Cloud-based steel structure big data management system and method

Abstract

The invention relates to the technical field of big data management, in particular to a cloud-based steel structure big data management system and method, and the system comprises a data collection and preprocessing module, a structure performance analysis module, a crack prediction and warning module and a maintenance management support module. According to the method, through real-time data collection and dynamic data analysis, the timeliness and accuracy of data are enhanced, and the data quality is enhanced through timestamp sorting and data integrity verification, so that the system can quickly identify the structure problem, thereby effectively early warning in advance, and the analysis of historical and real-time data is combined, so that the accuracy of crack prediction is improved, and the accuracy of crack prediction is improved. The method has the advantages that the early warning response speed is increased, accurate data analysis supports enterprises to make decisions quickly, resource allocation is optimized, maintenance cost is reduced, structural safety is improved remarkably, dependence on traditional physical detection is lowered remarkably, working pressure is relieved, and risks possibly caused by data processing errors are avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of big data management, and particularly to a cloud-based big data management system and method for steel structures. Background Art

[0002] The technical field of big data management involves technologies and methods for collecting, storing, analyzing, and processing large amounts of data from various sources. The core is to effectively manage and extract useful information from large-scale data sets to support decision-making, optimize business processes, and predict future trends. It covers multiple sub-fields such as data mining, machine learning, data warehousing, and cloud computing, aiming to improve the speed and efficiency of data processing while ensuring data security and privacy protection. With the progress of technology, big data management has become an indispensable part of enterprises and organizations, capable of processing multi-source information from social media, transaction records, sensor data, etc.

[0003] Among them, the big data management system for steel structures is a technical solution specifically designed for the data needs of the steel structure industry. By creating an integrated platform, it is used to collect, organize, analyze, and store a large amount of data related to steel structures, with the purpose of providing real-time data analysis support to help engineers and designers optimize the design and construction process of steel structures. Through big data technology, the system can predict structural behavior, analyze material properties, and ensure construction safety, thereby improving efficiency, reducing costs, and enhancing the overall sustainability of steel structure projects.

[0004] Although traditional big data management technologies cover a wide range, in specific applications such as steel structure data management, conventional processing logics and technical means have limitations in processing large-scale data with real-time updates. Especially in an environment lacking cloud computing support, conventional data processing methods may not be able to process or quickly access large amounts of data in real time, which is particularly critical in structural health monitoring. Additionally, it is often not effective enough in ensuring data integrity and processing time sensitivity, resulting in the inability to provide accurate and timely data analysis and early warnings in emergency situations. These deficiencies limit the timeliness and accuracy of decision-making and increase safety risks caused by insufficient monitoring. In the field of steel structure maintenance and safety management, delays and inaccuracies may lead to inappropriate maintenance decisions, thereby increasing the risk of structural failures and affecting project costs and schedules. Summary of the Invention

[0005] The purpose of the present invention is to solve the drawbacks existing in the prior art and propose a cloud-based big data management system and method for steel structures.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions: The cloud-based big data management system for steel structures includes:

[0007] The data collection and preprocessing module uses a cloud-accessed sensor array to collect steel structure data, screen key data points, sort the data by timestamp, and perform data integrity verification to generate a steel structure monitoring data set;

[0008] The structural performance analysis module uses the steel structure monitoring data set to perform stress calculation on the monitoring point data, identify stress abnormal points in the steel structure, compare the data differences of the monitoring points according to the identification results, analyze the data trends, evaluate the overall health status of the steel structure, and generate a steel structure performance health record;

[0009] The crack prediction and warning module uses the steel structure performance health record to extract stress abnormality data, combines historical crack development data to perform trend comparison, analyzes the possibility of steel structure crack development based on the trend comparison results, evaluates the abnormal pattern of the data, predicts future crack development trends, and outputs steel structure crack warning information;

[0010] The maintenance management support module optimizes the maintenance plan, reallocates maintenance resources and adjusts maintenance measures based on the steel structure crack warning information, combined with maintenance history records and resource allocation, to generate a steel structure maintenance management optimization plan.

[0011] As a further solution of the present invention, the steps of acquiring the steel structure monitoring data set are:

[0012] Install cloud-accessible sensor arrays at key locations of the steel structure to collect vibration, temperature, and stress data, transmit them to the cloud platform, and generate raw data sets for the steel structure;

[0013] The timestamp verification and integrity check of the original data set of the steel structure are performed to remove data outliers and incomplete parts, and the formula is used.

[0014]

[0015] Calculate the timestamp-checked weighted average X v , generate the verified monitoring data set, where t i represents the timestamp of the i-th data point, X i represents the monitoring value of the ith data point, and n is the total number of data points;

[0016] Analyze the verified monitoring data set, determine the key data points, and use the formula:

[0017]

[0018] Calculate the standard deviation D of the key monitoring data point set k , generate a steel structure monitoring data set, where X ci Represents the monitoring value of the i-th key data point, and m is the number of key data points.

