Component lifting operation safety monitoring and risk management method based on digital twinning

Abstract

The invention relates to a digital twinning-based component lifting operation safety monitoring and risk management method, belongs to the field of building industry, and aims to solve the problem of safety monitoring in lifting operation of prefabricated components. The method comprises the following steps: collecting real-time data of lifting operation, and establishing a digital twin system; adjusting the weight of at least one type of lifting operation real-time data in the real-time scene according to the influence of the multiple types of lifting operation historical data on lifting operation safety in different scenes; extracting risk feature data corresponding to the abnormal mode; predicting time series data in the lifting operation real-time data after weight adjustment to obtain first prediction data; according to the similarity between the feature data of the first prediction data and the risk feature data, risk early warning is judged; and carrying out risk labeling on the digital twin system. The intelligent safety management level of component lifting operation can be improved, and different lifting operation requirements can be met.

Description

Technical Field

[0001] The present invention relates to the field of construction industry, and in particular to a component lifting operation safety monitoring and risk management method based on digital twins. Background Art

[0002] With the rapid development of information technology, intelligent construction has become an important development trend in the construction industry. However, the current level of informatization of prefabricated components such as steel structures in the construction process is low, which limits intelligent construction. Digital twin technology can achieve high-fidelity simulation of physical space through virtual-reality interaction, thus providing a new direction for the intelligent construction of prefabricated components such as steel structures. The safety monitoring and risk management of lifting operations of traditional prefabricated components such as steel structures rely on experience and manual management. Their response speed, real-time safety and adaptability to different construction scenarios cannot meet the needs of modern production. Summary of the invention

[0003] In order to solve the problem of safety monitoring of prefabricated components during lifting operations, the inventors have made the present invention to provide a component lifting operation safety monitoring and risk management method based on digital twins.

[0004] In a first aspect, the present invention provides a component lifting operation safety monitoring and risk management method based on digital twins, comprising: Collect real-time data of lifting operations and establish a digital twin system; adjust the weight of at least one type of real-time data of lifting operations in real-time scenarios according to the impact of multiple types of historical data of lifting operations on the safety of lifting operations in different scenarios; extract the characteristics of the historical data of lifting operations corresponding to the abnormal mode to obtain risk characteristic data; predict the time series data in the real-time data of lifting operations after adjusting the weight to obtain first prediction data; judge the risk warning based on the similarity between the characteristic data of the first prediction data and the risk characteristic data; when it is determined to issue a risk warning, label the digital twin system with risks according to the corresponding first prediction data.

[0005] In some optional technical solutions, the component lifting operation safety monitoring and risk management method based on digital twins also includes the following steps: establishing a transfer matrix based on the real-time data of the lifting operation after adjusting the weight; determining the probability of risks occurring under different operations and the corresponding risks based on the transfer matrix; determining high-risk operations and low-risk operations with different risks based on the probability of risks occurring under different operations and the corresponding risks; and issuing operation adjustment suggestions based on the high-risk operations and low-risk operations with different risks.

[0006] In some optional technical solutions, the above-mentioned digital twin-based component lifting operation safety monitoring and risk management method also includes the following steps: predicting the next operation according to the operating standards and operating steps in the safety regulations; when the next operation is a high-risk operation, issuing a risk warning.

[0007] In some optional technical solutions, collecting real-time data of lifting operations includes the following steps: collecting different types of real-time data of lifting operations according to the initial collection frequency of each type of real-time data of lifting operations; when the collected real-time data of any type of lifting operations exceeds the collection frequency increase threshold, increasing the collection frequency of the corresponding type of real-time data of lifting operations; when the collected real-time data of any type of lifting operations is lower than the collection frequency decrease threshold, reducing the collection frequency of the corresponding type of real-time data of lifting operations.

[0008] In some optional technical solutions, collecting the real-time data of the lifting operation also includes the following steps: using wavelet transform to denoise the collected real-time data of the lifting operation.

[0009] In some optional technical solutions, the historical data of the lifting operation corresponding to the abnormal pattern is input into the first convolutional neural network for training to obtain risk characteristic data; the time series data in the real-time data of the lifting operation after adjusting the weight is input into the long short-term memory network for training and reasoning to obtain the first prediction data; the first prediction data is input into the second convolutional neural network for training to obtain the characteristic data of the first prediction data.

