Engineering Vibration Detection System Based on Machine Vision

Abstract

The present invention relates to the field of vibration detection technology, specifically an engineering vibration detection system based on machine vision. The engineering vibration detection system based on machine vision includes a high-speed image capture module, a vibration feature extraction module, a frequency response mapping module, and a data fusion detection module. In the present invention, by continuously capturing video data on the surface of an engineering structure and extracting time series data therefrom, high-precision monitoring of subtle vibration fluctuations is achieved. Time-frequency analysis can more accurately identify key vibration frequencies and amplitudes. By constructing a mapping relationship between frequency and amplitude, subtle changes in vibration frequency can be effectively revealed, enabling predictive maintenance and early fault identification, enhancing the reliability of data and the depth of analysis, comprehensively analyzing recognition deviations and abnormal patterns, greatly improving the accuracy and timeliness of structural health monitoring, and significantly reducing maintenance costs and downtime.

Description

Technical Field

[0001] The present invention relates to the technical field of vibration detection, and particularly to an engineering vibration detection system based on machine vision. Background Art

[0002] The technical field of vibration detection mainly involves using various sensors and devices to monitor and analyze the vibrations generated by mechanical equipment, structures, and systems during operation, which can identify abnormal vibrations and help maintenance personnel diagnose potential faults or damages. Vibration analysis can evaluate the health status of equipment by means of frequency analysis, amplitude evaluation, and waveform comparison to determine whether the equipment is within the normal operating range. In addition, vibration detection is also widely used in predictive maintenance strategies. By continuously monitoring and trend analysis, potential faults can be identified preventively, thereby reducing downtime and maintenance costs.

[0003] Among them, the engineering vibration detection system is an application of vibration detection technology, mainly used to monitor the vibration conditions of building structures, bridges, industrial equipment, etc. It can capture data in real time and evaluate the stability of the structure and the operating efficiency of the equipment by analyzing the characteristics of the vibration. The purpose is to ensure structural safety, optimize mechanical performance, and early warning of potential structural failure risks, thereby ensuring personnel safety and the long-term operation of the equipment.

[0004] Existing vibration detection technologies are physically limited in terms of data acquisition coverage and frequency resolution, making it difficult to capture tiny or high-frequency vibrations. Frequency analysis and waveform comparison in traditional technologies rely on post-data processing and cannot provide real-time feedback, which may lead to insufficient response to structural problems in engineering applications and increase safety risks. Due to the single data source, the accuracy and reliability of traditional methods are usually affected in complex or interference environments. The lack of real-time data analysis and comprehensive trend prediction also limits the application effect in predictive maintenance strategies and cannot fully utilize data-driven decision support, resulting in greater economic losses and safety accidents due to the failure to predict and handle potential faults in a timely manner. Summary of the Invention

[0005] The purpose of the present invention is to solve the disadvantages existing in the prior art, and to propose an engineering vibration detection system based on machine vision.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions: The engineering vibration detection system based on machine vision includes:

[0007] The high-speed image capture module continuously records the video data on the surface of the engineering structure through a camera, captures time-series data from the video stream, and generates vibration fluctuation signal data;

[0008] The vibration feature extraction module performs time-frequency analysis on the signal based on the vibration fluctuation signal data, identifies the key vibration frequencies, calculates the amplitude according to the identification results of the key vibration frequencies, extracts and classifies the differential vibration modes, and obtains the vibration feature data set;

[0009] The frequency response mapping module calculates the frequency value of each data point based on the vibration feature data set, constructs the mapping relationship between frequency and amplitude, reveals the frequency change situation according to the mapping relationship, corresponds with the amplitude, and updates and plots the frequency response diagram in real time;

[0010] The data fusion detection module fuses the frequency response diagram with the vibration sensor data, performs data comparison and historical trend analysis, identifies deviations according to the analysis results, determines the abnormal vibration mode, and generates the engineering structure health analysis result.

