A tower crane hoisting whole-process dynamic mapping and perception covering method and system

Abstract

The application relates to the technical field of intelligent control of engineering machinery, and provides a tower crane hoisting whole-process dynamic mapping and perception coverage method and system. The method comprises the following steps: installing two rotating radars on rotatable supports of a tower crane; acquiring a current hoisting operation stage of the tower crane in real time; switching the shooting angles of the two radar modules according to the hoisting operation stage, and controlling the two radar modules to acquire continuous point cloud data of the corresponding operation stage; collecting tower crane operation parameters, combining the continuous point cloud data acquired by the two radar modules to generate a continuous hook trajectory; identifying dynamic obstacles and static backgrounds contained in a target monitoring space from the continuous point cloud data, and constructing a construction environment point cloud model; performing space-time registration on the construction environment point cloud model and a physical coordinate system, and establishing a global dynamic map; fusing the continuous hook trajectory and the global dynamic map to construct a dynamic perception area coverage map, and realizing real-time perception of the space environment of the tower crane hoisting operation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent control of engineering machinery, and more particularly, to a tower crane hoisting whole-process dynamic mapping and perception coverage method and system. BACKGROUND

[0002] As a core heavy lifting equipment in the field of construction engineering, the tower crane is widely used in heavy vertical and horizontal transfer operations in scenarios such as high-rise buildings and bridge engineering. Due to the hoisting process involving high-altitude operation, dynamic transfer of heavy objects, and multi-device cooperation, there are multiple dynamic and static targets such as tower crane jibs, hooks, surrounding buildings, temporary construction machinery, and personnel in the operation space, and the operation safety and efficiency directly determine the project progress and personnel and property safety.

[0003] In the traditional technology, the visual observation of the on-site command personnel (such as the signalman) and the tower crane driver is mainly relied on, the information such as the hook position and the obstacle distance is transmitted through the intercom, the safety of the operation area is artificially judged, and the operation is guided. The field of vision of the operation personnel is easily blocked by the tower crane jib and the building to form a blind area, and is affected by physiological fatigue and subjective judgment errors, and the spatial environment of the hoisting operation process (such as lifting, transfer, and landing) cannot be continuously perceived. In addition, in some existing systems, cameras are arranged on the tower crane jib, the cab, and the operation site to collect two-dimensional image information of the hoisting area, which is transmitted to the display screen of the driving end to assist the driver in observing the hook state and the near-distance environment. However, the driver and the dispatching end can only rely on their own experience to judge the local relationship with the aid of the two-dimensional image information, and it is difficult to form a cognition of the overall spatial relationship, which may lead to decision-making errors in path planning, area blocking, and the like, and cause safety accidents.

[0004] In summary, the traditional tower crane operation mainly relies on artificial observation and experience judgment, and there are visual blind areas and lack of spatial judgment ability, which leads to safety hazards and low efficiency. SUMMARY

[0005] In this context, the embodiments of the present application aim to provide a tower crane hoisting whole-process dynamic mapping and perception coverage method and system, which can realize real-time perception of the spatial environment of the tower crane hoisting operation, and solve the technical problems of visual blind areas and lack of spatial judgment ability in the traditional tower crane.

[0006] In a first aspect of the embodiments of the present application, a tower crane hoisting whole-process dynamic mapping and perception coverage method is provided, comprising:

[0007] Two rotating radars are installed on the rotatable supports of the tower crane respectively to form a double-radar module, and the double-radar module is in communication connection with the control module;

[0008] The control module is used for acquiring a current hoisting operation stage of the tower crane in real time, the hoisting operation stage includes a lifting stage, a rotating stage, an amplitude changing stage and a descending stage; a shooting angle of the dual radar module is switched according to the hoisting operation stage, a key perception angle is used in the lifting stage and the descending stage, a layered perception angle is used in the rotating stage and the amplitude changing stage, and the dual radar module is synchronously controlled to acquire continuous point cloud data corresponding to the operation stage;

[0009] The tower crane operation parameters are collected, and the continuous point cloud data acquired by the dual radar module is combined to calculate a three-dimensional coordinate, a motion speed and a swing angle of a hook in real time, and generate a continuous hook trajectory;

[0010] A target perception algorithm is used to identify dynamic obstacles and static backgrounds contained in a target monitoring space from the continuous point cloud data, and construct a construction environment point cloud model;

[0011] The construction environment point cloud model is time-space registered with a physical coordinate system with the tower crane base as an origin, a global dynamic map containing the target monitoring space is established by using a time-space graph convolution network of a Gauss surface element mapping;

[0012] The continuous hook trajectory and the global dynamic map are fused, a dynamic perception area coverage map is constructed by using a trajectory differential algorithm and a dynamic ontology evolution algorithm, and the dynamic perception area coverage map is used to display the target monitoring space and the construction environment features in the target monitoring space which continuously change with each operation stage in the hoisting operation process.

[0013] In a second aspect of the embodiments of the present application, a tower crane hoisting whole-process dynamic mapping and perception coverage system is provided, comprising:

[0014] A construction unit is configured to install a plurality of rotating radars on rotatable supports of a tower crane to form a dual radar module, and the dual radar module is in communication connection with a control module;

[0015] A switching unit is configured to acquire a current hoisting operation stage of the tower crane in real time by using the control module, the hoisting operation stage includes a lifting stage, a rotating stage, an amplitude changing stage and a descending stage; a shooting angle of the dual radar module is switched according to the hoisting operation stage, a key perception angle is used in the lifting stage and the descending stage, a layered perception angle is used in the rotating stage and the amplitude changing stage, and the dual radar module is synchronously controlled to acquire continuous point cloud data corresponding to the operation stage;

[0016] A trajectory tracking unit is configured to collect tower crane operation parameters, combine the continuous point cloud data acquired by the dual radar module, calculate a three-dimensional coordinate, a motion speed and a swing angle of a hook in real time, and generate a continuous hook trajectory;

[0017] The point cloud modeling unit is configured to identify dynamic obstacles and static backgrounds contained in the target monitoring space from the continuous point cloud data by using a target perception algorithm, and construct a construction environment point cloud model.

[0018] The space-time registration unit is configured to take the tower base as the origin of a physical coordinate system, perform space-time registration between the construction environment point cloud model and the physical coordinate system, and establish a global dynamic map containing the target monitoring space by using a space-time graph convolution network based on a Gaussian surface element mapping method.

[0019] The coverage map generation unit is configured to fuse the continuous hook trajectory and the global dynamic map, and construct a dynamic perception area coverage map by using a trajectory differential algorithm and a dynamic ontology evolution algorithm. The dynamic perception area coverage map is used to display the target monitoring space and the construction environment features in the target monitoring space, which continuously change with each operation stage in the hoisting operation process.

[0020] In a third aspect of the embodiments of the present application, a terminal device is provided, which comprises at least one processor, a memory and an input-output unit. The memory is configured to store a computer program, and the processor is configured to invoke the computer program stored in the memory to execute the tower crane hoisting whole-process dynamic mapping and perception coverage method according to any one of the first aspect.

[0021] In a fourth aspect of the embodiments of the present application, a computer readable storage medium is provided, which comprises instructions that, when executed on a computer, cause the computer to execute the tower crane hoisting whole-process dynamic mapping and perception coverage method according to any one of the first aspect.

[0022] In a fifth aspect of the embodiments of the present application, a computer program product is provided, which comprises a computer program that, when executed by a processor, implements the tower crane hoisting whole-process dynamic mapping and perception coverage method according to any one of the first aspect.

[0023] According to an embodiment of the present application, a tower crane hoisting whole-process dynamic mapping and perception coverage method and system. First, two rotating radars are installed on the rotatable support of the tower crane to form a dual-radar module, and the dual-radar module is in communication connection with the control module. Then, the control module is used to obtain the current hoisting operation stage of the tower crane in real time, and the hoisting operation stage includes the lifting stage, the rotating stage, the amplitude changing stage and the descending stage. Next, the shooting angle of the dual-radar module is switched according to the hoisting operation stage, the key perception angle is used in the lifting stage and the descending stage, and the layered perception angle is used in the rotating stage and the amplitude changing stage, and the dual-radar module is controlled synchronously to obtain continuous point cloud data corresponding to the operation stage. The tower crane operation parameters are collected, the continuous point cloud data obtained by the dual-radar module is combined, the three-dimensional coordinates, the motion speed and the swing angle of the hook are calculated in real time, and the continuous hook trajectory is generated. A target perception algorithm is used to identify dynamic obstacles and static backgrounds contained in the target monitoring space from the continuous point cloud data, and a construction environment point cloud model is constructed. The construction environment point cloud model is time and space registered with the physical coordinate system with the tower crane base as the origin, a time and space convolution network of Gaussian surface element mapping is used to establish a global dynamic map containing the target monitoring space. Finally, the continuous hook trajectory and the global dynamic map are fused, a dynamic perception area coverage map is constructed through a trajectory differential algorithm and a dynamic ontology evolution algorithm, and the dynamic perception area coverage map is used to display the target monitoring space and the construction environment features in the target monitoring space which continuously change with each operation stage in the hoisting operation process. Through the cooperative operation of the dual-radar module and the dynamic mapping technology, the dynamic perception area coverage map is dynamically constructed, the whole-process coverage of the hoisting operation process is realized, the perception ability of the tower crane in the hoisting operation process is greatly improved, and the hoisting intelligent level and the operation safety redundancy capability of the construction site are further improved. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 A flowchart of a tower crane hoisting whole-process dynamic mapping and perception coverage method shown in the present application;

[0025] Figure 2 A schematic diagram of a tower crane hoisting whole-process dynamic mapping and perception coverage method shown in the present application;

[0026] Figure 3 A structural schematic diagram of a tower crane hoisting whole-process dynamic mapping and perception coverage system shown in the present application;

[0027] Figure 4 A structural schematic diagram of a medium of an embodiment of the present application is shown in the present application. DETAILED DESCRIPTION

[0028] Reference will now be made to Figure 1 ,Figure 1 A flowchart of a tower crane hoisting whole-process dynamic mapping and perception coverage method provided by an embodiment of the present application is shown. It should be noted that the embodiments of the present application can be applied to a tower crane operation scene.

