Linear intelligent monitoring system for bridge fabrication machine

Abstract

The invention relates to the technical field of intelligent traffic systems, in particular to a linear intelligent monitoring system for a bridge fabrication machine, which comprises a multi-source sensing module, a data processing module, a data processing module, a data processing module, a data processing module, a data processing module and a data processing module, and is characterized in that the multi-source sensing module comprises strain sensors, inclinometers, GPS positioning units and machine vision units; the edge computing node is internally provided with a self-adaptive data fusion algorithm and is used for denoising sensor data in real time; the cloud digital twin platform is used for synchronizing the BIM design model and generating a millimeter-level linear deviation early warning signal; the PLC control interface is used for receiving the early warning signal and automatically adjusting parameters of the template system and the walking mechanism; according to the method, environmental noise interference is effectively eliminated through multi-source sensor fusion and a self-adaptive data filtering technology, the monitoring precision is improved to a millimeter level, real-time comparison of construction data and a design model is realized based on a BIM and digital twinning dynamic interaction system, and the timeliness and accuracy of linear control are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent transportation systems, and in particular to a line intelligent monitoring system for a bridge-building machine. Background Art

[0002] At present, the linear control of bridge-building machines during construction mainly relies on manual measurement and manual adjustment, such as traditional measurement methods such as total stations and levels, combined with experience and judgment to adjust the template and running mechanism. Some advanced systems use a single type of sensor for local monitoring and perform limited adjustments through independent control systems. Some construction teams use BIM models for construction simulation, but the dynamic interaction ability between the model and actual construction data is weak, making it difficult to achieve real-time closed-loop control.

[0003] Traditional measurement methods have low worker efficiency and cannot meet the high-precision requirements of large-span bridge construction, especially during the concrete pouring stage, when structural deformation is time-varying and difficult to respond to in a timely manner. Single sensors have limited monitoring range and are easily affected by environmental noise, resulting in insufficient data reliability and possible misjudgment. BIM models are disconnected from on-site construction data and cannot dynamically correct construction deviations, resulting in delayed linear control. Existing early warning mechanisms are mostly passive alarms, lack intelligent decision-making support, and cannot automatically generate correction plans. They rely on manual experience adjustments and are prone to cumulative errors.

[0004] Therefore, in order to solve the above problems, the present invention proposes a line intelligent monitoring system for a bridge-building machine. Summary of the Invention

[0005] In order to overcome the problems of low measurement efficiency and low measurement accuracy of traditional measurement methods, the present invention proposes a line intelligent monitoring system for a bridge construction machine.

[0006] The technical solution of the present invention is: a line intelligent monitoring system for a bridge-building machine, comprising:

[0007] Multi-source sensing module, including strain sensor, inclinometer, GPS positioning unit and machine vision unit;

[0008] Edge computing nodes with built-in adaptive data fusion algorithms for real-time denoising and processing of sensor data;

[0009] The cloud-based digital twin platform synchronizes the BIM design model and generates millimeter-level linear deviation warning signals;

[0010] PLC control interface receives warning signals and automatically adjusts the template system and running mechanism parameters.

[0011] Preferably, the adaptive data fusion algorithm adopts a hybrid architecture of Kalman filtering and wavelet transform, wherein Kalman filtering is used to dynamically estimate the motion state of various parts of the bridge-building machine, and wavelet transform performs multi-scale decomposition on high-frequency noise, and dynamically adjusts the weight coefficient of each sensor according to different construction stages. In the walking stage, GPS positioning data is preferentially used to ensure trajectory accuracy, and in the concrete pouring stage, strain sensors are switched to be dominant to monitor structural stress changes. At the same time, a sliding time window mechanism is introduced to recalculate the confidence index of each sensor every 10 seconds. When the data mutation of a sensor exceeds 3 times the standard deviation, its weight is automatically reduced and redundancy check is triggered.

[0012] Preferably, the machine vision unit includes 6 groups of 20-megapixel industrial cameras arranged at the template joints, achieving full weld coverage with a resolution of 8mm / pixel, and adopting an improved YOLOv5 model. By adding a concrete surface texture feature dataset to the pre-trained model, combined with the temperature gradient data collected by the infrared thermal imaging module, an enhanced image is generated using a pixel-level fusion algorithm. When a crack with a width of ≥0.2mm or a temperature abnormality area is detected, the defect location is automatically marked and the expansion trend is calculated, and the corresponding area in the three-dimensional model is triggered to be highlighted.