[0011] As a further solution of the present invention, the steps for obtaining the vibration fluctuation signal data are as follows:

[0012] Configure cameras at key parts of the engineering structure to capture surface changes in real time and generate real-time video stream data;

[0013] Perform image analysis on the real-time video stream data, identify the motion information and changes in each frame, extract the time series, and obtain the time series data set;

[0014] Based on the time series data set, use the formula:

[0015]

[0016] Calculate the change amount Δs between frames to obtain the vibration fluctuation signal data, where frame i and frame i-1 respectively represent the i-th and i-1 key frames, and m represents the total number of key frames.

[0017] As a further solution of the present invention, the steps for identifying the key vibration frequencies are as follows:

[0018] Perform time-frequency conversion on the vibration fluctuation signal data, analyze the frequency components of the signal at different time points through short-time Fourier transform, and generate the signal time-frequency diagram;

[0019] According to the signal time-frequency diagram, perform peak detection to identify the frequency points with concentrated key energy, and generate a frequency candidate list; based on the frequency candidate list, perform energy threshold screening and statistical significance test, using the formula:

[0020]

[0021] Determine the key vibration frequencies and generate a list of key vibration frequencies. Here, f represents the frequency point, E(f) represents the energy value of the frequency point f, S(f) represents the statistical significance index value of the frequency point f, ∈ is a constant, and γ is the threshold of the energy-to-significance ratio.

[0022] As a further aspect of the present invention, the steps for obtaining the vibration feature dataset are as follows:

[0023] Based on the list of key vibration frequencies, use the formula:

[0024]

[0025] Calculate the amplitude value A(f) of each frequency point and generate a list of amplitude values. Here, P(f) represents the power spectral density of the frequency f, and 2π is a normalization constant used to adjust the scale factor for amplitude calculation;

[0026] Perform pattern recognition on the list of amplitude values to identify vibration patterns, including normal operation patterns and potential fault patterns, and output a set of classified vibration patterns;

[0027] Integrate the set of classified vibration patterns, record the key frequencies, amplitudes, and classification labels of all vibration patterns, and construct a vibration feature dataset.

[0028] As a further aspect of the present invention, the steps for constructing the mapping relationship between frequency and amplitude are as follows:

[0029] Extract the time series of each data point from the vibration feature dataset and use the formula:

[0030]

[0031] Calculate the frequency value F(f) of each data point and generate spectral data. Here, x(n) is the signal value at the nth time point, N is the total number of time points in the time series, and f is the frequency point;

[0032] Analyze the spectral data, identify the key frequency components, and record the peaks and corresponding amplitudes to obtain frequency peak and amplitude pairing data;

[0033] Utilize the frequency peak and amplitude pairing data to capture the relationship between frequency and amplitude and generate a mapping relationship table between frequency and amplitude.

[0034] As a further aspect of the present invention, the steps for obtaining the frequency response diagram are as follows:

[0035] According to the frequency and amplitude mapping relationship table, monitor new vibration data in real time. For each newly monitored frequency data point, query the mapping table to extract the corresponding amplitude value and generate real-time pairing data;

[0036] Based on the real-time pairing data, perform digital signal processing to remove noise and outliers in the data, and obtain the processed frequency and amplitude data;

[0037] For the processed frequency and amplitude data, dynamically display the data, reveal the changes in vibration characteristics in a graphical manner, and draw a frequency response diagram.

[0038] As a further solution of the present invention, the execution steps of the data comparison and historical trend analysis are as follows:

[0039] Collect vibration sensor data in real time, fuse it with the frequency response diagram, integrate the two data forms, and obtain a comprehensive data set;

[0040] Perform time series analysis on the comprehensive data set, extract the historical trend of the data, and obtain the historical trend analysis result;

[0041] Based on the historical trend analysis result, perform pattern matching and data comparison to identify abnormal fluctuations in the data, and obtain the data comparison and trend analysis result.