[0029] To solve the above at least one technical problem, the embodiments of the present application provide a tower crane hoisting whole-process dynamic mapping and perception coverage method and system. Specifically, for the combination of a double-radar module and a rotatable support, by monitoring the hoisting operation stage in real time and switching the double-radar view angle of the corresponding stage, the double-radar module can collect continuous point cloud data that continuously and without dead angle covers the operation range in the whole process of the tower crane operation. This dynamic adjustment strategy effectively avoids the technical problem of visual blind area in the traditional single-radar or fixed-view-angle multi-radar scheme. Secondly, combined with the fusion calculation of the tower crane operation parameters and the continuous point cloud data, the real-time calculation accuracy of the hook three-dimensional coordinates is further improved. The trajectory generation algorithm effectively suppresses noise interference through trajectory differential processing, ensuring the accuracy of the motion state prediction. The target perception algorithm intelligently classifies dynamic obstacles and static backgrounds, and cooperates with the Gaussian face element mapping technology of the spatio-temporal graph convolution network to realize the dynamic update of the construction environment point cloud model. Through the point cloud fusion strategy of error weight accumulation, the calculation complexity is reduced while the detail features are preserved, improving the calculation efficiency and helping to realize real-time perception. The spatio-temporal registration mechanism with the tower crane base as the origin effectively solves the spatial reference unification problem of multi-source data, so that the constructed global dynamic map can accurately reflect the environmental evolution of the hoisting operation area. The finally generated dynamic perception area coverage map realizes dynamic early warning and safety boundary adaptive adjustment of the dangerous area by fusing the hook trajectory and the three-dimensional environment model, which can greatly shorten the collision warning response time and improve the protection efficiency of the anti-collision system.

[0030] Figure 1 The flow of a tower crane hoisting whole-process dynamic mapping and perception coverage method provided by an embodiment of the present application is shown, which includes:

[0031] In step S101, two rotating radars are respectively installed on the rotatable supports of the tower crane to form a double-radar module, and the double-radar module is in communication connection with the control module.

[0032] In the embodiments of the present application, two groups of rotatable radar sensors are installed at key positions of the tower crane, and the radar angle is dynamically adjusted through the rotatable support (such as a motorized gimbal) to construct a double-radar cooperative perception hardware system (such as shown in the figure) with angle adjustment capability. Figure 2 The two groups of radars are respectively defined as a first rotating radar and a second rotating radar (non-fixed primary and secondary, dynamically switchable), and are in bidirectional communication with the control module (such as an industrial-grade PLC, an edge computing controller, integrated with or connected to multiple types of sensors) to realize data interaction and control instruction transmission.

[0033] Further optionally, in the dual-radar module, the two rotating radars can be synchronously scanned at a frequency of 10 Hz. For example, based on the technical parameters of the TF-ALS-LIDAR-01 (detection distance ≥ 150 meters, accuracy ≤ 3 cm, angle resolution ≤ 0.05°), a rotating support meeting the industrial protection standard (IP67) is selected, the support material needs to have corrosion resistance, and the rotating mechanism adopts a high-precision harmonic reducer to ensure a turning accuracy of 0.1 degrees. The dual radars are distributed and installed on both sides of the support and are rigidly connected to the tower crane arm frame through a customized structure. Further optionally, the installation angle needs to be optimized according to the three-dimensional modeling of the tower crane operation radius to ensure the overlapping range to eliminate the blind area.

[0034] For example, the communication connection adopts a CAN bus and an Ethernet dual-redundancy architecture, each radar is configured with an independent CAN channel to transmit point cloud data, and a gigabit-level data interaction between the control module and the radar is realized through an industrial switch. The control module selects a 32-bit ARM processor, runs a real-time operating system (RTOS), configures the radar parameters (such as scanning frequency, sampling rate) through the MODBUS-TCP protocol, and establishes a data buffer queue to process the point cloud data stream of more than 200,000 points per second. Dynamic calibration is required during installation. In the static state, a standard reflector (reflectivity 99%) is used to calibrate the angle deviation, and during dynamic testing, the tower crane inclination sensor data is synchronously collected to establish a real-time conversion model of the rotating support angle and the radar coordinate system, and to compensate for the measurement drift caused by mechanical vibration. Finally, a three-dimensional scanning system with a spatial resolution of 0.1 degrees is formed, which can generate more than 100,000 effective data points in a single scan, meeting the high-precision positioning requirements of the tower crane in the long-distance operation radius.

[0035] In step S102, the control module is used to obtain the current hoisting operation stage of the tower crane in real time. In the embodiments of the present application, the hoisting operation stage includes the lifting stage, the slewing stage, the luffing stage and the descending stage.

[0036] Specifically, step S102 realizes accurate identification and dynamic switching of the tower crane hoisting operation stage through multi-source sensor data fusion and kinematic model analysis. The control module establishes a stage feature discrimination model based on tower crane operation parameters, laser radar point cloud features, and environmental perception data. In the hoisting stage, the weight sensor mutation signal at the instant of hook-off-ground is monitored in real time. When the detected load weight reaches 90% of the rated value and the hoisting encoder speed exceeds 0.3 m / s, it is determined that the hoisting stage is entered. In this case, the dual radar module switches to the key perception mode, performs 180-degree sector scanning centered on the hook, generates more than 5000 three-dimensional point cloud data per second, combines the inclination data of the IMU module, and solves the spatial attitude of the hook through the Kalman filtering algorithm, with an error controlled within ±2 cm. For example, in the building material hoisting scene, when the hook carries 5 tons of steel reinforcement and is 0.5 meters off the ground, the sling stress analysis algorithm is automatically triggered, and whether the hoisted object has deviated is judged by the change of point cloud density.

[0037] After entering the slewing stage or the luffing stage, the control module establishes a dynamic model of the angular velocity and angular acceleration based on the rotation mechanism motor current fluctuation characteristics and the tower top displacement data. When the detected slewing support torque exceeds the set threshold and the tower displacement gradient slope absolute value is less than 0.05 degrees per second, it is determined that the slewing stage or the luffing stage is entered. At this time, the dual radar module adopts a layered scanning strategy, the upper radar performs environmental feature extraction with a resolution of 0.02 degrees, and the lower radar performs obstacle detection with a resolution of 0.1 degrees, and the scanning frequency is synchronously increased to 30 Hz. For example, when there are temporary materials piled up within a radius of 15 meters of the tower crane, obstacles with a diameter of more than 50 cm are identified through point cloud registration algorithm, triggering the path correction instruction to reduce the rotation speed to 65% of the rated value, and at the same time, the point cloud distortion caused by the rotation of the mechanical arm is eliminated through the vibration compensation algorithm.

[0038] The descent stage recognition is based on the hook height change rate and the ground feature matching degree. When the hoisting encoder detects that the descent speed is lower than 0.15 m / s and the ground feature matching similarity in the laser radar point cloud is more than 85%, it is determined that the descent stage is entered. At this time, the high-precision positioning mode is started, the RTK differential positioning data and the visual SLAM algorithm are fused, and the three-dimensional coordinate positioning accuracy of the hook is improved to ±3 cm. For example, in the precast component installation operation, when the hook approaches the target position by 20 cm, the precast plate edge profile is identified through the point cloud segmentation algorithm, the spatial conflict detection is performed in combination with the BIM model, and the drop point position is automatically adjusted to the safe area. The whole stage switching process is realized through the state machine model to achieve seamless transition, with an average state recognition delay of less than 50 ms, ensuring the continuity of perception coverage and the timeliness of control response in each link of the hoisting operation.

[0039] Step S103, switching the shooting angle of the dual-radar module according to the hoisting operation stage, synchronously controlling the dual-radar module to obtain continuous point cloud data corresponding to the operation stage. In the embodiment of the application, a key perception angle is used in the lifting stage and the descending stage, and a layered perception angle is used in the slewing stage and the luffing stage.

[0040] It can be understood that step S103 realizes the reinforcement of environmental perception in the whole process of hoisting operation through the cooperative optimization of the dynamic angle switching mechanism and the point cloud data acquisition strategy. In the lifting stage, the dual-radar module constructs a fan-shaped key perception area with the hook as the center, and the upper radar adopts a narrow beam scanning mode, which compresses the scanning angle to ±15 degrees, while improving the point cloud density to 200 measurement points per square meter, accurately capturing the spatial relative position changes of the hook and the hoisted object. For example, in the precast floor hoisting scene, when the hook carrying an 8-ton component is 0.3 meters away from the ground, the forward radar performs continuous scanning with a resolution of 0.03 degrees, generating millimeter-level point cloud data of the hoisted object edge profile, and combining with the real-time correction of the coordinate system by the inclination sensor, ensuring that when the hoisted object swing amplitude exceeds the safety threshold, the sound and light alarm is triggered. At this time, the rear radar synchronously performs low-frequency environmental scanning to capture the static obstacle features within a radius of 5 meters with a resolution of 0.1 degrees, forming a dual-channel data complement.

[0041] After entering the slewing stage or the luffing stage, the dual-radar module switches to a layered scanning mode, the upper radar performs 360-degree environmental modeling with a super-resolution of 0.01 degrees, generating more than 150,000 points of high-precision point cloud per second, and the lower radar adopts a step-by-step scanning strategy to focus on covering the area within a radius of 10 meters of the tower crane slewing radius with an interval of 0.5 degrees. For example, in the group tower cross-operation scene, when the tower crane rotates horizontally at a speed of 1.2r / min, the upper radar identifies the metal reflection features of the adjacent tower crane boom through multi-return fusion technology, constructs a three-dimensional point cloud model containing material properties, and at the same time, the lower radar detects the scaffold components on the ground, distinguishes the hard obstacles with a diameter of more than 30 cm through point cloud clustering algorithm, and triggers the path planning system to offset the rotation trajectory by 23 degrees to avoid collision. At this time, the control module establishes a time and space synchronization compensation model according to the scanning frequency difference of the radars, eliminating the point cloud distortion caused by the change of the slewing angular velocity.

[0042] The wide field of view configuration of the dual radar module is restored in the lowering phase, and the forward radar reconstructs the three-dimensional space of the hook landing area with a coverage range of ±30 degrees, and the backward radar synchronously collects ground elevation change data. For example, in a steel structure installation operation, when the hook is lowered to 50 cm from the target support, the forward radar analyzes the spatial posture of the steel structure connecting plate through phase modulation technology to generate a point cloud image containing 2000 feature points, which is matched with the BIM model at a millimeter level, while the backward radar detects an area with a ground subsidence of more than 5 mm, generates a local terrain compensation model through interpolation algorithm, and ensures that the final positioning error of the hook is controlled within ±3 cm. During the whole process, the switching delay of the dual radar view angle is controlled within 80 ms, and the point cloud data overlap rate is always maintained at a preset ratio to form a continuous and complete environment perception sequence.