[0013] Preferably, the system also includes an intelligent adjustment module, which includes 12 groups of outer mold horizontal hydraulic cylinders and 8 groups of vertical hydraulic cylinders, and realizes linkage adjustment through a PID controller; the inner mold inclined rigging is equipped with a spoke-type tension sensor with a range of 0-50kN and a sampling frequency of 100Hz, which provides real-time feedback to the hydraulic pump station to adjust the oil pressure; the bottom mold elevation adjustment mechanism adopts a servo motor to drive the ball screw, and cooperates with the inclinometer to form a closed-loop control. During the concrete pouring process, the template posture is fine-tuned every 30 seconds, and the cumulative adjustment amount does not exceed ±1.5mm of the design value.

[0014] Preferably, the system also includes a running system module, the reaction-force self-balancing track of which includes two H-shaped steel tracks arranged in parallel. The track on each side relies on a hydraulic cylinder to provide running power, and the encoder signal is transmitted through the CAN bus to realize synchronous control of the left and right slide speeds. When the speed difference exceeds 1%, the oil pressure is automatically compensated; 8 sets of laser rangefinders are installed on both sides of the track to detect changes in track spacing in real time. When a track misalignment ≥3mm or a local settlement ≥2mm is detected, the hydraulic power is immediately cut off and the wedge-shaped self-locking device is activated, and at the same time, an audible and visual alarm is used to prompt maintenance.

[0015] Preferably, the cloud-based digital twin platform integrates the ANSYS solver, and after receiving sensor data uploaded by the edge computing node every 30 seconds, it automatically updates the boundary conditions in the finite element model, calculates the theoretical deformation curve under the combined action of temperature gradient and concrete creep, and generates a dynamic error correction strategy library containing 12 typical working conditions. When the measured value deviates from the theoretical value by more than 2 mm, it automatically matches the optimal adjustment plan and generates a control instruction set including the hydraulic cylinder stroke adjustment amount and the travel speed correction value.

[0016] Preferably, the system includes a three-level early warning mechanism. When the first-level early warning is triggered, the system flashes a yellow warning icon on the HMI interface and saves a snapshot of the deviation data, and sends a lightweight notification to the responsible engineer's mobile phone; after the second-level early warning is activated, the concrete pumping or formwork action in the current area is automatically stopped, the relevant actuators are locked and the backup sensor calibration is started; when the third-level early warning occurs, the system automatically calls the emergency plan pre-stored in the BIM model, pushes the three-dimensional emergency plan animation to the on-site command terminal through the 4G / 5G network, and remotely unlocks the hydraulic locking mechanism of the foldable safety platform to quickly deploy the rescue channel.

[0017] As a preference, the foldable safety platform comprises four telescopic hydraulic columns, with X-shaped cross links between the columns to maintain stability. The platform pedal adopts a three-stage telescopic design, with anti-slip grids and embedded pressure sensors on the pedal surface. When a local load exceeding 300kg / m is detected, the platform pedal is 2 8 sets of ultrasonic sensors are arranged on the edge of the platform to monitor the deployment of the pedals in real time. When any section of the pedal is not fully extended, the system will prohibit the power output of the tensioning equipment and display a red no-operation prompt on the operation interface.

[0018] Preferably, in the edge-cloud collaborative architecture, the edge computing node adopts the NVIDIA Jetson AGX Orin module, runs the real-time control algorithm, deploys the LSTM neural network in the cloud, receives the minute-level data compressed by the edge node, analyzes the linear evolution law in the historical data, predicts the maximum possible deviation and high-risk areas in the next 2 hours, and superimposes the prediction results on the three-dimensional model in the form of a heat map.