[0042] As a further solution of the present invention, the steps for obtaining the engineering structure health analysis result are as follows:

[0043] Based on the data comparison and trend analysis result, perform deviation analysis to determine whether the deviation exceeds a preset threshold, identify abnormal points in the data, and generate a preliminary deviation identification result;

[0044] Perform cluster analysis on the preliminary deviation identification result, group the deviation data points according to vibration characteristics, identify common abnormal patterns, including periodic faults and random faults, and generate a cluster analysis result;

[0045] Using the cluster analysis result, record all abnormal vibration patterns and their potential impacts, formulate maintenance suggestions, and generate the engineering structure health analysis result.

[0046] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0047] In the present invention, by continuously capturing video data on the surface of the engineering structure and extracting time series data therefrom, high-precision monitoring of subtle vibration fluctuations is achieved. Time-frequency analysis can more accurately identify key vibration frequencies and amplitudes. By constructing a mapping relationship between frequency and amplitude, subtle changes in vibration frequency can be effectively revealed. Predictive maintenance and early fault identification can be carried out, enhancing the reliability of data and the depth of analysis. By comprehensively analyzing and identifying deviations and abnormal patterns, the accuracy and timeliness of structural health monitoring are greatly improved, and the maintenance cost and downtime are significantly reduced. Description of the Drawings

[0048] Figure 1 is the system flow chart of the present invention;

[0049] Figure 2 is the acquisition flow chart of the vibration fluctuation signal data of the present invention;

[0050] Figure 3 is the identification flow chart of the key vibration frequency of the present invention;

[0051] Figure 4 is the acquisition flow chart of the vibration feature data set of the present invention;

[0052] Figure 5 is the construction flow chart of the mapping relationship between the frequency and the amplitude of the present invention;

[0053] Figure 6 is the acquisition flow chart of the frequency response diagram of the present invention;

[0054] Figure 7 is the execution flow chart of the data comparison and historical trend analysis of the present invention;

[0055] Figure 8 is the acquisition flow chart of the engineering structure health analysis result of the present invention. Detailed implementation manner

[0056] In order to make the objectives, technical solutions and advantages of the present invention clearer, 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.

[0057] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the accompanying drawings, and is 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 therefore should not be construed as limiting the present invention. In addition, in the description of the present invention, "a plurality of" means two or more unless otherwise specifically defined.

[0058] Please refer to Figure 1 , the engineering vibration detection system based on machine vision includes:

[0059] The high-speed image capture module continuously records the video data on the surface of the engineering structure through the camera, captures the time series data from the video stream, and generates the vibration fluctuation signal data;

[0060] The vibration feature extraction module performs time-frequency analysis on the vibration fluctuation signal data, identifies the key vibration frequencies, calculates the amplitudes based on the identification results of the key vibration frequencies, extracts and classifies the differential vibration modes, and obtains the vibration feature data set;

[0061] The frequency response mapping module calculates the frequency values of each data point based on the vibration feature data set, constructs the mapping relationship between the frequency and the amplitude, reveals the frequency change situation according to the mapping relationship, corresponds with the amplitude, and updates and plots the frequency response diagram in real time;

[0062] The data fusion detection module fuses the frequency response diagram with the vibration sensor data, performs data comparison and historical trend analysis, identifies the deviation according to the analysis results, determines the abnormal vibration mode, and generates the engineering structure health analysis result.

[0063] The vibration fluctuation signal data includes vibration frequency data, time tags, and spatial positioning information. The vibration feature data set includes frequency distribution records, amplitude analysis results, and pattern recognition tags. The frequency response chart includes frequency values, amplitude mapping results, and real-time update diagrams. The engineering structure health analysis result includes deviation identification results and abnormal vibration mode analysis records.