[0043] Further optionally, in the key perception view angle, the first rotating radar is used to scan the three-dimensional space of the vertical lifting, lowering, and horizontal transfer of the hook, and the second rotating radar is used to scan the micro-environment space of the vertical path and landing area of the hook. In the hierarchical perception view angle, the first rotating radar is used to scan the medium and long distance space in front of the hoist arm, the side of the hoist arm, and the medium distance space where the hook is located. The second rotating radar is used to scan the near distance space where the hoisting core area is located. Based on this assumption, in step S103, the shooting view angle of the dual radar module is switched according to the hoisting operation stage, including: when the hoisting operation stage is in the lifting stage or the lowering stage, or when the hook lifting height is detected to be lower than the preset lifting height threshold and the horizontal movement range is less than the preset horizontal range threshold, the key perception view angle is switched to. When the hoisting operation stage is in the rotation stage or the amplitude changing stage, or when the hook is detected to be in long distance horizontal transfer or large range rotation operation, the hierarchical perception view angle is switched to.

[0044] Specifically, step S103 realizes fine environment perception of the whole hoisting operation process by dynamically adjusting the spatial scanning strategy of the dual radar module. In the key perception view angle, the first rotating radar constructs a three-dimensional scanning field with the hook as the core, and its scanning range covers the three-dimensional space of the vertical motion trajectory and horizontal transfer path of the hook, and it captures the attitude change of the hoisted object in real time through high frequency point cloud collection. For example, in a precast floor hoisting operation, when the hook carries an 8-ton component for vertical lifting, the first radar densely scans the area within a radius of 0.5 meters around the hoisted object with a resolution of 0.03 degrees to generate a three-dimensional point cloud image containing 2000 feature points, which is combined with the inclination sensor data to correct the coordinate system in real time, and ensures that the hoisted object triggers an audible and visual alarm when the swing amplitude exceeds the safety threshold. The second rotating radar focuses on the micro-environment space below the vertical path of the hook, and uses a narrow beam scanning mode to model the ground landing area at a sub-centimeter level, and when the hook approaches the target support, the radar can identify the protruding features of the ground embedded parts with an accuracy of ±2 cm, providing data support for precise hook landing.

[0045] After entering the hierarchical perception perspective, the first rotating radar switches to a medium and long-range environmental scanning mode, which expands its scanning range to 20 meters in front of the jib and 15 meters to the side. Through multi-echo fusion technology, a three-dimensional topological model of obstacles around the jib is constructed. For example, in a group tower intersection operation scenario, when the tower crane rotates horizontally at a speed of 1.2r / min, the first radar detects the metal reflection characteristics of the adjacent tower crane boom through Doppler effect and identifies the standard section profile of the tower crane 50 meters away. Combining the spatiotemporal graph convolution network, a dynamic obstacle trajectory prediction model is established. The second rotating radar focuses on the medium distance space where the hook is located and uses a step-by-step scanning strategy to cover the hoisting core area. When the hook moves horizontally for a long distance, the radar scans the area within a radius of 10 meters at an interval of 0.1 degrees, with each layer spaced 0.5 meters apart, forming a spatial feature dataset containing 5000 points, effectively identifying potential obstacles such as temporarily stacked scaffolding components.

[0046] The stage switching mechanism realizes intelligent decision-making through multi-source information fusion. When the hook lifting height is detected to be lower than 3 meters and the horizontal movement range is less than 5 meters, it automatically switches to the key perception perspective. At this time, the scanning frequency of the dual radar is synchronized to 40Hz, ensuring the dynamic tracking accuracy of the hoisted object during low-altitude operation. For example, in a steel structure installation operation, when the hook is lowered to 50cm from the target support, the second radar under the key perception perspective analyzes the spatial pose of the steel structure connecting plate through phase modulation technology, generating a point cloud chart containing 2000 feature points, and performing millimeter-level matching with the BIM model. When the tower crane performs large-scale slewing operation, through the fusion of slewing bearing torque sensor and encoder data, it is determined to enter the hierarchical perception mode. At this time, the scanning frequency of the first radar is increased to 60Hz, and the obstacles within 50 meters in front of the jib are dynamically modeled. Combined with the Kalman filtering algorithm, the motion trajectory is predicted, and the path correction instruction is triggered in advance to offset the rotation trajectory by a preset number of degrees to avoid collision.

[0047] As an optional embodiment, in step S103, a hybrid target perception network is adopted to identify the dynamic obstacles and the static background contained in the target monitoring space based on the adjusted continuous point cloud map, and to extract the geometric features of the dynamic obstacles, the geometric features of the static background, and the real-time motion parameters of the dynamic obstacles from the adjusted continuous point cloud map, including: processing the adjusted continuous point cloud map through a semantic perception feature enhancement module, adopting an improved BEBLID descriptor and a FasterNet backbone network, fusing multi-scale point cloud density information to construct an adaptive feature pyramid, enhancing the static structure boundary features through an edge perception loss function, and outputting an enhanced point cloud feature map. Based on the foregoing enhanced point cloud feature map, an ISS algorithm and a normal vector constraint are adopted to calculate the normal vector and the curvature of the point cloud, an ISS feature score is generated through neighborhood covariance analysis to filter static feature points, a DBSCAN clustering algorithm is combined to remove discrete noise points, a multi-view geometric constraint is introduced to remove dynamic interference points, and the curvature threshold is dynamically adjusted according to the weather information to extract the geometric features of the static background, and the static background geometric features are synchronously removed to obtain dynamic candidate point clouds. Based on the foregoing dynamic candidate point clouds, a hybrid architecture of YOLOv5s and PointNet is adopted to construct a dynamic target detection network, and the class, bounding box and point cloud clustering mask of the dynamic obstacle are output through a two-dimensional and three-dimensional joint detection head to extract the geometric features of the dynamic obstacle. The motion optical flow is calculated based on the Farneback optical flow method, and the initial speed is obtained in combination with the point cloud data, and the position, speed and acceleration parameters of the dynamic obstacle are optimized through the Kalman filtering algorithm to serve as the real-time motion parameters of the dynamic obstacle.

[0048] Step S103 realizes accurate separation of dynamic obstacles and static background and extraction of motion parameters through multi-level feature enhancement and hybrid architecture design. The semantic perception feature enhancement module first preprocesses the adjusted continuous point cloud map, adopts an improved BEBLID descriptor to construct a local feature matching framework, and enhances the local geometric description ability of the point cloud through the cosine similarity calculation between feature vectors. This module introduces a FasterNet backbone network to replace the traditional PointNet++ structure, uses its partial convolution (PConv) operation to reduce redundant calculation, and improves the calculation efficiency by more than 40% while maintaining the feature expression ability. The adaptive feature pyramid dynamically adjusts the contribution weight of different levels of features through a multi-scale point cloud density weighted fusion strategy, and still retains the key structure features in complex occlusion scenes. For example, in the tower crane group operation scene, when the adjacent tower crane booms form a dense metal reflection area, this module effectively suppresses background noise interference and accurately extracts the edge contour features of the obstacles within a range of 0.5 meters around the hook through the density-aware feature fusion strategy.

[0049] In the static background feature extraction stage, the ISS algorithm combines with the normal vector constraint to screen the feature points, and a three-dimensional space feature scoring model is constructed by calculating the curvature variance and the eigenvalue of the neighborhood covariance matrix of each point. When the curvature standard deviation is lower than the set threshold and the consistency of the principal curvature direction is more than 80%, it is determined as a static structure feature point. The DBSCAN clustering algorithm removes noise based on the density peak value, and the neighborhood radius parameter is dynamically adjusted according to the weather information. In rainy and foggy weather, the radius is expanded to 0.3 meters to compensate for the sparseness of the point cloud, and in sunny weather, it is reduced to 0.1 meters to improve the segmentation accuracy. The multi-view geometry constraint projects the point cloud to multiple virtual camera perspectives, and uses the perspective invariance feature to remove dynamic interference points, such as temporary objects such as protective nets in strong wind weather, which are automatically filtered when the multi-perspective projection feature difference value exceeds the threshold.

[0050] The dynamic target detection network adopts a hybrid architecture of YOLOv5s and PointNet, with YOLOv5s responsible for quickly generating two-dimensional candidate boxes in the front end, and PointNet extracting three-dimensional point cloud features in the candidate region in the back end. The two-dimensional detection head generates the initial target position through the anchor box mechanism, and the three-dimensional detection head optimizes the size of the bounding box combined with the normal vector and curvature information, and finally outputs the detection results including class probability, spatial coordinates and point cloud mask. In the precast floor hoisting scene, when the hoisting object passes through the scaffold area, the network detects the steel pipe protrusion with a diameter of 50 cm through the multi-scale feature pyramid, and accurately segments the obstacle geometry combined with the point cloud mask, improving the recognition accuracy. The motion parameter estimation module uses the Farneback optical flow method to calculate the motion vector field between adjacent frames of point clouds, and establishes the point cloud displacement relationship through dense optical flow matching. After the initial velocity field is optimized by Kalman filtering, it can effectively suppress the influence of sensor noise. For example, in the long-distance horizontal transfer operation of the tower crane, when the hook carries an 8-ton component at a speed of 0.5 m / s, the real-time calculation of the three-dimensional motion trajectory of the component can control the position prediction error within ±8 cm, and the velocity estimation fluctuation amplitude is less than 0.03 m / s, providing reliable data support for dynamic obstacle avoidance.

[0051] As an optional embodiment, in step S103, the initial velocity is obtained by calculating the motion optical flow based on the Farneback optical flow method and combining the point cloud data, and the position, velocity and acceleration parameters of the dynamic obstacle are obtained by optimizing through the Kalman filtering algorithm, which are used as the real-time motion parameters of the dynamic obstacle. In addition, the real-time motion parameters and geometric features of the dynamic obstacle can be used to extract the radial velocity of the dynamic obstacle by combining the Doppler feature coding, and a spatio-temporal correlation graph containing time stamp, coordinate and velocity vector is constructed, and the edge weight is the motion consistency probability. According to the speed of the obstacle, the detection sensitivity is dynamically adjusted, and when the dynamic obstacle approaches the extracted static background, the joint feature optimization process is triggered to realize the interactive optimization of the features of the static background and the dynamic obstacle.

