Status warning method for a safety protection device used in building construction

Abstract

The present application discloses a method for state warning of a safety protection device for building construction, belonging to the technical field of building construction. It includes: Step 1, real-time monitoring of the deformation amount, inclination angle and vibration acceleration of the protection device, and synchronously collecting the environmental data of the construction site; Step 2, preprocessing the collected data, segmenting the data by using the sliding time window method, and combining the wavelet transform technology to extract the characteristic parameters in each time window to generate a standardized feature vector set; Step 3, calculating the dynamic correlation strength between the features, and using a recursive neural network to analyze the temporal evolution characteristics between the indicators; Step 4, realizing the adaptive adjustment of the weights by minimizing the objective function of the warning error; Step 5, based on the obtained dynamic weight results, establishing a mapping relationship between the weights and the thresholds to realize the collaborative adjustment of the multi-level warning thresholds; Step 6, real-time evaluating the warning performance and intuitively displaying the state of the protection device and the warning information.

Description

Technical Field

[0001] The present application relates to the technical field of construction, and more specifically, to a method for warning the state of a safety protection device for construction. Background Art

[0002] Safety protection devices during construction are crucial for ensuring the safety of construction workers. At the construction site, the protection devices are often in a complex and dynamic environment and may be affected by external factors such as temperature, humidity, wind speed, and vibration, resulting in problems such as deformation, inclination, and vibration. If not detected and warned in time, serious safety accidents may occur. Therefore, how to monitor and evaluate the health status of the protection device in real time and early warn of its potential failure or instability risk is an important link to ensure construction safety.

[0003] Currently, traditional monitoring methods for protection devices usually rely on manual inspections and regular checks and cannot achieve real-time and dynamic state monitoring. Even some intelligent monitoring systems only rely on the monitoring results of a single sensor and cannot comprehensively and accurately evaluate the health status of the device. In the actual construction process, the state of the protection device is affected by multiple factors, especially the dynamic changes in aspects such as the force, deformation, inclination, and vibration of the structure, and these factors need to be comprehensively evaluated through complex data processing and correlation analysis.

[0004] Existing warning systems are mostly based on simple threshold settings and lack in-depth analysis of multi-parameter dynamic changes and adaptive adjustment of multi-level warning thresholds. Traditional methods usually use fixed warning criteria, which are difficult to cope with changes in different construction environments and different construction stages, resulting in warning results that may deviate from the actual situation, and false alarms or missed alarms occur frequently, affecting the accuracy and reliability of the warning system.

[0005] In summary, how to achieve real-time and accurate monitoring of the state of the safety protection device for construction and make coordinated adjustments to multi-level warning thresholds based on dynamic data has become a technical problem that urgently needs to be solved. Summary of the Invention

[0006] In order to overcome a series of defects existing in the prior art, the purpose of the present application is to provide a method for warning the state of a safety protection device for construction, including the following steps.

[0007] Step 1, real-time monitor the deformation amount, inclination angle, and vibration acceleration of the protection device, and synchronously collect the environmental data at the construction site.

[0008] Step 2, preprocess the collected data, segment the data using the sliding time window method, and extract the characteristic parameters within each time window in combination with wavelet transform technology to generate a standardized characteristic vector set.

[0009] Step 3, calculate the dynamic correlation strength between features, and analyze the temporal evolution characteristics among various indicators using a recurrent neural network.

[0010] Step 4, achieve the adaptive adjustment of weights by minimizing the objective function of the warning error.

[0011] Step 5, based on the obtained dynamic weight results, establish the mapping relationship between weights and thresholds, and achieve the coordinated adjustment of multi-level warning thresholds.

[0012] Step 6, evaluate the warning performance in real time and visually display the status of the protection device and warning information.

[0013] Furthermore, Step 1 includes the following steps.

[0014] Arrange a strain sensor array at the main stress points, deformation-sensitive areas, and key connection points to monitor the stress distribution and deformation state of the protection device in real time.

[0015] Install high-precision angle sensors at the corner nodes and tilt-sensitive areas of the protection device, adopt a two-axis or three-axis measurement scheme, continuously monitor the spatial attitude change of the protection device, and at the same time establish a tilt warning threshold system to achieve the real-time evaluation of the stability of the protection device.

[0016] Arrange a three-axis acceleration sensor network on the key components of the protection device, and capture the vibration response characteristics of the protection device during the construction process in real time through high-sampling-frequency data acquisition.

[0017] Deploy an environmental monitoring network at the construction site, including temperature sensors, humidity sensors, wind speed sensors, and rainfall sensors, establish an association analysis mechanism between environmental parameters and structural responses, and evaluate the impact of environmental factors on the performance of the protection device.

[0018] Build a data acquisition and transmission network based on the industrial Internet of Things, adopt a combination of wired and wireless communication methods to ensure the real-time and reliability of data transmission, and at the same time establish a data backup mechanism and network redundancy design to prevent data loss and communication interruption.

[0019] Furthermore, Step 2 includes the following steps.

[0020] Perform outlier detection and elimination on the collected raw data, and at the same time perform noise filtering processing and data normalization to ensure the consistency and comparability of the data.

[0021] Segment the processed data based on the sliding window method, and optimize the window size and sliding step length parameters according to the signal characteristics to ensure that the signal characteristics are accurately reflected.

[0022] Perform wavelet transform on the data sequence of each time window to obtain the multi-scale time-frequency decomposition result of the signal, forming a complete time-frequency domain feature representation.

[0023] Extract energy features, statistical features, and wavelet coefficients from the time-frequency domain feature representation to ensure that the features can comprehensively reflect the essential characteristics of the data.

[0024] Perform standardization processing on the extracted features to eliminate the influence of dimensions, and form feature vectors by combining the standardized features, finally forming a feature vector set.

[0025] Furthermore, segment the processed data based on the sliding window method, and optimize the window size and sliding step parameters according to the signal characteristics to ensure that the signal features are accurately reflected, including the following steps.

[0026] Set the window size W and sliding step S according to the time domain characteristics and frequency features of the signal.

[0027] Extract consecutive windows from the data sequence according to the set window size W and step S.

