BIM-based hoisting construction supervision optimization management system

Abstract

The invention discloses a BIM-based hoisting construction supervision optimization management system. The system comprises a sensing layer which collects multi-source heterogeneous data in real time; according to the decision-making layer, a bottom layer utilizes an incremental RRT # algorithm to generate candidate paths meeting crane kinematics constraints, a distributed Q-learning framework is embedded in an upper layer, all cranes serve as independent agents, collaborative learning is carried out through a shared experience pool, a path planning strategy is dynamically optimized according to environment sensing data and construction progress requirements, and a path planning strategy is established. Meanwhile, a fuzzy sliding mode control algorithm is developed to be combined with an LSTM-Transformer crane cart and trolley walking speed, a hook crane cart and trolley walking speed, a hook lifting speed and a steel wire rope disturbance prediction model to calculate crane motion compensation parameters, and an optimal control instruction is generated in advance based on predicted crane moving walking and lifting data; the execution layer is used for issuing the instruction generated by the decision-making layer to construction equipment; and the optimization layer is used for constructing a BIM model and feeding back an equipment execution result and structure safety monitoring data to the decision-making layer.

Description

Technical Field

[0001] The present invention relates to the technical field of large-unit hoisting safety for pumped-storage hydropower station units, and in particular to a BIM-based hoisting construction supervision and optimization management system. Background Art

[0002] Pumped-storage power station buildings have complex, multi-story structures. The harsh construction environment for hoisting large components such as stators, rotors, runners, top covers, inlet valves, and GCBs (generator breakers) makes this operation the most risk-intensive. The bridge crane (bridge crane) within the main building is the core lifting equipment for large-scale lifting. It consists of a bridge (also known as a trolley), a hoisting mechanism, a trolley, a trolley travel mechanism, an operator's cab, a trolley conductive device (auxiliary sliding wire), and a crane main power conductive device (main sliding wire). The trolley travels longitudinally along tracks laid on elevated structures on both sides, while the trolley travels horizontally along tracks laid on the bridge. Traditional supervision models rely on manual experience and offline planning, making them difficult to handle complex scenarios involving multiple variables, such as crane dynamic movement, lifting speed and wire rope disturbance, equipment coordination, and structural deformation. For example, hook swing caused by crane trolley movement, lifting, or stopping can lead to collisions. Path conflicts during cross-operation between cranes on the same or different working floors in the main powerhouse reduce construction efficiency. Furthermore, the uncertainty of multi-faceted operations during initial installation or unit maintenance at power plants is even more likely to pose safety hazards. The maturity of Building Information Modeling (BIM) technology provides a digital foundation for construction supervision. Its 3D visualization, data integration, and simulation analysis capabilities can address the shortcomings of traditional supervision in terms of real-time performance, accuracy, and predictiveness. By deeply integrating BIM with technologies such as the Internet of Things, artificial intelligence, and control theory, an intelligent supervision system covering the entire "perception-decision-making-execution-optimization" chain can be constructed. This system can implement core functions such as dynamic path planning, real-time compensation for crane travel and lifting speeds and wire rope disturbances, and structural safety warnings in the construction of hydropower stations, especially pumped-storage power stations. Ultimately, this significantly improves the safety margin and construction efficiency of lifting operations.

[0003] In existing technologies, the improved RRT* algorithm is tightly integrated with the BIM model. BIM models provide rich building information, and the improved RRT* algorithm, by leveraging this information, can more accurately plan paths that meet actual construction needs. However, while the improved RRT* algorithm takes into account the kinematic constraints of equipment, the dynamic environment during pumped-storage power station construction and electromechanical equipment installation is extremely dynamic (e.g., multiple working surfaces, temporary obstacles, etc.), and the algorithm's real-time response capability may be insufficient. This can cause construction equipment to continue moving along the original path, potentially deviating from actual working conditions and increasing collision risks. Therefore, a BIM-based hoisting construction supervision and optimization management system is proposed. Summary of the Invention

[0004] The purpose of the present invention is to solve the shortcomings of the existing technology and propose a BIM-based hoisting construction supervision and optimization management system.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A BIM-based hoisting construction supervision optimization management system, including:

[0007] Perception layer: 5G+ edge computing nodes are deployed to collect real-time heterogeneous data from multiple sources, including millimeter-wave radar, UWB+ lidar, micro lidar weather stations, and fiber Bragg grating sensor networks. A Kalman filter-particle filter fusion algorithm is used to construct a spatiotemporal map of dynamic obstacles and align the spatiotemporal data of multiple sensors. Furthermore, a fiber Bragg grating sensor network is embedded and a digital twin mirror interface is deployed to enable real-time monitoring of strain, tilt, and steel structure stress in key crane components, as well as synchronization of BIM models with physical entity states.