[0019] As a further solution of the present invention, the identification steps of the stress abnormal points are as follows:

[0020] Using the steel structure monitoring data set, calculate the stress of the original mechanical data of each monitoring point, using the formula

[0021]

[0022] Calculate the real-time stress value σ of each monitoring point to generate the real-time stress data of the monitoring point; where F represents the applied force and A represents the cross-sectional area;

[0023] According to the real-time stress data of the monitoring point, perform abnormal determination on each monitoring point, using the formula

[0024]

[0025] Judge whether the standard deviation multiple of the real-time stress value and the average stress exceeds the preset threshold to generate the stress abnormal identification result of the monitoring point, where σ represents the real-time stress value of the monitoring point, σ avg represents the average stress value, σ std represents the standard deviation of the stress, and T represents the threshold;

[0026] Based on the stress abnormal identification result of the monitoring point, extract all abnormal monitoring point data, record the actual position and stress value of the point, and generate the steel structure stress abnormal point record.

[0027] As a further solution of the present invention, the acquisition steps of the steel structure performance health record are as follows:

[0028] Based on the steel structure stress abnormal point record, compare the stress data of the monitoring point, analyze the stress value difference between the abnormal point and the normal point, and generate the data difference analysis result;

[0029] According to the data difference analysis result, use the time series analysis method to analyze the long-term trend and periodic fluctuation of the monitoring point data, determine the stress trend of the steel structure, and generate the data trend analysis result;

[0030] Integrate the steel structure stress abnormal point record, data difference analysis result and data trend analysis result, evaluate the overall health status of the steel structure, and generate the steel structure performance health record.

[0031] As a further solution of the present invention, the execution steps of the trend comparison are as follows:

[0032] Through the steel structure performance health record, extract the stress abnormal data and perform data point annotation processing to generate the stress abnormal data set;

[0033] Match the stress anomaly data set with the historical crack development data set to determine the abnormal stress points corresponding to the crack history, and obtain the associated abnormal stress data;

[0034] Perform time series analysis on the associated abnormal stress data, using the formula

[0035]

[0036] Calculate the trend consistency index T(x), analyze the current data and the historical crack development trend, and generate a trend comparison result, where x t represents the stress data at the current time point, Δx represents the change rate of the stress data, and x t-1 represents the data at the previous time point.

[0037] As a further solution of the present invention, the steps for obtaining the steel structure crack warning information are as follows:

[0038] Based on the trend comparison result, perform a quantitative analysis on the correlation between the abnormal stress data and the crack development, and generate a quantitative analysis result of the crack development possibility;

[0039] Use the quantitative analysis result of the crack development possibility to analyze the abnormal patterns in the data, identify the potential risks of the current steel structure, and generate an abnormal pattern evaluation record;

[0040] According to the abnormal pattern evaluation record, perform time series analysis, predict the future trend based on the current data and the historical crack development pattern, and combine the material properties of the steel structure to output the steel structure crack warning information.

[0041] As a further solution of the present invention, the steps for obtaining the steel structure maintenance management optimization plan are as follows:

[0042] Integrate the steel structure crack warning information and the maintenance history record, evaluate the duration and efficiency of the maintenance effect by analyzing the change in the structural performance and the time interval after the maintenance event, and obtain the structural maintenance record;

[0043] Based on the structural maintenance record, optimize the current resource allocation and maintenance requirements, using the formula

[0044]

[0045] Calculate the optimized resource allocation efficiency R, and generate a resource allocation optimization plan, where r i represents the improvement ratio of the i-th maintenance to the structural integrity, is the average value of all maintenance improvement ratios, used to evaluate the deviation degree of each maintenance, and s i represents the corresponding resource consumption, and H represents the number of maintenance times;

[0046] Using the resource allocation optimization plan, redesign the maintenance measures for the risk area to ensure that each maintenance activity targets the development trend of cracks and their potential risks, and generate an optimized steel structure maintenance management plan.