[0028] Extract signal features from the data within each window to obtain effective signal features.

[0029] Establish a validity verification mechanism for window data, perform quality checks on the data within each window, including data integrity, outlier ratio, and signal-to-noise ratio indicators, and automatically adjust the window parameters when the quality of the window data does not meet the requirements.

[0030] Design a dynamic optimization mechanism for window parameters, and regularly adjust the window size and sliding step by evaluating the feature extraction effect, computing resource occupancy, and response time indicators.

[0031] Furthermore, step 3 includes the following steps.

[0032] Construct a multi-dimensional time series correlation network, use each feature parameter as a network node, determine the initial connection relationship between nodes, and form a basic network structure.

[0033] Calculate the dynamic correlation strength between features within different time periods, quantify the degree of association between nodes, and use the correlation strength as the weight of the network edge to construct a weighted directed graph.

[0034] Use graph theory methods to analyze the network structure, calculate the degree centrality, betweenness centrality, and eigenvector centrality indicators of nodes, identify key nodes and important paths in the network, and reveal the conduction mechanism between main influencing factors.

[0035] Build a long short-term memory network model, take the key features identified in the time-series correlation network as input, and capture the non-linear time-series evolution features between various metrics through the deep learning ability of the recurrent neural network.

[0036] Furthermore, build a multi-dimensional time-series correlation network, take each feature parameter as a network node, determine the initial connection relationship between nodes, and form a basic network structure, including the following steps.

[0037] Map the preprocessed standardized feature parameters to network nodes, assign a unique identifier to each node, and establish a bijective mapping relationship between the feature parameters and the network nodes.

[0038] Calculate the Pearson correlation coefficient and mutual information value between all pairs of nodes.

[0039] Set the connection threshold λ as the determination criterion for the association strength. When the association strength between two nodes exceeds λ, record the corresponding connection relationship in the adjacency matrix.

[0040] Build the initial network topology structure based on the adjacency matrix, and set the edge weights according to the correlation coefficient or mutual information value.

[0041] Perform perturbation analysis within multiple time windows, evaluate the evolution characteristics of the network structure over time, and verify the temporal stability of the association relationship between nodes.

[0042] Furthermore, use graph theory methods to analyze the network structure, calculate the degree centrality, betweenness centrality, and eigenvector centrality metrics of nodes, identify the key nodes and important paths in the network, and reveal the conduction mechanism between the main influencing factors, including the following steps.

[0043] For each node in the network, count the number of directly connected edges to obtain the degree centrality of the node.

[0044] Obtain the betweenness centrality metric of the node by calculating the proportion of all shortest paths in the network that pass through the node.

[0045] Calculate the eigenvector centrality of the node based on the eigenvalue decomposition of the adjacency matrix.

[0046] Comprehensively compare the degree centrality, betweenness centrality, and eigenvector centrality to identify the key nodes in the network and analyze their characteristic attributes.

[0047] Identify the important propagation paths in the network by analyzing the shortest paths and key edges in the network.

[0048] Based on the analysis results of the key nodes and important paths, draw a map of the influence conduction mechanism to clearly show the conduction relationship between the main influencing factors, including the conduction direction, strength, and temporal characteristics.

[0049] Furthermore, step 4 includes the following steps.

[0050] Construct an early warning error objective function, and form an evaluation criterion for comprehensively measuring the early warning performance by weighted summation of the false negative rate and the false positive rate.

[0051] Introduce a forgetting factor mechanism to perform time decay processing on historical data, so that new data samples have high reference value, while maintaining a reasonable inheritance of historical patterns to ensure the dynamic balance of data influence.

[0052] Establish a constraint system for feature weights, including non - negative weight constraints, weight normalization constraints, and maximum step - size limits for weight changes, to ensure the rationality and stability of weight adjustment.

[0053] Design an adaptive weight optimization mechanism, optimize the objective function based on the gradient descent method, determine the direction and step - size of weight adjustment by calculating partial derivatives, and automatically update the weight parameters when new data samples arrive.

[0054] Furthermore, step 5 includes the following steps.

[0055] Based on the dynamic weight results, construct an early warning threshold optimization objective function, comprehensively consider the false positive rate, false negative rate, and early warning timeliness index of different early warning levels, and provide an optimization direction and evaluation criterion for the particle swarm algorithm.

[0056] Initialize the parameter configuration of the particle swarm algorithm, including the number of particles, the number of iterations, learning factors, and inertia weights, and assign initial positions and velocities to each particle, where the position represents the threshold combination of different early warning levels, ensuring that the initial particle swarm has good distribution and diversity.

[0057] Establish a mapping relationship between weights and thresholds. By analyzing the influence of changes in feature index weights on early warning thresholds, construct a mathematical model that can reflect the correlation between the two, providing a theoretical basis for the coordinated adjustment of thresholds.

[0058] Execute the particle swarm optimization process. In each iteration, update the individual optimal position and the global optimal position according to the fitness value of the particle, and combine the mapping relationship between weights and thresholds to dynamically adjust the search direction and step - size of the particle, ensuring rapid convergence to the optimal solution.

[0059] Implement a coordinated adjustment mechanism for multi - level early warning thresholds. When the feature index weights change, automatically update the thresholds of each early warning level based on the established mapping relationship, while ensuring a reasonable interval between different levels.

[0060] Furthermore, step 6 includes the following steps.

[0061] Construct a multi-dimensional early warning performance evaluation index system, covering early warning accuracy, timeliness of early warning, and resource consumption, and establish a standardized calculation method and evaluation criteria.

[0062] Continuously collect data of various performance indicators through data stream processing technology, conduct dynamic evaluation in combination with a sliding time window, and set performance early warning thresholds. When the indicators are abnormal, optimize suggestions are sent in a timely manner.

[0063] Construct an accurate digital model of the protection device based on three-dimensional modeling technology, accurately restore the geometric features and spatial relationships of the equipment, visually display the operating status of the equipment through visual elements, and add dynamic monitoring points at key parts to achieve real-time update of status information.

[0064] Overlay and display the early warning information on the three-dimensional model of the equipment in the form of floating labels or warning icons, supporting multi-angle viewing and information screening.

