A visual working monitoring method for tower crane based on digital twin

Abstract

The present invention relates to the technical field of crane work monitoring, and discloses a visual work monitoring method for a tower crane based on digital twins, comprising the following steps: S1, establishing a digital twin model of a tower crane, the digital twin model comprising a geometric model, a physical model and an environmental model, wherein the geometric model is constructed by a three-dimensional modeling tool; S2, collecting the working status data of the tower crane through a multi-sensor network, wherein the sensor comprises a force sensor, a position sensor, an acceleration sensor and an environmental sensor. Through the real-time collection of digital twin technology and multi-sensor data, a three-dimensional dynamic model of a tower crane can be constructed and displayed, and the operator can understand the working status of the equipment in real time through the system and obtain intelligent operation suggestions. If an abnormality is detected, the system automatically issues an early warning to prevent the equipment from malfunctioning due to problems such as overload and abnormal vibration, thereby significantly improving the working safety of the tower crane.

Description

Technical Field

[0001] The present invention relates to the technical field of crane operation monitoring, and in particular to a visual operation monitoring method for a tower crane based on digital twins. Background Art

[0002] Tower cranes are widely used lifting equipment in modern construction sites. They have become key equipment on construction sites because of their flexible operation and strong lifting capacity. As tower cranes are increasingly used, problems such as the safety of lifting equipment in actual work, monitoring of equipment status and prevention of faults have been exposed.

[0003] The existing tower crane monitoring system relies on sensor alarms and is unable to build a real-time three-dimensional model of the equipment. The operator can only obtain crane status information through local sensor data, resulting in the inability to intuitively display the overall operating status of the crane. In addition, the existing technology lacks intelligent operation suggestions. If an abnormal situation occurs, it relies on manual intervention and judgment, which increases the burden on the operator, easily leads to misoperation, and increases the risk of accidents.

[0004] In the comparative document CN115303946B, monitoring is carried out through digital twins combined with sensors, but the focus is on the digital model of the construction site and dynamic object monitoring, and there is a lack of real-time three-dimensional modeling and visualization of the tower crane equipment itself.

[0005] Comparative document CN115303946B performs real-time positioning and correlation operations on the positions of tower cranes and their components, but does not make effective monitoring and predictions on the wear of key components of tower cranes, making it impossible to carry out fault prevention and maintenance plans in advance; in comparative document CN115303946B, dynamic object positioning and correlation detection with the position of the crane are mentioned, but in the face of changing construction site environments, such as extreme weather and complex working conditions, there is a lack of comprehensive risk management methods, and it fails to provide multi-dimensional real-time safety monitoring and active emergency response mechanisms, and there are potential safety management defects.

[0006] Therefore, those skilled in the art provide a visual work monitoring method for tower cranes based on digital twins to solve the problems raised in the above background technology. Summary of the invention

[0007] In view of the shortcomings of the prior art, the present invention provides a visual work monitoring method for a tower crane based on digital twin to solve the problems raised in the above background technology.

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions: A tower crane visual work monitoring method based on digital twins, comprising the following steps:

[0009] S1. Establish a digital twin model of a tower crane, where the digital twin model includes a geometric model, a physical model and an environmental model, and the geometric model is constructed by a three-dimensional modeling tool;

[0010] S2. Collecting the working status data of the tower crane through a multi-sensor network, wherein the sensors include force sensors, position sensors, acceleration sensors and environmental sensors;

[0011] S3. Process sensor data in real time through the Kalman filter algorithm, and dynamically monitor the working state of the tower crane based on the data stream collected by the sensor, so as to detect whether the key data of the tower crane deviates from the normal working state and respond to abnormal situations;

[0012] Assume that the normal mean of the sensor data stream is , the actual observed value is x, then the mean shift detection formula is:

[0013] ,

[0014] in, is the drift of the mean, is the sensor data at the i-th moment, is the expected mean value under normal working conditions, is the number of observed samples;