[0047] A cloud-based big data management method for steel structures includes the following steps:

[0048] S1: Install a sensor array, collect data and transmit it to the cloud platform, perform timestamp verification and integrity check, determine key data points, and generate a steel structure monitoring data set;

[0049] S2: Using the steel structure monitoring data set, calculate the stress of the original mechanical data at each monitoring point, determine anomalies at each monitoring point, identify stress anomalies, record the actual positions and stress values of the anomaly points, and generate a record of stress anomaly points in the steel structure;

[0050] S3: Based on the record of stress anomaly points in the steel structure, analyze the stress value differences between anomaly points and normal points, analyze the long-term trends and periodic fluctuations of the data at the monitoring points, determine the stress trend of the steel structure, evaluate the overall health status of the steel structure, and generate a record of the performance health of the steel structure;

[0051] S4: Through the record of the performance health of the steel structure, extract stress anomaly data, perform data point annotation processing, match it with the historical crack development data set, determine the anomaly stress points corresponding to the crack history, analyze the current data and the historical crack development trend, and generate a trend comparison result;

[0052] S5: Based on the trend comparison result, conduct a quantitative analysis of the correlation between the anomaly stress data and the crack development, analyze the anomaly patterns in the data, identify the potential risks of the current steel structure, predict the future trend based on the current data and the historical crack development pattern, and output steel structure crack warning information;

[0053] S6: Integrate the steel structure crack warning information with the maintenance history record, analyze the structural performance changes and time intervals after maintenance events, optimize the current resource allocation and maintenance requirements, redesign the maintenance measures for the risk area, and generate an optimized steel structure maintenance management plan.

[0054] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0055] In the present invention, through real-time data collection and dynamic data analysis, the timeliness and accuracy of data are enhanced. Timestamp sorting and data integrity verification improve data quality, enabling the system to quickly identify structural problems, thus effectively providing early warnings. The analysis combining historical and real-time data not only improves the accuracy of crack prediction but also accelerates the response speed of early warnings. Precise data analysis supports enterprises in making rapid decisions, optimizing resource allocation, reducing maintenance costs, and significantly enhancing structural safety. It significantly reduces the dependence on traditional physical inspections, alleviates work pressure, and avoids risks that may be caused by data processing errors. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 is the system flowchart of the present invention;

[0057] Figure 2 is the flowchart for obtaining the steel structure monitoring data set of the present invention;

[0058] Figure 3 is the flowchart for identifying stress abnormal points of the present invention;

[0059] Figure 4 is the flowchart for obtaining the steel structure performance health record of the present invention;

[0060] Figure 5 is the flowchart for executing trend comparison of the present invention;

[0061] Figure 6 is the flowchart for obtaining the steel structure crack early warning information of the present invention;

[0062] Figure 7 is the flowchart for obtaining the steel structure maintenance management optimization plan of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0063] In order to make the objectives, technical solutions and advantages of the present invention more clear and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0064] In the description of the present invention, it should be understood that the orientation or positional relationships indicated by the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. are based on the orientation or positional relationships shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as limiting the present invention. In addition, in the description of the present invention, the meaning of "a plurality" is two or more unless otherwise specifically defined.

[0065] Please refer to Figure 1 , the cloud-based steel structure big data management system includes:

[0066] The data collection and preprocessing module uses a cloud-connected sensor array to collect steel structure data, screening key data points, sorting the data by timestamp and performing data integrity verification to generate a steel structure monitoring dataset;

[0067] The structural performance analysis module uses the steel structure monitoring dataset to calculate the stress of the data at the monitoring points, identify stress anomaly points in the steel structure, compare the data differences at the monitoring points according to the identification results, analyze the data trend, evaluate the overall health status of the steel structure, and generate a steel structure performance health record;

[0068] The crack prediction and warning module uses the steel structure performance health record to extract stress anomaly data, combines historical crack development data to perform trend comparison, analyzes the possibility of steel structure crack development according to the trend comparison results, evaluates the abnormal pattern of the data, predicts the future crack development trend, and outputs steel structure crack warning information;

[0069] The maintenance management support module optimizes the maintenance plan, reallocates maintenance resources and adjusts maintenance measures according to the steel structure crack warning information, combined with the maintenance history record and resource allocation, and generates an optimized steel structure maintenance management plan.

[0070] The steel structure monitoring dataset includes sensor identification records, data quality rating results and key time point screening records, the steel structure performance health record includes stress value records, health status ratings and monitoring point difference records, the steel structure crack warning information includes warning levels, potential risk areas and predicted crack paths, and the optimized steel structure maintenance management plan includes resource allocation tables, maintenance cycle adjustment records and preventive measure lists.

[0071] Please refer to Figure 2 , the acquisition steps of the steel structure monitoring dataset are as follows:

[0072] Install a cloud-connected sensor array at key parts of the steel structure to collect vibration, temperature and stress data, transmit it to the cloud platform, and generate a steel structure original dataset;

[0073] Install a sensor array with cloud access at key parts of the steel structure. Select specific sensor types and distribution locations. According to the design parameters of the steel structure and the expected monitoring requirements, layout the sensors. The position of each sensor must ensure that it can capture key structural stress points and potential deformation areas. For this purpose, it is necessary to analyze the maximum pressure and temperature difference that each part may withstand based on the original performance data of the steel structure, and select sensors that can work stably under extreme working conditions, so as to collect data covering the entire structure and ensure the timeliness and integrity of the data. The data synchronization between each sensor installation position also needs to be strictly controlled to avoid data errors caused by time differences in data acquisition. The data output of the sensors needs to meet the technical requirements of real-time transmission and be transmitted to the cloud platform in real time. This step ensures the high quality and high reliability of the monitoring data and generates the original steel structure data set.