[0065] Compared with the prior art, the present application has the following beneficial effects.

[0066] The present application improves the accuracy of status monitoring and the effectiveness of early warning of safety protection devices in building construction by real-time monitoring the deformation, tilt, and vibration data of the protection device, and comprehensively analyzing in combination with environmental parameters, and realizing the optimized adjustment of multi-level early warning thresholds through dynamic weight adjustment and recurrent neural network. Description of the Drawings

[0067] Figure 1 It is a schematic flowchart of a method for early warning the status of a safety protection device for building construction disclosed in an embodiment of the present application. Detailed Embodiments

[0068] To make the purpose, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be described in more detail below with reference to the accompanying drawings in the embodiments of the present invention. In the drawings, the same or similar reference numerals represent the same or similar elements or elements with the same or similar functions from beginning to end. The described embodiments are some, but not all, of the embodiments of the present invention.

[0069] Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0070] The embodiments described below with reference to the accompanying drawings and directional terms are exemplary only and are intended to explain the present invention and should not be construed as limiting the present invention.

[0071] As Figure 1 shown, a method for early warning the status of a safety protection device for building construction includes the following steps.

[0072] Step 1: Real-time monitor the deformation amount, tilt angle, and vibration acceleration of the protection device, and synchronously collect the environmental data at the construction site.

[0073] Step 2: Preprocess the collected data. Use the sliding time window method to segment the data, and combine wavelet transform technology to extract the characteristic parameters within each time window, generating a standardized feature vector set.

[0074] Step 3: Calculate the dynamic correlation strength between the features, and use a recurrent neural network to analyze the temporal evolution characteristics among the various indicators.

[0075] Step 4: Achieve the adaptive adjustment of the weights by minimizing the objective function of the warning error.

[0076] Step 5: Based on the obtained dynamic weight results, establish the mapping relationship between the weights and the thresholds, and realize the collaborative adjustment of the multi-level warning thresholds.

[0077] Step 6: Real-time evaluate the warning performance and intuitively display the status of the protection device and the warning information.

[0078] The core of Step 1 is to real-time monitor the key physical parameters of the protection device and the construction environment. The deformation amount, tilt angle, and vibration acceleration of the protection device are important indicators for evaluating its structural stability and stress conditions, which are directly related to whether the protection device can effectively perform its protective role. The deformation amount monitoring can reflect the deformation degree of the safety protection device under external loads, the tilt angle can indicate whether the stress is balanced, and the vibration acceleration provides a basis for revealing whether the safety protection device generates unsafe vibrations due to dynamic loads during the construction process. At the same time, environmental data (such as temperature, humidity, wind speed, etc.) can provide the influence of external conditions on the performance of the protection device. For example, extreme weather may cause excessive deformation or vibration of the structure. Therefore, the synchronous collection of environmental data provides additional information for comprehensively evaluating the status of the protection device.

[0079] The data preprocessing in Step 2 is the basis for ensuring the accuracy and effectiveness of subsequent analysis. The sliding time window method is used to segment the collected data, which can divide the data into several segments according to the time sequence, thus realizing the dynamic analysis of the data. The wavelet transform technology can effectively extract different frequency components in the signal, helping to identify the minute changes of the protection device in different time periods, and then more accurately analyze the possible potential safety hazards. Through wavelet transform, not only can noise be removed, but also the sensitivity to abnormal situations can be enhanced. In addition, the generation of the standardized feature vector set unifies the dimensions of different feature data, avoiding the analysis errors caused by different data dimensions, and laying a foundation for subsequent machine learning analysis.

[0080] The technical highlight of Step 3 lies in capturing the temporal evolution relationship between features through a Recurrent Neural Network (RNN). RNN is particularly suitable for processing time series data and can automatically identify the temporal patterns in the data. Therefore, it can effectively analyze the state evolution of the protection device. By calculating the dynamic correlation strength between features, the interaction between different physical quantities and its impact on the device's safety can be revealed. For example, there may be a certain temporal dependence relationship between the tilt angle and vibration acceleration of the protection device, and RNN can capture these non-linear temporal features, thereby improving the accuracy and reliability of the early warning system.

[0081] Step 4 optimizes the prediction performance of the system by minimizing the early warning error. The design of the objective function can automatically adjust the weights of each feature according to the actual early warning effect of historical data. Through iterative optimization, the influence degree of feature parameters can be flexibly adjusted according to the actual situation to reduce the false alarm rate and missed alarm rate. For example, when a false early warning is issued, the objective function will adjust the corresponding weights to make the next round of prediction more accurate. The adaptive adjustment mechanism can maintain a high early warning performance under different construction sites and different environmental conditions.

[0082] In Step 5, based on the dynamic weight adjustment in the previous step, a mapping relationship between the weight and the early warning threshold is established. This means that under different construction stages or different safety levels, the early warning threshold can be dynamically adjusted to ensure that it can respond to potential dangers in a timely manner without overreacting. Through the coordinated adjustment of multi-level thresholds, adaptable and refined safety protection can be provided at multiple levels. For example, at some key construction nodes (such as high-altitude operations, large machinery operations, etc.), a lower early warning threshold can be set, while in other conventional operating environments, the threshold can be appropriately increased to reduce unnecessary early warnings.

[0083] In Step 6, the real-time evaluation of the early warning performance and the intuitive display of the state of the protection device and the early warning information are to facilitate construction workers and safety management personnel to timely understand the current safety situation. This link displays the early warning results through a graphical interface or an alarm system to help the staff make a quick response. The real-time evaluation function can further optimize the early warning mechanism and continuously improve the system performance through a feedback mechanism. This measure not only improves the safety management efficiency but also enhances the safety awareness of on-site workers.

[0084] Generally speaking, through the integration of various advanced technologies such as the sliding time window method, wavelet transform, and recurrent neural network, this status warning method realizes the dynamic monitoring and warning of safety protection devices in building construction. The technical measures at each step cooperate with each other, forming a complete closed-loop mechanism from data collection, preprocessing, feature analysis, to warning model optimization and real-time feedback. The application of this method can not only effectively improve the safety during the construction process and reduce the possibility of accidents, but also ensure that the device can maintain an efficient and reliable working state in a complex and changeable construction environment through adaptive adjustment and multi-level warning mechanisms, with high practical value and technical prospects.