[0008] Decision-making layer: The bottom layer uses an incremental RRT# algorithm, combined with a non-uniform sampling strategy and an intelligent pruning mechanism, to generate candidate paths that meet the crane's kinematic constraints. The upper layer embeds a distributed Q-learning framework. The crane acts as an independent intelligent agent and conducts collaborative learning through a shared experience pool. It dynamically optimizes the path planning strategy based on environmental perception data and construction schedule requirements. At the same time, a fuzzy sliding mode control algorithm is developed, combined with an LSTM-Transformer crane trolley and car travel speed, hook lifting speed, and wire rope disturbance prediction model to calculate the crane motion compensation parameters. Model predictive control is introduced to generate optimal control instructions in advance based on the predicted mobile travel speed, lifting speed, and wire rope disturbance data.

[0009] Execution layer: The path planning instructions and wire rope disturbance compensation control instructions generated by the decision layer are sent to the construction equipment via the OPC UA over TSN protocol;

[0010] Optimization layer: Build a BIM model, integrate geometric information, construction progress, structural response and equipment status, and feed back equipment execution results and structural safety monitoring data to the decision-making layer.

[0011] The above technical solution further includes:

[0012] Furthermore, the multi-source heterogeneous data collected by the perception layer include environmental perception data, structural perception data, equipment status data, and personnel and safety data. The environmental perception data includes millimeter wave radar data, UWB+ lidar data, and micro lidar weather station data. The structural perception data includes fiber grating sensor network data and digital twin mirror interface data. The equipment status data includes crane operation status data and elevator operation status data. The personnel and safety data includes personnel positioning data and security monitoring video data.

[0013] Furthermore, the perception layer adopts the Kalman filter-particle filter fusion algorithm to construct the spatiotemporal map of dynamic obstacles and align the spatiotemporal data of multiple sensors in the following specific steps:

[0014] Data preprocessing: Through data cleaning and data format unification, using threshold or statistical methods to detect and eliminate outliers, and defining a unified data structure, the raw sensor data is preliminarily processed to remove noise and outliers, and the data from different sensors is converted into a unified format;

[0015] Kalman filter: Through two steps of prediction and update, the state of the system is recursively estimated using the state transfer matrix and the observation matrix to perform optimal state estimation for linear Gaussian systems;

[0016] Particle filter: It uses a set of random samples to represent the state distribution of the system, and approximates the true posterior distribution through resampling and weight updating to complete the state estimation of nonlinear non-Gaussian systems;

[0017] Fusion: Use Kalman filtering to estimate the state of the linear part and particle filtering to estimate the state of the nonlinear part, and fuse the results of the two filters through weighted averaging or interactive multi-model algorithm.

[0018] Furthermore, the specific steps of the perception layer for performing digital twin mirror synchronization are:

[0019] Sensor deployment: Deploy fiber grating sensor networks on key components of the crane;

[0020] Integration of digital twin mirror interface and BIM model: Assign a unique identifier to each component of the BIM model, and also assign a unique identifier to each sensor in the sensor network. By establishing a mapping relationship between the identifiers, the sensor data is associated with the BIM model components. When the sensor data changes, the data is synchronized to the BIM model through the interface to drive the update of the model component information.

[0021] Furthermore, the bottom layer utilizes an incremental RRT# algorithm, combined with a non-uniform sampling strategy and an intelligent pruning mechanism, to generate candidate paths that satisfy the kinematic constraints of the crane, including the following steps;

[0022] Initialization: Set the starting point and target point in the BIM model;

[0023] Non-uniform sampling: Based on the risk heat map, sampling is performed in key areas. The sampling probability P(x) is calculated based on the risk value R(x) at position x and the obstacle density D(x). The calculation formula is P(x) = (R(x)·D(x)) / (∫ Ω R(x′)·D(x′)dx′), where Ω is the sampling space and the integral term is used to normalize the probability;

[0024] Path extension: Using the incremental RRT# algorithm, the path is randomly extended from the starting point, and the kinematic constraints of the crane are considered in each extension;

[0025] Intelligent pruning: removes redundant paths and retains paths with lower path costs and that satisfy kinematic constraints;

[0026] Path optimization: smoothing the retained path;

[0027] Output candidate paths: Output candidate paths that satisfy all constraints.

