Equivalent pulsating wind load carrier frequency domain risk early warning method
By deploying sensors at key structural locations, performing signal preprocessing and frequency domain analysis, and combining this with a risk assessment model, the problems of untimely monitoring of pulsating wind loads and insufficient risk assessment in existing technologies have been solved. This enables real-time and accurate assessment and early warning of wind load risks, ensuring the safety of the structure.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-27
- Publication Date
- 2026-03-27
AI Technical Summary
Existing technologies are insufficient for real-time and accurate monitoring and assessment of the frequency domain characteristics of pulsating wind loads in building structures, bridge engineering, and large-scale infrastructure. This results in untimely capture of dynamic changes in wind loads, a lack of effective risk warning models, and difficulty in providing timely and effective decision-making basis for engineers.
Multiple types of sensors are deployed at key structural locations to collect response data in real time. After preprocessing such as wavelet denoising and filtering, frequency domain analysis is performed. Combined with frequency domain characteristic parameters and risk assessment models, the wind load risk level is assessed in real time, and early warnings are sent to engineers through audible and visual alarms, SMS notifications, and other means.
It enables real-time and accurate monitoring and risk warning of pulsating wind loads, provides a guarantee for the safe operation of the structure, improves the accuracy and timeliness of wind load risk assessment, and ensures the safety of the engineering structure.
Smart Images
Figure CN121747290A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the field of structural engineering wind-resistant technology, in particular to an equivalent fluctuating wind load frequency domain risk early warning method. BACKGROUND
[0002] In building structures, bridge engineering and various large infrastructure construction, wind load is one of the key factors affecting the safety and stability of the structure. As an important part of wind load, fluctuating wind has randomness and complexity, and its dynamic effect often causes great damage to the structure. For example, in large-span roof structures, wind-induced vibration response caused by fluctuating wind may cause excessive deformation or even damage in local areas of the roof; in high-rise power transmission towers, fluctuating wind load may cause fatigue damage to the tower, reducing its service life.
[0003] Traditional wind load monitoring and analysis methods focus more on the study of average wind load, and have limited ability to fine-tune monitoring and risk assessment of fluctuating wind load. Although some existing monitoring systems can obtain some structural response data, they have obvious shortcomings in frequency domain analysis of these data to accurately assess the risk of fluctuating wind load. On the one hand, it is not possible to convert the collected structural response signals into equivalent fluctuating wind load frequency domain characteristics in real time and accurately, resulting in a delay in capturing the dynamic changes of wind load; on the other hand, there is a lack of effective risk early warning model, which cannot give the risk level faced by the structure in time according to the frequency domain analysis results, making it difficult to provide timely and effective decision-making basis for engineers. SUMMARY
[0004] The purpose of the present application is to provide an equivalent fluctuating wind load frequency domain risk early warning method, which can monitor the fluctuating wind load on the structure in real time and accurately, and provide a powerful guarantee for the safe operation of the structure by early warning the wind load risk faced by the structure through frequency domain analysis.
[0005] To achieve the above purpose, the present application provides the following technical solution: an equivalent fluctuating wind load frequency domain risk early warning method, comprising the following steps:
[0006] A plurality of types of sensors are arranged at key parts of the structure to collect response data of the structure under the action of wind load in real time.
[0007] The collected original response signals are preprocessed by denoising, filtering and amplifying.
[0008] The frequency domain analysis method is used to convert the preprocessed time domain response signals into frequency domain signals and calculate the frequency domain characteristic parameters.
[0009] A risk assessment model based on frequency domain features is established, the risk level threshold corresponding to different frequency domain feature parameters is determined by combining the mechanical properties of the structure, the design parameters and the historical wind load data, and the current calculated frequency domain feature parameters are compared with the risk level threshold to assess the wind load risk level currently faced by the structure.
[0010] When the assessed wind load risk reaches or exceeds the set warning threshold, warning information is sent to relevant engineering personnel through the ways of sound-light alarm, short message notification and email push.