[0015] If the mean drift exceeds a certain set threshold T, it is judged as abnormal: >T triggers abnormal detection, where T is the threshold set by the system based on historical data and models. Exceeding this threshold means that the working status of the crane is abnormal, and the system will trigger an alarm or operation suggestion;

[0016] S4. Use machine learning algorithms to perform intelligent analysis and fault prediction on processed data;

[0017] S5. The intelligent visualization system displays the spatial position data of the tower crane boom in real time and provides operation suggestions and early warning information;

[0018] S6. Generate predictive maintenance plans based on failure prediction and provide emergency response management.

[0019] Preferably, in step S1, the physical model includes:

[0020] The mechanical model is expressed by the following formula:

[0021] ,

[0022] in, is the force on the boom, is the mass of the boom, is the acceleration;

[0023] The torque model is used to describe the rotational motion of the slewing platform of the tower crane and satisfies the following formula:

[0024] ,

[0025] in, is the torque on the rotary platform, is the moment of inertia, is the angular acceleration;

[0026] The vibration model is used to describe the elastic vibration behavior of the boom when lifting weights. The model is a second-order differential equation:

[0027] ,

[0028] in, is the mass of the boom, is the damping coefficient, is the stiffness coefficient, is the time-varying external force, is the vibration displacement.

[0029] Preferably, in the step S1, the environmental model is used to describe the dynamic characteristics of the tower crane operating environment. The following is a modeling formula of the environmental model based on physical and statistical methods:

[0030] ,

[0031] in, is the environmental state at time t, is the functional relationship of environmental factors, are the wind speed and direction at time t, is the ambient temperature at time t, is the humidity at time t, is the vibration amplitude at time t, are other environmental factors at time t, is the model error.

[0032] Preferably, in step S3, the sensor data is processed in real time by a Kalman filter algorithm, and the Kalman filter processing steps include:

[0033] Status prediction: ,

[0034] Measurement Update: ,

[0035] in, is the state prediction at time k, is the state transfer matrix, is the input quantity, is the Kalman gain, is the measured value, is the measurement matrix, B is the control input matrix, For at the moment Status The estimated value of is the state at time k The estimated value of .

[0036] Preferably, in step S6, when an abnormality is detected in the predictive maintenance plan based on fault prediction, the system provides operation suggestions to the operator through an intelligent visual interface, and the system needs to use an algorithm to analyze the working status of the crane;

[0037] The optimization algorithm generates adjustment measures based on the working status of the crane, and the constraint optimization model is used to calculate the operation suggestions:

[0038] Assume the objective function is , x represents the current operating parameters, then the objective function is expressed as:

[0039] ,

[0040] in, is the risk function of the crane in its current state, is the operation energy consumption function, and is the weight parameter of risk and energy consumption;

[0041] Constraints:

[0042] Rotation Angle Limit: , lifting speed limit: ,

[0043] in, and is the minimum and maximum value of the boom rotation angle, and is the minimum and maximum value of the lifting speed;

[0044] Risk Function It is defined by the following formula:

[0045] ,

[0046] in, It's wind power. is the lifting weight, is the current rotation angle and the reference angle The deviation between and is the weight parameter for adjusting the risk function;

[0047] Energy consumption function The energy consumption of the crane under different operating parameters is expressed by the following formula:

[0048] ,

[0049] in, is the power consumption of the crane, The time to complete the operation under the current operating parameters.

[0050] Preferably, in step S6, the predictive maintenance plan for fault prediction is based on a support vector machine and a long short-term memory network algorithm. The long short-term memory network is used to analyze time series data and predict the long-term behavior of the tower crane. The prediction formula is: ,

[0051] in, is the current cell state, For the Gate of Forgetfulness, is the input gate, is the weight matrix, For the previous moment The cell state, is the hyperbolic tangent function, For the previous moment The hidden state of is the input at the current time t, is the bias term.