[0052] Step S103 realizes high-precision analysis and adaptive perception optimization of the dynamic obstacle motion state through multi-modal feature fusion and spatio-temporal modeling technology. After obtaining the position, velocity and acceleration parameters of the dynamic obstacle, its Doppler feature code is further extracted, and the radial velocity component of the obstacle relative to the radar is accurately calculated by analyzing the Doppler frequency shift characteristics of the millimeter wave radar echo signal. For example, in the scene of crossing operation of adjacent tower cranes, when the adjacent tower cranes move towards each other at a speed of 2.5 m / s, the Doppler feature coding module calculates the relative radial velocity of 4.8 m / s by analyzing the phase change of the echo signal. After weighting fusion of this data and the point cloud motion parameters, a spatio-temporal correlation graph node containing timestamp, three-dimensional coordinates and velocity vector is formed. The edge weight between nodes is calculated by the difference between the predicted value by Kalman filter and the actual observation, and the motion consistency probability is calculated. When the difference exceeds 15%, the edge weight attenuation mechanism is triggered to ensure the timeliness of the dynamic topology structure.

[0053] For the scene where the obstacle approaches the static background, an adaptive sensitivity adjustment algorithm is used to dynamically adjust the detection parameters by analyzing the angle between the obstacle velocity vector and the normal vector of the static background. When the angle is less than 30° and the relative speed exceeds 1 m / s, the joint feature optimization process is triggered. Specifically, first, the point cloud features of the dynamic obstacle and the ISS features of the static background are matched across modalities, and the feature similarity weight is calculated using a graph attention network. Second, the motion features of adjacent nodes are aggregated by a spatio-temporal graph convolution network to establish a spatio-temporal correlation model containing velocity trend prediction. Finally, an adversarial training strategy is used to optimize the feature fusion weight. In the case of steel structure workshop hoisting, this mechanism improves the accuracy of the hook and steel beam collision warning. For example, when the prefabricated wallboard is hoisted to 0.8 meters away from the concrete column, the angle between the wallboard edge point cloud and the column surface normal vector is suddenly changed, triggering the multi-scale feature enhancement module. Through the expansion convolution network, the contact surface geometric features with a precision of 0.1 meters are extracted, combined with the BIM model to verify the collision risk, and finally a dynamic safety plan containing the avoidance path is generated.

[0054] Step S104, collect the tower crane operation parameters, combine the continuous point cloud data obtained by the dual radar module, and calculate the three-dimensional coordinates, motion speed and swing angle of the hook in real time to generate a continuous hook trajectory.

[0055] It can be understood that step S104 realizes real-time tracking of the hook movement state and trajectory generation through multi-source data fusion and kinematic modeling technology. First, the tower crane operation parameters are collected, including the basic movement parameters such as the hoisting mechanism speed, slewing bearing angular velocity, trolley walking encoder pulse signal, and continuous point cloud data generated by the dual radar module scanning is obtained at the same time. The dual radar adopts a heterogeneous layout scheme, the forward radar covers the vertical movement trajectory of the hook with a narrow beam, and the backward radar monitors the horizontal swing range with a wide field of view. In the data fusion stage, the tower crane operation parameters and radar point cloud data are synchronized to millisecond level accuracy through a time stamp alignment algorithm, eliminating the error caused by mechanical transmission delay. For example, in the precast floor hoisting scene, when the hook carries an 8-ton component for vertical lifting, the forward radar analyzes the relative position offset between the hook center point and the sling through the phase modulation technology, combines the linear velocity data output by the tower crane hoisting encoder, and uses a motion compensation algorithm to eliminate the point cloud distortion caused by the sling swing, and calculates the three-dimensional coordinates of the hook in real time. At this time, the backward radar synchronously scans the 0.5-meter radius area around the hook, analyzes the swing amplitude of the hook through the point cloud density change, and triggers the dynamic filtering mechanism when the swing angular velocity exceeds the preset threshold, and uses the particle filtering algorithm to optimize the swing angle estimate.

[0056] The trajectory generation module constructs the motion state equation based on the Kalman filter framework, recursively calculates the three-dimensional coordinates of the hook, the velocity vector, and the angular displacement parameters. In complex working conditions, the tower crane structure parameters are introduced to establish the motion constraint model, for example, in the group tower crossing operation, the geometric constraint conditions are constructed by the tower crane jib length and the slewing center coordinates, and the hook movement trajectory is corrected in real time. For example, when the tower crane performs large arm slewing operation, the backward radar detects that the horizontal offset of the hook caused by inertia exceeds the safety threshold, immediately calls the historical trajectory data and the current angular velocity parameter, calculates the actual swing center point coordinates of the hook through the kinematic inverse algorithm, combines the tower crane structure characteristics in the BIM model, and generates a dynamic obstacle avoidance trajectory containing a safety buffer zone.

[0057] In step S105, a target perception algorithm is used to identify dynamic obstacles and static backgrounds contained in the target monitoring space from the continuous point cloud data, and a construction environment point cloud model is constructed.

[0058] It can be understood that step S105 realizes dynamic obstacle identification and spatial structure modeling of the construction environment through multi-modal perception fusion and three-dimensional modeling technology. The target perception algorithm in this scenario specifically refers to a multi-layer perception architecture that fuses dual radar point clouds and visual features, and its core is to realize dynamic obstacle semantic segmentation and static background feature extraction through an improved deep learning model. The construction environment point cloud model is a three-dimensional spatial topology constructed based on time series point cloud data, which contains dynamic obstacle motion trajectories, static facility geometric shapes, and digital representations of environmental physical properties.

[0059] For example, the point cloud data collected by the dual-radar module is first aligned in perspective and optimized for fusion. For example, in the scene of tower crane group cross operation, the forward radar scans the 5-meter range around the hook with an angular resolution of 0.1°, and the backward radar covers a radius of 20 meters in wide-angle mode. Through voxel grid downsampling and feature-level fusion algorithm, the multipath effect interference is eliminated. When detecting rain and fog weather, the real-time precipitation intensity data is obtained by calling the weather API, and the fusion weight is dynamically adjusted based on the point cloud density decay model. For example, the forward radar weight is increased to 0.7 in rainy weather to compensate for the decrease in visibility, and decreased to 0.3 in sunny weather to highlight visual features.

[0060] The hybrid target perception network adopts the heterogeneous architecture of YOLOv7 and PointNet++, and the front-end YOLOv7 detects the candidate obstacle area through the multi-scale feature pyramid, and the back-end PointNet++ extracts the normal vector and curvature features of the candidate area. In the precast component hoisting scene, when the hook passes through the scaffold area, the network focuses on the edge features of the metal pipe through the attention mechanism, and combines the point cloud density information after weather compensation to accurately identify the steel pipe protrusion with a diameter of 30 cm, with a boundary box positioning error controlled within ±5 cm. The dynamic obstacle motion parameters are optimized by the optical flow method and Kalman filter, for example, in strong wind environment, when the protective net swings are detected, the optical flow displacement vector and radar speed measurement results are weighted and fused, so that the speed estimation fluctuation amplitude is reduced to 0.05 m / s.

[0061] The geometric feature extraction stage uses an improved GICP algorithm for feature registration, and describes the surface structure characteristics of the point cloud by constructing a local covariance matrix. In the steel structure factory, when the hook approaches the steel beam, the point cloud cluster within 0.2 meters of the contact surface is extracted, the principal curvature and normal vector distribution features are calculated, and a geometric descriptor containing 128-dimensional feature vectors is generated. Static background modeling uses the TSDF algorithm to construct an implicit surface of the environment, and dynamically adjusts the voxel resolution, such as using 0.1-meter voxel size in the key area (such as within 5 meters of the tower crane rotation radius), and expanding to 0.5 meters in the peripheral area, so that the BIM model matching accuracy reaches ±3 cm. Finally, the construction environment point cloud model is dynamically updated by the space-time association graph, the node contains obstacle geometric features, motion state and environment semantic label, and the edge weight reflects the space-time consistency probability, providing a high-precision environment semantic map for intelligent obstacle avoidance and path planning.

[0062] Further optionally, in step S105, the continuous point cloud data collected by the dual-radar module is fused and optimized according to the shooting angle and the respective corresponding collection angle, to obtain a continuous point cloud map of the target monitoring space; weather information of a geographical area where the tower crane is located is acquired in real time, the point cloud weight under different weather conditions is adjusted through a point cloud density attenuation model; a hybrid target perception network is used to identify dynamic obstacles and static backgrounds contained in the target monitoring space based on the adjusted continuous point cloud map, and to extract geometric features of the dynamic obstacles, geometric features of the static backgrounds, and real-time motion parameters of the dynamic obstacles from the adjusted continuous point cloud map; an improved generalized iterative closest point (GICP) algorithm and a dynamic fusion type truncated signed distance function (TSDF) algorithm are used to construct the extracted geometric features and real-time motion parameters into the construction environment point cloud model.

[0063] Specifically, in step S105, in the dual-radar point cloud fusion stage, a heterogeneous sensor space-time alignment algorithm is used to map the point cloud data of the forward narrow-beam radar and the backward wide-angle radar to a unified coordinate system. The forward radar captures high-precision motion trajectories in the vertical direction through the narrow-beam characteristic, and the backward radar covers the horizontal space with a wide angle, and the multi-path effect interference is eliminated through voxel grid downsampling and feature-level fusion strategy. A weather self-adaptive adjustment mechanism is introduced, the sensor weight is dynamically adjusted based on meteorological data, the point cloud quality loss under adverse weather is compensated through an exponential attenuation model, and the perception robustness in complex environments is ensured.

[0064] The hybrid target perception network uses a heterogeneous architecture of YOLOv7 and PointNet++ to realize feature complementation, the front-end YOLOv7 detects candidate obstacle regions through a multi-scale feature pyramid, and the back-end PointNet++ extracts normal vector and curvature features of the candidate regions. The network dynamically focuses on key geometric features through a channel attention mechanism, combines the point cloud density information after weather compensation, realizes accurate identification and classification of dynamic obstacles. The motion parameters of the dynamic obstacles are optimized through a combination of optical flow method and Kalman filter, the environmental interference is eliminated by using a motion compensation algorithm, and a dynamic correlation model of the motion state and the spatial position of the obstacles is established.