[0019] Preferably, the system also includes an AI-assisted decision-making module, which pre-generates a template adjustment parameter curve through the BIM design model, including a time series of theoretical values ​​of the outer mold opening and closing degree, bottom mold elevation and inner mold cable tension in different pouring stages. During the concrete pouring process, the measured data of multi-source sensors are compared with the BIM theoretical value every 5 minutes, and then the three-dimensional space comprehensive deviation is calculated. When the cumulative deviation exceeds the threshold, the correction mechanism is triggered. When the cumulative deviation exceeds the threshold, the AI-assisted decision-making module is started to generate a correction plan. Finally, the correction effect simulation is displayed through a three-dimensional visualization interface and executed after manual confirmation.

[0020] Beneficial effects of the present invention:

[0021] 1. Multi-source sensor fusion and adaptive data filtering technology effectively eliminate environmental noise interference, improve monitoring accuracy to the millimeter level, and realize real-time comparison of construction data and design models based on the dynamic interactive system of BIM and digital twins, significantly improving the timeliness and accuracy of linear control.

[0022] 2. The intelligent template adjustment and travel control system realizes one-click automatic adjustment of the construction process through hydraulic synchronous drive and closed-loop feedback mechanism, greatly reducing the need for manual intervention and operational errors, and improving measurement efficiency.

[0023] 3. The three-level early warning mechanism combined with AI-assisted decision-making can quickly identify construction anomalies and generate optimization plans, thereby effectively preventing quality accidents. While ensuring real-time control response, the edge-cloud collaborative architecture optimizes construction processes through big data analysis, improves construction efficiency, reduces linear deviations, and reduces the risks and labor costs of high-altitude operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 Shown is a schematic diagram of the system framework of the present invention;

[0025] Figure 2 What is shown is a schematic diagram of the workflow of the present invention. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of the present invention.

[0027] See also Figure 1 The present invention provides an embodiment: a line intelligent monitoring system for a bridge construction machine, comprising:

[0028] Multi-source sensing module, including strain sensor, inclinometer, GPS positioning unit and machine vision unit;

[0029] Edge computing nodes with built-in adaptive data fusion algorithms for real-time denoising and processing of sensor data;

[0030] The cloud-based digital twin platform synchronizes the BIM design model and generates millimeter-level linear deviation warning signals;

[0031] PLC control interface receives warning signals and automatically adjusts the template system and running mechanism parameters.

[0032] The adaptive data fusion algorithm adopts a hybrid architecture of Kalman filtering and wavelet transform. Kalman filtering is used to dynamically estimate the motion state of various parts of the bridge-building machine, while wavelet transform performs multi-scale decomposition of high-frequency noise. The weight coefficients of each sensor are dynamically adjusted according to different construction stages. GPS positioning data is preferentially used to ensure trajectory accuracy during the walking stage. During the concrete pouring stage, strain sensors are switched to the dominant position to monitor structural stress changes. At the same time, a sliding time window mechanism is introduced to recalculate the confidence index of each sensor every 10 seconds. When the data mutation of a sensor exceeds three times the standard deviation, its weight is automatically reduced and a redundancy check is triggered.

[0033] The machine vision unit includes six groups of 20-megapixel industrial cameras arranged at the template joints, achieving full weld coverage with a resolution of 8mm / pixel. It adopts an improved YOLOv5 model, adds a concrete surface texture feature dataset to the pre-trained model, and uses a pixel-level fusion algorithm to generate enhanced images in combination with temperature gradient data collected by the infrared thermal imaging module. When a crack with a width of 0.2mm or an abnormal temperature area is detected, the defect location is automatically marked and the expansion trend is calculated, while triggering the corresponding area in the three-dimensional model to be highlighted.

[0034] The system also includes an intelligent adjustment module, which contains 12 sets of outer mold horizontal hydraulic cylinders and 8 sets of vertical hydraulic cylinders, and realizes linkage adjustment through a PID controller; the inner mold inclined rigging is equipped with a spoke-type tension sensor with a range of 0-50kN and a sampling frequency of 100Hz, which provides real-time feedback to the hydraulic pump station to adjust the oil pressure; the bottom mold elevation adjustment mechanism uses a servo motor to drive the ball screw, and cooperates with the inclinometer to form a closed-loop control. During the concrete pouring process, the template posture is fine-tuned every 30 seconds, and the cumulative adjustment amount does not exceed ±1.5mm of the design value.