[0010] In some optional technical solutions, the computing tasks of the digital twin system are divided into low-latency tasks and cloud tasks according to local computing resources, and low-latency tasks are calculated locally, including: parallel calculation of the training process of the first convolutional neural network and the second convolutional neural network and the training and inference process of the long short-term memory network through the local GPU; and using the local TPU for matrix calculations.

[0011] In the second aspect, the present invention provides a component lifting operation safety monitoring and risk management platform based on digital twins, including: a simulation module and a risk assessment module. Among them, the simulation module is used to collect real-time data of lifting operations and establish a digital twin system; the risk assessment module is used to adjust the weight of at least one type of real-time data of lifting operations in real-time scenarios according to the impact of multiple types of historical data of lifting operations on the safety of lifting operations in different scenarios; extract the characteristics of the historical data of lifting operations corresponding to the abnormal mode to obtain risk characteristic data; predict the time series data in the real-time data of lifting operations after adjusting the weight to obtain the first prediction data; judge the risk warning according to the similarity between the characteristic data of the first prediction data and the risk characteristic data; when it is determined to perform a risk warning, the digital twin system is risk-labeled according to the corresponding first prediction data.

[0012] Based on the same inventive concept, the present invention also provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the aforementioned digital twin-based component lifting operation safety monitoring and risk management method is implemented.

[0013] Based on the same inventive concept, the present invention also provides a computer storage medium, in which computer executable instructions are stored. When the computer executable instructions are executed, the aforementioned digital twin-based component lifting operation safety monitoring and risk management method is implemented.

[0014] The beneficial effects of the above technical solution provided by the present invention include at least: The present invention can collect and process various types of data in the lifting operation in real time, ensure that the system monitors every link in the lifting process in real time, can quickly respond to any changes on site, and adjust the lifting operation plan in time, greatly improving the safety of the operation. Not only can it identify current safety hazards, but it can also analyze possible risk points in the future based on real-time data and historical data, and issue early warnings to help operators avoid risks in time. By continuously accumulating on-site data and historical experience, it can continuously improve its adaptability to complex construction environments in practical applications, ensuring that more reliable safety guarantees are provided in future lifting operations. By introducing digital twin technology and intelligent management methods, the present invention can greatly improve the level of intelligent safety management of component lifting operations and adapt to different lifting operation requirements.

[0015] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0016] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 This is a flow chart of a component lifting operation safety monitoring and risk management method based on digital twins in an embodiment of the present invention; Figure 2 A flowchart of collecting real-time data of lifting operations and establishing a digital twin system in an embodiment of the present invention; Figure 3A flowchart of simulation and emulation in an embodiment of the present invention; Figure 4 It is a block diagram of a component lifting operation safety monitoring and risk management platform based on digital twins in an embodiment of the present invention; Figure 5 This is an example interface diagram of a component lifting operation safety monitoring and risk management platform based on digital twins in an embodiment of the present invention; Figure 6 The figure is a schematic diagram of the structure of an electronic device in an embodiment of the present invention. DETAILED DESCRIPTION

[0018] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0019] In order to solve the problems existing in the prior art, an embodiment of the present invention provides a component lifting operation safety monitoring and risk management method based on digital twin, a component lifting operation safety monitoring and risk management platform based on digital twin, corresponding electronic equipment, and corresponding computer storage medium.

[0020] The embodiment of the present invention provides a component lifting operation safety monitoring and risk management method based on digital twins, and its process is as follows: Figure 1 As shown, the following steps are included: Step S1: Collect real-time data of lifting operations and establish a digital twin system.

[0021] In some optional embodiments, step S1 includes steps S11 to S13, and the process is as follows: Figure 2 shown.

[0022] Step S11: Collect relevant data such as personnel and equipment in the on-site lifting operation.

[0023] Before the lifting operation begins, RFID (Radio Frequency Identification) tags and sensors are installed on the hydraulic hoist, tower crane equipment, prefabricated components and lifting area.

[0024] Through RFID tags and various sensors, various real-time data in lifting operations are collected, such as the status of tower crane equipment (such as load, lifting angle, wind speed, etc.), component information (such as stress, displacement, etc.), personnel status (such as location information, operating behavior, etc.) and surrounding environment parameters (such as temperature, humidity, etc.).