[0011] As preferred, the multiple types of sensors are at least two of an acceleration sensor, a strain sensor and a displacement sensor, and the range, resolution, measurement accuracy and sampling frequency of each sensor meet the requirement of accurately collecting response data of the structure under the action of fluctuating wind load; wherein the arrangement of the sensors follows the principle of mechanical analysis, and the specific installation position of the sensors is determined according to the wind vibration response sensitive area calculated by the structural dynamics model.
[0012] As preferred, the wavelet denoising algorithm is used for the denoising processing of the original response signal, the wavelet basis function is selected, the decomposition layer is determined, and the threshold function is used to process the wavelet coefficients; after the denoising processing, the signal smoothing processing step is performed, and the moving average algorithm is used to smooth the denoised signal to eliminate the residual slight fluctuations.
[0013] As preferred, the frequency domain analysis method is Fourier transform or wavelet transform, the fast Fourier transform algorithm is used when the Fourier transform is used, and the appropriate data truncation method is selected; when the wavelet transform is used, the appropriate mother wavelet is selected and the decomposition scale range is determined; before the frequency domain conversion, the preprocessed time domain signal is normalized to make the signal amplitude in a specific interval and improve the accuracy of the frequency domain analysis.
[0014] As preferred, the frequency domain feature parameter is at least one of power spectral density, autocorrelation function and cross-correlation function, and the corresponding appropriate algorithm and parameter setting are used when calculating the power spectral density, autocorrelation function and cross-correlation function; the calculated frequency domain feature parameter also needs to go through the feature screening step, removes the redundant features based on the principal component analysis method, and retains the key frequency domain feature parameters which have significant influence on risk assessment.
[0015] As preferred of the present application, the establishment of the risk assessment model further comprises the step of training and optimizing the historical data by a machine learning algorithm, the machine learning algorithm optimizes the model parameters by setting the number of trees, the maximum depth, and selecting appropriate feature selection criteria, and utilizes cross-validation.
[0016] As preferred of the present application, the sending mode of the early warning information is sound and light alarm, short message notification, and email push, wherein the sound and light alarm has a specific sound intensity and flashing frequency, the short message notification contains key early warning information, and the email push is attached with detailed risk assessment materials; meanwhile, the sending of the early warning information is also linked with an emergency management system to automatically trigger the corresponding emergency plan process according to the risk level.
[0017] As preferred of the present application, before the preprocessing of the original response signal, the method further comprises the step of validity verification of the collected data, which judges whether the data is abnormal by checking the data missing condition and comparing the data statistics with the statistical characteristics of the historical normal data; when there is a small amount of missing data, linear interpolation or spline interpolation method is used for data filling.
[0018] As preferred of the present application, after the wind load risk level assessment, the method further comprises the step of storing and analyzing the assessment results and related data for optimizing the risk assessment model, the stored data includes original response data, preprocessed data, frequency domain feature parameters, and risk assessment results, and the time series analysis method is used to adjust the risk level threshold according to the data change trend; in the analysis process, the wind speed and direction data in the weather forecast are introduced, and the joint analysis is carried out combined with the structure response data to optimize the prediction accuracy of the risk assessment model.
[0019] The structure is a building structure, a bridge structure, or a large infrastructure structure, and the corresponding key monitoring points are determined according to the mechanical characteristics and wind-affected parts of different structure types; for complex structures, the finite element simulation method is used to simulate and analyze the response of the structure under different wind load conditions, thereby assisting in determining the arrangement scheme of the key monitoring points.
[0020] Compared with the prior art, the present application has the following advantages. BRIEF DESCRIPTION OF DRAWINGS
[0021] Figure 1 The present application is a flowchart. DETAILED DESCRIPTION
[0022] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.
[0023] Please refer to Figure 1 The present application provides a technical solution: an equivalent fluctuating wind load frequency domain risk early warning method, comprising the following steps:
[0024] A plurality of types of sensors are arranged at key parts of the structure to collect response data of the structure under the action of wind load in real time.
