Lifting control system for steel platform

Abstract

The invention relates to the technical field of steel platform lifting control, and discloses a steel platform lifting control system which comprises the steps that a sensor is used for collecting data of a steel platform construction site in real time, a data preprocessing strategy is executed, and the collected data is fused; executing a deep learning simulation judgment strategy according to the fused data, and judging whether an obstacle is detected at the moment t; if the obstacle is detected, executing a filtering optimization strategy, and optimizing the position coordinate of the obstacle; executing a digital twinning simulation strategy, and updating the positions of the steel platform and the obstacle in the digital twinning model; executing an edge calculation strategy, and generating a feedback signal; acquiring a feedback signal generated by executing an edge calculation strategy, executing a self-adaptive optimization control strategy, and acquiring an optimal control signal; and executing a fault diagnosis and maintenance strategy, and carrying out fault self-diagnosis and self-repair on the steel platform. And the intelligence, adaptability, accuracy and safety of the steel platform system are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of steel platform lifting control, and specifically provides a control system for steel platform lifting. Background Art

[0002] A steel platform generally refers to a structure used to support equipment, facilities or as a working platform in fields such as construction, construction, oil exploration, and ocean engineering. Steel platforms are widely used in complex working environments due to their high strength, good durability, and ability to withstand heavy loads. Especially in places such as high-altitude operations, ship platforms, steel production, and energy extraction, steel platforms play a crucial role as the infrastructure for construction and operation.

[0003] Traditional steel platform lifting systems have problems such as insufficient obstacle detection ability, single data processing, poor environmental adaptability, fixed control strategies, maintenance dependence on manual labor, and response delays. Existing systems usually rely on basic sensors, lack accurate obstacle recognition and real-time data fusion, and cannot cope with complex weather and environmental changes, resulting in low control accuracy and easy occurrence of failures. In addition, the maintenance of traditional systems mainly relies on manual labor, and potential problems cannot be diagnosed and repaired in real time, increasing operation risks and maintenance costs.

[0004] This solution proposes a control system for steel platform lifting. By introducing advanced technologies such as multi-sensor data fusion, digital twin, edge computing, deep learning, and adaptive control, it can effectively solve these shortcomings, improve the intelligence, adaptability, accuracy, and safety of the steel platform system, and meet the working requirements in modern complex environments. Summary of the Invention

[0005] The present invention provides a control system for steel platform lifting, which helps to solve the problems mentioned in the above background art.

[0006] In a first aspect, the present application provides a control system for steel platform lifting, adopting the following technical solution: A control system for steel platform lifting includes: A data acquisition and preprocessing module: Using sensors to collect data of the steel platform construction site in real time, executing a data preprocessing strategy, fusing the collected data, and representing the fused data as s(t); An obstacle detection module: According to the fused data, execute a deep learning simulation judgment strategy to judge whether an obstacle is detected at time t; If an obstacle is detected, execute a filtering and optimization strategy to optimize the position coordinates of the obstacle; A digital twin module: Establish a three-dimensional coordinate system with the ground position where the steel platform is located as the origin; Obtain the coordinates of the center of the workbench of the steel platform in real time; Obtain the position coordinates of the obstacle in real time; Execute the digital twin simulation strategy to update the positions of the steel platform and the obstacle in the digital twin model; Edge computing module: Obtain the position of the steel platform generated by executing the digital twin simulation strategy; Execute the edge computing strategy to generate a feedback signal for making decision control of the steel platform; Adaptive control module: Obtain the feedback signal generated by executing the edge computing strategy, execute the adaptive optimization control strategy to obtain the optimal control signal; use the optimal control signal to adjust the steel platform; Fault diagnosis and maintenance module: Execute the fault diagnosis and maintenance strategy to perform self-diagnosis and self-repair of the faults of the steel platform.

[0007] Preferably, use sensors to collect data on the construction site of the steel platform in real time, execute the data preprocessing strategy, and fuse the collected data, including: Record the data collected by the mechanical sensor as s 1 '(t); Record the data collected by the optical sensor as s' 2 (t); Normalization processing: Where s 1 (t), μ 1 , σ 1 Are the normalized mechanical sensor data, the mean and standard deviation of the mechanical sensor data collection respectively; Where s 2 (t), μ 2 , σ 2 Are the normalized optical sensor data, the mean and standard deviation of the optical sensor data collection respectively; Use the feature extraction function to extract the features of the data s 1 (t) to obtain s 3 (t); Use the feature extraction function to extract the features of the data s 2 (t) to obtain s 4 (t); Use the feature-level fusion algorithm to fuse s 3 (t) and s 4 (t) to obtain s(t).