[0025] Specifically, RFID tags are used to identify the identity information of components and equipment, and sensors are used to monitor the tower crane's lifting weight, angle, wind speed and other parameters in real time, as well as the stress, displacement and other data of prefabricated components.

[0026] In some optional embodiments, collecting real-time data of lifting operations includes the following steps: collecting different types of real-time data of lifting operations according to the initial collection frequency of each type of real-time data of lifting operations; when any type of real-time data of lifting operations collected exceeds the collection frequency increase threshold, increasing the collection frequency of the corresponding type of real-time data of lifting operations; when any type of real-time data of lifting operations collected is lower than the collection frequency decrease threshold, reducing the collection frequency of the corresponding type of real-time data of lifting operations. Through the aforementioned adaptive frequency data collection, the data collection frequency is automatically adjusted according to the real-time changes of the working environment. On the basis of ensuring data accuracy, it is possible to reduce unnecessary data redundancy, reduce the amount of data that needs to be transmitted and processed, and improve the input data quality of the digital twin system. For example, when parameters such as wind speed and hoisting angle exceed the preset threshold, the digital twin system will increase the collection frequency and optimize the data collection density through the sensor network.

[0027] In some optional embodiments, collecting the real-time data of the lifting operation also includes the following steps: using wavelet transform to denoise the collected real-time data of the lifting operation.

[0028] The specific steps of the wavelet transform denoising method are as follows: According to the characteristics of the real-time data of the lifting operation, select a suitable wavelet as the wavelet basis for wavelet decomposition, such as the Meyer wavelet basis. Different wavelet bases can be selected for different types of real-time data of lifting operations, and wavelet transform denoising can be performed separately. According to the real-time requirements of the digital twin system, the number of decomposition layers N is determined. According to the selected wavelet basis, the real-time data of the lifting operation is decomposed into N layers of wavelets to obtain the wavelet coefficients of each layer; the wavelet coefficients are adjusted, specifically by selecting a wavelet coefficient threshold, retaining the wavelet coefficients with a modulus greater than the wavelet coefficient threshold, and setting the wavelet coefficients with a modulus less than the wavelet coefficient threshold to zero. According to the adjusted wavelet coefficients of each layer, wavelet reconstruction is performed to obtain the denoised real-time data of the lifting operation. Through wavelet denoising, noise data is automatically filtered out to ensure that the key data collected has high quality and accuracy.

[0029] Step S12: Construct a digital twin system with geometric, physical, behavioral and rule elements.

[0030] In the digital twin system, the geometric model, physical model, behavioral model and rule model of the lifting operation are constructed based on the collected real-time data.

[0031] Geometric model: used to describe the three-dimensional spatial structure and dynamic position relationship of lifting equipment and prefabricated components to ensure that the model is consistent with the physical site. For example, the geometric model of the lifting operation is modeled using BIM (Building Information Modeling) software to establish a three-dimensional structural model of the prefabricated components.

[0032] Physical model: used to simulate the stress state of prefabricated components during the lifting process, simulate the changes in their physical properties (such as stress, strain, etc.), and help predict the stability of components in actual operation. For example, based on the geometric model, the stress state of the components is simulated using finite element analysis software, and risk assessment is performed in combination with actual working conditions. The physical model ensures that every link of the construction site can be perceived in real time by comprehensively monitoring the tower crane equipment, lifting components, operators and environmental parameters. The monitoring system in the physical model uses a high-precision sensor network that can collect the tower crane's lifting capacity, tilt angle, stress distribution of lifting components, and climatic conditions of the surrounding environment during the lifting process (such as wind speed, temperature, etc.) in real time, and upload it to the cloud through the Internet of Things technology, ensuring efficient and safe management of the lifting process.

[0033] Behavior model: used to simulate the possible operation behaviors and operation paths during the lifting process. Specifically, based on the real-time data of the lifting operation on site, combined with the operation steps of the personnel and the movement trajectory of the components, the dynamic behavior prediction of the lifting operation is realized.

[0034] Rule model: This model sets a set of safety specifications for lifting operations based on building safety standards and relevant regulations, and combines historical data with real-time on-site data to determine whether the operation complies with safety specifications.