[0085] Furthermore, Step 1 includes the following steps.

[0086] Arrange a strain sensor array at the main stress points, deformation-sensitive areas, and key connection points to continuously monitor the stress distribution and deformation state of the protection device.

[0087] Install high-precision angle sensors at the corner nodes and tilt-sensitive areas of the protection device, adopt a two-axis or three-axis measurement scheme, continuously monitor the spatial attitude change of the protection device, and at the same time establish a tilt warning threshold system to realize the real-time evaluation of the stability of the protection device.

[0088] Arrange a three-axis acceleration sensor network on the key components of the protection device, and through high-sampling-frequency data collection, capture the vibration response characteristics of the protection device during the construction process in real time.

[0089] Deploy an environmental monitoring network at the construction site, including temperature sensors, humidity sensors, wind speed sensors, and rainfall sensors, establish a correlation analysis mechanism between environmental parameters and structural responses, and evaluate the impact of environmental factors on the performance of the protection device.

[0090] Build a data collection and transmission network based on the industrial Internet of Things, adopt a combination of wired and wireless communication methods to ensure the real-time and reliability of data transmission, and at the same time establish a data backup mechanism and network redundancy design to prevent data loss and communication interruption.

[0091] In summary, by arranging a strain sensor array at the main stress points, deformation-sensitive areas, and key connection points to monitor the stress distribution and deformation state of the protection device in real time; installing high-precision angle sensors at the corner nodes and tilt-sensitive areas to continuously monitor the spatial attitude changes and establish a tilt warning threshold system; arranging a three-axis acceleration sensor network on the key components to capture the vibration response characteristics in real time; deploying an environmental monitoring network to evaluate the impact of environmental factors on the performance of the protection device; and building a data acquisition and transmission network based on the industrial Internet of Things to ensure the real-time and reliability of data transmission, this series of technical measures jointly construct a comprehensive, efficient, and reliable status monitoring and warning system for the building construction safety protection device. Each part conducts refined monitoring and analysis on the safety, stability, and reliability of the protection device. Through real-time data acquisition, dynamic evaluation, and warning mechanisms, it can effectively prevent the occurrence of safety accidents and ensure the lives of construction workers and the quality of the project.

[0092] Furthermore, the vibration response characteristics of the protection device during the construction process are expressed by the formula: , where a total (t) represents the comprehensive acceleration signal of the protection device in all directions, which is the synthesized value of the three-axis acceleration signals and is used to measure the overall vibration response; a x (t) represents the acceleration signal of the protection device in the x-axis direction; a y (t) represents the acceleration signal of the protection device in the y-axis direction; a z (t) represents the acceleration signal of the protection device in the z-axis direction.

[0093] In summary, the comprehensive acceleration signal can comprehensively reflect the vibration response of the protection device during the construction process. This signal can not only effectively capture the vibration changes in all directions but also provide data support for vibration source localization, dynamic load identification, and structural safety assessment. With the help of this formula, the state of the protection device can be analyzed in real time, and a warning can be issued when the vibration exceeds the safety threshold, thus providing effective protection for construction safety.

[0094] Furthermore, Step 2 includes the following steps.

[0095] Detect and remove outliers from the collected raw data, perform noise filtering, and execute data normalization to ensure the consistency and comparability of the data.

[0096] Segment the processed data based on the sliding window method and optimize the window size and sliding step parameters according to the signal characteristics to ensure that the signal characteristics are accurately reflected.

[0097] Perform wavelet transform on the data sequence of each time window to obtain the multi-scale time-frequency decomposition result of the signal and form a complete time-frequency domain feature representation.

[0098] Extract energy features, statistical features, and wavelet coefficients from the time-frequency domain feature representation to ensure that the features can comprehensively reflect the essential characteristics of the data.

[0099] Standardize the extracted features to eliminate the influence of dimensions, and form feature vectors by combining the standardized features, ultimately forming a feature vector set.

[0100] In summary, in step 2, through detailed preprocessing, segmentation, transformation, and feature extraction of the collected data, high-quality data support is provided for subsequent analysis and prediction. By removing outliers and noise, the data quality is guaranteed; the sliding time window method and wavelet transform effectively extract valuable time-frequency domain features from the signal; and feature standardization ensures the consistency between different features, laying a solid foundation for further pattern recognition and modeling. Finally, after this series of processes, the obtained standardized feature vector set can fully reflect the dynamic characteristics of the protection device and support subsequent early warning and risk assessment work.

[0101] Furthermore, based on the sliding window method, segment the processed data, and optimize the window size and sliding step parameters according to the signal characteristics to ensure that the signal features are accurately reflected, including the following steps.

[0102] Set the window size W and sliding step S according to the time-domain characteristics and frequency features of the signal.

[0103] Extract consecutive windows from the data sequence according to the set window size W and step S.

[0104] Extract signal features from the data within each window to obtain effective signal features.

[0105] Establish a validity verification mechanism for window data to conduct quality inspections on the data within each window, including data integrity, outlier ratio, and signal-to-noise ratio indicators. When the quality of the window data does not meet the requirements, automatically adjust the window parameters.

[0106] Design a dynamic optimization mechanism for window parameters, and regularly adjust the window size and sliding step by evaluating the feature extraction effect, computing resource occupancy, and response time indicators.

[0107] In summary, by segmenting the data based on the sliding window method and optimizing the window size and sliding step parameters, the accurate extraction of signal features is achieved. First, the window size and sliding step are set according to the time-domain and frequency-domain characteristics of the signal. Then, consecutive windows are extracted from the data sequence, and the features of the data within each window are extracted. At the same time, a window data validity verification mechanism is established to check the data quality. If the requirements are not met, the window parameters are automatically adjusted. In addition, a dynamic optimization mechanism is designed to regularly adjust the window parameters according to indicators such as feature extraction effect, computing resource occupancy, and response time, thereby ensuring the accuracy and efficiency of signal feature extraction while optimizing the use of computing resources.