[0035] The system also includes a running system module, the reaction-force self-balancing track of which consists of two parallel H-shaped steel tracks. Each side of the track relies on a hydraulic cylinder to provide running power, and the encoder signal is transmitted through the CAN bus to achieve synchronous control of the left and right slide speeds. When the speed difference exceeds 1%, the oil pressure is automatically compensated; 8 sets of laser rangefinders are installed on both sides of the track to detect changes in track spacing in real time. When a track misalignment of ≥3mm or local settlement of ≥2mm is detected, the hydraulic power is immediately cut off and the wedge self-locking device is activated. At the same time, an audible and visual alarm is used to prompt maintenance.

[0036] The cloud-based digital twin platform integrates the ANSYS solver. After receiving sensor data uploaded by the edge computing node every 30 seconds, it automatically updates the boundary conditions in the finite element model, calculates the theoretical deformation curve under the combined action of temperature gradient and concrete creep, and generates a dynamic error correction strategy library containing 12 typical working conditions. When the measured value deviates from the theoretical value by more than 2 mm, it automatically matches the optimal adjustment plan and generates a control instruction set including the hydraulic cylinder stroke adjustment amount and the travel speed correction value.

[0037] The system includes a three-level early warning mechanism. When the first-level warning is triggered, the system flashes a yellow warning icon on the HMI interface and saves a snapshot of the deviation data, while sending a lightweight notification to the responsible engineer's mobile phone. When the second-level warning is activated, the concrete pumping or formwork action in the current area is automatically stopped, the relevant actuators are locked, and the backup sensor calibration is started. When the third-level warning occurs, the system automatically calls the emergency plan pre-stored in the BIM model, pushes a three-dimensional emergency plan animation to the on-site command terminal via the 4G / 5G network, and remotely unlocks the hydraulic locking mechanism of the foldable safety platform to quickly deploy a rescue channel.

[0038] The foldable safety platform consists of four telescopic hydraulic columns, with X-shaped cross links between the columns to maintain stability. The platform pedal adopts a three-stage telescopic design, with anti-slip grids and embedded pressure sensors on the pedal surface. When the local load exceeds 300kg / m 2 8 sets of ultrasonic sensors are arranged on the edge of the platform to monitor the deployment of the pedals in real time. When any section of the pedal is not fully extended, the system will prohibit the power output of the tensioning equipment and display a red no-operation prompt on the operation interface.

[0039] In the edge-cloud collaborative architecture, the edge computing nodes use NVIDIA Jetson AGX Orin modules to run real-time control algorithms, and the LSTM neural network is deployed in the cloud. It receives minute-level data compressed by the edge nodes, analyzes the linear evolution patterns in the historical data, and predicts the maximum possible deviation and high-risk areas in the next two hours. The prediction results are superimposed on the three-dimensional model in the form of a heat map.

[0040] The system also includes an AI-assisted decision-making module, which pre-generates a template adjustment parameter curve through the BIM design model, including a time series of theoretical values ​​of the outer mold opening and closing degree, bottom mold elevation, and inner mold cable tension at different pouring stages. During the concrete pouring process, the measured data of multi-source sensors is compared with the BIM theoretical value every 5 minutes, and then the three-dimensional space comprehensive deviation is calculated. When the cumulative deviation exceeds the threshold, the correction mechanism is triggered. When the cumulative deviation exceeds the threshold, the AI-assisted decision-making module is started to generate a correction plan. Finally, the correction effect simulation is displayed through a three-dimensional visualization interface and executed after manual confirmation.

[0041] See also Figure 2 , further, the specific workflow of the present invention is described in detail:

[0042] The system uses distributed strain sensors, high-precision inclinometers, RTK-GPS positioning modules and industrial vision camera arrays to collect multi-dimensional construction data such as bridge-building machine template deformation, travel trajectory, structural stress and surface defects in real time at a sampling frequency of 100Hz. All sensor data is transmitted to the edge computing node via industrial Ethernet.

[0043] The edge computing node runs an adaptive data fusion algorithm, performs wavelet noise reduction on the raw data, and uses dynamic weighted Kalman filtering to eliminate the influence of GPS signal drift and strain gauge temperature drift. At the same time, it uses timestamp synchronization technology to ensure the time alignment of multi-source data. The processed data packets are synchronously uploaded to the cloud digital twin platform via the 5G private network.