[0008] Through multi-sensor data acquisition and fusion technology, the working environment and status information of the steel platform can be obtained in real time and accurately. In particular, the data of optical sensors and mechanical sensors can be monitored in multiple dimensions in space and time. By preprocessing the data and using methods such as normalization and feature extraction, noise interference can be effectively reduced, and the effectiveness and accuracy of the data can be improved. By fusing the data of multiple sensors such as optical and mechanical sensors, more comprehensive and accurate sensing information can be obtained, providing reliable input for subsequent obstacle detection and decision-making control. In this way, the system can not only make real-time responses in complex environments, but also improve the stability of the system, reduce the occurrence of misjudgments and missed judgments, and improve the safety and efficiency during the lifting process of the platform.

[0009] Preferably, according to the fusion data, execute a deep learning simulation judgment strategy to judge whether an obstacle is detected at time t, including: Obtain the deep learning model f 1 (·) for classifying the fusion data to judge whether there is an obstacle at time t; Obtain the result of f 1 (s(t)); If the result output is 1, an obstacle is detected; If the result output is 0, no obstacle is detected.

[0010] Preferably, if an obstacle is detected, execute a filtering optimization strategy to optimize the position coordinates of the obstacle, including: when the output result of the deep learning model is 1: Obtain the position coordinates P 0 (t - 1) = (x 0 (t - 1), y 0 (t - 1), z 0 (t - 1)), where x 0 (t - 1), y 0 (t - 1), z 0 (t - 1) are the coordinates of the obstacle on the x-axis, y-axis, and z-axis respectively; Use the Kalman filter optimization strategy, specifically: Predict the predicted position P 0 (t|t - 1) = P 0 (t - 1); Map the predicted position to the observed position: Obtain the observation matrix Calculate the Kalman gain K(t) = Q(t|t - 1) × H T × (H × Q(t|t - 1) × H T + R) -1, where \(Q(t|t - 1)\), \(H\) T , \(R\) are the predicted error covariance matrix, the transpose of the observation matrix, and the observation noise covariance matrix respectively; Calculate the observed position \(P\) of the obstacle at time \(t\) 0 (t|t)=P 0 (t|t - 1)+K(t)×(s(t)-H×P 0 (t|t - 1)); Update the error covariance matrix \(Q(t|t)=(I - K(t)×H)×Q(t|t - 1)\), where \(I\) is the identity matrix.

[0011] Through the deep learning model and the Kalman filter optimization strategy, the position of the obstacle can be detected in real time and accurately tracked. The deep learning model can intelligently identify whether there is an obstacle and react by classifying and analyzing sensor data. The Kalman filter further improves the accuracy of position estimation by combining the prediction of the obstacle position and the observed values. Especially in a dynamic environment, it can effectively eliminate the uncertainty caused by sensor errors or data noise. Through this real-time and accurate obstacle detection and optimization, the steel platform can avoid collisions or impacts with obstacles, thereby improving the operational safety during the lifting process of the platform and ensuring the efficiency and accuracy of the system, reducing potential risks.

[0012] Preferably, the implementation of the digital twin simulation strategy to update the positions of the steel platform and the obstacle in the digital twin model includes: Record the coordinates of the center of the workbench of the steel platform at time \(t\) as \(P\) 1 (t)=(x 1 (t), y 1 (t), z 1 (t)), where \(x\) 1 (t), \(y\) 1 (t), \(z\) 1 (t) are the coordinates of the workbench on the \(x\)-axis, \(y\)-axis, and \(z\)-axis respectively; Record the position coordinates of the obstacle at time \(t\) as \(P\) 0 (t)=(x 0 (t), y 0 (t), z 0 (t)); Obtain the digital twin model generation function \(f\) 2 (·); Use the digital twin model to calculate the position coordinates \(P\) of the virtual steel platform 3 (t)=f 2 (s(t), p 0 (t), p 1(t), φ), where φ is the relevant data in the digital twin model; Calculate the position coordinates P of the virtual obstacle using the digital twin model 4 (t) = f 2 (s(t), p 1 (t), p 0 (t), φ), where φ is the relevant data in the digital twin model; Obtain the update step size Δt of the digital twin model 0 ; Calculate the coordinates of the updated virtual steel platform and virtual obstacle; where v 1 (t), a 1 (t) are the velocity and acceleration of the virtual steel platform respectively; where v 2 (t), a 2 (t) are the velocity and acceleration of the virtual obstacle respectively.