[0025] Specifically, the plurality of types of sensors are at least two of an acceleration sensor, a strain sensor and a displacement sensor, and the range, resolution, measurement accuracy and sampling frequency of each sensor meet the requirement of accurately collecting response data of the structure under the action of fluctuating wind load; wherein the arrangement of the sensors follows the principle of mechanical analysis, and the specific installation position of the sensors is determined according to the wind vibration response sensitive area calculated by the structural dynamics model.
[0026] The collected original response signal is subjected to denoising, filtering and amplification preprocessing operations.
[0027]
[0027] Specifically, the wavelet denoising algorithm is used for the denoising processing of the original response signal, the wavelet basis function is selected, the decomposition layer number is determined, and the threshold function is used to process the wavelet coefficients; after the denoising processing, the signal smoothing processing step is performed, and the moving average algorithm is used to smooth the denoised signal to eliminate residual small fluctuations.
[0028] Specifically, before the preprocessing of the original response signal, the step of verifying the effectiveness of the collected data is further included, whether the data is abnormal is judged by checking the data missing condition and comparing the data statistics with the historical normal data statistical characteristics; when there is a small amount of missing data, the linear interpolation or spline interpolation method is used for data filling.
[0029] The frequency domain analysis method is used to convert the preprocessed time domain response signal into a frequency domain signal, and the frequency domain characteristic parameters are calculated.
[0030] Specifically, the frequency domain analysis method is Fourier transform or wavelet transform, the fast Fourier transform algorithm is used when the Fourier transform is used, and the appropriate data truncation method is selected; when the wavelet transform is used, the appropriate mother wavelet is selected and the decomposition scale range is determined; before the frequency domain conversion, the preprocessed time domain signal is subjected to normalization processing to make the signal amplitude in a specific interval, thereby improving the accuracy of the frequency domain analysis.
[0031] The risk assessment model based on frequency domain characteristics is established, the mechanical properties of the structure, the design parameters and the historical wind load data are combined to determine the risk level threshold corresponding to different frequency domain characteristic parameters, and the current calculated frequency domain characteristic parameters are compared with the risk level threshold to assess the wind load risk level currently faced by the structure.
[0032] Specifically, the frequency domain characteristic parameters are at least one of power spectral density, autocorrelation function and cross-correlation function, and appropriate algorithms and parameter settings are used when calculating the power spectral density, autocorrelation function and cross-correlation function; the calculated frequency domain characteristic parameters also need to go through a feature selection step, based on principal component analysis method, to remove redundant features and retain key frequency domain characteristic parameters that have significant impact on risk assessment.
[0033] Specifically, the establishment of the risk assessment model also includes the step of training and optimizing the historical data by a machine learning algorithm, the machine learning algorithm sets the number of trees, the maximum depth, and selects appropriate feature selection criteria, and uses cross-validation to optimize the model parameters; in the training process, data augmentation techniques are used to translate, scale and transform the historical data, expand the data set, and improve the generalization ability of the model.
[0034] Specifically, after the wind load risk level assessment, the step of storing and analyzing the assessment results and related data for optimizing the risk assessment model is included, the stored data includes original response data, preprocessed data, frequency domain characteristic parameters and risk assessment results, and the time series analysis method is used to adjust the risk level threshold according to the data trend; in the analysis process, the wind speed and direction data in the weather forecast are introduced, combined with the structure response data for joint analysis, to optimize the prediction accuracy of the risk assessment model.
[0035] When the assessed wind load risk reaches or exceeds the set warning threshold, the warning information is sent to the relevant engineering personnel through the ways of sound and light alarm, short message notification and email push.
[0036] Specifically, the ways of sending warning information are sound and light alarm, short message notification and email push, the sound and light alarm has a specific sound intensity and flashing frequency, the short message notification contains key warning information, and the email push is attached with detailed risk assessment materials; at the same time, the warning information sending is also linked to the emergency management system, and the corresponding emergency plan process is automatically triggered according to the risk level.