[0074] Perform timestamp verification and integrity check on the original steel structure data set, eliminate data outliers and incomplete parts, and use the formula

[0075]

[0076] Calculate the weighted average X after timestamp verification v , and generate the verified monitoring data set. Among them, t i represents the timestamp of the i-th data point, X i represents the monitoring value of the i-th data point, and n is the total number of data points;

[0077] Collect data: X1 = 35 units, X2 = 40 units, n is the total number of data points, n = 2; The calculation process is as follows:

[0078] Calculate the product of the sum of the squares of the timestamps and the monitoring values:

[0079]

[0080] Calculate the sum of the timestamps:

[0081] ∑t i = 1500 + 1525 = 3025

[0082] Calculate the weighted average:

[0083]

[0084] The results show that considering the weighting of the timestamps, the weighted average in the verified steel structure monitoring data set is 56811.57 units. This calculation process shows the comprehensive effect of the data under the influence of timestamps and provides an accurate basic data index for further analysis and use.

[0085] Analyze the verified monitoring data set, determine the key data points, and use the formula

[0086]

[0087] Calculate the standard deviation D of the set of key monitoring data points k , generate a steel structure monitoring data set, where X ci represents the monitoring value of the i-th key data point, and m is the number of key data points;

[0088] Collect data: The key data point X c1 = 50 units, X c2 = 55 units, m is the number of key data points, m = 2; The calculation process is as follows:

[0089] Calculate the square of the ratio of each key data point to the number of points:

[0090]

[0091] Calculate the square root:

[0092]

[0093] The results show that the standard deviation of the set of key monitoring data points is 37.16, indicating the fluctuation of the monitoring data at the key points and providing important statistical information for the safety assessment of the structure.

[0094] Please refer to Figure 3 , the steps for identifying stress anomaly points are:

[0095] Using the steel structure monitoring data set, calculate the stress of the original mechanical data of each monitoring point, and use the formula

[0096]

[0097] Calculate the real-time stress value σ of each monitoring point to generate real-time stress data of the monitoring point; where F represents the applied force and A represents the cross-sectional area;

[0098] Collect the applied force of a certain monitoring point as 1000N and the cross-sectional area as 50 cm 2 (i.e., 0.005 m 2 ), then the stress calculation process is as follows:

[0099]

[0100] The results show that the current stress of this monitoring point is 200000 Pascal. This numerical result indicates the stress state of the monitoring point at a given moment and is an important basis for judging the structural safety. If this stress value exceeds the allowable stress of the material, further evaluation or reinforcement measures may be required.

[0101] Based on the real-time stress data of the monitoring points, an abnormality determination is made for each monitoring point. Using the formula,

[0102]

[0103] judge whether the multiple of the standard deviation of the real-time stress value and the average stress exceeds the preset threshold, and generate the recognition result of the stress abnormality of the monitoring point. Among them, σ represents the real-time stress value of the monitoring point, σ avg represents the average stress value, σ std represents the standard deviation of the stress, and T represents the threshold;

[0104] Combined with the aforementioned calculation result that σ is 200,000 Pa, σ avg and σ std represent the average value and the standard deviation of the stress value of the monitoring point respectively. Calculate to obtain that the average stress is 180,000 Pa and the standard deviation is 15,000 Pa. Set the threshold T to 2. The calculation process is as follows:

[0105]

[0106] Since 1.33 is less than the set threshold of 2, it indicates that although the stress of this monitoring point has increased, it has not reached the abnormal state. This result shows that although the current stress state of this monitoring point is on the high side, it is still within the safe range and does not require immediate structural intervention or repair, which can help judge the immediate safety of the steel structure and contribute to formulating a more accurate maintenance and inspection plan.

[0107] Based on the recognition result of the stress abnormality of the monitoring point, extract the data of all abnormal monitoring points, record the actual position and stress value of the points, and generate a record of the stress abnormal points of the steel structure;

[0108] According to the recognition result of the stress abnormality, make a detailed record of the abnormal points. First, it is necessary to classify all the monitoring points marked as abnormal and group them according to the severity and location of the stress exceeding the standard. This process needs to combine the specific layout and design requirements of the structure to determine which areas have the most serious stress exceeding the standard and need to be processed first. For each abnormal monitoring point, record detailed information including its specific position, real-time stress value, and deviation from the average stress value. The data summary will support the long-term trend analysis of the structural health monitoring system and the decision-making process for necessary maintenance and reinforcement measures for the structure. The generated record of the stress abnormal points in the steel structure provides a basis for the subsequent structural assessment.