[0052] Preferably, in the step S5, the intelligent visualization system is constructed using WebGL or Unity 3D technology, and can display the three-dimensional model of the tower crane and its working status in real time.

[0053] Preferably, in step S5, the spatial position data of the tower crane boom, including the rotation angle of the boom, is displayed in real time through an intelligent visualization system. , elevation angle and stretch length , based on the spatial position data of the boom, the boom position in the three-dimensional model is calculated and updated in real time. The specific formula is:

[0054] ,

[0055] ,

[0056] ,

[0057] in, is the rotation angle of the boom, is the elevation angle of the boom, is the extended length of the boom;

[0058] Real-time position update of the crane arm and real-time data collected by sensors , , Will be used to update the position of the boom in the 3D visualization model in real time:

[0059] ,

[0060] ,

[0061] ,

[0062] in, Used to update the current position of the boom in the 3D model.

[0063] Preferably, in step S5, the force distribution of the boom is calculated in real time by an intelligent visualization system, and a thermal diagram of the force distribution is displayed:

[0064] The force distribution of the boom is calculated by the following formula:

[0065] ,

[0066] in, is the force at a certain point x on the boom, For lifting weight, is the total length of the boom.

[0067] Preferably, the thermal map shows: Converted into color mapping, the intelligent visualization system is used to display the real-time state of the crane arm force in the form of a heat map. The color mapping formula is defined as:

[0068] ,

[0069] in, is the color value corresponding to the load distribution, and are the minimum and maximum values ​​of the force, and The minimum and maximum values ​​for the color map.

[0070] The present invention provides a tower crane visual work monitoring method based on digital twins, which has the following beneficial effects:

[0071] 1. The present invention can construct and display a three-dimensional dynamic model of a tower crane through digital twin technology and real-time collection of multi-sensor data. The operator can understand the working status of the equipment in real time through the system and obtain intelligent operation suggestions. If an abnormality is detected, the system will automatically issue an early warning to prevent the equipment from malfunctioning due to overload, abnormal vibration and other problems, thereby significantly improving the working safety of the tower crane. At the same time, the system's data-driven operation optimization function can maximize the equipment's working efficiency while ensuring safety.

[0072] 2. The present invention combines real-time data with historical data, uses machine learning algorithms to predict faults, and generates predictive maintenance plans to monitor and predict the wear of key components of tower cranes. The system can identify potential faults in advance and reduce unplanned downtime of equipment. Compared with traditional post-maintenance methods, predictive maintenance effectively reduces maintenance costs, extends the service life of equipment, and improves the full life cycle management capabilities of cranes.

[0073] 3. The present invention integrates multiple sensor data and external environment data to build a multi-dimensional safety management system, so that the system can dynamically monitor the working status of the tower crane in a complex environment, and automatically trigger the emergency response mechanism when potential risks are detected, and automatically shut down or alarm. In addition, combined with intelligent visualization and augmented reality technology, the operator can intuitively understand the operating conditions and safety risks of the equipment, effectively reduce the safety hazards caused by human errors, and significantly improve the working safety and reliability of the tower crane. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0075] In order to make the technical personnel in the technical field understand the scheme of the present invention, the technical scheme in the embodiment of the present invention will be clearly and completely described below in combination with the drawings in the embodiment of the present invention. Obviously, the described embodiment is a partial embodiment of the present invention, not a complete embodiment. Based on the embodiment of the present invention, other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present invention.

[0076] The present invention is described in detail below in conjunction with the accompanying drawings:

[0077] Example:

[0078] Please see attached Figure 1 , an embodiment of the present invention provides a tower crane visual work monitoring method based on digital twin, comprising the following steps:

[0079] S1. Establish a digital twin model of the tower crane. The digital twin model includes a geometric model, a physical model and an environmental model. The geometric model is constructed using a 3D modeling tool.