[0044] The cloud platform maps the pre-processed data with the BIM design model in real time, calculates the theoretical deformation value through the finite element analysis engine, and uses the point cloud matching algorithm to generate a three-dimensional deviation field. When a lateral deviation of >1.5mm or a vertical deviation of >2mm is detected, the early warning mechanism is automatically triggered and the abnormal area is marked in the twin model.

[0045] The AI ​​decision-making module calls the historical working condition database, combines the current ambient temperature, concrete age and other parameters, and uses the LSTM neural network to predict the deformation trend in the next 30 minutes. It generates an optimal control plan that includes template adjustment and travel speed correction values, and verifies the feasibility of the plan through Monte Carlo simulation.

[0046] After confirmation by the engineer, the system decomposes the control instructions into PLC executable code and sends it to the hydraulic cylinder group through the PROFINET bus. At the same time, it activates the laser tracker to monitor the execution effect in real time, forming a closed-loop control.

[0047] The 3D visualization platform displays measured data (red point cloud), design model (blue grid) and correction effect prediction (green surface) in an AR format, supports multi-perspective zooming and cross-sectional analysis, and all operation logs and adjustment parameters are automatically archived to the blockchain notarization system.

[0048] The cloud-based big data platform periodically analyzes construction data and optimizes control strategy parameters through deep reinforcement learning.

[0049] Through the above steps, multi-source sensor fusion and adaptive data filtering technology are used to effectively eliminate environmental noise interference, improve monitoring accuracy to the millimeter level, and realize real-time comparison of construction data and design models based on the dynamic interactive system of BIM and digital twins, which significantly improves the timeliness and accuracy of linear control, and solves the problems of low measurement efficiency and low measurement accuracy of traditional measurement methods.

Claims

1. A line intelligent monitoring system for a bridge-building machine, characterized in that: Includes: Multi-source sensing module, including strain sensor, inclinometer, GPS positioning unit and machine vision unit; Edge computing nodes with built-in adaptive data fusion algorithms for real-time denoising and processing of sensor data; The cloud-based digital twin platform synchronizes the BIM design model and generates millimeter-level linear deviation warning signals; PLC control interface receives warning signals and automatically adjusts the template system and running mechanism parameters.

2. The intelligent line monitoring system for a bridge-building machine according to claim 1, characterized in that: The adaptive data fusion algorithm adopts a hybrid architecture of Kalman filtering and wavelet transform. Kalman filtering is used to dynamically estimate the motion state of various parts of the bridge-building machine, while wavelet transform performs multi-scale decomposition of high-frequency noise. The weight coefficients of each sensor are dynamically adjusted according to different construction stages. GPS positioning data is preferentially used to ensure trajectory accuracy during the walking stage. During the concrete pouring stage, strain sensors are switched to the dominant position to monitor structural stress changes. At the same time, a sliding time window mechanism is introduced to recalculate the confidence index of each sensor every 10 seconds. When the data mutation of a sensor exceeds three times the standard deviation, its weight is automatically reduced and a redundancy check is triggered.

3. The intelligent line monitoring system for a bridge-building machine according to claim 1, characterized in that: The machine vision unit includes six groups of 20-megapixel industrial cameras arranged at the template joints, achieving full weld coverage with a resolution of 8mm / pixel. It adopts an improved YOLOv5 model, adds a concrete surface texture feature dataset to the pre-trained model, and uses a pixel-level fusion algorithm to generate enhanced images in combination with temperature gradient data collected by the infrared thermal imaging module. When a crack with a width of 0.2mm or an abnormal temperature area is detected, the defect location is automatically marked and the expansion trend is calculated, while triggering the corresponding area in the three-dimensional model to be highlighted.

4. The intelligent line monitoring system for a bridge-building machine according to claim 1, characterized in that: The system also includes an intelligent adjustment module, which contains 12 sets of outer mold horizontal hydraulic cylinders and 8 sets of vertical hydraulic cylinders, and realizes linkage adjustment through a PID controller; the inner mold inclined rigging is equipped with a spoke-type tension sensor with a range of 0-50kN and a sampling frequency of 100Hz, which provides real-time feedback to the hydraulic pump station to adjust the oil pressure; the bottom mold elevation adjustment mechanism uses a servo motor to drive the ball screw, and cooperates with the inclinometer to form a closed-loop control. During the concrete pouring process, the template posture is fine-tuned every 30 seconds, and the cumulative adjustment amount does not exceed ±1.5mm of the design value.