[0109] Please refer to Figure 4 , the steps for obtaining the performance health record of the steel structure are as follows:

[0110] Based on the records of stress abnormal points of the steel structure, compare the stress data of the monitoring points, analyze the stress value differences between the abnormal points and the normal points, and generate the data difference analysis results;

[0111] Using the records of stress abnormal points of the steel structure, first extract the stress data of each monitoring point, especially pay attention to the monitoring points marked as abnormal, screen out the stress values of all monitoring points and classify and store them. Subsequently, use the Pandas library in Python to process the data, calculate the statistical indicators of the stress values of all monitoring points, including the mean, standard deviation and variance. This analysis helps to determine the distribution and fluctuation of the stress values. By comparing the data distributions of the abnormal points and the non-abnormal points, determine the stress concentration areas in the structure. The final data difference analysis results are the comparison reports of the stress between the monitoring points.

[0112] According to the data difference analysis results, use the time series analysis method to analyze the long-term trend and periodic fluctuation of the monitoring point data, determine the stress trend of the steel structure, and generate the data trend analysis results;

[0113] Using the data difference results, conduct time series analysis, select an appropriate time window, such as the data of the past 12 months, perform seasonal adjustment and trend decomposition, apply the autoregressive moving average (ARMA) model, first determine the order of the model, then estimate the parameters, and use the autoregressive coefficient and the moving average coefficient to describe the long-term trend and periodic fluctuation of the data, predict the possible stress change trends in the next few months, so as to provide a time window for formulating preventive maintenance measures. The output of the analysis is a time series analysis report, including trend charts and prediction results.

[0114] Integrate the records of stress abnormal points of the steel structure, the data difference analysis results and the data trend analysis results, evaluate the overall health status of the steel structure, and generate the steel structure performance health records;

[0115] Integrate the results of anomaly identification, data difference and trend analysis, load data, material properties, historical fatigue data, etc. in real time, and calculate and process the data. Evaluate the remaining life and safety factor of the steel structure in real time, assess the current health status and potential risks of the structure. The output report details the safety factors of each part, predicts the remaining life, and gives the structural parts that need attention or maintenance. The comprehensive evaluation results provide a scientific basis for future maintenance strategies and decisions, ensuring the safe operation of the structure.

[0116] Please refer to Figure 5 , the execution steps for trend comparison are as follows:

[0117] Extract the stress abnormal data through the steel structure performance health records, and perform data point marking processing to generate the stress abnormal data set;

[0118] Through the performance health record of the steel structure, after the automatic screening of the preliminary data, all stress data exceeding the safety threshold are marked, including the original readings collected from various sensors. If the readings exceed the preset threshold, they are automatically recorded as abnormal. The acquisition of these data depends on the accurate arrangement of the sensors and the real-time transmission of the data. According to the collected data, the abnormal data points are marked and processed, and the resulting is a stress anomaly dataset containing all marked abnormal points, which will provide the basis for subsequent analysis.

[0119] Match the stress anomaly dataset with the historical crack development dataset to determine the abnormal stress points corresponding to the crack history and obtain the associated abnormal stress data;

[0120] After obtaining the stress anomaly dataset, perform data correlation analysis by matching the abnormal data with the crack development data stored in the historical database. In this process, the key data processing steps include data cleaning and format standardization. Through the comparative analysis of the abnormal stress data and the historical crack data, ensure the accurate identification of the stress anomaly points corresponding to the historical cracks, so as to obtain the abnormal stress data that can directly reflect the historical crack trend. The final matching result provides the basis for crack warning.

[0121] Perform time series analysis on the associated abnormal stress data using the formula

[0122]

[0123] Calculate the trend consistency index T(x), analyze the current data and the historical crack development trend, and generate a trend comparison result. Among them, x t represents the stress data at the current time point, Δx represents the change rate of the stress data, and x t-1 represents the data at the previous time point;

[0124] Collect x t is 5.2 MPa (the stress data at the current time point), Δx is -0.8 MPa (the difference between the current and the previous data points), and x t-1 is 6.0 MPa (the stress data at the previous time point).

[0125] Substitute into the formula for calculation:

[0126]

[0127] The result shows that according to the trend comparison analysis, the trend of the current data point has a high consistency (2.07) with the historical crack development trend, which has an important guiding role for future structural safety monitoring and warning systems.