[0080] S2. Collecting the working status data of the tower crane through a multi-sensor network, wherein the sensors include force sensors, position sensors, acceleration sensors and environmental sensors;

[0081] S3. Process sensor data in real time through the Kalman filter algorithm, and dynamically monitor the working state of the tower crane based on the data stream collected by the sensor, so as to detect whether the key data of the tower crane deviates from the normal working state and respond to abnormal situations;

[0082] Assume that the normal mean of the sensor data stream is , the actual observed value is x, then the mean shift detection formula is:

[0083] ,

[0084] in, is the drift of the mean, is the sensor data at the i-th moment, is the expected mean value under normal working conditions, is the number of observed samples;

[0085] If the mean drift exceeds a certain set threshold T, it is judged as abnormal: >T triggers abnormal detection, where T is the threshold set by the system based on historical data and models. Exceeding this threshold means that the working status of the crane is abnormal, and the system will trigger an alarm or operation suggestion;

[0086] S4. Use machine learning algorithms to perform intelligent analysis and fault prediction on processed data;

[0087] S5. The intelligent visualization system displays the spatial position data of the tower crane boom in real time and provides operation suggestions and early warning information;

[0088] S6. Generate predictive maintenance plans based on failure prediction and provide emergency response management.

[0089] Benefits of S1: It can build an accurate virtual copy of the tower crane, realize the synchronous mapping of the physical equipment and the virtual model, so that the working status of the equipment can be intuitively presented in the virtual environment, provide global visual monitoring of the equipment operation, and help improve the operator's real-time control of the equipment status;

[0090] Benefits of S2: High-precision operation information can be obtained, including boom position, load, vibration and external environment data. Multi-source data provides comprehensive equipment operation information, laying the foundation for subsequent analysis and monitoring.

[0091] The benefits of S3 are that it can process data in real time on local devices, reduce data transmission delays, and improve the accuracy and stability of sensor data through denoising and data fusion technology. At the same time, real-time anomaly detection can promptly identify abnormal conditions of equipment and issue early warnings before problems occur, preventing equipment damage or safety accidents.

[0092] Benefits of S4 Intelligent analysis of historical and real-time data of equipment to identify potential failure trends and abnormal situations. Based on situation analysis, the system can make early failure predictions to help operators take preventive measures before problems occur, reduce the incidence of equipment failures, and reduce unplanned downtime;

[0093] The benefits of S5 allow operators to intuitively understand the operation of the equipment. In addition, the system can provide operation suggestions and early warning information based on the equipment status, helping operators make quick and accurate decisions and avoid human error.

[0094] The benefits of S6 are that it can generate customized predictive maintenance plans, arrange maintenance and care of key components in advance, and avoid downtime and repair costs caused by sudden failures. At the same time, the system can provide emergency response management and automatically take measures in the event of major risks to effectively ensure the safety of equipment and personnel.

[0095] In step S1, the physical model includes:

[0096] The mechanical model is expressed by the following formula:

[0097] ,

[0098] in, is the force on the boom, is the mass of the boom, is the acceleration;

[0099] The torque model is used to describe the rotational motion of the slewing platform of the tower crane and satisfies the following formula:

[0100] ,

[0101] in, is the torque on the rotary platform, is the moment of inertia, is the angular acceleration;

[0102] The vibration model is used to describe the elastic vibration behavior of the boom when lifting weights. The model is a second-order differential equation:

[0103] ,

[0104] in, is the mass of the boom, is the damping coefficient, is the stiffness coefficient, is the time-varying external force, is the vibration displacement.

[0105] In step S1, the environmental model is used to describe the dynamic characteristics of the tower crane operating environment. The following is the modeling formula of the environmental model based on physical and statistical methods:

[0106] ,

[0107] in, is the environmental state at time t, is the functional relationship of environmental factors, are the wind speed and direction at time t, is the ambient temperature at time t, is the humidity at time t, is the vibration amplitude at time t, are other environmental factors at time t, is the model error.