[0064] Please refer to Figure 2 , and the steps for obtaining the vibration fluctuation signal data are as follows:

[0065] Configure cameras at the key parts of the engineering structure to capture the surface changes in real time and generate real-time video stream data;

[0066] The process of configuring cameras at the key parts of the engineering structure involves determining the optimal installation angles and positions of the cameras so as to capture the key dynamic changes of the structure. The technology for capturing the surface changes in real time relies on high-resolution video capture devices. The steps for the device to record the subtle structural changes and obtain a continuous video stream include using software for real-time transmission and storage of the video stream, requiring sufficient storage space and good transmission bandwidth. During the process of generating real-time video stream data, the integrity and real-time nature of the data should be ensured to guarantee the accuracy and effectiveness of subsequent analysis, and high-quality video data that can be directly used for analysis is obtained.

[0067] Perform image analysis on the real-time video stream data, identify the motion information and changes in each frame, extract the time series, and obtain the time series data set;

[0068] The process of image analysis from real-time video stream data begins with the preprocessing of video data, including noise reduction and resolution adjustment, to improve the efficiency and accuracy of subsequent processing. During the process of extracting time series through edge detection and object tracking algorithms and identifying motion information and changes in each frame, the algorithm needs to be adjusted to adapt to different environmental and lighting conditions to ensure that key data can be accurately extracted from dynamic videos, obtaining a set of time series data to ensure the reliability of the data and the accuracy of subsequent analysis.

[0069] Based on the set of time series data, the formula:

[0070]

[0071] is used to calculate the change amount Δs between frames, obtaining vibration fluctuation signal data, where frame i and frame i-1 represent the i-th and (i - 1)-th key frames respectively, and m represents the total number of key frames;

[0072] There are three frames of data, and each frame of data represents a digital quantization value of a structural change. The data is as follows:

[0073] The differential quantization value from the 1st frame to the 2nd frame is 1000.

[0074] The differential quantization value from the 2nd frame to the 3rd frame is 1500.

[0075] Then the calculation process is as follows:

[0076]

[0077] The calculated change amount between frames is 1802.78, representing the magnitude of the vibration fluctuation signal extracted from the video data. This value reflects the amplitude of the dynamic changes in the structure due to various external or internal factors during the observation period. A higher vibration fluctuation signal value may indicate that the structure has experienced significant physical changes or been subjected to large external disturbances during the monitoring period, which is crucial for engineering safety monitoring and maintenance and can help engineers identify possible structural problems or areas that require further inspection.

[0078] Please refer to Figure 3 , and the steps for identifying the key vibration frequency are as follows:

[0079] Perform time-frequency conversion on the vibration fluctuation signal data, analyze the frequency components of the signal at different time points through short-time Fourier transform, and generate a signal time-frequency diagram;

[0080] Before performing spectral analysis on the vibration fluctuation signal data, the data is first processed through a digital filter to filter out the noise within the non-target frequency range and enhance the frequency band of interest in the signal. Butterworth or Chebyshev filters are used depending on their order and type to adapt to different signal conditions. Subsequently, the data undergoes a short-time Fourier transform (STFT). During this process, the signal is decomposed into a combination of a series of sine and cosine waves, each wave corresponding to a specific frequency, and its amplitude and phase describe the contribution of that frequency in the original signal. The implementation of STFT is achieved by multiplying point by point with a complex exponential function and accumulating to obtain the spectrum, which details the energy distribution from the lowest to the highest measurable frequencies, providing a detailed view for further analysis and enabling the subsequent frequency identification process to more accurately find those frequency components that dominate the structural response.