[0128] Please refer to Figure 6 , the steps for obtaining the steel structure crack warning information are as follows:

[0129] Based on the trend comparison results, quantitatively analyze the correlation between abnormal stress data and crack development to generate the quantitative analysis results of crack development possibility;

[0130] Based on the trend comparison results, unify the formats and synchronize the time of abnormal stress data from different sensors with historical crack records. Then, through the Pearson correlation coefficient formula, calculate the correlation coefficient between crack data and stress abnormal data to determine the direct correlation strength between them. The level of the coefficient directly reflects the strength of the linear relationship between the two. Subsequently, apply linear regression analysis to establish an impact model of stress data on crack development. This estimates the parameters of the regression line by the least squares method to determine the linear equation that best represents the data trend, thereby predicting the possibility of crack development and generating a detailed report containing a potential crack risk assessment.

[0131] Utilize the quantitative analysis results of crack development possibility to analyze the abnormal patterns in the data, identify the potential risks of the current steel structure, and generate an abnormal pattern assessment record;

[0132] Based on the existing analysis results of crack development possibility, use the support vector machine (SVM) in machine learning to classify and identify abnormal patterns. First, extract training samples from historical data, including selecting representative abnormal cases and normal cases, and perform feature engineering on the data, including normalization and principal component analysis, to reduce the computational complexity and improve the generalization ability of the model. Then, in the supervised learning framework, use the training samples to train the SVM model. After the model training is completed, classify the new data to verify the prediction accuracy and efficiency of the model. Through this method, the abnormal behavior patterns of the current structure can be effectively identified, and finally a comprehensive assessment record of data abnormal patterns is generated.

[0133] According to the abnormal pattern assessment record, conduct time series analysis, predict the future trend based on the current data and historical crack development patterns, and combine with the material properties of the steel structure to output steel structure crack warning information;

[0134] Use the ARIMA time series prediction model to predict the future crack development trend based on historical crack development data and current abnormal patterns. Select appropriate ARIMA model parameters, such as the number of seasonal differences, autoregressive terms, and moving average terms. These parameters are determined by automated model selection methods such as the Akaike Information Criterion (AIC) to ensure the best fit of the model to historical data. Then apply the model to the current data for future value prediction. During the prediction process, regularly update the model parameters with new data to improve the prediction accuracy. Finally, combine the prediction results with environmental factors and material properties to evaluate the development probability and possible impacts of future cracks, thereby generating a detailed crack development prediction report and crack warning information.

[0135] Please refer to Figure 7 , the steps to obtain the optimized steel structure maintenance management plan are as follows:

[0136] Integrate the steel structure crack warning information with the maintenance history records. By analyzing the changes in structural performance and time intervals after maintenance events, evaluate the duration and efficiency of the maintenance effect to obtain the structural maintenance records.

[0137] Collect crack warning information, including the width, depth, and location of the cracks, and integrate it with the historical maintenance records extracted from the maintenance database. The historical records include maintenance time, materials used, and structural performance evaluation data before and after maintenance. By statistically analyzing the information, calculate the efficiency of each maintenance (using the percentage improvement in structural performance divided by the maintenance cost and time), thereby obtaining a maintenance efficiency index that reflects the average effectiveness of all past maintenance activities. This data set will be directly applied to the next step of resource allocation optimization analysis as a basis for evaluating the rationality of resource allocation.

[0138] Based on the structural maintenance records, optimize the current resource allocation and maintenance requirements, using the formula,

[0139]

[0140] Calculate the optimized resource allocation efficiency R and generate an optimized resource allocation plan. Among them, r i represents the improvement ratio of the i-th maintenance to the structural integrity, is the average value of all maintenance improvement ratios, used to evaluate the deviation degree of each maintenance, and s i represents the corresponding resource consumption, and H represents the number of maintenance times;

[0141] The number of maintenance times H = 5, and the improvement ratios r i for each maintenance are 0.1, 0.2, 0.15, 0.18, 0.16 respectively, and the resource consumptions s i are 500 units, 400 units, 600 units, 500 units, 450 units respectively.

[0142] Calculate That is, the average improvement ratio, we get:

[0143]

[0144] Calculate the deviation between each improvement ratio and the average value, we get:

[0145] ...

[0147] And so on, substituting into the formula, we get:

[0148]

[0149] The results show that the optimized resource allocation efficiency is 0.00323, which means that under the consumption of unit resources, the average improvement ratio of the structural integrity brought by the maintenance activities is 0.323%, which is still relatively low and needs to be improved.