[0065] In the geometric feature extraction stage, an improved generalized iterative closest point algorithm is used for feature registration, a local covariance matrix is constructed to describe the surface structure characteristics of the point cloud, and principal component analysis is used to extract geometric feature descriptors. The dynamic truncated signed distance function algorithm is used for static background modeling, an adaptive voxel resolution is used to construct an implicit surface of the environment, high-resolution voxels are used in key areas to ensure detailed features, and low-resolution voxels are used in peripheral areas to balance the calculation efficiency. The environment model is dynamically updated through a space-time correlation graph, the nodes contain obstacle geometric features, motion states and environment semantic labels, the edge weights reflect the space-time consistency probability, and three-dimensional semantic modeling and dynamic evolution of the construction environment are realized.

[0066] The finally constructed construction environment point cloud model integrates multi-source perception data, and realizes digital representation of the geometric shape, motion state and environment structure of the obstacle through feature fusion and semantic labeling. The model adopts a hierarchical storage structure, and dynamic obstacle information and static background data are stored in independent layers, supporting real-time query and dynamic update. The model provides a high-precision environment semantic map for the intelligent obstacle avoidance system, and supports real-time decision-making of path planning and collision warning.

[0067] In step S106, the construction environment point cloud model is spatiotemporally registered with the physical coordinate system with the tower base as the origin of the physical coordinate system. A spatiotemporal graph convolution network based on Gaussian cell mapping is used to establish a global dynamic map containing the target monitoring space.

[0068] As an optional embodiment, in step S106, a three-dimensional physical coordinate system is established with the geometric center of the tower base as the coordinate origin, the installation position coordinates and attitude parameters of the dual-radar module are obtained, and the original coordinates of each point in the construction environment point cloud model are mapped to the physical coordinate system through a coordinate conversion matrix. Based on the timestamps of the point cloud data collected by each rotating radar, a time interpolation algorithm is used to time-synchronize the point cloud data at different times, eliminating the spatiotemporal deviation caused by radar data acquisition delay, so that the construction environment point cloud model and the physical coordinate system are spatiotemporally aligned. The aligned point cloud model is discretized into a Gaussian cell grid using the Gaussian cell mapping method. Each Gaussian cell grid stores the mean value of the static background geometric features, the probability distribution of the dynamic obstacle motion parameters and the point cloud density confidence of the corresponding spatial region. A spatiotemporal graph convolution network is constructed, the Gaussian cell grid is taken as a graph node, the stored geometric feature mean value, motion parameter probability distribution and point cloud density confidence are uniformly represented through a Gaussian distribution model, and the node features of each graph node include the geometric parameters, timestamps and dynamic attribute labels of the cell. The edge weight is set based on the spatial distance of the cell and the time correlation. The spatiotemporal graph convolution network is used to fuse and operate the Gaussian cell node features in continuous time sequence, the motion parameter distribution of the cell where the dynamic obstacle is located is updated in real time, and the geometric features of the static background cell are retained. A global dynamic map containing time dimension evolution information is generated, and the global dynamic map dynamically updates the environment features of the target monitoring space with the progress of the hoisting operation.

[0069] Specifically, step S106 realizes global perception and dynamic evolution modeling of the construction environment through spatio-temporal registration and dynamic modeling technology. A three-dimensional physical coordinate system is constructed with the geometric center of the tower crane base as the origin, and a coordinate conversion relationship is established through the installation parameters of the dual radar module to map each data point in the point cloud model to a unified coordinate system. This process uses a quaternion rotation and translation compensation algorithm to eliminate sensor installation deviations, for example, when the dual radar is installed at a 90-degree angle, a rotation matrix is calculated using the Rodrigues formula, and spatial alignment is completed in combination with the Euler angle parameters of the base coordinate system. To address the time asynchronous problem of radar data, a cubic spline interpolation algorithm is used to time-resample point clouds at different times. When the hook moves at a speed of 2.5 m / s, the time interpolation accuracy can reach milliseconds, effectively eliminating the spatio-temporal misalignment caused by differences in sensor refresh rates.

[0070] The Gaussian cell mapping module discretizes the registered point cloud space into regular grids, and each grid represents the environmental characteristics through a Gaussian distribution function. In dynamic obstacle dense areas, the grid resolution is increased to 0.1 meters to capture motion details, and in static areas, it is expanded to 0.5 meters to balance computational efficiency. Each cell stores the velocity vector probability distribution of dynamic obstacles, the geometric feature mean of static backgrounds, and the point cloud density confidence, for example, within a 5-meter radius of the tower crane's turning radius, the cell geometric feature mean includes normal vector and curvature information, and the dynamic parameter distribution is updated through Kalman filtering. The spatio-temporal graph convolution network takes the cells as graph nodes, the node features integrate Gaussian distribution parameters and timestamp information, and the edge weights are calculated through spatial Euclidean distance and time difference, forming a dynamic spatio-temporal correlation structure.

[0071] The network adopts a hierarchical convolution architecture, the spatial layer aggregates neighborhood node features through graph attention mechanism, and the temporal layer captures temporal dependency using dilated causal convolution. In the group tower crossing operation scene, when the adjacent tower crane booms enter the monitoring area, the network enhances the relevance of nodes in the interaction area through dynamic edge weight adjustment mechanism, and updates the obstacle motion parameter distribution in real time. Static background cells use a feature freezing strategy, only allowing the geometric feature mean to adjust slowly with environmental changes, while the motion parameter distribution of dynamic cells is updated every 0.5 seconds, ensuring real-time and stability. The final generated global dynamic map realizes the deep fusion of obstacle trajectory prediction and static environment semantics through the recursive calculation of the spatio-temporal graph convolution network, providing an environment model containing spatio-temporal evolution characteristics for tower crane intelligent obstacle avoidance.

[0072] Step S107, fuse the continuous hook trajectory and the global dynamic map, construct a dynamic perception area coverage map through trajectory differentiation algorithm and dynamic ontology evolution algorithm.

[0073] In the embodiment of the present application, the dynamic perception area coverage map is used to show the target monitoring space which continuously changes with each operation stage in the hoisting operation process and the construction environment features in the target monitoring space.

[0074] It can be understood that step S107 constructs a perception coverage map reflecting the spatio-temporal evolution characteristics of the construction process by fusing the dynamic trajectory of the hoisting operation and the three-dimensional environment model. The continuous hook trajectory is generated by Kalman filtering and spline interpolation algorithm from the hook position data collected by the double radar module, containing a sequence of three-dimensional coordinates marked with time stamps, reflecting the continuous motion path of the hook in the hoisting operation. The global dynamic map integrates point cloud density, obstacle motion parameters and environment semantic information through the spatio-temporal graph convolution network, forming a spatio-temporal correlation model containing static facility geometric features and dynamic obstacle probability distribution.

[0075] In the embodiment of the present application, the trajectory differential algorithm adopts a linear tracking differentiator to analyze the kinematics of the hook trajectory, and converts discrete position data into a continuous and derivable motion parameter sequence through the construction of a state observer. The trajectory differential algorithm suppresses high-frequency noise interference through pole placement strategy, extracts derivative features such as velocity and acceleration while maintaining the smoothness of the trajectory, for example, it can still accurately output an acceleration estimate value of the order of 0.1 m / s² in the hook emergency stop working condition. The dynamic ontology evolution algorithm is based on the change propagation mechanism, identifies the entity attributes that need to be updated in the environment model through ontology difference calculation, generates a change path with the minimum influence range using a heuristic search strategy, and ensures the synchronous adjustment of the associated semantic labels when updating the position of the dynamic obstacle.

[0076] The dynamic perception area coverage map realizes multi-dimensional information fusion through a rasterized spatio-temporal coordinate system, and each grid cell stores dynamic obstacle contact probability, static facility damage level and environmental parameter distribution features. In the group tower operation scene, the hook trajectory and the global dynamic map are spatio-temporally aligned, the intersection features of the trajectory and the obstacle area are extracted using convolution operation, and the coverage heat map is generated through dynamic threshold segmentation. The dynamic perception area coverage map adopts a four-dimensional spatio-temporal coding structure, including spatial coordinates, time stamps and environmental state vectors, and supports real-time query of the environmental parameter change trajectory of a specific area in the historical period. Thus, the environment perception delay is reduced, the obstacle recognition accuracy is improved, and the intelligent decision basis with spatio-temporal consistency is provided for the tower group collaborative operation.

[0077] As an optional embodiment, in step S107, the continuous hook trajectory is spatiotemporally aligned with the global dynamic map, so that the hook trajectory coordinates are consistent with the physical coordinate system of the global dynamic map, and the hook movement timing is synchronously associated with the real-time movement parameters of the dynamic obstacles. The trajectory differential algorithm is used to perform differential operation on the continuous hook trajectory, to extract the speed change rate, acceleration change rate and swing angle change trend of the hook at different operation stages, and to generate hook movement characteristic parameters. The dynamic ontology evolution algorithm is used to construct an ontology model containing the hoisting operation stage, hook movement characteristic parameters, static background geometric features and real-time movement parameters of dynamic obstacles, and to dynamically update the associated weights of various elements in the ontology model according to the operation stage switching. The dynamic perception boundary is generated by the ontology model, taking the current position of the hook as the center, combining the hook movement characteristic parameters and the environmental features in the global dynamic map. The dynamic perception boundary is adjusted in real time with the movement of the hook, and covers the target monitoring space within the preset safety range of the hook. The dynamic perception boundary is superimposed on the static background and dynamic obstacles in the global dynamic map, to generate the dynamic perception area coverage map. The dynamic perception area coverage map displays the spatial position relationship between the hook trajectory and the environmental features, the early warning information of the dynamic obstacles entering the dynamic perception boundary, and the perception area changes in different operation stages.

[0078] Specifically, step S107 realizes global perception and intelligent early warning of hoisting operation through spatiotemporal alignment and dynamic modeling technology. First, the continuous hook trajectory is spatiotemporally registered with the global dynamic map, and the sensor installation deviation and data asynchronous problem are eliminated through the coordinate conversion matrix, to ensure that the three-dimensional coordinates of the hook are strictly corresponding to the physical coordinate system. In this process, a time interpolation algorithm is used to resample the hook trajectory at sub-second level, to synchronously associate the real-time position and movement parameters of the dynamic obstacles, and to establish a spatiotemporal consistency association model.