[0013] Through digital twin technology, it is possible to simulate the state of the steel platform and the position change of the obstacle in real time in a virtual environment. This technology can dynamically update the virtual model, ensure the state synchronization between the virtual steel platform and the actual steel platform, and provide more accurate working data. In addition, combined with edge computing technology, it is possible to process data in real time near the sensor and generate feedback signals, greatly reducing the latency of information transmission, improving the response speed and decision-making efficiency of the system. The low-latency feature of edge computing enables the system to respond quickly when facing real-time control requirements, ensuring that the steel platform can adjust its lifting operation according to the latest environmental changes and state information, thereby improving the operation efficiency of the platform and reducing the risks brought by latency.

[0014] Preferably, the execution of the edge computing strategy to generate a feedback signal for decision control of the steel platform includes: Obtain the position P of the steel platform generated by the digital twin simulation strategy 3 (t) = f 0 (s(t), p 0 (t), p 1 (t), φ); Obtain the edge computing function f 3 (·); Execute the edge computing function to generate an estimated value f of the feedback signal 3 (P 3 (t), s(t)); Calculate the control signal according to the estimated value of the feedback signal: f 4 (f 3 (P3 (t), s(t))), where f 4 (·) is a control decision generation model for generating a control signal according to a feedback signal, and the control signal is a parameter for adjusting the steel platform equipment.

[0015] Preferably, obtaining the feedback signal generated by executing the edge computing strategy, executing the adaptive optimization control strategy, and obtaining the optimal control signal includes: Obtaining the wind speed v 0 (t) and temperature T(t) at time t; Obtaining the inclination θ(t) of the working surface of the steel platform; Obtaining the actual lifting speed v 1 (t) of the steel platform; Calculating the adaptive adjustment value Δu(t) = α 1 ×v 0 (t) + α 2 ×T(t) + α 3 ×θ(t) + α 4 ×v 1 (t), where α 1 , α 2 , α 3 , α 4 are the wind speed weight, temperature weight, inclination weight, and lifting speed weight respectively; Updating the control signal f 4 (f 3 (P 3 (t), s(t)), Δu(t)).

[0016] By introducing the adaptive control strategy, the system can dynamically adjust the control signal according to real-time environmental changes, such as factors like wind speed, temperature, platform inclination, and lifting speed. This control method not only considers the changes in actual working conditions but also can perform feedback adjustment according to the actual performance of the system, thereby improving the control accuracy and the adaptability of the system. Especially in complex environments or variable weather conditions, it can optimize the strategy in real time to ensure the stable operation and efficient operation of the steel platform under various working conditions. The intelligent optimization control enables the system to not only have the self-regulation ability but also improve the long-term stability and working efficiency of the system through learning and adaptation.

[0017] Preferably, implementing the fault diagnosis and maintenance strategy for self-diagnosis and self-repair of the steel platform includes: Obtaining historical fusion data; Calculating the fault probability for the historical fusion data using a probability function; Setting a fault probability threshold; Comparing the fault probability with the fault probability threshold: If the failure probability < the failure probability threshold, it is determined that the steel platform is not faulty; If the failure probability ≥ the failure probability threshold, it is determined that the steel platform has failed, and the steel platform is repaired.

[0018] Through the fault diagnosis and predictive maintenance module, the system can timely detect potential fault hazards and perform self-repair, thus reducing the dependence on manual maintenance. This self-diagnosis ability enables the steel platform to automatically identify the fault source and execute repair when problems occur, reducing downtime and maintenance costs. The fault diagnosis combines historical sensor data to predict the failure probability, enabling early warning and guiding maintenance work, preventing the expansion of faults and their impact on platform operation. At the same time, this predictive maintenance mode can significantly improve the reliability of the platform, reduce safety hazards caused by equipment failures, and provide strong guarantee for the continuous and stable operation of the steel platform.

[0019] The present invention has the following beneficial effects: 1. For the steel platform lifting control system, through the multi-sensor data acquisition and fusion technology, the system can obtain multi-dimensional data of the steel platform working environment in real time, ensuring comprehensive monitoring of the platform operation status and the surrounding environment. The combination of optical sensors and mechanical sensors can monitor in different dimensions, making up for the limitations of a single sensor and providing more comprehensive and accurate information. Data preprocessing strategies, such as normalization and feature extraction, can effectively eliminate noise and redundant information, improve data quality, and make the fused data more accurate and reliable. Through this multi-sensor data fusion method, the system can make a quick response in a dynamic environment and improve the accuracy of the decision-making process. This comprehensive monitoring method not only ensures a more stable lifting process of the steel platform but also reduces the risk of system misjudgment and missed judgment in a complex construction environment, ultimately improving the operation efficiency and safety of the platform.