[0150] Utilize the resource allocation optimization plan to redesign the maintenance measures for the risk areas to ensure that each maintenance activity targets the crack development trend and its potential risks, and generate an optimized steel structure maintenance management plan;

[0151] Utilize the new resource allocation optimization plan to redesign the maintenance measures for the identified high-risk areas, which includes adjusting the priority of resource investment, such as allocating more monitoring equipment and repair teams to the areas where cracks develop rapidly, and at the same time adjusting the maintenance cycle and specific technical measures according to the crack warning information to ensure that each measure can minimize potential risks. The finally generated optimized maintenance management plan details the specific allocation plans and schedules of various resources (personnel, materials, equipment), as well as the expected maintenance effects. The plan will be continuously updated according to real-time data to cope with possible risk changes.

[0152] A cloud-based big data management method for steel structures, including the following steps:

[0153] S1: Install a sensor array, collect data and transmit it to the cloud platform, perform timestamp verification and integrity check, determine key data points, and generate a steel structure monitoring data set;

[0154] S2: Utilize the steel structure monitoring data set to calculate the stress of the original mechanical data at each monitoring point, determine anomalies at each monitoring point, identify stress anomalies, record the actual positions and stress values of the anomaly points, and generate a record of steel structure stress anomaly points;

[0155] S3: Based on the records of the abnormal stress points of the steel structure, analyze the stress value differences between the abnormal points and the normal points, analyze the long-term trends and periodic fluctuations of the data at the monitoring points, determine the stress trend of the steel structure, evaluate the overall health status of the steel structure, and generate a performance health record of the steel structure;

[0156] S4: Extract the abnormal stress data through the performance health record of the steel structure, perform data point marking processing, match it with the historical crack development data set, determine the abnormal stress points corresponding to the crack history, analyze the current data and the historical crack development trend, and generate a trend comparison result;

[0157] S5: Based on the trend comparison result, conduct a quantitative analysis of the correlation between the abnormal stress data and the crack development, analyze the abnormal patterns in the data, identify the potential risks of the current steel structure, predict the future trend based on the current data and the historical crack development pattern, and output the steel structure crack warning information;

[0158] S6: Integrate the steel structure crack warning information with the maintenance history record, analyze the structural performance changes and time intervals after the maintenance events, optimize the current resource allocation and maintenance requirements, redesign the maintenance measures for the risk areas, and generate an optimized steel structure maintenance management plan.

[0159] The above are only the preferred embodiments of the present invention, and do not limit the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. Cloud-based big data management system for steel structures, characterized in that: The system comprises: The data collection and preprocessing module uses a cloud-accessed sensor array to collect steel structure data, screen key data points, sort the data by timestamp, and perform data integrity verification to generate a steel structure monitoring data set; The structural performance analysis module uses the steel structure monitoring data set to perform stress calculation on the monitoring point data, identify stress abnormal points in the steel structure, compare the data differences of the monitoring points according to the identification results, analyze the data trends, evaluate the overall health status of the steel structure, and generate a steel structure performance health record; The crack prediction and warning module uses the steel structure performance health record to extract stress abnormality data, combines historical crack development data to perform trend comparison, analyzes the possibility of steel structure crack development based on the trend comparison results, evaluates the abnormal pattern of the data, predicts future crack development trends, and outputs steel structure crack warning information; The maintenance management support module optimizes the maintenance plan, reallocates maintenance resources and adjusts maintenance measures based on the steel structure crack warning information, combined with maintenance history records and resource allocation, to generate a steel structure maintenance management optimization plan.

2. The cloud-based steel structure big data management system according to claim 1, characterized in that The steps for obtaining the steel structure monitoring data set are: Install cloud-accessible sensor arrays at key locations of the steel structure to collect vibration, temperature, and stress data, transmit them to the cloud platform, and generate raw data sets for the steel structure; The timestamp verification and integrity check of the original data set of the steel structure are performed to remove data outliers and incomplete parts, and the formula is used. Calculate the weighted average X after timestamp verification v , and generate a verified monitoring data set, where t i represents the timestamp of the i-th data point, and X i represents the monitoring value of the i-th data point, and n is the total number of data points; Analyze the verified monitoring data set, determine the key data points, and use the formula: Calculate the standard deviation D of the set of key monitoring data points k , generate a steel structure monitoring data set, where X ci represents the monitoring value of the i-th key data point, and m is the number of key data points.

3. The cloud-based steel structure big data management system according to claim 2, wherein, The steps for identifying the stress abnormal point are as follows: Using the steel structure monitoring data set, the original mechanical data of each monitoring point is used to calculate the stress, using the formula: Calculate the real-time stress value σ of each monitoring point to generate the real-time stress data of the monitoring point; where F represents the applied force and A represents the cross-sectional area; According to the real-time stress data of the monitoring point, each monitoring point is judged abnormally, using the formula: Determine whether the multiple of the standard deviation of the real-time stress value and the average stress exceeds a preset threshold, and generate an identification result of stress anomaly at the monitoring point, where σ represents the real-time stress value of the monitoring point, and σ avg represents the average stress value, and σ std represents the standard deviation of the stress, and T represents the threshold; Based on the stress anomaly identification result of the monitoring point, all abnormal monitoring point data are extracted, the actual position and stress value of the point are recorded, and the stress anomaly point record of the steel structure is generated.