[0081] Based on the signal time-frequency diagram, peak detection is performed to identify the frequency points where the key energy is concentrated, generating a list of frequency candidates;

[0082] After obtaining the spectrum distribution diagram, the next step is to use advanced signal processing techniques to identify the key vibration frequency points in the diagram. The key to this step lies in using a local peak detection algorithm, which locates the key frequencies by searching for the frequency points in the spectrum where the energy is significantly higher than the surrounding energy. For this purpose, a dynamic threshold is set, which is dynamically adjusted according to the average energy and standard deviation of the signal to adapt to the variability of different signals. The detection of key frequencies not only depends on the energy magnitude of the frequencies but also needs to consider their persistence in the spectrum and the energy comparison with neighboring frequency points to ensure that the identified frequency points are physically interpretable and technically reliable. Each frequency point identified as key will be recorded and used for further analysis, such as predicting possible structural defects or performance degradation through the amplitude and phase information of these frequency points, providing real-time monitoring and early warning of the structural health status.

[0083] Based on the list of frequency candidates, energy threshold screening and statistical significance testing are performed, using the formula:

[0084]

[0085] Determine the key vibration frequencies, generating a list of key vibration frequencies, where f represents the frequency point, E(f) represents the energy value of the frequency point f, S(f) represents the statistical significance index value of the frequency point f, ∈ is a constant to avoid division by zero errors, and γ is the threshold of the energy-to-significance ratio;

[0086] In a certain actual measurement, for the frequency point f = 500 Hz, the obtained energy E(f) = 200 unit energies, and the statistical significance S(f) = 20. Set ∈ = 0.001 and γ = 0.05. Then substitute into the formula for calculation:

[0087]

[0088] The result shows that when the ratio of the energy at the frequency point of 500 Hz to the statistical significance is greater than 0.05, this frequency point is identified as a key vibration frequency. The judgment process helps to further screen out the main frequencies affecting the dynamic characteristics of the system, which is of great significance for the monitoring and maintenance of the system.

[0089] Please refer to Figure 4 , the steps for obtaining the vibration characteristic data set are as follows:

[0090] Based on the list of key vibration frequencies, use the formula:

[0091]

[0092] Calculate the amplitude value A(f) of each frequency point to generate a list of amplitude values. Among them, P(f) represents the power spectral density of frequency f, and 2π is a normalization constant used to adjust the scale factor for amplitude calculation;

[0093] At the point of 500 Hz, the power spectral density is 50 units. Then insert specific values for calculation:

[0094]

[0095] It shows that the amplitude at 500 Hz is approximately 2.82 units, which helps to evaluate the vibration intensity at this frequency.

[0096] Perform pattern recognition on the list of amplitude values to identify vibration patterns, including normal operation patterns and potential fault patterns, and output the classified set of vibration patterns;

[0097] After obtaining the amplitude data from the key frequency points, these amplitudes are classified through pattern recognition methods to identify different vibration patterns. This process involves using machine learning algorithms, especially clustering algorithms such as K-means or hierarchical clustering, to analyze the amplitude data. The clustering algorithm evaluates the similarity between data points and groups data with similar amplitude values into the same class. Such classification helps to identify common operating vibration patterns and potential fault patterns. For example, a normally operating device usually exhibits a consistent vibration pattern, while a faulty device may show an abnormal vibration pattern. This step not only includes the application of clustering algorithms but also requires tuning of algorithm parameters such as selecting the appropriate number of clusters and distance metrics to ensure the effectiveness and accuracy of the clustering results. Through this analysis, valuable diagnostic information can be extracted from the vibration data.

[0098] Integrate the classified set of vibration patterns, record the key frequencies, amplitudes, and classification labels of all vibration patterns, and construct a vibration characteristic data set;

[0099] Constructing a vibration feature dataset is a crucial step in integrating the analysis results. This dataset includes the classification information obtained from the previous steps, with each category containing different vibration frequencies, corresponding amplitudes, and identified vibration modes. The construction process of this dataset involves data collation and formatting to ensure that each piece of data accurately reflects the actual vibration situation. Each record in the dataset is based on the vibration data obtained from actual measurements, and the data is processed through advanced signal processing techniques to ensure the quality and reliability of the data. In addition, this dataset is also applied to the training of machine learning models for predicting and identifying various vibration modes that may occur in the future. This not only helps to monitor the device status in real time but also enables preventive measures to be taken before potential problems occur, thereby reducing unplanned downtime and improving device efficiency and productivity.