5. The intelligent line monitoring system for a bridge-building machine according to claim 1 is characterized in that: The system also includes a running system module, the reaction-force self-balancing track of which consists of two parallel H-shaped steel tracks. Each side of the track relies on a hydraulic cylinder to provide running power, and the encoder signal is transmitted through the CAN bus to achieve synchronous control of the left and right slide speeds. When the speed difference exceeds 1%, the oil pressure is automatically compensated; 8 sets of laser rangefinders are installed on both sides of the track to detect changes in track spacing in real time. When a track misalignment of ≥3mm or local settlement of ≥2mm is detected, the hydraulic power is immediately cut off and the wedge self-locking device is activated. At the same time, an audible and visual alarm is used to prompt maintenance.

6. The intelligent line monitoring system for a bridge-building machine according to claim 1, characterized in that: The cloud-based digital twin platform integrates the ANSYS solver. After receiving sensor data uploaded by the edge computing node every 30 seconds, it automatically updates the boundary conditions in the finite element model, calculates the theoretical deformation curve under the combined action of temperature gradient and concrete creep, and generates a dynamic error correction strategy library containing 12 typical working conditions. When the measured value deviates from the theoretical value by more than 2 mm, it automatically matches the optimal adjustment plan and generates a control instruction set including the hydraulic cylinder stroke adjustment amount and the travel speed correction value.

7. The intelligent line monitoring system for a bridge-building machine according to claim 1, characterized in that: The system includes a three-level early warning mechanism. When the first-level warning is triggered, the system flashes a yellow warning icon on the HMI interface, saves a snapshot of the deviation data, and sends a lightweight notification to the responsible engineer's mobile phone. When the second-level warning is activated, the system automatically stops concrete pumping or formwork operation in the current area, locks the relevant actuators, and initiates backup sensor calibration. When a level 3 warning occurs, the system automatically retrieves the emergency plan pre-stored in the BIM model, pushes a three-dimensional emergency plan animation to the on-site command terminal via the 4G / 5G network, and remotely unlocks the hydraulic locking mechanism of the foldable safety platform to quickly deploy a rescue channel.

8. The intelligent line monitoring system for a bridge-building machine according to claim 7, characterized in that: The foldable safety platform consists of four telescopic hydraulic columns, with X-shaped cross links between the columns to maintain stability. The platform pedal adopts a three-stage telescopic design, with anti-slip grids and embedded pressure sensors on the pedal surface. When the local load exceeds 300kg / m 2 8 sets of ultrasonic sensors are arranged on the edge of the platform to monitor the deployment of the pedals in real time. When any section of the pedal is not fully extended, the system will prohibit the power output of the tensioning equipment and display a red no-operation prompt on the operation interface.

9. The intelligent line monitoring system for a bridge-building machine according to claim 1, characterized in that: In the edge-cloud collaborative architecture, the edge computing nodes use NVIDIA Jetson AGX Orin modules to run real-time control algorithms, and the LSTM neural network is deployed in the cloud. It receives minute-level data compressed by the edge nodes, analyzes the linear evolution patterns in the historical data, and predicts the maximum possible deviation and high-risk areas in the next two hours. The prediction results are superimposed on the three-dimensional model in the form of a heat map.

10. The intelligent line monitoring system for a bridge-building machine according to claims 1-9, characterized in that: The system also includes an AI-assisted decision-making module, which pre-generates a template adjustment parameter curve through the BIM design model, including a time series of theoretical values ​​of the outer mold opening and closing degree, bottom mold elevation, and inner mold cable tension at different pouring stages. During the concrete pouring process, the measured data of multi-source sensors is compared with the BIM theoretical value every 5 minutes, and then the three-dimensional space comprehensive deviation is calculated. When the cumulative deviation exceeds the threshold, the correction mechanism is triggered. When the cumulative deviation exceeds the threshold, the AI-assisted decision-making module is started to generate a correction plan. Finally, the correction effect simulation is displayed through a three-dimensional visualization interface and executed after manual confirmation.

Images