[0108] The function of the mechanical model formula is to calculate the force F on the boom of the tower crane, where m is the mass of the boom and a is the acceleration of the boom. The mechanical model formula describes the external force on the boom when it is lifting or moving. It is a basic mechanical equation used to analyze the motion state of the boom. Through this formula, the system can calculate the force on the boom under different loads and operating conditions, help predict the risk of stress overload or structural deformation, and thus ensure the safety and stability of the equipment;

[0109] The torque model formula describes the rotational motion of the slewing platform of a tower crane, where is the torque of the slewing platform, I is the moment of inertia of the platform, and α is the angular acceleration. Through the torque model formula, the system can calculate the torque of the tower crane during slewing. When the platform is overloaded or rotating at high speed, monitoring the change of torque can effectively prevent the platform from being overloaded or unstable. It plays an important role in optimizing the control of the slewing angle and speed during operation, ensuring the stability of the platform during rotation.

[0110] The vibration model formula describes the elastic vibration behavior of the tower crane boom when lifting weights, where m is the mass of the boom, c is the damping coefficient, k is the stiffness coefficient, is the time-varying external force, x is the vibration displacement, and the vibration model formula is used to analyze the vibration characteristics of the boom when lifting and moving loads. Through real-time monitoring of the vibration state, the system can identify whether the boom is vibrating excessively or resonating, and then help optimize operating parameters, such as reducing the boom movement speed, adjusting the load position, and avoiding structural damage to the equipment caused by vibration.

[0111] In step S3, the sensor data is processed in real time by a Kalman filter algorithm, and the processing steps of the Kalman filter include:

[0112] Status prediction: ,

[0113] Measurement Update: ,

[0114] in, is the state prediction at time k, is the state transfer matrix, is the input quantity, is the Kalman gain, is the measured value, is the measurement matrix, B is the control input matrix, For at the moment Status The estimated value of is the state at time k The estimated value of .

[0115] The main function of the prediction formula is to predict the status of each component of the tower crane based on the known status at the previous moment and the current external input, provide a preliminary estimate of the current status, and lay the foundation for subsequent sensor data correction and noise filtering;

[0116] The main function of the measurement update formula is to correct the predicted state based on the sensor measurement value to generate a more accurate state estimate. The sensor measurement value usually contains noise, and the Kalman gain can determine how to weigh the predicted state and the sensor measurement value, thereby effectively filtering out the noise and obtaining a more accurate tower crane state.

[0117] In step S6, when the predictive maintenance plan based on fault prediction detects an abnormality, the system provides operation suggestions to the operator through an intelligent visual interface, and the system needs to use an algorithm to analyze the working status of the crane;

[0118] The optimization algorithm generates adjustment measures based on the working status of the crane, and the constraint optimization model is used to calculate the operation suggestions:

[0119] Assume the objective function is , x represents the current operating parameters, then the objective function is expressed as:

[0120] ,

[0121] in, is the risk function of the crane in its current state, is the operation energy consumption function, and is the weight parameter of risk and energy consumption;

[0122] Constraints:

[0123] Rotation Angle Limit: , lifting speed limit: ,

[0124] in, and is the minimum and maximum value of the boom rotation angle, and is the minimum and maximum value of the lifting speed;

[0125] Risk Function It is defined by the following formula:

[0126] ,

[0127] in, It's wind power. is the lifting weight, is the current rotation angle and the reference angle The deviation between and is the weight parameter for adjusting the risk function;

[0128] Energy consumption function The energy consumption of the crane under different operating parameters is expressed by the following formula:

[0129] ,

[0130] in, is the power consumption of the crane, The time to complete the operation under the current operating parameters.

[0131] The objective function is to optimize the operating parameters of the crane and generate the best operation suggestions by considering both the operation risk and energy consumption. The system calculates safe and efficient operation parameters by minimizing the objective function according to the current working status.