[0035] Geometric models, physical models, behavioral models, and rule models are models at different levels in the digital twin system. They are essentially associative databases at different levels, and are displayed in real time in a three-dimensional and intuitive manner in the form of three-dimensional models.

[0036] Step S13: Monitor the personnel, equipment, components and environmental data in the lifting operation, and transmit these data to the digital twin system in real time.

[0037] By integrating this data into an IoT sensor network, all dynamic information on site is collected in real time and uploaded to the cloud through wireless networks such as 5G and LoRa (Long Range Radio), ensuring that the data can be quickly transmitted to the digital twin system.

[0038] Step S2: Perform dynamic simulation and simulation in the digital twin system to simulate the various states and risk changes of the lifting process, and perform two-way information transmission through the virtual-real information interaction module.

[0039] In some optional embodiments, step S2 includes steps S21 to S23, and the process is as follows: Figure 3 shown.

[0040] Step S21: According to the impact of multiple types of lifting operation historical data on lifting operation safety in different scenarios, the weight of at least one type of lifting operation real-time data in the real-time scenario is adjusted. For example, under extreme weather conditions, the weight of meteorological factors will be increased to improve the risk prediction ability of the algorithm.

[0041] In some optional embodiments, a safety evaluation is performed according to the lifting operation state corresponding to the lifting operation historical data to obtain the safety data corresponding to the lifting operation historical data. The change of safety data when various types of lifting operation data change is analyzed. Under the same change rate of the lifting operation data, the types of lifting operation data are sorted from large to small according to the change of safety data, and the weight of the first type of lifting operation data in the sorting order is increased, or the weight of multiple types of lifting operation data in the sorting order is increased by several positions, or the weight of the last type of lifting operation data in the sorting order is reduced, or the weight of multiple types of lifting operation data in the sorting order is reduced by several positions. According to the aforementioned adjusted weight, the real-time data of the lifting operation of the corresponding class is adjusted. The above process can be understood as, according to the degree of influence of different types of lifting operation data on the safety of the lifting operation, the class of the lifting operation data is classified into two major categories, each major category includes one or more types of lifting operation data, the weight of each type of lifting operation data under one major category is adjusted, and the weight of each type of lifting operation data under another major category is reduced, so as to achieve more accurate safety monitoring. Therefore, a classifier algorithm can be selected to adjust the weight of the real-time data of the lifting operation.

[0042] Furthermore, in some optional embodiments, the weight of the real-time data of the lifting operation is adjusted by a weighted random forest algorithm. The weighted random forest is an improved version of the random forest algorithm, which is particularly suitable for processing data sets with imbalanced categories. The specific steps are as follows: (1) Preprocessing the real-time data of the lifting operation, including missing value filling, feature scaling, etc. (2) Extracting random samples (with replacement sampling) from the preprocessed real-time data of the lifting operation, and using the extracted samples to form multiple training sets. A classification regression tree is established for each training set to obtain multiple decision trees. Multiple decision trees constitute a forest. In each decision tree, the best attribute among multiple attributes is randomly selected to generate branch nodes. (3) The generated multiple decision trees are retrained, and corresponding weights are set according to the classification effect of the decision tree. Specifically, an initial voting weight is set for each node of the decision tree, and a training set is input into each decision tree. The weight of the node through which the training set sample passes is changed to the ratio of the number of samples judged correctly to the total number of samples reaching this node. The weight of each node is modified according to this method until all samples are classified. Through the above-mentioned weighted random forest algorithm, the weight of the real-time data of the lifting operation can be adjusted, which will help to make subsequent predictions more accurate, continuously monitor the safety risks in the lifting process, and perform self-optimization.

[0043] Furthermore, in some optional embodiments, after the real-time data of the new lifting operation is input into the weighted random forest, the weighted voting of the nodes is used, such as in the form of majority voting, to predict whether the lifting operation is safe.

[0044] Step S22: extract the characteristics of the lifting operation historical data corresponding to the abnormal pattern to obtain risk characteristic data; predict the time series data in the real-time data of the lifting operation after adjusting the weight to obtain the first prediction data; and judge the risk warning according to the similarity between the characteristic data of the first prediction data and the risk characteristic data. Among them, the abnormal pattern refers to the situation where the lifting operation has risks.