[0020] 2. For the steel platform lifting control system, through the obstacle detection and optimization strategy combining deep learning and Kalman filtering, the system can achieve efficient and accurate obstacle recognition and position update. The deep learning model can, through training and classification, intelligently identify the presence of obstacles and provide real-time feedback, providing reliable obstacle information for the lifting process of the steel platform. Kalman filtering, through the dynamic prediction and optimization of the obstacle position, makes the estimation of the obstacle position more accurate. Especially in a complex environment and when there are errors in sensors, it can effectively reduce the deviation of position estimation. With this strategy, the steel platform can track the position change of obstacles in real time, avoid collisions or interferences with obstacles, and thus improve the operation safety and the efficiency of the platform. The advantage of this solution is that it not only enhances the platform's environmental adaptability but also reduces the need for human intervention, ensuring the stable operation and safe operation of the platform.

[0021] 3. For the steel platform lifting control system, through real-time simulation of the steel platform's state and obstacle positions in a virtual environment, digital twin technology can achieve synchronous updates between the platform and the actual environment. This virtualization modeling technology can provide real-time dynamic monitoring of the steel platform and the surrounding environment, offering more accurate feedback information for decision-making. Combined with edge computing, the system can perform real-time calculations at the data source close to the sensors, reducing data transmission latency and ensuring fast response. The feedback signals generated through edge computing strategies can not only adjust the lifting control of the steel platform in real time but also improve the system's processing efficiency and reduce latency during information transmission. Through the combination of these two, the steel platform can make decisions quickly in complex and dynamic environments, ensuring the real-time nature and efficiency of operations, and enhancing the overall operation accuracy and safety.

[0022] 4. For the steel platform lifting control system, the adaptive control strategy dynamically adjusts control signals according to real-time environmental changes to ensure the stable operation of the steel platform in different working environments. Through real-time monitoring of factors such as wind speed, temperature, and platform inclination, adaptive control can perform intelligent optimization during the platform lifting process, reducing the impact of external environmental changes on platform operations. The system can provide real-time feedback and adjust control parameters to optimize the lifting speed and path of the platform, ensuring efficient operation even under harsh environmental conditions. This strategy not only enhances the platform's adaptability to dynamic environments but also improves the overall work efficiency and safety. Through the application of a reinforcement learning model, the platform can continuously self-optimize during long-term operations, adjusting operation strategies based on historical data, thereby improving operation accuracy and reducing resource waste. Ultimately, intelligent optimization control enables the steel platform to better cope with complex and changing environments and improve the stability and efficiency of long-term operation.

[0023] 5. For the steel platform lifting control system, the fault diagnosis and predictive maintenance strategy provides strong support for the long-term stable operation of the steel platform. By analyzing historical sensor data, the system can identify potential faults in advance and predict possible fault situations. The system calculates the fault probability and compares it with a threshold to detect fault risks in a timely manner and initiate a self-repair mechanism. This not only reduces the need for manual intervention but also avoids the spread of faults through early warning, reducing maintenance costs and downtime. The self-diagnosis ability can automatically identify problems and perform repair operations when a fault occurs, ensuring the continuous and efficient operation of the platform. Through this strategy, the steel platform can achieve "preventive maintenance", avoiding unnecessary downtime and repairs during conventional maintenance, enhancing the system's reliability and safety, and thus ensuring stability and cost-effectiveness during long-term operation. Description of the Drawings

[0024] Figure 1Schematic diagram of the method of the present invention.

[0025] Figure 2 Schematic diagram of the module of the present invention. Detailed implementation manners

[0026] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0027] Embodiment 1. Refer to Figure 1 , a lifting control system for a steel platform, including: a data acquisition and preprocessing module: using sensors to collect data of the steel platform construction site in real time, executing a data preprocessing strategy, fusing the collected data, and representing the fused data as s(t); An obstacle detection module: According to the fused data, execute a deep learning simulation judgment strategy to judge whether an obstacle is detected at time t; If an obstacle is detected, execute a filtering and optimization strategy to optimize the position coordinates of the obstacle; A digital twin module: Establish a three-dimensional coordinate system with the ground position where the steel platform is located as the origin; Obtain the coordinates of the center of the workbench of the steel platform in real time; Obtain the position coordinates of the obstacle in real time; Execute a digital twin simulation strategy to update the positions of the steel platform and the obstacle in the digital twin model; An edge computing module: Obtain the position of the steel platform generated by executing the digital twin simulation strategy; Execute an edge computing strategy to generate a feedback signal for making decision control on the steel platform; An adaptive control module: Obtain the feedback signal generated by executing the edge computing strategy, execute an adaptive optimization control strategy to obtain the best control signal; use the best control signal to adjust the steel platform; A fault diagnosis and maintenance module: Execute a fault diagnosis and maintenance strategy to perform self-diagnosis and self-repair of the faults of the steel platform.