[0108] Furthermore, perform wavelet transform on the data sequence of each time window to obtain the multi-scale time-frequency decomposition result of the signal, forming a complete time-frequency domain feature representation, including the following steps.

[0109] For the data sequence within each time window, perform multi-scale decomposition using the selected wavelet basis function to decompose the signal into approximation components and detail components of different frequency components.

[0110] Set the decomposition level according to the analysis requirements and obtain the wavelet coefficients corresponding to each decomposition level.

[0111] Reorganize and arrange the obtained wavelet coefficients in two dimensions of time and frequency to construct a complete time-frequency feature representation matrix, realizing the joint time-frequency domain representation of the signal.

[0112] In summary, through wavelet transform on the data sequence of each time window, the multi-scale time-frequency decomposition of the signal can effectively extract different frequency components in the signal, including both the low-frequency steady trend and the high-frequency instantaneous changes. Wavelet transform decomposes the signal into approximation components and detail components, enabling the time information and frequency information of the signal to be fully represented at multiple scales. By setting the decomposition level according to the analysis requirements and extracting the corresponding wavelet coefficients, the characteristics of the signal at different scales can be captured more accurately. Finally, these wavelet coefficients are reorganized and arranged in the time and frequency dimensions to form a complete time-frequency domain feature representation matrix, realizing the joint time-frequency representation of the signal. This extraction and representation of time-frequency domain features can not only effectively improve the recognition accuracy of complex signals but also provide rich information support for subsequent signal analysis, anomaly detection, and early warning.

[0113] Furthermore, the energy feature is expressed by the formula: , where E j is the energy of the j-th layer wavelet transform; W j (k) is the value of the k-th wavelet coefficient in the j-th layer wavelet transform; N is the number of signal sampling points after wavelet transform.

[0114] The statistical features are expressed by the formula: , where is the kurtosis of the wavelet coefficients at the j-th layer, representing the sharpness of the data distribution; σ j is the variance of the wavelet coefficients at the j-th layer, representing the degree of dispersion or the fluctuation range of the data.

[0115] The wavelet coefficients are expressed by the formula: , where x(n) is the original signal; is the wavelet basis function.

[0116] In summary, through the analysis of the above formulas, the energy features, statistical features, and wavelet coefficients provide rich information for the analysis of vibration signals. The energy features can reflect the intensity changes of the signals, the statistical features reveal the distribution patterns and fluctuation characteristics of the signals, and the wavelet coefficients capture the time-frequency features of the signals through multi-scale decomposition. The extraction of these features provides necessary support for subsequent pattern recognition, anomaly detection, and early warning, helping to identify potential risks and fault hazards during the construction process, thus ensuring the safety of the protection device.

[0117] Furthermore, step 3 includes the following steps.

[0118] Construct a multi-dimensional time-series correlation network, use each feature parameter as a network node, determine the initial connection relationship between the nodes, and form a basic network structure.

[0119] Calculate the dynamic correlation strength between the features within different time periods, quantify the degree of association between the nodes, and use the correlation strength as the weight of the network edges to construct a weighted directed graph.

[0120] Apply graph theory methods to analyze the network structure, calculate the degree centrality, betweenness centrality, and eigenvector centrality indicators of the nodes, identify the key nodes and important paths in the network, and reveal the conduction mechanism between the main influencing factors.

[0121] Construct a long short-term memory network model, use the key features identified in the time-series correlation network as inputs, and capture the non-linear time-series evolution features between the indicators through the deep learning ability of the recurrent neural network.

[0122] In summary, through the technical means of constructing a multi-dimensional time-series correlation network and combining graph theory analysis in Step 3, the dynamic correlations among the characteristics of the protection device and the conduction mechanism of their impact on safety are effectively revealed. By calculating the dynamic correlation strength between characteristics and constructing a weighted graph, the relationships between characteristics can be quantified, and then key characteristics and important paths can be identified. Combining with the long short-term memory network model, complex non-linear time-series evolution characteristics can be captured, providing strong technical support for the state warning of the protection device. This method has important application value in real-time monitoring, risk prediction, and structural safety assessment, and can provide more accurate safety guarantees for building construction.

[0123] Furthermore, to construct a multi-dimensional time-series correlation network, each characteristic parameter is used as a network node, and the initial connection relationship between nodes is determined to form a basic network structure, including the following steps.

[0124] Map the preprocessed standardized characteristic parameters to network nodes, and assign a unique identifier to each node to establish a bijective mapping relationship between the characteristic parameters and the network nodes.

[0125] Calculate the Pearson correlation coefficient and mutual information value between all pairs of nodes.

[0126] Set the connection threshold λ as the criterion for judging the correlation strength. When the correlation strength between two nodes exceeds λ, record the corresponding connection relationship in the adjacency matrix.

[0127] Based on the adjacency matrix, construct the initial network topology structure, and set the weight of the edge according to the correlation coefficient or mutual information value.

[0128] Conduct perturbation analysis within multiple time windows to evaluate the evolution characteristics of the network structure over time and verify the temporal stability of the correlation relationship between nodes.

[0129] In summary, by constructing a multi-dimensional time-series correlation network and using each characteristic parameter as a network node, the correlation relationships among various characteristic parameters can be effectively captured, forming a dynamic network structure. By mapping the standardized characteristic parameters to network nodes and assigning a unique identifier to each node, the precise correspondence between nodes and characteristics is ensured, facilitating subsequent analysis. After calculating the Pearson correlation coefficient and mutual information value between node pairs, setting an appropriate connection threshold ensures that only those nodes with significant correlations are connected, thus avoiding over-correlation and maintaining the simplicity of the network. By establishing an adjacency matrix and using the correlation coefficient or mutual information value as the weight of the edge, the topological structure of the network can truly reflect the strong and weak correlation relationships between characteristics. Further, by evaluating the evolution characteristics of the network over time through perturbation analysis, it helps to understand the temporal stability of the correlation relationship between nodes, providing profound insights at the temporal level for anomaly detection and prediction.