[0100] Please refer to Figure 5 , and the steps for constructing the mapping relationship between frequency and amplitude are as follows:

[0101] Extract the time series of each data point from the vibration feature dataset and use the formula:

[0102]

[0103] Calculate the frequency value F(f) of each data point to generate spectral data, where x(n) is the signal value at the nth time point, N is the total number of points in the time series, and f is the frequency point;

[0104] There is a simple time series data x(n) = [0, 1, 2, 3], and N = 4.

[0105] Calculate F(1), that is, the spectrum when f = 1:

[0106] When n = 0, x(0) = 0, and the contribution is 0.

[0107] When n = 1, x(1) = 1, and the contribution is 1×e -i·2π·1 / 4 = 1×(0.707 - 0.707i).

[0108] When n = 2, x(2) = 2, and the contribution is 2×e -i·2π·2 / 4 = 2×(-1).

[0109] When n = 3, x(3) = 3, and the contribution is 3×e -i·2π·3 / 4 = 3×(-0.707 - 0.707i).

[0110] F(1) = 0 + (0.707 - 0.707i) + (-2) + (-2.121 - 2.121i)

[0111] F(1) = -3.293 - 2.828i

[0112] The results show that the component with a frequency of 1 has a relatively large amplitude and a complex phase, indicating that the contribution of this frequency component in the original signal is significant. This numerical result further reflects the complex variation of the original signal at frequency 1.

[0113] Analyze the spectral data, identify the key frequency components, and record the peaks and corresponding amplitudes to obtain the paired data of frequency peaks and amplitudes.

[0114] Analyze the frequency distribution in the spectral data to identify the main frequency components. First, set an amplitude threshold and compare whether the amplitude of each frequency point exceeds this threshold to effectively screen out the significant frequency points in the signal. Then, record these main frequency points and their corresponding amplitude values to provide the basic data for establishing the mapping relationship. This process not only involves threshold comparison but also includes classifying the frequency points. Consider the frequency points above the threshold as the main frequency components, and at the same time, record the specific frequency values and relative amplitudes of each main frequency point in detail to ensure the accuracy and reliability of these data, preparing for constructing a detailed frequency-amplitude relationship diagram. The recorded data will be directly applied to the next mathematical modeling analysis.

[0115] Use the paired data of frequency peaks and amplitudes to capture the relationship between frequency and amplitude and generate a mapping relationship table between frequency and amplitude.

[0116] Use the collected paired data of frequency peaks and amplitudes to construct a mathematical model to describe the mapping relationship between frequency and amplitude. This modeling process uses the non-linear regression method to calculate the parameters of the regression model by the least squares method according to the collected data points to ensure the prediction accuracy and generalization ability of the model. This kind of model can predict the corresponding amplitude according to the input frequency. The construction of the model is not only based on statistical analysis but also considers the physical rationality to ensure that the relationship between frequency and amplitude can accurately reflect the essence of the actual physical phenomenon. The finally obtained mapping relationship table provides a reliable basis for further analysis of vibration characteristics. The mapping relationship helps to better understand and predict the response behavior of the system at different frequencies.

[0117] Please refer to Figure 6 , the steps to obtain the frequency response diagram are as follows:

[0118] According to the frequency-amplitude mapping relationship table, monitor the new vibration data in real time. For each newly monitored frequency data point, query the mapping table to extract the corresponding amplitude value and generate real-time paired data.