[0079] The trajectory differential algorithm performs kinematic analysis on the hook trajectory based on an adaptive tracking differentiator, and converts the discrete position sequence into a continuous and derivable movement parameter sequence through a state observer. This algorithm uses a pole placement strategy to suppress high-frequency noise, and can accurately extract the speed change rate and acceleration change rate under the conditions of hook sudden stop or turning, for example, when the hook is lifted at an acceleration of 3 m / s², the algorithm can output an estimated jerk value of the order of 0.05 m / s³ in real time. The dynamic ontology evolution algorithm identifies the entity attributes that need to be updated in the environmental model through ontology difference calculation, generates a change path with the smallest influence range using a heuristic search strategy, and ensures synchronous adjustment of the associated semantic labels when the position of the dynamic obstacle is updated, for example, when a new incoming tower crane is detected, the algorithm automatically enhances the associated weight of its movement parameters in the environmental model.

[0080] The dynamic perception area coverage map realizes multi-dimensional information fusion through rasterizing the space-time coordinate system, and each grid unit stores dynamic obstacle contact probability, static facility damage level and environmental parameter distribution characteristics. Taking the current position of the hook as the center, the motion characteristic parameters extracted by the trajectory differential algorithm and the environmental characteristics in the global dynamic map are combined to generate a dynamic perception boundary through the ontology model. The boundary uses a probability density function to describe the safety area around the hook. When a dynamic obstacle is detected entering the preset threshold area, a multi-level warning mechanism is triggered, and the obstacle entry probability distribution is displayed in the form of a heat map on the coverage map. In the group tower intersection operation scene, the coverage map identifies the changes in the perception area in different operation stages through a dynamic threshold segmentation algorithm, for example, the perception radius is automatically expanded to 5 meters during the hoisting and positioning stage, and is contracted to 2 meters during the slewing process to improve the calculation efficiency.

[0081] Therefore, the environment perception delay can be reduced to within 150 ms, the obstacle collision warning accuracy is improved, and the dynamic perception boundary adjustment response time is shortened to 50 ms. Through the space-time alignment and ontology evolution technology, the semantic characteristics and dynamic evolution law of the hoisting operation environment can be reflected in real time, providing intelligent decision support for tower group collaborative operation with space-time consistency, and significantly reducing the collision risk and operation error under complex working conditions.

[0082] As an optional embodiment, after step S107, the real-time motion parameters of the hook and the real-time motion parameters of the dynamic obstacles can also be called to calculate the Euclidean distance between the hook and each obstacle and the relative motion trend. A preset collision probability model is used to quantify the risk value, and the safety risk level is marked in the dynamic perception area coverage map. The high-risk area uses dense markers, the medium-risk area uses sparse markers, and the low-risk area uses safety markers. The collision probability model is constructed based on a distance threshold and a relative speed coupling algorithm. The obstacle classification features are extracted from the construction environment point cloud model, including the labeled geometric size of static obstacles, fixed coordinates, real-time position of dynamic obstacles, and motion trajectory vector. In the dynamic perception area coverage map, the differential contour symbols are superimposed and marked. Static obstacles are marked with solid figures, dynamic obstacles are marked with dynamic figures, the position coordinates of dynamic obstacles are updated synchronously, and the motion trajectory curve of dynamic obstacles connected by continuous multiple position coordinate points is generated.

[0083] Specifically, after step S107, the depth of construction safety warning is enhanced by dynamic risk quantification and visualization technology. The motion parameters of the hook and dynamic obstacles are obtained in real time, the Euclidean distance and relative motion vector between the two are calculated based on the time-space aligned coordinate system, the distance and speed coupling algorithm is used to construct the collision probability model, the risk factor and spatial geometric relationship are mapped through a nonlinear function, for example, when the obstacle enters the swing radius of the hook, the model automatically increases the collision probability weight. The risk level division adopts a multi-level threshold strategy, combined with historical accident data and engineering specifications to set different color coding rules, and through the difference in grid density and filling mode, the risk intensity is directly reflected, so that the operator can quickly identify high-risk areas.

[0084] The obstacle classification feature extraction module separates the static and dynamic entity attributes from the point cloud model. The static obstacle extracts the geometric contour and coordinate information through the region growing algorithm, and the dynamic obstacle generates the motion vector using the trajectory prediction algorithm. In the visualization stage, different types of obstacles are rendered differently using semiotic coding rules. Static obstacles are characterized by fixed color blocks to represent their geometric shapes, while dynamic obstacles show their motion state through continuous trajectory segments, combined with motion direction arrows and speed vector lines to enhance dynamic perception. Risk level labeling uses dynamic penetration technology. When the obstacle risk parameters change, smooth transition is achieved through color gradient and transparency adjustment to avoid misjudgment caused by visual mutations.

[0085] Thus, the response time of risk warning is shortened, and the collision false alarm rate is reduced. Through multi-dimensional information fusion display, the operator's perception efficiency of complex working conditions is improved, especially in the group tower intersection operation scene, the trajectory prediction accuracy of dynamic obstacles is improved. The differential labeling strategy improves the visualization accuracy of the spatial distribution of safety risks, effectively supports real-time decision-making and path optimization for hoisting operations, and provides an intelligent safety protection system with adaptive ability for large construction projects.

[0086] Exemplarily, the dynamic risk visualization technology is used to realize the intuitive presentation of the construction safety situation. Based on the real-time calculation of the spatial relationship and motion state of the hook and obstacles, a multi-level risk quantification model is used to dynamically partition and label the environment. In the dense red grid area, the system uses high-frequency point cloud sampling and motion trajectory prediction algorithm to update the obstacle contact probability every second, and combines Kalman filter to optimize the relative speed estimation accuracy, ensuring the real-time warning capability of high-risk areas. The yellow sparse grid area uses an adaptive sampling strategy. When the obstacle enters this area, the point cloud density is automatically increased to 0.2 meters resolution, and the motion prediction line in the next 3 seconds is generated by the trajectory extrapolation algorithm, providing buffer decision time for the operator. The green grid area maintains the baseline features of the environment through spatial interpolation algorithm, and uses historical data caching technology to reduce the sensor data acquisition frequency and reduce system load.

[0087] Thus, the risk area identification response time is compressed, and the false positive rate is controlled low. The differentiated labeling strategy improves the accuracy of the operator's identification of risk areas, especially in the group tower intersection operation scene, and the dynamic grid adjustment mechanism can automatically expand the coverage range of the high-risk area to 8 meters in radius. Through the cooperative optimization of grid density and color coding, the GPU resource consumption is reduced while maintaining a rendering speed of 50 frames per second. The deep integration of the three-dimensional dynamic perception coverage map and the driving end AR display module enables the hoisting operation personnel to quickly locate the risk source through the visual focus, shortens the operation instruction response delay, and thus improves the operation safety and efficiency under complex conditions.

[0088] As an optional embodiment, after step S107, the rated turning radius of the tower crane, the lifting height range, and the obstacle distribution in the dynamic perception area coverage map are obtained, and a fast random expansion tree algorithm is used to calculate the passable path of the hook from the current position to the target landing point. In the dynamic perception area coverage map, the optimal passable path is marked with a solid line, the potential passable path is marked with a dashed line, and the unreachable area blocked by obstacles or beyond the performance of the tower crane is marked with another type of solid line. The position coordinates of the path key nodes are marked in each passable path, and the speed limit threshold of the path key nodes is marked based on the hook motion parameters. The dynamic perception area coverage map is updated synchronously at a preset frequency to automatically refresh the safety risk level, obstacle distribution, and each passable path. The dynamic perception area coverage map is transmitted to the dispatching end and the driving end, the dispatching end interface supports the superposition of the dynamic perception area coverage map of multiple tower cranes, and the superimposed dynamic perception area coverage map is projected into a three-dimensional display space for assisting users in global path planning and high-risk area blocking decision-making. The driving end interface shows the user the safety risk level of the key area around the hook based on the dynamic perception area coverage map, enlarges the display of the risk label and the passable path, and provides operation assistance prompts to the user.

[0089] Specifically, after step S107, global optimization and safety control of tower crane operation are achieved through intelligent path planning and dynamic coordination technology. Based on the rated turning radius and lifting height parameters of the tower crane, a motion constraint boundary is constructed, combined with the obstacle distribution characteristics in the dynamic perception area coverage map, and an improved fast random expansion tree algorithm is used for path search. This algorithm introduces a tower crane kinematic constraint model based on the traditional RRT, balances the path optimality and computational efficiency through a dynamic step adjustment strategy, and automatically reduces the sampling area in narrow channel scenarios to improve obstacle avoidance accuracy. The path generation module divides the passable path into two categories: optimal path and potential path. The optimal path is evaluated by a cost function that considers path length, curvature, and safety margin, while the potential path retains suboptimal but valuable emergency routes. Both are marked on the coverage map with different line types.

[0090] The path key nodes are identified by a kinematic parameter extraction algorithm, including amplitude turning points, rotation limit angles, and lifting speed mutation points. Each node stores position coordinates and speed limit thresholds. When the hook approaches a node, the system predicts the speed change curve within the next 0.5 seconds based on the current motion state, dynamically adjusts the threshold range to match the actual working conditions. The dynamic perception area coverage map uses an incremental update mechanism, refreshing the obstacle position, risk level, and path state every 200 milliseconds. The difference algorithm only updates the changed area to reduce the computational load. The dispatch interface integrates a multi-tower crane coverage map projection function, which maps each tower crane's working area to a unified spatial coordinate system using three-dimensional coordinate transformation. The administrator can adjust the transparency and overlay the layers to analyze global conflict hotspots, such as identifying spatial interference areas when multiple tower cranes rotate simultaneously and automatically generating blocking suggestions.

[0091] The driving interface uses augmented reality technology to integrate dynamic perception data. The viewpoint following algorithm keeps the hook position always in the center of the field of view. The key area risk labeling uses dynamic scaling and color gradient enhancement to improve display effectiveness. The operation assistance prompt module generates semantic instructions based on the path key nodes and the current hook motion state, such as triggering voice alarms and displaying recommended deceleration curves when approaching speed limit thresholds. Technically, this scheme compresses the path planning response time to within 300 milliseconds, reduces the conflict rate of multi-tower crane collaborative work, and reduces the operation error rate of the driving end. Through the deep integration of three-dimensional situational awareness and intelligent path guidance, the system improves the safety and efficiency of operations in complex construction scenarios, providing a reliable intelligent control solution for tower crane groups in super high-rise building and bridge engineering.

[0092] For example, the driving end triggers sound and light alarms, and simultaneously displays obstacle avoidance suggestions in the interface pop-up window. The dispatch end automatically pushes high-risk area coordinates to the on-site safety terminal and sends temporary restriction instructions to the tower crane control cabinet, pausing the hook's movement in the dangerous direction until the risk level decreases to medium or low risk.