[0130] Furthermore, the degree of association between nodes is expressed by the formula: , where W(X, Y, μ) represents the weighted dynamic association strength between feature X and feature Y within time period μ; T is the total length of the time period; α is the decay factor, which controls the influence degree of data in a relatively distant time period on the current association strength; a T-τ is the dynamic adjustment of the decay factor α at time τ, where: when τ is close to T, a T-τ will approach 1, meaning that data at a relatively recent time point has a greater impact on the current association strength; when τ is far from T, a T-τ will approach 0, meaning that data at a relatively distant time point has a smaller impact on the current association strength; r(X τ , Y τ ) is the Pearson correlation coefficient calculated within time period τ.

[0131] In summary, the weighted dynamic association strength between nodes provides a flexible and efficient method to quantify the temporal relationship between features. By introducing a decay factor to down-weight data in a distant time period, the influence of data in different periods on the current association strength can be controlled more precisely, which is particularly important in the dynamic analysis of time-series data. Combining with the Pearson correlation coefficient, the linear dependence relationship between features can be effectively quantified, thereby identifying key features and influence paths. This method not only improves the timeliness and accuracy of the status warning of the protection device, but also provides a theoretical basis for more complex time-series data analysis.

[0132] Furthermore, use graph theory methods to analyze the network structure, calculate the degree centrality, betweenness centrality, and eigenvector centrality indicators of nodes, identify key nodes and important paths in the network, and reveal the conduction mechanism between main influencing factors, including the following steps.

[0133] For each node in the network, count the number of edges directly connected to it to obtain the degree centrality of the node.

[0134] Obtain the betweenness centrality indicator of the node by calculating the proportion of all shortest paths in the network that pass through this node.

[0135] Based on the eigenvalue decomposition of the adjacency matrix, calculate the eigenvector centrality of the node.

[0136] Comprehensively compare the degree centrality, betweenness centrality, and eigenvector centrality to identify key nodes in the network and analyze their characteristic attributes.

[0137] Identify important propagation paths in the network by analyzing the shortest paths and key edges in the network.

[0138] Based on the analysis results of key nodes and important paths, draw a map of the influencing conduction mechanism to clearly show the conduction relationship between the main influencing factors, including the conduction direction, intensity, and timing characteristics.

[0139] In summary, by using graph theory methods to analyze the network structure, key nodes and important paths in the network can be deeply identified, and the conduction mechanism between the main influencing factors can be revealed. By calculating the degree centrality, betweenness centrality, and eigenvector centrality of nodes, the importance of nodes in the network and their impact on the overall structure can be comprehensively evaluated. Degree centrality reflects the direct influence of a node in the network by counting the number of edges directly connected to the node; betweenness centrality reveals the importance of a node as an information dissemination intermediary by calculating the node's participation in the shortest paths; eigenvector centrality measures the global importance of a node in the network through the eigenvalue decomposition of the adjacency matrix. The combined use of these metrics enables the accurate identification of key nodes in the network, providing data support for further analysis. In addition, analyzing the shortest paths and critical edges helps to identify the most important propagation paths in the network, thereby revealing the conduction mechanism between the influencing factors. Finally, by drawing a map of the influencing conduction mechanism, the conduction direction, intensity, and timing characteristics can be clearly shown, providing a strong basis for network optimization, anomaly warning, and information dissemination management.

[0140] Furthermore, step 4 includes the following steps.

[0141] Construct an early warning error objective function, and form an evaluation criterion for comprehensively measuring the early warning performance by weighted summation of the false negative rate and false positive rate.

[0142] Introduce a forgetting factor mechanism to perform time decay processing on historical data, making new data samples highly valuable for reference while maintaining a reasonable inheritance of historical patterns to ensure the dynamic balance of data influence.

[0143] Establish a constraint system for feature weights, including non - negative weight constraints, weight normalization constraints, and maximum step - size limits for weight changes, to ensure the rationality and stability of weight adjustment.

[0144] Design an adaptive weight optimization mechanism to optimize the objective function based on the gradient descent method, determine the direction and step - size of weight adjustment by calculating partial derivatives, and automatically update the weight parameters when new data samples arrive.

[0145] In summary, in Step 4, by constructing an early warning error objective function and introducing a forgetting factor mechanism, the early warning performance is effectively optimized. The objective functions of weighted false alarm rate and missed alarm rate provide a comprehensive evaluation criterion, which helps to find the best balance between different error types. The introduction of the forgetting factor ensures that the impact of new data on the model is more prominent, while maintaining a reasonable inheritance of historical patterns, guaranteeing the adaptability and stability of the model. The constraint system of feature weights and the adaptive weight optimization mechanism further enhance the robustness and self-learning ability of the model, enabling it to respond to changes in real time and optimize the prediction results.

[0146] Furthermore, Step 5 includes the following steps.

[0147] Based on the dynamic weight results, construct an early warning threshold optimization objective function, comprehensively consider the false alarm rate, missed alarm rate, and early warning timeliness index of different early warning levels, and provide an optimization direction and evaluation criterion for the particle swarm algorithm.

[0148] Initialize the parameter configuration of the particle swarm algorithm, including the number of particles, the number of iterations, learning factors, and inertia weight, and assign initial positions and velocities to each particle, where the position represents the threshold combination of different early warning levels, ensuring that the initial particle swarm has good distribution and diversity.

[0149] Establish a mapping relationship between weights and thresholds. By analyzing the impact of changes in feature index weights on early warning thresholds, construct a mathematical model that can reflect the correlation between the two, providing a theoretical basis for the coordinated adjustment of thresholds.

[0150] Execute the particle swarm optimization process. In each iteration, update the individual optimal position and the global optimal position according to the fitness value of the particle, and combine the mapping relationship between weights and thresholds to dynamically adjust the search direction and step size of the particle, ensuring that it can quickly converge to the optimal solution.

[0151] Implement a coordinated adjustment mechanism for multi-level early warning thresholds. When the feature index weights change, automatically update the thresholds of each early warning level based on the established mapping relationship, while ensuring a reasonable interval between different levels.