[0028] Furthermore, the upper layer embeds a distributed Q-learning framework. Each crane acts as an independent intelligent agent and performs collaborative learning through a shared experience pool. It dynamically optimizes the path planning strategy based on environmental perception data and construction progress requirements, including the following steps:

[0029] Initialization: The agent obtains the current state information from the BIM model and initializes the Q value table;

[0030] Experience collection: Each agent collects experience during the interaction process and stores it in a shared experience pool;

[0031] Collaborative learning: The agent randomly extracts a batch of experiences from the experience pool for learning, and uses the Q-learning algorithm to update the Q-value table based on the extracted experiences;

[0032] Path planning: Based on the learned Q-value table, the agent selects the optimal action for path planning;

[0033] Dynamic adjustment: During the lifting process, the intelligent agent selects the optimal action based on environmental perception data and construction progress requirements (such as floor construction progress adjustment). The intelligent agent executes the selected action and updates its own position. Based on the results of the action, the intelligent agent receives reward signals from the environment.

[0034] Furthermore, the developed fuzzy sliding mode control algorithm combines the LSTM-Transformer crane trolley and car travel speeds, hook lifting speeds, and wire rope disturbance prediction model to calculate the crane motion compensation parameters, introduces model predictive control, and generates optimal control instructions in advance based on the predicted mobile travel speed, lifting speed, and wire rope disturbance data, including the following steps:

[0035] Fuzzy sliding mode control algorithm: The sliding surface function s(t) is defined to represent the deviation between the system state and the desired state. The sliding surface function is expressed as Where, e(t) is the position or angle error, is the error change rate, c is the design parameter, and the control law u(t) is designed to make the system state reach the sliding surface in a finite time and move to the equilibrium point along the sliding surface. The control law is expressed as u(t) = u eq (t)+u sw (t), where u eq (t) is the equivalent control term, u sw (t) is the switching control item, and fuzzy logic is used to adjust the gain of the switching control item to adapt to different working conditions;

[0036] Build an LSTM-Transformer model to predict the travel speed of the crane trolley and carriage, the hook lifting speed, and wire rope disturbances. The LSTM layer processes time series data to capture short-term changes in movement, lifting speed, and wire rope disturbances. The Transformer encoder processes spatial data to capture the spatial correlation between movement speed, lifting speed, and wire rope disturbances of large objects at different locations. The model is trained using historical movement speed, lifting speed, and wire rope disturbance data, and the loss function is optimized to enable the model to predict lifting speed and wire rope disturbance data for a period of time in the future.

[0037] Integrated model predictive control: Based on the LSTM-Transformer crane trolley and carriage travel speed, hook lifting speed, and wire rope disturbance prediction model, rolling optimization and feedback correction, the system inputs the current lifting speed and large-piece position data and outputs the future mobile travel speed, lifting speed, and wire rope disturbance prediction values. In each control cycle, the finite time domain optimal control problem is solved and the optimal control problem is solved using the optimization algorithm to obtain the optimal control sequence. The first element of the optimal control sequence is applied to the crane system, thereby generating the optimal control instruction in advance.

[0038] Furthermore, the OPC UA over TSN protocol is used to send the data to the construction equipment, including the following steps:

[0039] Instruction encoding and encapsulation: Encode the path planning results and compensation parameters into OPC UA data structures;

[0040] OPC UA over TSN network transmission: Build an OPC UA-based construction equipment information model, define the nodes and methods of equipment, sensors, and actuators, configure TSN network time synchronization and traffic scheduling parameters, and send encoded instructions to construction equipment via the OPC UA over TSN protocol;

[0041] Device-side instruction decoding and execution: An OPC UA client is set up on the construction equipment to receive and decode instructions from the decision-making layer. The OPC UA client parses the instruction parameters in the OPC UA data structure based on the specific model and configuration of the equipment.