[0037] The structure is a building structure, a bridge structure, or a large infrastructure structure. According to the mechanical characteristics and wind-affected parts of different structure types, the corresponding key monitoring points are determined. For complex structures, the finite element simulation method is used to simulate and analyze the response of the structure under different wind load conditions, which helps to determine the layout scheme of the key monitoring points.
[0038] The equivalent fluctuating wind load frequency domain risk early warning method focuses on the goal of real-time and accurate evaluation of structure wind load risk, and realizes effective early warning through multi-link close cooperation.
[0039] In the data acquisition link, at least two of acceleration sensors, strain sensors, and displacement sensors are accurately arranged at the key parts of building structures, bridge structures, or large infrastructure structures. The wind-induced vibration sensitive area is determined based on the structure dynamics model to determine the sensor installation location, ensuring that the collected response data of the structure under wind load is true and effective, laying a foundation for subsequent analysis.
[0040] After the original data is collected, the signal preprocessing stage is entered. First, the data is verified for effectiveness by checking for missing data and comparing statistical characteristics to determine if the data is abnormal. When there is a small amount of missing data, linear interpolation or spline interpolation is used to fill in the missing data. Then, wavelet denoising algorithm is used to remove noise, and moving average algorithm is used for smoothing processing, followed by filtering, amplification, and other operations to ensure signal quality and improve the accuracy of subsequent analysis.
[0041] In the frequency domain analysis link, the preprocessed time domain signal is converted into a frequency domain signal using Fourier transform or wavelet transform. Fast Fourier transform algorithm or selected appropriate mother wavelet and decomposition scale range are used to improve analysis accuracy, and power spectral density, autocorrelation function, cross-correlation function, and other frequency domain characteristic parameters are calculated. Key features are selected through principal component analysis to deeply explore the frequency domain characteristics of wind load.
[0042] Based on the frequency domain characteristics, a risk assessment model is established, and the risk level threshold is determined based on the structure mechanics characteristics, design parameters, and historical wind load data. Machine learning algorithms are used to train and optimize historical data, data augmentation techniques are used to expand the data set, and model parameters are optimized through cross-validation to accurately evaluate the risk level of structure wind load.
[0043] When the evaluated risk reaches or exceeds the set threshold, the early warning module immediately sends warning information to engineering personnel through sound and light alarms, SMS notifications, email pushes, and other methods, and triggers the emergency plan through the emergency management system. The evaluation results and related data are stored and analyzed, weather forecast data is introduced for joint analysis, and the risk assessment model is continuously optimized to form a closed-loop process of data acquisition, processing, analysis, early warning, and optimization, providing reliable protection for the safe operation of structures.
[0044] For example, the application is applied in a large bridge engineering. Acceleration sensors and strain sensors are arranged at key positions of the bridge, such as the bridge pier, the bridge tower and the main beam. During a strong wind weather process, the system collects the response data of the bridge structure in real time, and the frequency domain characteristic parameters of the equivalent fluctuating wind load are calculated through signal preprocessing and frequency domain conversion. The risk assessment model judges that the bridge currently faces a high wind load risk according to these parameters, and immediately issues a warning signal when the assessed wind load risk reaches or exceeds the set warning threshold. After receiving the warning, the engineering personnel quickly start the emergency plan, conduct traffic control on the bridge, and organize professional personnel to inspect the bridge structure, so that the loosening problems in some connecting positions are found and handled in time, and the safety of the bridge is effectively ensured.
[0045] Although embodiments of the application have been shown and described, it is to be understood that various modifications, substitutions, replacements and changes can be made to these embodiments without departing from the principles and spirit of the application, and the scope of the application is defined by the appended claims and their equivalents.
Claims
1. A frequency domain risk early warning method for equivalent pulsating wind load, characterized in that, Includes the following steps: Multiple types of sensors are deployed at key parts of the structure to collect real-time response data of the structure under wind load. The acquired raw response signal is preprocessed by denoising, filtering, and amplification. Frequency domain analysis methods are used to convert the preprocessed time-domain response signal into a frequency-domain signal, and frequency domain characteristic parameters are calculated. Establish a risk assessment model based on frequency domain characteristics, combine the mechanical properties of the structure, design parameters and historical wind load data, determine the risk level thresholds corresponding to different frequency domain characteristic parameters, compare the currently calculated frequency domain characteristic parameters with the risk level thresholds, and assess the wind load risk level currently faced by the structure. When the assessed wind load risk reaches or exceeds the set warning threshold, warning information is sent to relevant engineering personnel through audible and visual alarms, SMS notifications, and email pushes.