[0152] In summary, in Step 5, by constructing an objective function for early warning threshold optimization, initializing the parameters of the particle swarm optimization algorithm, and through the mapping relationship between weights and thresholds and the execution of the particle swarm optimization process, a coordinated adjustment mechanism for multi-level early warning thresholds is successfully implemented. This process can ensure that the early warning automatically updates the early warning thresholds under different changes in feature weights, thereby improving the response speed and accuracy. Through the application of the particle swarm optimization algorithm, the optimal combination of early warning thresholds can be efficiently searched, further enhancing the status monitoring and early warning capabilities of the protection device, and thus enhancing the safety guarantee level of building construction.

[0153] Further, step 6 includes the following steps.

[0154] Construct a multi-dimensional early warning performance evaluation index system, covering early warning accuracy, early warning timeliness, and resource consumption, and establish a standardized calculation method and evaluation criteria.

[0155] Continuously collect data of various performance indicators through data stream processing technology, conduct dynamic evaluation in combination with a sliding time window, and set performance early warning thresholds. When the indicators are abnormal, optimize suggestions are sent out in a timely manner.

[0156] Based on three-dimensional modeling technology, construct an accurate digital model of the protection device, accurately restore the geometric features and spatial relationships of the equipment, intuitively display the operating status of the equipment through visual elements, and add dynamic monitoring points at key positions to achieve real-time update of status information.

[0157] Overlay and display the early warning information on the three-dimensional model of the equipment in the form of floating labels or warning icons, supporting multi-angle viewing and information filtering.

[0158] In summary, step 6 provides comprehensive technical support for the status monitoring and early warning of the protection device through constructing a multi-dimensional early warning performance evaluation index system, real-time data stream processing and dynamic evaluation, accurate three-dimensional modeling technology, and intuitive early warning information display. The application of these technologies not only improves the early warning accuracy and timeliness, but also optimizes resource consumption and enhances the visual management of the equipment. By integrating these technologies, the construction site personnel can monitor and manage the protection device more efficiently and accurately, ensuring the maximization of construction safety and operation efficiency.

[0159] Finally, it should be pointed out that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for state warning of a safety protection device for building construction, characterized in that, It includes the following steps: Step 1: Monitor the deformation amount, tilt angle, and vibration acceleration of the protective device in real time, and synchronously collect the environmental data at the construction site; Step 2: Preprocess the collected data, segment the data using the sliding time window method, extract the characteristic parameters within each time window in combination with wavelet transform technology, and generate a standardized feature vector set; Step 3: Calculate the dynamic correlation strength between the features, and analyze the temporal evolution characteristics among the indicators using a recurrent neural network; Step 4: Achieve the adaptive adjustment of the weights by minimizing the objective function of the warning error; Step 5: Based on the obtained dynamic weight results, establish the mapping relationship between the weights and the thresholds, and realize the coordinated adjustment of the multi-level warning thresholds; Step 6: Evaluate the warning performance in real time and visually display the status of the protective device and the warning information; Step 1 includes the following steps: Arrange a strain sensor array at the main stress points, deformation-sensitive areas, and key connection points to monitor the stress distribution and deformation state of the protective device in real time; Install high-precision angle sensors at the corner nodes and tilt-sensitive areas of the protective device, adopt a two-axis or three-axis measurement scheme, continuously monitor the spatial attitude change of the protective device, and at the same time establish a tilt warning threshold system to realize the real-time evaluation of the stability of the protective device; Arrange a three-axis acceleration sensor network on the key components of the protective device, and capture the vibration response characteristics of the protective device during the construction process in real time through high-sampling-frequency data acquisition; Deploy an environmental monitoring network at the construction site, including temperature sensors, humidity sensors, wind speed sensors, and rainfall sensors, establish an association analysis mechanism between environmental parameters and structural responses, and evaluate the impact of environmental factors on the performance of the protective device; Build a data acquisition and transmission network based on the industrial Internet of Things, adopt a combination of wired and wireless communication methods to ensure the real-time and reliability of data transmission, and at the same time establish a data backup mechanism and network redundancy design to prevent data loss and communication interruption; The vibration response characteristics of the protection device during construction are expressed by the formula: , where a total (t) represents the comprehensive acceleration signal of the protection device in all directions. It is the synthesized value of the triaxial acceleration signals and is used to measure the overall vibration response; a x (t) represents the acceleration signal of the protection device in the x-axis direction; a y (t) represents the acceleration signal of the protection device in the y-axis direction; a z (t) represents the acceleration signal of the protection device in the z-axis direction; Step 2 includes the following steps: Detect and remove outliers from the collected raw data, perform noise filtering processing at the same time, and perform data normalization to ensure the consistency and comparability of the data; Segment the processed data based on the sliding window method, and optimize the window size and sliding step parameters according to the signal characteristics to ensure that the signal features are accurately reflected; Perform wavelet transform on the data sequence of each time window to obtain the multi-scale time-frequency decomposition result of the signal, and form a complete time-frequency domain feature representation; Extract energy features, statistical features, and wavelet coefficients from the time-frequency domain feature representation to ensure that the features can comprehensively reflect the essential characteristics of the data; Standardize the extracted features to eliminate the influence of dimensions, and form a feature vector by combining the standardized features, and finally form a feature vector set; Segment the processed data based on the sliding window method, and optimize the window size and sliding step parameters according to the signal characteristics to ensure that the signal features are accurately reflected, including the following steps: Set the window size W and sliding step S according to the time-domain characteristics and frequency features of the signal; Extract continuous windows from the data sequence according to the set window size W and step S; Extract signal features from the data within each window to obtain effective signal features; Establish a validity verification mechanism for window data to perform quality checks on the data within each window, including data integrity, outlier ratio, and signal-to-noise ratio metrics. When the quality of the window data does not meet the requirements, automatically adjust the window parameters; Design a dynamic optimization mechanism for window parameters to regularly adjust the window size and sliding step by evaluating the feature extraction effect, computing resource occupancy, and response time metrics. Perform wavelet transform on the data sequence of each time window to obtain the multi-scale time-frequency decomposition result of the signal, forming a complete time-frequency domain feature representation, including the following steps: For the data sequence within each time window, perform multi-scale decomposition using the selected wavelet basis function to decompose the signal into approximate components and detail components of different frequency components; Set the decomposition level according to the analysis requirements to obtain the wavelet coefficients corresponding to each decomposition level; Reorganize and arrange the obtained wavelet coefficients in two dimensions of time and frequency to construct a complete time-frequency feature representation matrix, realizing the joint time-frequency domain representation of the signal. The energy feature is expressed by the formula: , where E j is the energy of the j-th layer wavelet transform; W j (k) is the value of the k-th wavelet coefficient in the j-th layer wavelet transform; N is the number of signal sampling points after wavelet transform; The statistical features are expressed by the formula: , where is the kurtosis of the wavelet coefficients at the j-th layer, representing the sharpness of the data distribution; σ j is the variance of the wavelet coefficients at the j-th layer, representing the degree of dispersion or the range of fluctuations of the data. The wavelet coefficients are expressed by the formula: , where x(n) is the original signal; is the wavelet basis function; Step 3 includes the following steps: Construct a multi-dimensional time series correlation network, use each feature parameter as a network node, determine the initial connection relationship between nodes, and form a basic network structure; Calculate the dynamic correlation strength between features within different time periods, quantify the correlation degree between nodes, and use the correlation strength as the weight of the network edge to construct a weighted directed graph; Use graph theory methods to analyze the network structure, calculate the degree centrality, betweenness centrality, and eigenvector centrality indicators of nodes, identify the key nodes and important paths in the network, and reveal the conduction mechanism between the main influencing factors; Construct a long short-term memory network model, use the key features identified in the time series correlation network as inputs, and capture the non-linear time series evolution features between each indicator through the deep learning ability of the recurrent neural network. Construct a multi-dimensional time series correlation network, use each feature parameter as a network node, determine the initial connection relationship between nodes, and form a basic network structure, including the following steps: Map the preprocessed standardized feature parameters to network nodes, assign a unique identifier to each node, and establish a bijective mapping relationship between the feature parameters and network nodes; Calculate the Pearson correlation coefficient and mutual information value between all pairs of nodes; Set the connection threshold λ as the determination criterion for the correlation strength. When the correlation strength between two nodes exceeds λ, record the corresponding connection relationship in the adjacency matrix; Construct the initial network topology structure based on the adjacency matrix, and set the weight of the edge according to the correlation coefficient or mutual information value; Perform perturbation analysis within multiple time windows to evaluate the evolution characteristics of the network structure over time and verify the temporal stability of the correlation relationship between nodes. Using graph theory methods to analyze the network structure, calculate the degree centrality, betweenness centrality, and eigenvector centrality indicators of nodes, identify the key nodes and important paths in the network, and reveal the conduction mechanism between the main influencing factors, including the following steps: For each node in the network, count the number of directly connected edges to obtain the degree centrality of the node; obtain the betweenness centrality indicator of the node by calculating the proportion of all shortest paths in the network that pass through the node; calculate the eigenvector centrality of the node based on the eigenvalue decomposition of the adjacency matrix; comprehensively compare the degree centrality, betweenness centrality, and eigenvector centrality to identify the key nodes in the network and analyze their characteristic attributes; identify the important propagation paths in the network by analyzing the shortest paths and key edges in the network; based on the analysis results of the key nodes and important paths, draw a map of the influence conduction mechanism to clearly show the conduction relationship between the main influencing factors, including the conduction direction, intensity, and timing characteristics.