[0132] Constraints are used to ensure that the operation suggestions are within the physical limits. There are clear safety ranges for rotation and lifting speeds. During the optimization process, the system needs to avoid the calculated operation parameters exceeding the limits. Through constraints, the operation suggestions generated by the system optimize safety and energy consumption, and ensure the feasibility of actual operations.

[0133] The risk function mainly describes the risk of operation through two items: one is the ratio of wind force to load, and the other is the deviation of rotation angle. Excessive wind force or excessive deviation of rotation angle will lead to increased operation risk. By calculating this function, the system can quantify the current operation risk and then adjust the operation parameters to reduce the risk. In particular, when the wind speed is high or the load is heavy, it is recommended to reduce the rotation speed or adjust the angle.

[0134] The energy consumption function reflects the impact of operating parameters on the energy consumption of the crane. The power consumption of the crane will be different at different operating speeds and angles. Through this formula, the system can calculate the energy consumption required to complete the operation under different parameter combinations and minimize energy consumption during the optimization process. For example, a slower lifting speed reduces energy consumption, while a higher speed increases energy consumption. This formula ensures that the system considers safety and the energy efficiency of the equipment when generating operation recommendations.

[0135] In step S6, the predictive maintenance plan for fault prediction is based on the support vector machine and long short-term memory network algorithm. The long short-term memory network is used to analyze time series data and predict the long-term behavior of the tower crane. The prediction formula is:

[0136] ,

[0137] in, is the current cell state, For the Gate of Forgetfulness, is the input gate, is the weight matrix, For the previous moment The cell state, is the hyperbolic tangent function, For the previous moment The hidden state of is the input at the current time t, is the bias term.

[0138] The cell state in the LSTM network is the core part of the network's "memory", which is used to store long-term dependencies on information about past time steps. In tower crane failure prediction, this formula is used to maintain memory of the crane's historical state, which can capture long-term behavior patterns and make state predictions based on historical data;

[0139] The main function of the forget gate is to decide which historical information needs to be retained and which information can be ignored. In the state prediction of tower cranes, the forget gate can automatically determine whether to forget long-standing data.

[0140] The input gate determines the degree of influence of the input data at the current moment on the cell state. In the tower crane prediction, the input gate can decide whether to update the current prediction model based on the current state, ensuring that the model can make reasonable adjustments based on the latest data;

[0141] The output gate determines how the current state affects the output and prediction at the next moment. In the long-term state prediction of tower cranes, the output gate controls how to convert the memory in the cell state into the predicted output at the current moment. Through reasonable adjustment, the system can determine the changing trend of the future state based on the current state, and thus detect possible faults or abnormalities in advance.

[0142] In step S5, the intelligent visualization system is built using WebGL or Unity 3D technology, which can display the three-dimensional model of the tower crane and its working status in real time;

[0143] In step S5, the spatial position data of the tower crane boom, including the rotation angle of the boom, is displayed in real time through the intelligent visualization system. , elevation angle and stretch length , based on the spatial position data of the boom, the boom position in the three-dimensional model is calculated and updated in real time. The specific formula is:

[0144] ,

[0145] ,

[0146] ,

[0147] in, is the rotation angle of the boom, is the elevation angle of the boom, is the extended length of the boom;

[0148] Real-time position update of the crane arm and real-time data collected by sensors , , Will be used to update the position of the boom in the 3D visualization model in real time:

[0149] ,

[0150] ,

[0151] ,

[0152] in, Used to update the current position of the boom in the 3D model.