2. The frequency domain risk early warning method for equivalent pulsating wind load according to claim 1, characterized in that: The various types of sensors include at least two of the following: acceleration sensors, strain sensors, and displacement sensors. The range, resolution, measurement accuracy, and sampling frequency of each sensor meet the requirements for accurately acquiring the response data of the structure under pulsating wind loads.
3. The frequency domain risk early warning method for equivalent pulsating wind load according to claim 1, characterized in that: The original response signal is denoised using a wavelet denoising algorithm, which involves selecting a suitable wavelet basis function, determining the number of decomposition levels, and using a threshold function to process the wavelet coefficients. After the denoising process, a signal smoothing process is performed, which uses a moving average algorithm to smooth the denoised signal in order to eliminate residual minor fluctuations.
4. The frequency domain risk early warning method for equivalent pulsating wind load according to claim 1, characterized in that: The frequency domain analysis method is Fourier transform or wavelet transform. When using Fourier transform, the fast Fourier transform algorithm is used, and an appropriate data truncation method is selected. When using wavelet transform, select an appropriate mother wavelet and determine the decomposition scale range; Before performing frequency domain conversion, the preprocessed time domain signal is normalized to ensure that the signal amplitude is within a specific range, thereby improving the accuracy of frequency domain analysis.
5. The frequency domain risk early warning method for equivalent pulsating wind load according to claim 1, characterized in that: The frequency domain characteristic parameters are at least one of power spectral density, autocorrelation function, and cross-correlation function, and appropriate algorithms and parameter settings are used when calculating power spectral density, autocorrelation function, and cross-correlation function. The calculated frequency domain feature parameters still need to undergo a feature screening step. Based on the principal component analysis method, redundant features are removed, and key frequency domain feature parameters that have a significant impact on risk assessment are retained.
6. The frequency domain risk early warning method for equivalent pulsating wind load according to claim 1, characterized in that: The establishment of the risk assessment model also includes the step of training and optimizing historical data through machine learning algorithms. The machine learning algorithms optimize the model parameters by setting the number of trees, the maximum depth, and selecting appropriate feature selection criteria, and using cross-validation. During training, data augmentation techniques are used to translate and scale historical data to expand the dataset and improve the model's generalization ability.
7. The frequency domain risk early warning method for equivalent pulsating wind load according to claim 1, characterized in that: The warning information is sent through audible and visual alarms, SMS notifications, and email push notifications. The audible and visual alarms have specific sound intensities and flashing frequencies, the SMS notifications contain key warning information, and the email push notifications include detailed risk assessment data. At the same time, the warning information is also linked to the emergency management system, which automatically triggers the corresponding emergency response procedures based on the risk level.
8. The frequency domain risk early warning method for equivalent pulsating wind load according to claim 1, characterized in that: Before the raw response signal is preprocessed, the method also includes a step of validating the collected data. This involves checking for missing data and comparing the data statistics with the statistical characteristics of historical normal data to determine whether the data is abnormal. When there are a few missing data points, linear interpolation or spline interpolation methods are used to fill in the missing data.
9. The frequency domain risk early warning method for equivalent pulsating wind load according to claim 1, characterized in that: After the wind load risk level assessment, the method also includes storing and analyzing the assessment results and related data to optimize the risk assessment model. The stored data includes the original response data, preprocessed data, frequency domain characteristic parameters, and risk assessment results. The risk level threshold is adjusted according to the data change trend using time series analysis. The analysis process incorporates wind speed and direction data from weather forecasts, and combines them with structural response data for joint analysis to optimize the predictive accuracy of the risk assessment model.