above-described artificial intelligence-based tower crane trajectory planning method, this application embodiment also provides a computer-readable storage medium storing machine-executable instructions. When the machine-executable instructions are invoked and executed by a processor, the machine-executable instructions cause the processor to perform the steps of the above-described artificial intelligence-based tower crane trajectory planning method.

[0121] The artificial intelligence-based tower crane trajectory planning device provided in this application embodiment can be specific hardware on the device or software or firmware installed on the device. The implementation principle and technical effects of the device provided in this application embodiment are the same as those in the foregoing method embodiments. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the foregoing method embodiments. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can all be referred to the corresponding processes in the above method embodiments, and will not be repeated here.

[0122] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0123] For example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0124] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0125] In addition, the functional units in the embodiments provided in this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0126] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause an electronic device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the artificial intelligence-based tower crane trajectory planning method described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0127] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. In addition, the terms "first", "second", "third", etc. are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0128] Finally, it should be noted that the above-described embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and not to limit them. The protection scope of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this application; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application. All should be covered within the protection scope of this application. Therefore, the protection scope of this application should be determined by the protection scope of the claims.

Claims

1. A tower crane trajectory planning method based on artificial intelligence, characterized in that, The method is implemented by a shared situational blackboard and a cluster of agents communicating with it, wherein the cluster of agents includes at least a trajectory planning agent and a spatiotemporal conflict negotiation agent, and the method includes: The trajectory planning agent acquires the hoisting task and construction site model, generates candidate hoisting paths and their estimated time occupancy, and publishes them to the shared situation blackboard; The spatiotemporal conflict negotiation agent reads the information in the shared situation blackboard, detects the spatiotemporal occupancy conflict between the candidate hoisting path and the other tower cranes' planned optimal hoisting paths published on the shared situation blackboard, generates a conflict resolution instruction, and writes it into the shared situation blackboard; The trajectory planning agent adjusts the candidate hoisting paths according to the conflict resolution instructions, and publishes the conflict-free optimal hoisting path to the shared situation blackboard to guide the tower crane to execute the hoisting path.

2. The method according to claim 1, characterized in that, The intelligent agent cluster also includes environmentally perceptive intelligent agents; The environmental perception agent establishes the construction site model and obtains the real-time status of each tower crane. It publishes the construction site model and the real-time status to the shared situation blackboard, so that the trajectory planning agent can obtain the construction site model and the real-time status from the shared situation blackboard and generate candidate hoisting paths in combination with the hoisting task.

3. The method according to claim 2, characterized in that, The environmental perception agent establishes a construction site model, including: A three-dimensional mesh model of the construction site in a cylindrical coordinate system is established, and a set of mesh nodes is generated. The angular coordinates are divided equally with the central axis of the tower crane as the vertical axis, and the angular coordinates are used as the first index, while the radius coordinates and height coordinates are used as the second index.

4. The method according to claim 1, characterized in that, The trajectory planning agent generates candidate hoisting paths, including: A heuristic search algorithm is used to maintain a first list and a second list. The hoisting starting point is placed in the first list. If the first list is not empty, the node with the minimum total cost is taken as the current node and added to the second list. The adjacent reachable nodes of the current node are calculated to obtain the child nodes. After processing each child node accordingly, the nodes are repeatedly extracted until the hoisting endpoint is reached, and candidate hoisting paths are obtained by backtracking based on the parent node.

5. The method according to claim 4, characterized in that, The calculation of the total cost incorporates a vertical priority weight function, a regional suppression weight function, and a horizontal guidance weight function, and the weights of each weight function are dynamically adjusted according to environmental factors. The regional suppression weight function is calculated based on the horizontal distance from the current node to the hoisting start point, the total horizontal distance from the hoisting start point to the hoisting end point, and the preset reference height. The horizontal guidance weight function is calculated based on the coordinate relationship between the current node and the hoisting endpoint.

6. The method according to claim 1, characterized in that, The conflict resolution instructions include at least one of the following: discarding the conflict node, delaying the waiting for the conflict node, or replanning the local path.

7. The method according to claim 1, characterized in that, The spatiotemporal conflict negotiation agent detects spatiotemporal occupancy conflicts, including: Obtain the estimated occupancy time of nodes in the candidate hoisting path, and the occupancy time of the node in other tower cranes' planned optimal hoisting paths; Calculate the absolute value of the difference between the estimated occupancy time and the occupancy time; When the absolute value is less than a preset time threshold, it is determined that there is a time-space occupancy conflict; The preset time threshold is dynamically determined based on at least one of wind speed, the swing amplitude of the suspended object, and the site congestion.

8. The method according to claim 1, characterized in that, Also includes: When the trajectory planning agent publishes the optimal hoisting path to the shared situation blackboard, it locks the time occupancy commitment of each node in the optimal hoisting path as a constraint for other tower crane path planning.

9. The method according to claim 1, characterized in that, The agent cluster also includes an executive guiding agent; The execution guidance agent obtains the optimal hoisting path from the shared situation blackboard, guides the current tower crane to perform hoisting according to the optimal hoisting path, and updates the execution progress to the shared situation blackboard. The intelligent agent cluster also includes an anomaly monitoring intelligent agent; The anomaly monitoring agent obtains the real-time status of the tower crane and the execution progress of the optimal hoisting path from the shared situation blackboard. When the execution deviation is detected to exceed a preset threshold, a replanning trigger event is issued to the shared situation blackboard. The trajectory planning agent and the spatiotemporal conflict negotiation agent respond to the replanning trigger event and replan the path.

10. A tower crane trajectory planning device based on artificial intelligence, characterized in that, include: A shared situational awareness blackboard, and a trajectory planning unit and a spatiotemporal conflict negotiation unit respectively communicatively connected to the shared situational awareness blackboard; wherein... The trajectory planning unit is used to acquire the hoisting task and construction site model, generate candidate hoisting paths and their estimated time occupancy, and publish them to the shared situation blackboard; The spatiotemporal conflict negotiation unit is used to read the information in the shared situation blackboard, detect the spatiotemporal occupancy conflict between the candidate hoisting path and the other tower cranes' planned optimal hoisting paths published on the shared situation blackboard, generate a conflict resolution instruction, and write it into the shared situation blackboard. The trajectory planning unit is also used to adjust the candidate hoisting paths according to the conflict resolution instructions, and publish the conflict-free optimal hoisting path to the shared situation blackboard to guide the tower crane to execute.

Images