[0042] The present invention has the following beneficial effects:

[0043] In the present invention, the bottom layer of the decision-making layer generates candidate paths that meet the kinematic constraints of the crane through a non-uniform sampling strategy and an intelligent pruning mechanism, thereby improving the path search efficiency. The upper layer of the decision-making layer is embedded in a distributed Q-learning framework. The crane, as an independent intelligent agent, conducts collaborative learning through a shared experience pool, and dynamically optimizes the path planning strategy based on environmental perception data and construction progress requirements. When the environment suddenly changes, it can quickly converge to the optimal path. At the same time, a fuzzy sliding mode control algorithm is developed in combination with the LSTM-Transformer crane trolley and car walking speed, hook lifting speed, and wire rope disturbance prediction model to calculate the crane motion compensation parameters. Based on the predicted mobile walking speed, lifting speed, and wire rope disturbance data, the optimal control instructions are generated in advance to shorten the equipment response delay. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is a system block diagram of a BIM-based hoisting construction supervision and optimization management system proposed in this invention. DETAILED DESCRIPTION

[0045] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0046] See also Figure 1 As shown, the present invention is a BIM-based hoisting construction supervision optimization management system, including:

[0047] Perception layer: 5G+ edge computing nodes are deployed to collect real-time heterogeneous data from multiple sources, including millimeter-wave radar (300m detection range, 50Hz refresh rate), UWB+ lidar (3D positioning accuracy of 2cm), micro lidar weather stations (10Hz sampling frequency), and fiber Bragg grating sensor networks (strain resolution of 1με, tilt measurement accuracy of 0.01°). A Kalman filter-particle filter fusion algorithm is used to construct a spatiotemporal map of dynamic obstacles and align the spatiotemporal data of multiple sensors. Furthermore, a fiber Bragg grating sensor network is embedded and a digital twin mirror interface is deployed to enable real-time monitoring of strain, tilt, and steel structure stress in key crane components, as well as synchronization of BIM models with physical entity states.

[0048] Decision-making layer: The bottom layer uses an incremental RRT# algorithm, combined with a non-uniform sampling strategy and an intelligent pruning mechanism, to generate candidate paths that meet the crane's kinematic constraints. The upper layer embeds a distributed Q-learning framework. The crane acts as an independent intelligent agent and conducts collaborative learning through a shared experience pool. It dynamically optimizes the path planning strategy based on environmental perception data and construction progress requirements. At the same time, a fuzzy sliding mode control algorithm is developed, combined with an LSTM-Transformer crane trolley and car travel speed, hook lifting speed, and wire rope disturbance prediction model to calculate the crane motion compensation parameters. Model predictive control (MPC) is introduced to generate optimal control instructions in advance based on the predicted mobile travel speed, lifting speed, and wire rope disturbance data. A physical constraint neural network is constructed, and CFD simulation results are used as supervision signals to train a 3D-CNN+Graph Transformer model. The model parameters are dynamically updated based on real-time monitoring data.

[0049] Execution layer: The path planning instructions, lifting speed, and wire rope disturbance compensation control instructions generated by the decision layer are sent to construction equipment such as cranes and elevators via the OPC UA over TSN protocol. A dual PLC hot backup safety redundancy mechanism is deployed. When the main controller fails, the backup controller performs failover.

[0050] Optimization layer: A BIM model was constructed to integrate geometric information, construction progress, structural response, and equipment status. Extreme working conditions were simulated through a digital twin test bed, and equipment execution results and structural safety monitoring data were fed back to the decision-making layer to update the environmental perception model, path planning algorithm, moving walking and lifting speed, and wire rope disturbance compensation control strategy.

[0051] In one embodiment, the multi-source heterogeneous data collected by the perception layer include environmental perception data, structural perception data, equipment status data, and personnel and safety data. The environmental perception data includes millimeter wave radar data (through the deployed millimeter wave radar (detection distance up to 300m, update frequency up to 50Hz), real-time acquisition of dynamic information such as distance, speed, direction, etc. of obstacles around the construction site), UWB+ lidar data (combining ultra-wideband (UWB) positioning technology and the three-dimensional scanning capability of lidar to obtain high-precision three-dimensional positioning data) and micro lidar weather station data (the micro lidar weather station deployed at the construction site monitors meteorological parameters such as wind speed, wind direction, temperature, humidity, etc. in real time). It includes fiber grating sensor network data (the fiber grating sensor network embedded in the key components of the crane monitors the strain, tilt and stress of the components and other state parameters of the steel structure in real time) and digital twin mirror interface data (through the digital twin mirror interface, the BIM model and the status data of the physical entity (such as crane, elevator, etc.) are synchronized in real time, including geometric information, construction progress, equipment status, etc.). The equipment status data includes crane operation status data (including the crane's moving speed, lifting speed, boom acceleration, lifting height, load weight and other operating parameters) and elevator operation status data (elevator operation speed, load capacity, position, etc.). The personnel and safety data includes personnel positioning data and security monitoring video data.