[0045] In some optional embodiments, the historical data of the lifting operation corresponding to the abnormal pattern is input into a first convolutional neural network (CNN) for training to obtain risk characteristic data; the time series data in the real-time data of the lifting operation after adjusting the weight is input into a long short-term memory network (LSTM) for training and reasoning to obtain first prediction data; the first prediction data is input into a second convolutional neural network for training to obtain characteristic data of the first prediction data.

[0046] According to the similarity between the characteristic data of the first prediction data and the risk characteristic data, the risk warning is judged, specifically including: selecting methods such as cosine similarity, Euclidean distance, or Pearson correlation coefficient, determining the similarity between the characteristic data of the first prediction data and the risk characteristic data, and setting a risk warning threshold. When the similarity exceeds the risk warning threshold, a risk warning is issued. For example, if the cosine similarity method is adopted, the characteristic data of the first prediction data and the risk characteristic data are respectively converted into two vectors, and the cosine value of the angle between the two vectors is calculated. The similarity between the characteristic data of the first prediction data and the risk characteristic data is measured by the cosine value of the angle.

[0047] In some optional embodiments, the computing tasks of the digital twin system are divided into low-latency tasks and cloud tasks according to local computing resources. Based on the edge computing architecture, computing tasks with high response requirements are performed locally, tasks that exceed the response time of local computing resources are transmitted to the cloud, and the task results are transmitted back to the local to improve the response speed. Among them, local calculation of low-latency tasks includes: parallel calculation of the training process of the first convolutional neural network and the second convolutional neural network and the training and reasoning process of the long short-term memory network through the local GPU; and matrix calculation using the local TPU (Tensor Processing Unit).

[0048] Step S23: When it is determined to issue a risk warning, the digital twin system is risk-labeled according to the corresponding first prediction data.

[0049] Since a digital twin system integrating geometric, physical, behavioral and rule elements is constructed in step S12, the location information corresponding to the risk can be determined based on the first prediction data corresponding to the risk warning, and automatically marked in the digital twin system, so that users can locate the risk in time.

[0050] Real-time two-way data transmission can be achieved between the physical model and the virtual model through efficient wireless data transmission networks such as 5G, LoRa, NB-IoT (Narrow Band Internet of Things), etc. This wireless transmission network can ensure that the environmental and operation data collected by all physical sensors during the lifting process can be transmitted to the virtual model without delay. At the same time, the risk prediction results and optimization suggestions of the virtual model for the lifting operation can also be quickly fed back to the on-site operators, thereby achieving seamless synchronization between the physical and virtual models.

[0051] Based on the Markov property, given the current state and all past states, the conditional probability distribution of the future state of a random process depends only on the current state. Therefore, in some optional embodiments, the component lifting operation safety monitoring and risk management method based on digital twins also includes the following steps: Construct a transfer matrix based on the real-time data of the lifting operation after adjusting the weight. By establishing the transfer matrix of each risk state in the operation process, a multi-dimensional state Markov chain is formed to model and analyze the multi-dimensional data in the lifting process. According to the transfer matrix, determine the probability of risk and the corresponding risk under different operations; according to the probability of risk and the corresponding risk under different operations, determine the high-risk operations and low-risk operations of different risks; according to the high-risk operations and low-risk operations of different risks, issue operation adjustment suggestions. Among them, according to the probability of risk, the operation is divided into high-risk operation and low-risk operation. When the probability of risk under the current operation exceeds the set probability threshold, the high-risk operation and / or low-risk operation corresponding to this risk is prompted, so that the operator can make a more reasonable and safer next operation.

[0052] In some optional embodiments, the component lifting operation safety monitoring and risk management method based on digital twins further includes the following steps: According to the operating steps in the operating standards and safety regulations, the next operation is predicted; when the next operation is a high-risk operation, a risk warning is issued. For example, according to the operating steps in the operating standards and safety regulations, the next operation can be determined, but when the determined next operation is judged as a high-risk operation by the digital twin system under special circumstances, a risk warning is issued to remind the operator to make further judgments before the next operation to avoid risks caused by the mismatch between the operating standards or safety regulations and the construction site.