[0093] In this way, the multi-level alarm linkage and radar perception enhancement technology realize active safety protection for tower crane operations. When the driving end triggers sound and light alarms, the system uses a hierarchical warning strategy to convey risk level information through different color LED light groups and buzzer frequency combinations, such as red flashing LEDs with high-frequency buzzer sounds to indicate emergency avoidance. The interface pop-up window uses dynamic arrows to mark the obstacle direction and recommended turning angle. When the dispatch end receives high-risk signals, it pushes coordinate data to the on-site safety terminal through industrial communication protocols, triggers sound and light warning columns and electronic fence equipment, and sends PWM modulation instructions to the tower crane control cabinet to dynamically adjust the motor output power through the frequency converter to reduce the hook's movement speed until the risk parameters return to the safety threshold.

[0094] Further optionally, the radar parameter setting of the hierarchical perception visual angle is that the first rotating radar scanning frequency is not less than 10 Hz, the second rotating radar scanning frequency is not less than 15 Hz, and the third rotating radar scanning frequency is not less than 20 Hz and the scanning accuracy reaches millimeter level. For example, each rotating radar adopts an equal-interval microstrip antenna array, and an integrated clutter suppression module is used to reduce the interference of the tower crane structure itself on the detection signal. Specifically, the hierarchical perception radar system adopts a multi-dimensional detection architecture to improve the environmental perception accuracy. The hierarchical rotating radar is deployed at different positions of the tower crane jib. The first layer radar adopts a 10 Hz scanning frequency to cover a 360° horizontal field of view. The clutter interference reflected by the metal structure of the tower crane is suppressed by a digital beam forming technology. The second layer radar focuses on the key operation area with a 15 Hz frequency. The FMCW modulation technology is used to achieve a 0.5 meter level ranging accuracy. The third layer millimeter wave radar analyzes the motion vector of the obstacle by the Doppler effect and predicts its future trajectory by the Kalman filtering algorithm. The core area radar system adopts a 20 Hz high frequency scanning and a ±90° pitch angle coverage. The composite control of electronic scanning and mechanical scanning is realized by a microstrip antenna array. In a high temperature and dust environment, the temperature compensation module based on a PT100 thermistor corrects the radar phase shift in real time to ensure that the 0.1 millimeter level ranging accuracy is not affected by environmental fluctuations.

[0095] Further optionally, the radar parameter setting of the core area precise perception visual angle is that the first rotating radar scanning frequency is not less than 20 Hz and the three-dimensional space coverage pitch angle range reaches ±90°, and the second rotating radar scanning frequency is not less than 25 Hz and the scanning accuracy reaches 0.1 millimeter level. Further, each rotating radar is integrated with a temperature compensation module to adapt to the high temperature and dust environment in the construction scene. Specifically, the hierarchical radar system cooperates with multi-band detection to improve the obstacle recognition accuracy, especially in the group tower cross operation scene. The millimeter wave radar can penetrate the tower crane jib to detect the rear obstacle. The temperature compensation module makes the ranging stability error of the radar in the environment of minus 20℃ to 60℃ less than 2 millimeters. The dynamic trajectory prediction algorithm predicts the potential collision risk 2.5 seconds in advance. Through the deep coupling of alarm linkage and motion control, the system can automatically limit the speed of the hook to below the safety threshold in dangerous working conditions, so as to reduce the accident rate of the group tower cooperative operation and provide a reliable safety protection system for the super high-rise building and port hoisting operation.

[0096] In the embodiments of the present application, through the cooperative operation of the dual radar module and the dynamic mapping technology, a dynamic perception area coverage map is dynamically constructed, the whole process coverage of the hoisting operation process is realized, the perception ability of the tower crane in the hoisting operation process is greatly improved, and the hoisting intelligent level and the operation safety redundancy capability of the construction site are further improved.

[0097] After introducing the method of the example embodiments of the present application, next, reference is made to Figure 3A tower crane hoisting whole process dynamic mapping and perception coverage system of an exemplary embodiment of the present application is described, which comprises: a construction unit configured to install a plurality of rotating radars on a rotatable support of a tower crane to form a dual-radar module, and the dual-radar module is in communication connection with a control module; a switching unit configured to acquire a current hoisting operation stage of the tower crane in real time through the control module, the hoisting operation stage comprising a lifting stage, a rotating stage, an amplitude changing stage and a descending stage; switch the shooting angle of the dual-radar module according to the hoisting operation stage, use a key perception angle in the lifting stage and the descending stage, and use a layered perception angle in the rotating stage and the amplitude changing stage; synchronously control the dual-radar module to acquire continuous point cloud data of the corresponding operation stage; a trajectory tracking unit configured to acquire tower crane operation parameters, combine the continuous point cloud data acquired by the dual-radar module, calculate a hook three-dimensional coordinate, a motion speed and a swing angle in real time, and generate a continuous hook trajectory; a point cloud modeling unit configured to use a target perception algorithm to identify dynamic obstacles and static backgrounds contained in a target monitoring space from the continuous point cloud data, and construct a construction environment point cloud model; a space-time registration unit configured to take a tower crane base as an origin of a physical coordinate system, perform space-time registration on the construction environment point cloud model and the physical coordinate system, and use a space-time graph convolution network of a Gaussian surface element mapping to establish a global dynamic map containing the target monitoring space; and a coverage map generation unit configured to fuse the continuous hook trajectory and the global dynamic map, construct a dynamic perception area coverage map through a trajectory differential algorithm and a dynamic ontology evolution algorithm, and the dynamic perception area coverage map is used to display the target monitoring space and construction environment features in the target monitoring space which continuously change with each operation stage in the hoisting operation process. The above system can implement each step described in the above method embodiment, and the specific implementation of each step is not repeated here.

[0098] After introducing the method and system of the exemplary embodiment of the present application, next, a terminal device of an exemplary embodiment of the present application is described, which can implement each step described in the above method embodiment, and the specific implementation of each step is not repeated here.

[0099] After introducing the method, system and terminal device of the exemplary embodiment of the present application, next, with reference to Figure 4 A computer readable storage medium of an exemplary embodiment of the present application is described, please refer to Figure 4 The computer readable storage medium shown is an optical disc 30, and a computer program (i.e. program product) is stored on the optical disc 30, and the computer program will implement each step described in the above method embodiment when being run by a processor. The specific implementation of each step is not repeated here.

[0100] It should be noted that the above-described embodiments are merely specific implementations of the present application, and are used to explain the technical solutions of the present application, but not to limit the same. The protection scope of the present application is not limited to this. Although the present application is described in detail with reference to the foregoing embodiments, it should be understood by those skilled in the art that any person skilled in the art can make modifications or easy changes to the technical solutions recorded in the foregoing embodiments, or make equivalent replacements to some of the technical features within the technical scope disclosed by the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for dynamic mapping and sensing coverage of the entire tower crane hoisting process, characterized in that, The method includes: Two rotating radars are installed on the rotatable support of the tower crane to form a dual radar module, and the dual radar modules are respectively connected to the control module. The control module obtains the current hoisting operation stage of the tower crane in real time. The hoisting operation stage includes the hoisting stage, slewing stage, luffing stage and descent stage. The shooting angle of the dual radar module is switched according to the hoisting operation stage. The key perception angle is used in the hoisting and descent stages, and the layered perception angle is used in the slewing and luffing stages. The dual radar module is controlled to acquire continuous point cloud data of the corresponding operation stage in a synchronous manner. The tower crane's operating parameters are collected, and combined with the continuous point cloud data obtained by the dual radar modules, the three-dimensional coordinates, movement speed, and swing angle of the hook are calculated in real time to generate a continuous hook trajectory. A target perception algorithm is used to identify dynamic obstacles and static backgrounds contained in the target monitoring space from the continuous point cloud data, and a point cloud model of the construction environment is constructed. Using the tower crane base as the origin of the physical coordinate system, the point cloud model of the construction environment is spatiotemporally registered with the physical coordinate system, and a global dynamic map containing the target monitoring space is established using a spatiotemporal graph convolutional network with Gaussian surface mapping. By integrating the continuous hook trajectory with the global dynamic map, a dynamic sensing area coverage map is constructed using a trajectory differentiation algorithm and a dynamic ontology evolution algorithm. The dynamic sensing area coverage map is used to display the target monitoring space that changes continuously with each operation stage during the hoisting operation, as well as the construction environment characteristics within the target monitoring space.

2. The method for dynamic mapping and sensing coverage of the entire tower crane hoisting process according to claim 1, characterized in that, In the key sensing perspective, the first rotating radar is used to scan the three-dimensional space when the hook is lifted vertically, lowered, and transferred horizontally; the second rotating radar is used to scan the micro-environment space of the hook's vertical path and landing area. In the layered perception perspective, the first rotating radar is used to scan the front end of the boom, the medium-to-long-range space on the side of the boom, and the medium-range space where the hook is located. The second rotating radar is used to scan the close-range space where the hoisting core area is located; The switching of the shooting angle of the dual radar modules according to the hoisting operation stage includes: When the hoisting operation is in the lifting or lowering phase, or when the hook lifting height is detected to be lower than the preset lifting height threshold and the horizontal movement range is less than the preset horizontal range threshold, switch to the key perception perspective. When the hoisting operation is in the slewing or luffing phase, or when the hook is detected to be performing a long-distance horizontal transfer or a large-scale slewing operation, the viewpoint is switched to the layered sensing perspective.

3. The method for dynamic mapping and sensing coverage of the entire tower crane hoisting process according to claim 1, characterized in that, The step involves employing a target perception algorithm to identify dynamic obstacles and static backgrounds within the target monitoring space from the continuous point cloud data, and constructing a point cloud model of the construction environment, including: According to the shooting angle and their respective corresponding acquisition angle, the continuous point cloud data acquired by the dual radar modules are fused and optimized to obtain a continuous point cloud map of the target monitoring space. Real-time weather information for the geographical area where the tower crane is located is obtained, and the point cloud weights are adjusted under different weather conditions through a point cloud density attenuation model. A hybrid target perception network is used to identify dynamic obstacles and static backgrounds contained in the target monitoring space based on the adjusted continuous point cloud map, and to extract the geometric features of the dynamic obstacles, the geometric features of the static background, and the real-time motion parameters of the dynamic obstacles from the adjusted continuous point cloud map. An improved generalized iterative nearest point (GICP) algorithm and a dynamically fused truncated symbolic distance function (TSDF) algorithm are used to construct the point cloud model of the construction environment using the extracted geometric features and real-time motion parameters.