[0119] The real-time monitoring system relies on an established mapping relation table to respond to changes in frequency data. In this process, after the system first detects new frequency data, it immediately obtains the predicted amplitude value by quickly querying the mapping relation table. The query operation is based on an efficient data structure to optimize the retrieval speed to cope with frequent data updates and query requirements. The mapping relation table is obtained by training a machine learning model with a historical data set, ensuring an accurate mapping from frequency to amplitude. This real-time query and response process not only enhances the system's monitoring ability for dynamic changes but also enables the relationship between frequency and amplitude to be updated immediately, providing real-time and accurate input data for subsequent data analysis and processing.

[0120] Based on real-time paired data, digital signal processing is carried out to remove noise and outliers in the data, obtaining processed frequency and amplitude data.

[0121] In the process of processing the frequency and amplitude data obtained in real time, advanced digital signal processing techniques are adopted. First, the sliding average algorithm is implemented to smooth the data to eliminate occasional noise and short-term fluctuations. In addition, outliers are clipped by setting a dynamic threshold. These outliers often originate from sensor errors or external interference. The data processing at this stage not only improves the stability of the data but also ensures the accuracy of data analysis, providing high-quality data for visualization, ensuring the continuity and efficiency of the data processing flow, and at the same time providing technical support for the system's anomaly detection and response.

[0122] For the processed frequency and amplitude data, the data is dynamically displayed, and the changes in vibration characteristics are revealed in a graphical way to draw a frequency response diagram.

[0123] The processed data is transmitted in real time to the graphical user interface (GUI) for drawing a dynamic frequency response diagram. The visualization tool used can immediately display the relationship between frequency and amplitude. The real-time updated graphical display helps users intuitively understand and analyze the changes in vibration data. This chart is implemented through advanced graphical drawing techniques, including data-driven graphical rendering and dynamic zooming functions, which can process large-scale data streams and display complex data relationships. This kind of real-time graphical display is an indispensable part of the monitoring system, not only enhancing the user interaction experience but also strengthening the real-time nature of data monitoring and the effectiveness of decision-making. The system can quickly respond to any vibration anomaly.

[0124] Please refer to Figure 7 , and the execution steps for data comparison and historical trend analysis are as follows:

[0125] Vibration sensor data is collected in real time and fused with the frequency response diagram, integrating the two data forms to obtain a comprehensive data set.

[0126] Fuse real-time vibration data with existing frequency response diagrams to form a comprehensive data set. The process involves integrating sensor data with frequency data, using data synchronization techniques to ensure timestamp alignment, and unifying the format to ensure data type consistency. The comprehensive data set allows for the analysis of time-domain and frequency-domain information within a unified framework. This fusion process is to better reflect the current working state of the device. Through this method, the accuracy of subsequent analysis can be ensured, providing a data basis for identifying potential vibration anomalies.

[0127] Perform time series analysis on the comprehensive data set to extract the historical trend of the data and obtain the historical trend analysis results.

[0128] Utilize the comprehensive data set to perform historical trend analysis. Adopt time series analysis methods such as moving average and exponential smoothing techniques to extract and highlight the main trends in the data. The analysis process calculates the moving average to smooth out data fluctuations and strengthens the weight of recent data through the exponential smoothing method to more sensitively capture possible change trends. This technology helps to separate the true trend from random fluctuations, providing a basis for subsequent anomaly pattern recognition. By comparing the long-term and short-term behaviors of the data, future behavior patterns can be predicted, and those data patterns that deviate from the norm can be identified, which is crucial for the early diagnosis of potential equipment problems.

[0129] Based on the historical trend analysis results, perform pattern matching and data comparison to identify abnormal fluctuations in the data and obtain the data comparison and trend analysis results.

[0130] Based on the data analysis and trend recognition results of the foregoing steps, conduct the identification of abnormal patterns and the generation of engineering structure health analysis results. This includes using statistical anomaly detection techniques such as Z-score and standard deviation analysis to identify abnormal fluctuations in the data, which can accurately mark the abnormal points in the data, indicating potential faults in the device. Then, combined with the historical maintenance records and engineering parameters of the device, recommend the optimal maintenance plan, which not only improves the reliability of the device but also optimizes the maintenance cost. Through the comprehensive analysis results, engineers can timely adjust the operation strategy and intervene in advance to prevent serious equipment failures and ensure the continuity of production and the long-term operation of the equipment.