[0028] Using sensors to collect data of the steel platform construction site in real time, execute a data preprocessing strategy, and fuse the collected data, including: Record the data collected by the mechanical sensor as s 1 '(t); Denote the data collected by the optical sensor as s' 2 (t); Normalization processing: where s 1 (t), μ 1 , σ 1 are the normalized data of the mechanical sensor, the mean and standard deviation of the data collected by the mechanical sensor respectively; where s 2 (t), μ 2 , σ 2 are the normalized data of the optical sensor, the mean and standard deviation of the data collected by the optical sensor respectively; Use the feature extraction function to extract the features of the data s 1 (t), and obtain s 3 (t); Use the feature extraction function to extract the features of the data s 2 (t), and obtain s 4 (t); Use the feature-level fusion algorithm to fuse s 3 (t) and s 4 (t) to obtain s(t).

[0029] Through the multi-sensor data acquisition and fusion technology, the system can comprehensively obtain various data of the steel platform working environment, including the information provided by the optical sensor and the mechanical sensor. This data fusion scheme effectively overcomes the limitations that may exist in a single sensor, ensuring the comprehensiveness and accuracy of the collected data. By preprocessing the sensor data and using techniques such as normalization, noise and redundant information in the data can be eliminated, making the data more stable and reliable. Through feature extraction, the system can extract effective information from a large amount of raw data, providing a more accurate basis for subsequent analysis. Data fusion not only improves the signal-to-noise ratio of the data, but also can improve the accuracy of the data through cross-verification of multiple sensor information. Through this scheme, the steel platform can grasp the dynamic changes of the working environment in real time, make timely responses, and ensure the operation stability and safety of the platform. This multi-dimensional and multi-level data acquisition and fusion method can not only effectively prevent misjudgment and missed judgment problems that may occur in the traditional single-sensor method, but also greatly enhance the operation flexibility and emergency response ability of the steel platform in a complex environment. Finally, through in-depth analysis and accurate judgment of the data, the system can improve the platform operation efficiency, reduce the probability of failures, and ensure the project progress and safety.

[0030] According to the fused data, execute the deep learning simulation judgment strategy to judge whether an obstacle is detected at time t, including: Obtain the deep learning model f 1 (·), which is used to classify the fused data and determine whether there is an obstacle at time t; Obtain f 1 (s(t)); If the result output is 1, an obstacle is detected; If the result output is 0, no obstacle is detected.

[0031] If an obstacle is detected, execute the filtering optimization strategy to optimize the position coordinates of the obstacle, including: When the output result of the deep learning model is 1: Obtain the position coordinates P 0 (t - 1)=(x 0 (t - 1), y 0 (t - 1), z 0 (t - 1)), where x 0 (t - 1), y 0 (t - 1), z 0 (t - 1) are the coordinates of the obstacle on the x-axis, y-axis, and z-axis respectively; Use the Kalman filtering optimization strategy, specifically: Predict the predicted position P 0 (t|t - 1)=P 0 (t - 1); Map the predicted position to the observed position: Obtain the observation matrix Calculate the Kalman gain K(t)=Q(t|t - 1)×H T ×(H×Q(t|t - 1)×H T +R) -1 , where Q(t|t - 1), H T , R are the predicted error covariance matrix, the transpose of the observation matrix, and the observation noise covariance matrix respectively; Calculate the observed position P 0 (t|t)=P 0 (t|t - 1)+K(t)×(s(t)-H×P 0 (t|t - 1)); Update the error covariance matrix Q(t|t)=(I - K(t)×H)×Q(t|t - 1), where I is the identity matrix.

[0032] Through the combination of a deep learning model and Kalman filtering, the system can efficiently detect obstacles during the lifting process of the steel platform and optimize their positions in real time. The deep learning model can accurately determine the presence of obstacles through training and classification, and immediately feedback the obstacle detection results to the system. Compared with traditional rule-based obstacle detection methods, the deep learning method has stronger adaptability and accuracy, and can effectively identify different types of obstacles in complex working environments. Kalman filtering can optimize the position estimation by predicting and correcting the error of the obstacle position. Especially in the presence of measurement noise and uncertainty, it can still maintain high accuracy. By optimizing the obstacle position, the system can effectively reduce the operation risk caused by the obstacle position estimation error and ensure the safety during the lifting process of the steel platform. In addition, the data processing by Kalman filtering can also reduce the system calculation amount and improve the real-time performance, enabling the steel platform to respond quickly when encountering obstacles. The advantage of this solution is that the system can not only dynamically update the obstacle position to avoid collisions or interferences, but also significantly improve the work efficiency, reduce the need for manual intervention, and increase the automation level. By combining deep learning and Kalman filtering, the obstacle detection ability of the system in complex environments has been significantly improved, ensuring the safe operation of the platform.