2. The state warning method of a safety protection device for building construction according to claim 1, characterized in that, Step 4 includes the following steps: Construct an early warning error objective function, and form an evaluation criterion for comprehensively measuring the early warning performance by weighted summation of the false negative rate and false positive rate. Introduce a forgetting factor mechanism to perform time decay processing on historical data, so that new data samples have high reference value, while maintaining a reasonable inheritance of historical patterns to ensure the dynamic balance of data influence. Establish a constraint system for feature weights, including non-negative weight constraints, weight normalization constraints, and maximum step size limits for weight changes, to ensure the rationality and stability of weight adjustment. Design an adaptive weight optimization mechanism, optimize the objective function based on the gradient descent method, determine the direction and step size of weight adjustment by calculating partial derivatives, and automatically update the weight parameters when new data samples arrive.

3. The status warning method of a safety protection device for building construction according to claim 1, characterized in that, Step 5 includes the following steps: Based on the dynamic weight results, construct an early warning threshold optimization objective function, comprehensively consider the false positive rate, false negative rate, and early warning timeliness indicators of different early warning levels, and provide an optimization direction and evaluation criterion for the particle swarm algorithm. Initialize the parameter configuration of the particle swarm algorithm, including the number of particles, the number of iterations, learning factors, and inertia weights, and assign initial positions and velocities to each particle, where the position represents the threshold combination of different early warning levels, ensuring that the initial particle swarm has good distribution and diversity. Establish a mapping relationship between weights and thresholds, and construct a mathematical model that can reflect the correlation between the two by analyzing the influence of changes in feature index weights on early warning thresholds, providing a theoretical basis for the coordinated adjustment of thresholds. Execute the particle swarm optimization process. In each iteration, update the individual optimal position and global optimal position according to the fitness value of the particle, and combine the mapping relationship between weights and thresholds to dynamically adjust the search direction and step size of the particle to ensure rapid convergence to the optimal solution. Implement a coordinated adjustment mechanism for multi-level early warning thresholds. When the feature index weights change, automatically update the thresholds of each early warning level based on the established mapping relationship, while ensuring a reasonable interval between different levels.

4. The state warning method of a safety protection device for building construction according to claim 1, characterized in that, Step 6 includes the following steps: Construct a multi-dimensional early warning performance evaluation index system, covering early warning accuracy, timeliness of early warning and resource consumption, and establish a standardized calculation method and evaluation criteria; Continuously collect data of various performance indicators through data stream processing technology, conduct dynamic evaluation in combination with a sliding time window, and set performance early warning thresholds. When the indicators are abnormal, optimize suggestions are sent out in a timely manner; Construct an accurate digital model of the protection device based on three-dimensional modeling technology, accurately restore the geometric features and spatial relationships of the equipment, visually display the operating state of the equipment through visual elements, and add dynamic monitoring points at key positions to realize real-time update of status information; Overlay and display the early warning information on the three-dimensional model of the equipment in the form of floating labels or warning icons, supporting multi-angle viewing and information screening.

Images