[0052] In one embodiment, the perception layer uses the Kalman filter-particle filter fusion algorithm to perform the following steps to construct a spatiotemporal map of dynamic obstacles and align the spatiotemporal data of multiple sensors:

[0053] Data preprocessing: Through data cleaning and data format unification, using threshold or statistical methods to detect and eliminate outliers, and defining a unified data structure, the raw sensor data is preliminarily processed to remove noise and outliers, and the data from different sensors is converted into a unified format;

[0054] Kalman filter: Through two steps of prediction and update, the state of the system is recursively estimated using the state transfer matrix and the observation matrix, and the optimal state estimation is performed for the linear Gaussian system. Prediction: State prediction: Use the state transfer matrix F and the optimal estimate of the previous moment To predict the current state Covariance prediction: using the state transfer matrix F and the covariance P of the previous moment k-1 To predict the covariance P at the current moment k :P k =FP k-1 F T +Q, where Q is the covariance matrix of the process noise; Update: Calculate the Kalman gain: Use the predicted covariance Pk and the covariance R of the observation noise to calculate the Kalman gain K k :K k =P k H T (HP k H T +R) -1 , where H is the observation matrix; state update: use Kalman gain K k , predicted value z k and predicted status To update the optimal state estimate at the current moment Covariance update: using Kalman gain K k and the predicted covariance P k To update the covariance Pk at the current moment: Pk=(IK k H)P k , where I is the identity matrix; Iteration: The optimal state estimate at the current moment is The and covariance Pk are used as the initial state estimate and covariance at the next moment, and the prediction and update are repeated;

[0055] Particle filtering: It uses a set of random samples (particles) to represent the state distribution of the system, and approximates the true posterior distribution through resampling and weight updating to complete the state estimation of nonlinear non-Gaussian systems.

[0056] Fusion: Use Kalman filtering for state estimation of the linear part and particle filtering for state estimation of the nonlinear part. The results of the two filters are fused through weighted averaging or interactive multi-model algorithm.

[0057] In one embodiment, the specific steps of the perception layer performing digital twin image synchronization are:

[0058] Sensor deployment: Fiber Bragg grating sensor networks are deployed on key components of the crane, such as the main beam, trolley beam, and winch drum;

[0059] Integration of digital twin mirror interface and BIM model: Assign a unique identifier to each component of the BIM model, and also assign a unique identifier to each sensor in the sensor network. By establishing a mapping relationship between the identifiers, the sensor data is associated with the BIM model components. When the sensor data changes, the data is synchronized to the BIM model through the interface, driving the update of information such as the geometry, position and attributes of the model components.

[0060] In one embodiment, the bottom layer utilizes an incremental RRT# algorithm, combined with a non-uniform sampling strategy and an intelligent pruning mechanism, to generate candidate paths that satisfy the kinematic constraints of the crane, including the following steps:

[0061] Initialization: Set the starting point (ground) and target point (50th floor) in the BIM model;

[0062] Non-uniform sampling: Based on the risk heat map (generated based on historical data, expert knowledge or real-time monitoring data, reflecting the risk level of different locations), sampling is performed in key areas. The sampling probability P(x) is calculated based on the risk value R(x) and obstacle density D(x) at location x. The calculation formula is expressed as P(x) = (R(x)·D(x)) / (∫ Ω R(x′)·D(x′)dx′), where Ω is the sampling space and the integral term is used to normalize the probability;

[0063] Path extension: Using the incremental RRT# algorithm, the path is randomly extended from the starting point, and the kinematic constraints of the crane are considered in each extension;

[0064] Intelligent pruning: removes redundant paths and retains paths with lower path costs and that satisfy kinematic constraints;

[0065] Path optimization: Smoothe the retained path to reduce angle mutations and acceleration fluctuations;

[0066] Output candidate paths: Output candidate paths that satisfy all constraints.