[0153] In step S5, the force distribution of the boom is calculated in real time through the intelligent visualization system, and a thermal diagram of the force distribution is displayed:

[0154] Calculation of the force distribution of the boom. The change of force distribution along the boom is calculated by the following formula:

[0155] ,

[0156] in, is the force at a certain point x on the boom, For lifting weight, is the total length of the boom;

[0157] Heat map display: Converted into color mapping, the intelligent visualization system is used to display the real-time state of the crane arm force in the form of a heat map. The color mapping formula is defined as:

[0158] ,

[0159] in, is the color value corresponding to the load distribution, and are the minimum and maximum values ​​of the force, and The minimum and maximum values ​​for the color map.

[0160] The function of the boom 3D space coordinate calculation formula is to convert the actual physical position of the tower crane boom into the space coordinates in the 3D visualization system in real time, so that the operator can see the exact position of the boom in the 3D model and track the movement trajectory of the boom. In actual operation, it is very important to monitor the dynamic changes of the boom such as rotation, extension and lifting.

[0161] The boom real-time position update formula plays a key role in real-time updating of the boom position, ensuring that the boom position in the 3D visualization system is always synchronized with the actual equipment, so that the operator can intuitively view the current working status of the tower crane and make operational decisions based on the actual situation to avoid collisions between the boom and obstacles;

[0162] The boom force distribution formula is used to calculate the gradient distribution of the boom force, showing the change in force from the boom root to the lifting point. It is essential for the operator to understand the force conditions of the boom under different loads and positions, helping to avoid overload operation or structural fatigue of the boom.

[0163] The function of the heat map color mapping formula is to visualize the force information at different positions on the boom through colors to generate a heat map. The heat map display helps the operator quickly identify uneven load distribution or local overload conditions and take corresponding operational measures.

[0164] Although 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 the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A tower crane visual work monitoring method based on digital twin, characterized in that: The steps include: S1. Establish a digital twin model of a tower crane, where the digital twin model includes a geometric model, a physical model and an environmental model, and the geometric model is constructed by a three-dimensional modeling tool; S2. Collecting the working status data of the tower crane through a multi-sensor network, wherein the sensors include force sensors, position sensors, acceleration sensors and environmental sensors; S3. Process sensor data in real time through the Kalman filter algorithm, and dynamically monitor the working state of the tower crane based on the data stream collected by the sensor, so as to detect whether the key data of the tower crane deviates from the normal working state and respond to abnormal situations; Assume that the normal mean of the sensor data stream is , the actual observed value is x, then the mean shift detection formula is: ， in, is the drift of the mean, is the sensor data at the i-th moment, is the expected mean value under normal working conditions, is the number of observed samples; If the mean drift exceeds a certain set threshold T, it is judged as abnormal: >T triggers abnormal detection, where T is the threshold set by the system based on historical data and models. Exceeding this threshold means that the working status of the crane is abnormal, and the system will trigger an alarm or operation suggestion; S4. Use machine learning algorithms to perform intelligent analysis and fault prediction on processed data; S5. The intelligent visualization system displays the spatial position data of the tower crane boom in real time and provides operation suggestions and early warning information; S6. Generate predictive maintenance plans based on failure prediction and provide emergency response management.

2. A tower crane visual work monitoring method based on digital twin according to claim 1, characterized in that: In the step S1, the physical model includes: The mechanical model is expressed by the following formula: ， in, is the force on the boom, is the mass of the boom, is the acceleration; The torque model is used to describe the rotational motion of the slewing platform of the tower crane and satisfies the following formula: ， in, is the torque on the rotary platform, is the moment of inertia, is the angular acceleration; The vibration model is used to describe the elastic vibration behavior of the boom when lifting weights. The model is a second-order differential equation: ， in, is the mass of the boom, is the damping coefficient, is the stiffness coefficient, is the time-varying external force, is the vibration displacement.

3. A tower crane visual work monitoring method based on digital twin according to claim 1, characterized in that: In the step S1, the environmental model is used to describe the dynamic characteristics of the tower crane operating environment. The following is the modeling formula of the environmental model based on physical and statistical methods: ， in, is the environmental state at time t, is the functional relationship of environmental factors, are the wind speed and direction at time t, is the ambient temperature at time t, is the humidity at time t, is the vibration amplitude at time t, are other environmental factors at time t, is the model error.