[0033] Execute the digital twin simulation strategy to update the positions of the steel platform and obstacles in the digital twin model, including: Record the coordinates of the workbench center of the steel platform at time t as P 1 (t) = (x 1 (t), y 1 (t), z 1 (t)), where x 1 (t), y 1 (t), z 1 (t) are the coordinates of the workbench on the x-axis, y-axis, and z-axis respectively; Record the position coordinates of the obstacle at time t as P 0 (t) = (x 0 (t), y 0 (t), z 0 (t)); Obtain the digital twin model generation function f 2 (·); Use the digital twin model to calculate the position coordinates P 3 (t) = f 2 (s(t), p 0 (t), p 1 (t), φ), where φ is the relevant data in the digital twin model; Use the digital twin model to calculate the position coordinates P 4$(t)=f 2 (s(t),p 1 (t),p 0 (t),φ), where φ is the relevant data in the digital twin model; Obtain the update step Δt of the digital twin model 0 ; Calculate the coordinates of the updated virtual steel platform and virtual obstacles; where v 1 (t), a 1 (t) are the velocity and acceleration of the virtual steel platform respectively; where v 2 (t), a 2 (t) are the velocity and acceleration of the virtual obstacle respectively.

[0034] Through digital twin technology, the system can establish a virtual steel platform model in real time and simulate its relative position with obstacles in the virtual environment. This real-time synchronous update method not only improves the system's response speed to the actual environment but also helps operators predict potential risks in advance. Digital twin technology can provide a real-time virtual image of the steel platform, enabling the system to obtain detailed information about the platform state and environment. This allows the system to optimize control strategies through the virtual model to ensure the stable operation of the steel platform in a complex environment. Combining the capabilities of edge computing, data processing can be performed in real time near the data source, reducing delays in the information transmission process, improving the response speed, and the real-time nature of decision-making. Through edge computing, the system can quickly calculate and generate feedback signals to achieve immediate control of the steel platform. The combination of this edge computing and digital twin enables the steel platform to quickly respond and adjust in a highly dynamic and complex construction environment, improving the operation efficiency and decision-making accuracy. At the same time, when the system performs real-time feedback, it can also ensure low latency and high performance, further improving the working efficiency and safety of the platform. Through the combination of these two technologies, the real-time monitoring ability and environmental adaptation ability of the steel platform have been greatly enhanced, optimizing the operation efficiency of the platform.

[0035] Execute the edge computing strategy to generate feedback signals for decision control of the steel platform, including: Obtain the position P of the steel platform generated by the digital twin simulation strategy 3 (t)=f 0 (s(t),p 0 (t),p 1 (t),φ); Obtain the edge computing function f 3 (·); Execute the edge computing function to generate an estimated value f of the feedback signal 3 (P 3 (t), s(t)); Calculate the control signal based on the estimated value of the feedback signal: f 4 (f 3 (P 3 (t), s(t))), where f 4 (·) is a control decision generation model for generating a control signal based on the feedback signal, and the control signal is a parameter for adjusting the steel platform equipment.

[0036] Obtain the feedback signal generated by executing the edge computing strategy, execute the adaptive optimization control strategy, and obtain the optimal control signal, including: Obtain the wind speed v 0 (t) and temperature T(t) at time t; Obtain the inclination θ(t) of the working surface of the steel platform; Obtain the actual lifting speed v 1 (t) of the steel platform; Calculate the adaptive adjustment value Δu(t) = α 1 ×v 0 (t) + α 2 ×T(t) + α 3 ×θ(t) + α 4 ×v 1 (t), where α 1 , α 2 , α 3 , α 4 are the wind speed weight, temperature weight, inclination weight and lifting speed weight respectively; Update the control signal f 4 (f 3 (P 3 (t), s(t)), Δu(t)).