[0067] In one embodiment, the upper layer embeds a distributed Q-learning framework, and the crane acts as an independent intelligent agent, performs collaborative learning through a shared experience pool, and dynamically optimizes the path planning strategy based on environmental perception data and construction progress requirements, including the following steps:

[0068] Initialization: The agent obtains the current state information from the BIM model and initializes the Q value table;

[0069] Experience collection: Each agent collects experience during the interaction process and stores it in a shared experience pool;

[0070] Collaborative learning: The agent randomly extracts a batch of experiences from the experience pool for learning, and uses the Q-learning algorithm to update the Q-value table based on the extracted experiences;

[0071] Path planning: Based on the learned Q-value table, the agent selects the optimal action for path planning to avoid collisions and conflicts;

[0072] Dynamic adjustment: During the lifting process of large units, the intelligent agent selects the optimal action based on environmental perception data (such as wind speed changes) and construction progress requirements (such as the construction progress of the generator layer or the adjustment of the installation progress of large units). The intelligent agent executes the selected action and updates its own position. Based on the results of the action, the intelligent agent receives reward signals from the environment.

[0073] In one embodiment, the fuzzy sliding mode control algorithm is developed in combination with the LSTM-Transformer crane trolley and carriage travel speed, hook lifting speed, and wire rope disturbance prediction model to calculate the crane motion compensation parameters, introduce model predictive control (MPC), and generate optimal control instructions in advance based on the predicted mobile travel speed, lifting speed, and wire rope disturbance data, including the following steps:

[0074] Fuzzy sliding mode control algorithm: The deviation between the system state and the desired state is represented by the sliding surface function s(t), which is expressed as Where, e(t) is the position or angle error, is the error change rate, c is the design parameter, and the control law u(t) is designed to make the system state reach the sliding surface in a finite time and move to the equilibrium point along the sliding surface. The control law is expressed as u(t) = u eq (t)+u sw (t), where u eq (t) is the equivalent control term, u sw (t) is the switching control term, and fuzzy logic is used to adjust the gain of the switching control term to adapt to different working conditions, thereby effectively handling the uncertainty and nonlinearity in the crane system;

[0075] Build an LSTM-Transformer model to predict the travel speed of the crane trolley and trolley, the hook lifting speed, and wire rope disturbances. The LSTM layer processes time series data to capture short-term changes in wind speed and direction. The Transformer encoder processes spatial data to capture the spatial correlation between lifting speeds at different locations and wire rope disturbances. The model is trained using historical data on travel speed, lifting speed, and wire rope disturbances, and the loss function is optimized to enable the model to predict wire rope disturbance data for a period of time in the future.

[0076] Integrated Model Predictive Control (MPC): Based on the LSTM-Transformer crane trolley and carriage travel speed, hook lifting speed, and wire rope disturbance prediction model, rolling optimization and feedback correction, the current mobile travel speed and lifting speed data are input, and the future wire rope disturbance prediction value is output. In each control cycle, the finite time domain optimal control problem is solved, and the optimal control problem is solved using optimization algorithms (such as gradient descent and genetic algorithm) to obtain the optimal control sequence. The first element of the optimal control sequence is applied to the crane system. The accuracy of the prediction model is ensured through feedback correction, so that the optimal control instructions are generated in advance.

[0077] In one embodiment, the OPC UA over TSN protocol is used to send data to construction equipment such as cranes and construction elevators, including the following steps:

[0078] Instruction encoding and encapsulation: Encode the path planning results and compensation parameters into OPC UA data structures;

[0079] OPC UA over TSN network transmission: Build an OPC UA-based construction equipment information model, define the nodes and methods of equipment, sensors, and actuators, configure TSN network time synchronization and traffic scheduling parameters, and send encoded instructions to construction equipment via the OPC UA over TSN protocol;

[0080] Device-side instruction decoding and execution: An OPC UA client is set up on the construction equipment to receive and decode instructions from the decision-making layer. The OPC UA client parses the instruction parameters in the OPC UA data structure based on the specific model and configuration of the equipment.