4. The cloud-based steel structure big data management system according to claim 3, characterized in that, The steps for obtaining the steel structure performance health record are as follows: Based on the abnormal stress point records of the steel structure, the stress data of the monitoring points are compared, the stress value difference between the abnormal point and the normal point is analyzed, and the data difference analysis result is generated; According to the data difference analysis results, use the time series analysis method to analyze the long-term trend and periodic fluctuation of the monitoring point data, determine the stress trend of the steel structure, and generate data trend analysis results; By combining the abnormal stress point records of the steel structure, the data difference analysis results and the data trend analysis results, the overall health status of the steel structure is evaluated and a performance health record of the steel structure is generated.

5. The cloud-based steel structure big data management system according to claim 4, wherein: The execution steps of the trend comparison are: Extract stress abnormality data through the steel structure performance health record, perform data point annotation processing, and generate a stress abnormality data set; Matching the stress anomaly data set with the historical crack development data set, determining the abnormal stress point corresponding to the crack history, and obtaining associated abnormal stress data; Perform time series analysis on the associated abnormal stress data using the formula Calculate the trend consistency index T(x), analyze the current data and the historical crack development trend, and generate a trend comparison result, where x t represents the stress data at the current time point, Δx represents the change rate of the stress data, and x t-1 represents the data at the previous time point.

6. The cloud-based steel structure big data management system according to claim 5, wherein, The steps for obtaining the steel structure crack warning information are as follows: Based on the trend comparison result, quantitatively analyze the correlation between the abnormal stress data and crack development, and generate a quantitative analysis result of crack development possibility; Use the quantitative analysis result of crack development possibility to analyze the abnormal patterns in the data, identify the potential risks of the current steel structure, and generate an abnormal pattern evaluation record; According to the abnormal pattern evaluation record, perform time series analysis, predict the future trend based on the current data and historical crack development patterns, and combine with the material properties of the steel structure to output the steel structure crack warning information.

7. The cloud-based steel structure big data management system according to claim 6, wherein The steps for obtaining the steel structure maintenance management optimization plan are as follows: Integrate the steel structure crack warning information and maintenance history records, evaluate the duration and efficiency of the maintenance effect by analyzing the structural performance changes and time intervals after maintenance events, and obtain the structural maintenance records; Based on the structural maintenance records, optimize the current resource allocation and maintenance requirements, using the formula Calculate the optimized resource allocation efficiency R and generate an optimized resource allocation plan, where r i represents the improvement ratio of the structural integrity for the i-th maintenance, is the average value of all maintenance improvement ratios and is used to evaluate the deviation degree of each maintenance, s i represents the corresponding resource consumption, and H represents the number of maintenances; Use the resource allocation optimization plan to redesign the maintenance measures for the risk areas, ensure that each maintenance activity targets the crack development trend and its potential risks, and generate a steel structure maintenance management optimization plan.

8. Cloud-based big data management method for steel structures, characterized in that, Execute according to the cloud-based steel structure big data management system described in any one of claims 1-7, including the following steps: Install a sensor array, collect data and transmit it to the cloud platform, perform timestamp verification and integrity check, determine the key data points, and generate a steel structure monitoring data set; Use the steel structure monitoring data set to calculate the stress of the original mechanical data at each monitoring point, determine the abnormality at each monitoring point, identify the stress abnormality, record the actual position and stress value of the abnormal point, and generate a steel structure stress abnormal point record; Based on the steel structure stress abnormal point record, analyze the stress value difference between the abnormal point and the normal point, analyze the long-term trend and periodic fluctuation of the monitoring point data, determine the stress trend of the steel structure, evaluate the overall health status of the steel structure, and generate a steel structure performance health record; Extract the stress abnormal data through the steel structure performance health record, perform data point annotation processing, match it with the historical crack development data set, determine the abnormal stress points corresponding to the crack history, analyze the current data and historical crack development trends, and generate a trend comparison result; Based on the trend comparison result, quantitatively analyze the correlation between the abnormal stress data and crack development, analyze the abnormal patterns in the data, identify the potential risks of the current steel structure, predict the future trend based on the current data and historical crack development patterns, and output the steel structure crack warning information; Integrate the steel structure crack warning information and maintenance history records, analyze the structural performance changes and time intervals after maintenance events, optimize the current resource allocation and maintenance requirements, and redesign the maintenance measures for the risk areas to generate a steel structure maintenance management optimization plan.

Images