4. The method for dynamic mapping and sensing coverage of the entire tower crane hoisting process according to claim 3, characterized in that, The method employs a hybrid target perception network to identify dynamic obstacles and static backgrounds within the target monitoring space based on an adjusted continuous point cloud map. It then extracts the geometric features of the dynamic obstacles, the geometric features of the static background, and the real-time motion parameters of the dynamic obstacles from the adjusted continuous point cloud map, including: The adjusted continuous point cloud map is processed by the semantic awareness feature enhancement module. An improved BEBLID descriptor and FasterNet backbone network are used to construct an adaptive feature pyramid by fusing multi-scale point cloud density information. The static structural boundary features are enhanced by the edge-aware loss function, and the enhanced point cloud feature map is output. Based on the enhanced point cloud feature map, the ISS algorithm and normal vector constraints are used to calculate the normal vector and curvature of the point cloud. ISS feature scores are generated through neighborhood covariance analysis and static feature points are screened. The DBSCAN clustering algorithm is combined to remove discrete noise points. Multi-view geometric constraints are introduced to remove dynamic interference points. The curvature threshold is dynamically adjusted according to the weather information to extract the geometric features of the static background. The static background geometric features are removed simultaneously to obtain dynamic candidate point clouds. Based on the dynamic candidate point cloud, a dynamic target detection network is constructed using a hybrid architecture of YOLOv5s and PointNet. The geometric features of the dynamic obstacle are extracted by outputting the category, bounding box and point cloud clustering mask of the dynamic obstacle through a joint two-dimensional and three-dimensional detection head. The motion optical flow is calculated based on the Farneback optical flow method and the initial velocity is obtained by combining it with point cloud data. The position, velocity and acceleration parameters of the dynamic obstacle are obtained by optimizing the Kalman filter algorithm, which serve as the real-time motion parameters of the dynamic obstacle.

5. The method for dynamic mapping and sensing coverage of the entire tower crane hoisting process according to claim 4, characterized in that, The process of calculating the motion optical flow based on the Farneback optical flow method and obtaining the initial velocity by combining it with point cloud data, then optimizing it using the Kalman filter algorithm to obtain the position, velocity, and acceleration parameters of the dynamic obstacle as the real-time motion parameters of the dynamic obstacle, further includes: By using the real-time motion parameters and geometric features of dynamic obstacles, and combining Doppler feature encoding to extract the radial velocity of the dynamic obstacles, a spatiotemporal correlation graph containing timestamps, coordinates, and velocity vectors is constructed, with the edge weights being the motion consistency probability. The detection sensitivity is dynamically adjusted based on the obstacle's speed, and a joint feature optimization process is triggered when a dynamic obstacle approaches the extracted static background.

6. The method for dynamic mapping and sensing coverage of the entire tower crane hoisting process according to claim 1, characterized in that, The construction environment point cloud model is spatiotemporally registered with the physical coordinate system, using the tower crane base as the origin. A spatiotemporal graph convolutional network with Gaussian element mapping is then used to establish a global dynamic map containing the target monitoring space, including: A three-dimensional physical coordinate system is established with the geometric center of the tower crane base as the origin. The installation position coordinates and attitude parameters of the dual radar modules are obtained. The original coordinates of each point in the point cloud model of the construction environment are mapped to the physical coordinate system through a coordinate transformation matrix. Based on the timestamps of the point cloud data collected by each rotating radar, a time interpolation algorithm is used to synchronize the point cloud data at different times, eliminate the spatiotemporal deviation caused by radar data acquisition delay, and align the construction environment point cloud model with the physical coordinate system in time and space. The aligned point cloud model is discretized into Gaussian grids using the Gaussian grid mapping method. Each Gaussian grid stores the mean static background geometric features of the corresponding spatial region, the probability distribution of dynamic obstacle motion parameters, and the confidence level of point cloud density. A spatiotemporal graph convolutional network is constructed, using Gaussian surface meshes as graph nodes. The stored geometric feature mean, motion parameter probability distribution, and point cloud density confidence are uniformly represented by a Gaussian distribution model and used as node features for each graph node. By using a spatiotemporal graph convolutional network to fuse the features of Gaussian surface nodes in a continuous time sequence, the motion parameter distribution of the surface elements containing dynamic obstacles is updated in real time, while the geometric features of static background surface elements are preserved, generating a global dynamic map containing time-dimensional evolution information. The global dynamic map dynamically updates the environmental features of the target monitoring space as the hoisting operation progresses.

7. The method for dynamic mapping and sensing coverage of the entire tower crane hoisting process according to claim 1, characterized in that, The process of fusing the continuous hook trajectory with the global dynamic map, and constructing a dynamic sensing region coverage map using a trajectory differentiation algorithm and a dynamic ontology evolution algorithm, includes: The continuous hook trajectory is spatiotemporally aligned with the global dynamic map to ensure that the hook trajectory coordinates are consistent with the physical coordinate system of the global dynamic map, and the hook movement sequence is synchronously associated with the real-time movement parameters of dynamic obstacles. Based on the trajectory differential algorithm, the continuous hook trajectory is differentially calculated, and the velocity change rate, acceleration change rate and swing angle change trend of the hook in different operation stages are extracted to generate hook motion characteristic parameters. An ontology model is constructed using a dynamic ontology evolution algorithm, which includes the lifting operation stage, hook motion characteristic parameters, static background geometric features, and real-time motion parameters of dynamic obstacles. The association weights of each element in the ontology model are dynamically updated according to the operation stage. Centered on the current position of the hook, and combining the hook motion characteristic parameters and environmental features in the global dynamic map, a dynamic perception boundary is generated through the ontology model. The dynamic perception boundary is adjusted in real time with the hook motion, covering the target monitoring space within the preset safety range of the hook. The dynamic sensing boundary is overlaid with the static background and dynamic obstacles in the global dynamic map to generate the dynamic sensing area coverage map. The dynamic sensing area coverage map displays in real time the spatial positional relationship between the hook trajectory and environmental features, early warning information of dynamic obstacles entering the dynamic sensing boundary, and changes in the sensing area at different operation stages.

8. The method according to claim 1, characterized in that, After fusing the continuous hook trajectory with the global dynamic map, and constructing a dynamically perceived region coverage map using a trajectory differentiation algorithm and a dynamic ontology evolution algorithm, the method further includes: Call the real-time motion parameters of the hook and the real-time motion parameters of the dynamic obstacles to calculate the Euclidean distance and relative motion trend between the hook and each obstacle; A preset collision probability model is used to quantify the risk value, and the safety risk level is marked on the coverage map of the dynamic sensing area. High-risk areas are marked with dense markings, medium-risk areas with sparse markings, and low-risk areas with safe markings. The collision probability model is constructed based on a distance threshold and relative velocity coupling algorithm. Obstacle classification features are extracted from the point cloud model of the construction environment. The obstacle classification features include the labeled geometric dimensions of static obstacles, fixed coordinates, the real-time position of dynamic obstacles, and motion trajectory vectors. Differentiated contour symbols are superimposed and labeled in the dynamic perception area coverage map. Static obstacles are labeled with solid graphics, and dynamic obstacles are labeled with dynamic graphics. The position coordinates of dynamic obstacles are updated synchronously, and the motion trajectory curve of dynamic obstacles is generated by connecting the position coordinate points of multiple consecutive frames.

9. The method according to claim 1, characterized in that, After fusing the continuous hook trajectory with the global dynamic map, and constructing a dynamically perceived region coverage map using a trajectory differentiation algorithm and a dynamic ontology evolution algorithm, the method further includes: The tower crane's rated slewing radius, lifting height range, and obstacle distribution in the dynamic sensing area coverage map are obtained. The fast random extended tree algorithm is used to calculate the passable path of the hook from the current position to the target landing point. In the dynamic sensing area coverage map, the optimal passable path is marked with a solid line, the potential passable path is marked with a dashed line, and the inaccessible area blocked by obstacles or beyond the tower crane's performance is marked with another type of solid line. Mark the location coordinates of key nodes in each passable path, and mark the speed limit threshold of key nodes based on hook motion parameters; The dynamic sensing area coverage map is updated synchronously at a preset frequency, and the safety risk level, obstacle distribution, and each passable path are automatically refreshed. The dynamic sensing area coverage map is transmitted to the dispatching terminal and the driving terminal. The dispatching terminal interface supports the overlay of dynamic sensing area coverage maps of multiple tower cranes. The overlaid dynamic sensing area coverage map is projected onto the three-dimensional display space. The driver interface displays the safety risk level of key areas around the hook to the user based on the dynamic perception area coverage map, and enlarges the risk markings and passable paths, and provides operation assistance prompts to the user.

10. A dynamic mapping and sensing coverage system for monitoring the entire process of tower crane hoisting, characterized in that, include: The building unit is used to install multiple rotating radars on the rotatable support of the tower crane to form a dual radar module. The dual radar modules are respectively connected to the control module for communication. The switching unit is used to obtain the current hoisting operation stage of the tower crane in real time through the control module. The hoisting operation stage includes the hoisting stage, slewing stage, luffing stage and descent stage. The unit switches the shooting angle of the dual radar modules according to the hoisting operation stage. The key perception angle is used in the hoisting and descent stages, and the layered perception angle is used in the slewing and luffing stages. The unit synchronously controls the dual radar modules to obtain continuous point cloud data of the corresponding operation stage. The trajectory tracking unit is used to collect tower crane operating parameters and, combined with the continuous point cloud data obtained by the dual radar modules, calculate the three-dimensional coordinates, movement speed and swing angle of the hook in real time to generate a continuous hook trajectory. The point cloud modeling unit is used to identify dynamic obstacles and static backgrounds contained in the target monitoring space from the continuous point cloud data using a target perception algorithm, and to construct a point cloud model of the construction environment. The spatiotemporal registration unit is used to perform spatiotemporal registration of the construction environment point cloud model with the physical coordinate system, with the tower crane base as the origin of the physical coordinate system, and to establish a global dynamic map containing the target monitoring space using a spatiotemporal graph convolutional network with Gaussian surface mapping. The coverage map generation unit is used to integrate the continuous hook trajectory with the global dynamic map, and construct a dynamic sensing area coverage map through trajectory differentiation algorithm and dynamic ontology evolution algorithm. The dynamic sensing area coverage map is used to display the target monitoring space that changes continuously with each operation stage during the hoisting operation, as well as the construction environment characteristics within the target monitoring space.

Images