[0037] The adaptive control strategy can automatically adjust the operating parameters of the steel platform according to real-time environmental changes, thus maintaining the stable operation of the platform. In actual work, factors such as wind speed, temperature, platform inclination, and lifting speed will directly affect the lifting efficiency and stability of the steel platform. By monitoring these environmental and platform states in real time, the system can make dynamic adjustments to ensure that the platform always maintains the best working state under changing environmental conditions. The adaptive control strategy optimizes the lifting path and rate of the platform through comprehensive analysis of multiple factors, reducing the impact of the external environment on platform operation. This strategy also incorporates the idea of reinforcement learning, which can continuously learn and optimize after multiple operations, gradually improving the accuracy and efficiency of platform control. Ultimately, the steel platform can achieve efficient, safe, and stable operation in a complex construction environment, avoiding operation errors or efficiency reduction caused by environmental fluctuations. Through this intelligent optimization control, the system not only improves the adaptability of the steel platform to environmental changes but also continuously optimizes control decisions, enhancing the long-term working efficiency and safety of the platform.

[0038] Implement a fault diagnosis and maintenance strategy to perform self-diagnosis and self-repair of the steel platform, including: Obtain historical fusion data; Calculate the fault probability using a probability function for the historical fusion data; Set a fault probability threshold; Compare the fault probability with the fault probability threshold: If the fault probability < the fault probability threshold, it is determined that the steel platform is not faulty; If the fault probability ≥ the fault probability threshold, it is determined that the steel platform has a fault, and the steel platform is repaired.

[0039] The fault diagnosis and predictive maintenance strategy can effectively improve the operating reliability of the steel platform and reduce downtime. By analyzing historical sensor data, the system can identify potential fault risks and give early warnings. This mechanism for early fault identification can provide the platform with sufficient time for maintenance, thus avoiding the high costs and project delays caused by emergency repairs after traditional faults occur. By calculating the fault probability and comparing it with a preset threshold, the system can determine whether there is a fault risk in the steel platform and perform self-diagnosis on potential problems. The self-diagnosis system can perform self-repair through a feedback mechanism to ensure the normal operation of the platform and reduce the need for manual intervention. The fault diagnosis and predictive maintenance plan can not only prevent the occurrence of major faults in advance but also significantly reduce the maintenance cost and downtime of the platform through regular maintenance and self-repair functions, enhancing the long-term stability and economic benefits of the system. This intelligent maintenance method can greatly reduce the passive response and cost waste in the traditional maintenance mode, improving the overall efficiency and reliability of the system.

[0040] It should be noted that in this text, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprising", "including" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device.

[0041] The above are only the preferred embodiments of the present invention. It should be pointed out that for those of ordinary skill in the art, without departing from the technical principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A control system for lifting a steel platform, characterized in that: include: Data acquisition and preprocessing module: Use sensors to collect data from the steel platform construction site in real time, execute data preprocessing strategies, fuse the collected data, and represent the fused data as s(t); Obstacle detection module: Based on the fused data, a deep learning simulation judgment strategy is executed to determine whether an obstacle is detected at time t; If an obstacle is detected, a filtering optimization strategy is executed to optimize the position coordinates of the obstacle; Digital Twin Module: A three-dimensional coordinate system is established with the ground position of the steel platform as the origin; Obtain the coordinates of the workbench center of the steel platform in real time; Obtain the location coordinates of obstacles in real time; Execute the digital twin simulation strategy and update the positions of the steel platform and obstacles in the digital twin model; Edge computing module: Obtain the location of the steel platform generated by executing the digital twin simulation strategy; Execute edge computing strategies and generate feedback signals for decision-making and control of the steel platform; Adaptive Control Module: Obtain the feedback signal generated by executing the edge computing strategy, execute the adaptive optimization control strategy, and obtain the best control signal; use the best control signal to adjust the steel platform; Fault diagnosis and maintenance module: Execute fault diagnosis and maintenance strategies, and conduct self-diagnosis and self-repair of steel platform faults.

2. The steel platform lifting control system according to claim 1, characterized in that: The method of using sensors to collect data from the steel platform construction site in real time, executing a data preprocessing strategy, and fusing the collected data includes: The data collected using the mechanical sensor is recorded as s1'(t); The data collected using the optical sensor is recorded as s'2(t); Normalization: Among them, s1(t), μ1, σ1 are the normalized mechanical sensor data, and the mean and standard deviation of the mechanical sensor data are collected; Among them, s2(t), μ2, σ2 are the normalized optical sensor data, and the mean and standard deviation of the collected optical sensor data are respectively; Use the feature extraction function to extract the features of data s1(t) to obtain s3(t); Use the feature extraction function to extract the features of data s2(t) to obtain s4(t); Use the feature-level fusion algorithm to fuse s3(t) and s4(t) to obtain s(t).