4. A tower crane visual work monitoring method based on digital twin according to claim 1, characterized in that: In step S3, the sensor data is processed in real time by a Kalman filter algorithm, and the processing steps of the Kalman filter include: Status prediction: , Measurement Update: , in, is the state prediction at time k, is the state transfer matrix, is the input quantity, is the Kalman gain, is the measured value, is the measurement matrix, B is the control input matrix, For at the moment Status The estimated value of is the state at time k The estimated value of .

5. The method for visualizing the operation of a tower crane based on digital twin according to claim 1 is characterized in that: In step S6, when the predictive maintenance plan based on fault prediction detects an abnormality, the system provides operation suggestions to the operator through an intelligent visual interface, and the system needs to use an algorithm to analyze the working status of the crane; The optimization algorithm generates adjustment measures based on the working status of the crane, and the constraint optimization model is used to calculate the operation suggestions: Assume the objective function is , x represents the current operating parameters, then the objective function is expressed as: ， in, is the risk function of the crane in its current state, is the operation energy consumption function, and is the weight parameter of risk and energy consumption; Constraints: Rotation Angle Limit: , lifting speed limit: , in, and is the minimum and maximum value of the boom rotation angle, and is the minimum and maximum value of the lifting speed; Risk Function It is defined by the following formula: ， in, It's wind power. is the lifting weight, is the current rotation angle and the reference angle The deviation between and is the weight parameter for adjusting the risk function; Energy consumption function The energy consumption of the crane under different operating parameters is expressed by the following formula: ， in, is the power consumption of the crane, The time to complete the operation under the current operating parameters.

6. The method for visualizing the operation of a tower crane based on digital twin according to claim 1 is characterized in that: In step S6, the predictive maintenance plan for fault prediction is based on the support vector machine and long short-term memory network algorithm. The long short-term memory network is used to analyze time series data and predict the long-term behavior of the tower crane. The prediction formula is: ， in, is the current cell state, For the Gate of Forgetfulness, is the input gate, is the weight matrix, For the previous moment The cell state, is the hyperbolic tangent function, For the previous moment The hidden state of is the input at the current time t, is the bias term.

7. The method for visualizing the operation of a tower crane based on digital twin according to claim 1 is characterized in that: In the step S5, the intelligent visualization system is constructed using WebGL or Unity 3D technology, and can display the three-dimensional model of the tower crane and its working status in real time.

8. The method for visualizing the operation of a tower crane based on digital twin according to claim 1 is characterized in that: In step S5, the spatial position data of the tower crane boom, including the rotation angle of the boom, is displayed in real time through the intelligent visualization system. , elevation angle and stretch length , based on the spatial position data of the boom, the boom position in the three-dimensional model is calculated and updated in real time. The specific formula is: ， ， ， in, is the rotation angle of the boom, is the elevation angle of the boom, is the extended length of the boom; Real-time position update of the crane arm and real-time data collected by sensors , , Will be used to update the position of the boom in the 3D visualization model in real time: ， ， ， in, Used to update the current position of the boom in the 3D model.

9. The method for visualizing the operation of a tower crane based on digital twin according to claim 1, characterized in that: In step S5, the force distribution of the boom is calculated in real time by the intelligent visualization system, and a thermal diagram of the force distribution is displayed: The force distribution of the boom is calculated by the following formula: ， in, is the force at a certain point x on the boom, For lifting weight, is the total length of the boom.

10. A tower crane visual work monitoring method based on digital twin according to claim 9, characterized in that: The heat map shows: Converted into a color map to display the load distribution of the boom in the visualization system. The color mapping formula is defined as: ， in, is the color value corresponding to the load distribution, and are the minimum and maximum values ​​of the force, and The minimum and maximum values ​​for the color map.

Images