[0053] In the above method of this embodiment, the present invention can collect and process various types of data in the lifting operation in real time, ensure that the system monitors every link in the lifting process in real time, can quickly respond to any changes on site, and adjust the lifting operation plan in time, greatly improving the safety of the operation. Not only can it identify current safety hazards, but it can also analyze possible risk points in the future based on real-time data and historical data, and issue early warnings to help operators avoid risks in time. By continuously accumulating on-site data and historical experience, it can continuously improve its adaptability to complex construction environments in practical applications, ensuring that more reliable safety guarantees are provided in future lifting operations. By introducing digital twin technology and intelligent management methods, the present invention can greatly improve the level of intelligent safety management of component lifting operations and adapt to different lifting operation requirements.

[0054] The virtual model of this method can not only reflect the real-time data of the physical site, but also has the function of simulation and deduction. Through the fusion analysis of historical data and current real-time data, the virtual model can simulate and predict the future progress of the lifting operation and identify possible construction risk points in advance. Through these simulation results, managers can adjust the lifting plan according to the system prompts to ensure that the lifting operation is carried out under the optimal conditions and avoid unnecessary safety accidents.

[0055] The digital twin system in the above method can automatically identify and manage multiple safety factors in lifting operations, including the load of the lifting equipment, the rationality of the lifting point setting, the stability of the components, and the safe operation behavior of the personnel. The system can not only monitor the potential risks such as equipment failure and personnel misoperation during the construction process, but also automatically manage the safety and provide intelligent early warning for each key operation link based on real-time monitoring data, thereby reducing the time of human intervention and improving the reliability of safety assurance.

[0056] Those skilled in the art can change the above sequence without departing from the protection scope of the present invention.

[0057] Another embodiment of the present invention provides a component lifting operation safety monitoring and risk management platform based on digital twins, such as Figure 4 As shown, it includes: a simulation module and a risk assessment module. Among them, the simulation module is used to collect real-time data of lifting operations and establish a digital twin system; the risk assessment module is used to adjust the weight of at least one type of real-time data of lifting operations in real-time scenarios according to the impact of multiple types of historical data of lifting operations on the safety of lifting operations in different scenarios; extract the characteristics of the historical data of lifting operations corresponding to the abnormal mode to obtain risk characteristic data; predict the time series data in the real-time data of lifting operations after adjusting the weight to obtain the first prediction data; judge the risk warning according to the similarity between the characteristic data of the first prediction data and the risk characteristic data; when it is determined to perform a risk warning, the digital twin system is risk-labeled according to the corresponding first prediction data.

[0058] Regarding the component lifting operation safety monitoring and risk management platform based on digital twin in the above-mentioned embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method and will not be elaborated here.

[0059] The above platform uses a visual interface to display the data monitored during the lifting process, risk assessment results and related safety management suggestions in real time. Furthermore, the above platform integrates data analysis, real-time monitoring and text generation functions to generate special risk management reports according to different user needs, helping managers to better formulate and adjust safety protection measures and ensure that on-site operators can obtain the latest safety guidance information at any time.

[0060] In this embodiment, the platform has functions such as data collection, simulation, risk assessment and early warning management, such as Figure 5 As shown in the figure, the platform displays various key data in the lifting process in real time through a visual interface. The platform can not only intuitively display the current risk status, but also generate historical risk reports and safety analysis results to help managers fully understand the safety status of lifting operations and make adjustments to operations based on the system's safety prompts. For example, managers can use the platform to view the real-time load status of tower cranes, the stress conditions of lifting components, and potential obstacles in the lifting path, so as to develop safer operation plans. Figure 5 This is only an example of a component lifting operation safety monitoring and risk management platform based on digital twins. The implementation of the present invention does not depend on Figure 5 The numbers, graphics, or text in the .

[0061] Based on the same inventive concept, an embodiment of the present invention further provides an electronic device, whose structure is as follows: Figure 6 As shown, it includes: a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the aforementioned digital twin-based component lifting operation safety monitoring and risk management method is implemented.

[0062] Based on the same inventive concept, an embodiment of the present invention also provides a computer storage medium, in which computer executable instructions are stored. When the computer executable instructions are executed by a processor, the aforementioned digital twin-based component lifting operation safety monitoring and risk management method is implemented.