3. The steel platform lifting control system according to claim 1, characterized in that: The step of executing a deep learning simulation judgment strategy based on the fusion data to judge whether an obstacle is detected at time t includes: Obtain the deep learning model f1(·) to classify the fused data and determine whether there is an obstacle at time t; Get the result of f1(s(t)); If the result output is 1, an obstacle is detected; If the result output is 0, no obstacle is detected.

4. The steel platform lifting control system according to claim 3 is characterized in that: If an obstacle is detected, a filtering optimization strategy is executed to optimize the position coordinates of the obstacle, including: When the deep learning model outputs a result of 1: Obtain the position coordinates of the obstacle at time t-1 P0(t-1)=(x0(t-1), y0(t-1), z0(t-1)), where x0(t-1), y0(t-1), z0(t-1) are the coordinates of the obstacle on the x-axis, y-axis, and z-axis respectively; Use Kalman filter optimization strategy, specifically: Predicted obstacle position at time t: P0(t|t-1)=P0(t-1); Map predicted locations to observed locations: Get the observation matrix Calculate the Kalman gain K(t) = Q(t|t-1) × H T ×(H×Q(t|t-1)×H T +R) -1 , where Q(t|t-1),H T , R are the prediction error covariance matrix, the transpose of the observation matrix and the observation noise covariance matrix respectively; Calculate the observed position of the obstacle at time t: P0(t|t)=P0(t|t-1)+K(t)×(s(t)-H×P0(t|t-1)); Update the error covariance matrix Q(t|t)=(IK(t)×H)×Q(t|t-1), where I is the identity matrix.

5. The steel platform lifting control system according to claim 4, characterized in that: The executing digital twin simulation strategy and updating the positions of the steel platform and obstacles in the digital twin model include: The coordinates of the center of the workbench of the steel platform at time t are marked as P1(t)=(x1(t), y1(t), z1(t)), where x1(t), y1(t), z1(t) are the coordinates of the workbench on the x-axis, y-axis and z-axis respectively; The position coordinates of the obstacle at time t are marked as P0(t) = (x0(t), y0(t), z0(t)); Get the digital twin model generation function f2(·); Use the digital twin model to calculate the position coordinates of the virtual steel platform P3(t)=f2(s(t),p0(t),p1(t),φ), where φ is the relevant data in the digital twin model; Use the digital twin model to calculate the position coordinates of the virtual obstacle P4(t)=f2(s(t),p1(t),p0(t),φ), where φ is the relevant data in the digital twin model; Get the update step size Δt0 of the digital twin model; calculating updated coordinates of the virtual steel platform and virtual obstacles; Among them, v1(t) and a1(t) are the velocity and acceleration of the virtual steel platform respectively; Among them, v2(t) and a2(t) are the velocity and acceleration of the virtual obstacle respectively.

6. The steel platform lifting control system according to claim 5, characterized in that: The execution of the edge computing strategy to generate a feedback signal for making decisions and controlling the steel platform includes: Get the position of the steel platform generated by the digital twin simulation strategy P3(t) = f0(s(t), p0(t), p1(t), φ); Get the edge computing function f3(·); Execute the edge computing function to generate an estimated value f3(P3(t), s(t)) of the feedback signal; The control signal is calculated according to the estimated value of the feedback signal: f4(f3(P3(t),s(t))), wherein f4(·) is a control decision generation model, which is used to generate a control signal according to the feedback signal, and the control signal is a parameter for adjusting the steel platform equipment.

7. The steel platform lifting control system according to claim 6, characterized in that: The step of obtaining a feedback signal generated by executing an edge computing strategy, executing an adaptive optimization control strategy, and obtaining an optimal control signal includes: Get the wind speed v0(t) and temperature T(t) at time t; Obtain the inclination θ(t) of the working surface of the steel platform; Get the actual lifting speed v1(t) of the steel platform; Calculate the adaptive adjustment value Δu(t) = α1×v0(t) + α2×T(t) + α3×θ(t) + α4×v1(t), where α1, α2, α3, and α4 are wind speed weight, temperature weight, inclination weight, and lifting speed weight, respectively; Update the control signal f4(f3(P3(t),s(t)),Δu(t)).

8. The steel platform lifting control system according to claim 1, characterized in that: The execution of the fault diagnosis and maintenance strategy to perform self-diagnosis and self-repair of the steel platform includes: Obtain historical fusion data; Calculate the failure probability using probability function on historical fusion data; Set a failure probability threshold; Compare the failure probability to the failure probability threshold: If the failure probability is less than the failure probability threshold, the steel platform is deemed not to have failed; If the failure probability ≧ the failure probability threshold, it is determined that the steel platform has failed and the steel platform is repaired.

Images