[0081] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A BIM-based hoisting construction supervision optimization management system, characterized by: include: Perception layer: 5G+ edge computing nodes are deployed to collect real-time heterogeneous data from multiple sources, including millimeter-wave radar, UWB+ lidar, micro lidar weather stations, and fiber Bragg grating sensor networks. A Kalman filter-particle filter fusion algorithm is used to construct a spatiotemporal map of dynamic obstacles and align the spatiotemporal data of multiple sensors. Furthermore, a fiber Bragg grating sensor network is embedded and a digital twin mirror interface is deployed to enable real-time monitoring of strain, tilt, and steel structure stress in key crane components, as well as synchronization of BIM models with physical entity states. Decision-making layer: The bottom layer uses an incremental RRT# algorithm, combined with a non-uniform sampling strategy and an intelligent pruning mechanism, to generate candidate paths that meet the crane's kinematic constraints. The upper layer embeds a distributed Q-learning framework. Each crane acts as an independent intelligent agent and conducts collaborative learning through a shared experience pool. The path planning strategy is dynamically optimized based on environmental perception data and construction schedule requirements. At the same time, a fuzzy sliding mode control algorithm is developed, combined with an LSTM-Transformer crane trolley and car travel speed, hook lifting speed, and wire rope disturbance prediction model to calculate the crane motion compensation parameters. Model predictive control is introduced to generate optimal control instructions in advance based on predicted movement, lifting speed, and wire rope disturbance data. Execution layer: The path planning instructions and wind disturbance compensation control instructions generated by the decision layer are sent to the construction equipment via the OPC UA over TSN protocol. Optimization layer: Build a BIM model, integrate geometric information, construction progress, structural response and equipment status, and feed back equipment execution results and structural safety monitoring data to the decision-making layer.

2. The BIM-based hoisting construction supervision optimization management system according to claim 1 is characterized in that: The multi-source heterogeneous data collected by the perception layer include environmental perception data, structural perception data, equipment status data, and personnel and safety data. The environmental perception data includes millimeter wave radar data, UWB+ lidar data, and micro lidar weather station data. The structural perception data includes fiber grating sensor network data and digital twin mirror interface data. The equipment status data includes crane operation status data and elevator operation status data. The personnel and safety data includes personnel positioning data and security monitoring video data.

3. The BIM-based hoisting construction supervision and optimization management system according to claim 1 is characterized in that: The perception layer uses the Kalman filter-particle filter fusion algorithm to construct a spatiotemporal map of dynamic obstacles and align the spatiotemporal data of multiple sensors. Data preprocessing: Through data cleaning and data format unification, using threshold or statistical methods to detect and eliminate outliers, and defining a unified data structure, the raw sensor data is preliminarily processed to remove noise and outliers, and the data from different sensors is converted into a unified format; Kalman filter: Through two steps of prediction and update, the state of the system is recursively estimated using the state transfer matrix and the observation matrix to perform optimal state estimation for linear Gaussian systems; Particle filter: It uses a set of random samples to represent the state distribution of the system, and approximates the true posterior distribution through resampling and weight updating to complete the state estimation of nonlinear non-Gaussian systems; Fusion: Use Kalman filtering to estimate the state of the linear part and particle filtering to estimate the state of the nonlinear part, and fuse the results of the two filters through weighted averaging or interactive multi-model algorithm.

4. The BIM-based hoisting construction supervision optimization management system according to claim 1 is characterized in that: The specific steps of the perception layer for digital twin mirror synchronization are: Sensor deployment: Deploy fiber grating sensor networks on key components of the crane; Integration of digital twin mirror interface and BIM model: Assign a unique identifier to each component of the BIM model, and also assign a unique identifier to each sensor in the sensor network. By establishing a mapping relationship between the identifiers, the sensor data is associated with the BIM model components. When the sensor data changes, the data is synchronized to the BIM model through the interface to drive the update of the model component information.

5. The BIM-based hoisting construction supervision optimization management system according to claim 1 is characterized in that: The bottom layer utilizes the incremental RRT# algorithm, combined with a non-uniform sampling strategy and an intelligent pruning mechanism, to generate candidate paths that satisfy the kinematic constraints of the crane, including the following steps: Initialization: Set the starting point and target point in the BIM model; Non-uniform sampling: Based on the risk heat map, sampling is performed in key areas. The sampling probability P(x) is calculated based on the risk value R(x) at position x and the obstacle density D(x). The calculation formula is P(x) = (R(x)·D(x)) / (∫ Ω R(x′)·D(x′)dx′), where Ω is the sampling space and the integral term is used to normalize the probability; Path extension: Using the incremental RRT# algorithm, the path is randomly extended from the starting point, and the kinematic constraints of the crane are considered in each extension; Intelligent pruning: removes redundant paths and retains paths with lower path costs and that satisfy kinematic constraints; Path optimization: smoothing the retained path; Output candidate paths: Output candidate paths that satisfy all constraints.