[0063] Any modification, supplement and equivalent replacement made within the scope of the principle of the present invention shall still fall within the scope of the patent of the present invention. "First", "second" and so on do not indicate a sequence, but only indicate the difference between different features.

Claims

1. A component lifting operation safety monitoring and risk management method based on digital twins, characterized in that: The following steps are involved: Collect real-time data of lifting operations and establish a digital twin system; According to the impact of multiple types of lifting operation historical data on lifting operation safety in different scenarios, the weight of at least one type of lifting operation real-time data in the real-time scenario is adjusted; Extract the features of the historical data of the lifting operation corresponding to the abnormal pattern to obtain risk feature data; Predicting the time series data in the real-time data of the weighted lifting operation to obtain first prediction data; Determine the risk warning according to the similarity between the characteristic data of the first prediction data and the risk characteristic data; When it is determined to issue a risk warning, the digital twin system is labeled for risk based on the corresponding first prediction data.

2. The method according to claim 1, characterized in that The component lifting operation safety monitoring and risk management method based on digital twins also includes the following steps: A transfer matrix is ​​established based on the real-time data of the lifting operation after adjusting the weights; According to the transfer matrix, determine the probability of risk and the corresponding risk under different operations; According to the probability of risk occurring under different operations and the corresponding risks, determine high-risk operations and low-risk operations of different risks; Issue operational adjustment suggestions based on high-risk operations and low-risk operations with different risks.

3. The method according to claim 2, characterized in that The component lifting operation safety monitoring and risk management method based on digital twins also includes the following steps: Predict the next operation based on the operating steps in the operating standards and safety regulations; When the next operation is a high-risk operation, a risk warning will be issued.

4. The method according to claim 1, characterized in that The collecting of real-time data of lifting operation comprises the following steps: According to the initial collection frequency of each type of lifting operation real-time data, collect different types of lifting operation real-time data respectively; When the collected real-time data of any type of lifting operation exceeds the collection frequency increase threshold, the collection frequency of the corresponding type of lifting operation real-time data is increased; When the collected real-time data of any type of lifting operation is lower than the collection frequency reduction threshold, the collection frequency of the corresponding type of lifting operation real-time data is reduced.

5. The method according to claim 1, characterized in that The collecting of real-time data of the lifting operation also includes the following steps: Wavelet transform is used to denoise the collected real-time data of lifting operations.

6. The method according to claim 1, characterized in that Input the lifting operation history data corresponding to the abnormal pattern into the first convolutional neural network for training to obtain risk feature data; The time series data in the real-time data of the weighted lifting operation is input into the long short-term memory network for training and reasoning to obtain the first prediction data; The first prediction data is input into the second convolutional neural network for training to obtain feature data of the first prediction data.

7. The method according to claim 6, characterized in that According to local computing resources, the computing tasks of the digital twin system are divided into low-latency tasks and cloud tasks. Local computing is performed on low-latency tasks, including: The training process of the first convolutional neural network and the second convolutional neural network and the training and reasoning process of the long short-term memory network are parallelly calculated through the local GPU; Use local TPU for matrix calculations.

8. A component lifting operation safety monitoring and risk management platform based on digital twins, characterized in that: include: A simulation module is used to collect real-time data of lifting operations and establish a digital twin system; A risk assessment module is used to adjust the weight of at least one type of real-time data of lifting operations in a real-time scenario according to the impact of multiple types of historical data of lifting operations on the safety of lifting operations in different scenarios; extract the characteristics of the historical data of lifting operations corresponding to the abnormal mode to obtain risk characteristic data; Predicting the time series data in the real-time data of the lifting operation after the weight adjustment to obtain first prediction data; judging the risk warning according to the similarity between the characteristic data of the first prediction data and the risk characteristic data; When it is determined to issue a risk warning, the digital twin system is labeled for risk based on the corresponding first prediction data.

9. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the digital twin-based component lifting operation safety monitoring and risk management method as described in any one of claims 1 to 7 when executing the computer program.

10. A computer storage medium, characterized in that: The computer storage medium stores computer executable instructions, which, when executed, implement the component lifting operation safety monitoring and risk management method based on digital twins as described in any one of claims 1 to 7.

Images