6. The BIM-based hoisting construction supervision optimization management system according to claim 5 is characterized in that: The upper layer embeds a distributed Q-learning framework. The crane acts as an independent intelligent agent and conducts collaborative learning through a shared experience pool. It dynamically optimizes the path planning strategy based on environmental perception data and construction progress requirements, including the following steps: Initialization: The agent obtains the current state information from the BIM model and initializes the Q value table; Experience collection: Each agent collects experience during the interaction process and stores it in a shared experience pool; Collaborative learning: The agent randomly extracts a batch of experiences from the experience pool for learning, and uses the Q-learning algorithm to update the Q-value table based on the extracted experiences; Path planning: Based on the learned Q-value table, the agent selects the optimal action for path planning; Dynamic adjustment: During the lifting process of large units, the intelligent agent selects the optimal action based on environmental perception data and construction progress requirements (such as the adjustment of the construction progress of a large unit on the generator floor or a certain unit). The intelligent agent executes the selected action and updates its own position. Based on the results of the action, the intelligent agent receives reward signals from the environment.

7. The BIM-based hoisting construction supervision optimization management system according to claim 6 is characterized in that: The developed fuzzy sliding mode control algorithm combines the LSTM-Transformer crane trolley and carriage travel speeds, hook lifting speeds, and wire rope disturbance prediction models to calculate crane motion compensation parameters. Model predictive control is introduced to generate optimal control instructions in advance based on the predicted mobile travel speed, lifting speed, and wire rope disturbance data, including the following steps: Fuzzy sliding mode control algorithm: The sliding surface function s(t) is defined to represent the deviation between the system state and the desired state. The sliding surface function is expressed as Where, e(t) is the position or angle error, is the error change rate, c is the design parameter, and the control law u(t) is designed to make the system state reach the sliding surface in a finite time and move to the equilibrium point along the sliding surface. The control law is expressed as u(t) = u eq (t)+u sw (t), where u eq (t) is the equivalent control term, u sw (t) is the switching control item, and fuzzy logic is used to adjust the gain of the switching control item to adapt to different working conditions; Build an LSTM-Transformer model to predict the travel speed of the crane trolley and crane carriage, the hook lifting speed, and wire rope disturbances. The LSTM layer processes time series data to capture short-term changes in movement, lifting or lowering speed, and wire rope disturbances. The Transformer encoder processes spatial data to capture the spatial correlation between movement speed, lifting speed, and wire rope disturbances at different locations. The model is trained using historical crane movement speed, lifting speed, and wire rope disturbance data, and the loss function is optimized to predict the lifting data of a specific large object over a period of time in the future. Integrated model predictive control: Based on the LSTM-Transformer crane trolley and carriage travel speed, hook lifting speed, and wire rope disturbance prediction model, rolling optimization and feedback correction, the system inputs the current lifting speed data and outputs the predicted values ​​of future lifting speed and wire rope disturbance. In each control cycle, the finite time domain optimal control problem is solved and the optimal control problem is solved using the optimization algorithm to obtain the optimal control sequence. The first element of the optimal control sequence is applied to the crane system, thereby generating the optimal control instructions in advance.

8. The BIM-based hoisting construction supervision and optimization management system according to claim 7 is characterized in that: The OPC UA over TSN protocol is used to send data to the construction equipment, including the following steps: Instruction encoding and encapsulation: Encode the path planning results and compensation parameters into OPC UA data structures; OPC UA over TSN network transmission: Build an OPC UA-based construction equipment information model, define the nodes and methods of equipment, sensors, and actuators, configure the time synchronization and traffic scheduling parameters of the TSN network, and send the encoded instructions to the construction equipment using the OPC UA over TSN protocol; Device-side instruction decoding and execution: An OPC UA client is set up on the construction equipment to receive and decode instructions from the decision-making layer. The OPC UA client parses the instruction parameters in the OPC UA data structure based on the specific model and configuration of the equipment.

Images