[0131] Please refer to Figure 8 , the steps to obtain the engineering structure health analysis results are as follows:

[0132] Based on the data comparison and trend analysis results, conduct deviation analysis to determine whether the deviation exceeds the preset threshold, identify the abnormal points in the data, and generate the preliminary deviation identification results.

[0133] Based on the data comparison and trend analysis results, perform deviation analysis techniques to identify outliers in the data. This operation includes calculating the deviation between the real-time value of each data point and its historical average, which is calculated from the data collected in the past under the same operating conditions. A dynamic threshold is set to evaluate these deviations. The setting of the threshold is based on past performance fluctuations and industry standards. Data points exceeding this threshold will be considered potential outliers. System deviation evaluation helps accurately identify those key data points that may indicate equipment performance degradation. Through precise deviation detection methods, preliminary deviation identification results can be generated, providing key input data for subsequent analysis steps.

[0134] Conduct cluster analysis on the preliminary deviation identification results, group the deviation data points according to vibration characteristics, identify common abnormal patterns, including periodic faults and random faults, and generate cluster analysis results;

[0135] Conduct in-depth cluster analysis on the data in the preliminary deviation identification results to clarify and classify the abnormal vibration patterns of the equipment. In this step, the data is grouped according to the deviation magnitude and frequency characteristics of the data points, with the aim of identifying specific vibration abnormal patterns, such as periodic faults or random faults. By analyzing the similarity of the data points in each group, the possible fault types of the equipment can be identified. This analysis not only helps to understand the current abnormal state of the equipment but also can predict potential future problems. The generated cluster analysis results provide a scientific basis for the maintenance and fault prevention of the equipment, enabling the maintenance team to formulate more effective maintenance strategies for specific abnormal patterns.

[0136] Utilize the cluster analysis results, record all abnormal vibration patterns and their potential impacts, formulate maintenance suggestions, and generate engineering structure health analysis results;

[0137] Integrate the abnormal vibration patterns identified through cluster analysis into the health analysis report of the engineering structure. This process includes detailed analysis of the cluster results and conversion into specific maintenance suggestions. The report details various abnormal vibration patterns and their potential impacts on equipment operation. In addition, it includes preventive measures and response strategies for each vibration pattern. In this way, not only a comprehensive view of the current state of the equipment is provided, but also potential future problems are predicted, enabling the operation and maintenance team to adjust operation strategies and maintenance plans in a timely manner according to the detailed information and suggestions in the report, and extending the service life of the project.

[0138] 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 relevant art may use the technical content disclosed above 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. An engineering vibration detection system based on machine vision, characterized in that, The system includes: The high-speed image capture module continuously records the video data of the surface of the engineering structure through a camera, captures the time series data from the video stream, and generates vibration fluctuation signal data; The vibration feature extraction module performs time-frequency analysis on the signal based on the vibration fluctuation signal data, identifies the key vibration frequencies, calculates the amplitude according to the identification result of the key vibration frequencies, extracts and classifies the different vibration modes, and obtains a vibration feature data set; The frequency response mapping module calculates the frequency value of each data point based on the vibration feature data set, constructs the mapping relationship between the frequency and the amplitude, reveals the frequency change situation according to the mapping relationship, corresponds with the amplitude, and updates and draws the frequency response diagram in real time; The data fusion detection module fuses the frequency response diagram with the vibration sensor data, performs data comparison and historical trend analysis, identifies the deviation according to the analysis result, determines the abnormal vibration mode, and generates the health analysis result of the engineering structure; The identification steps of the key vibration frequencies are as follows: Perform time-frequency conversion on the vibration fluctuation signal data, analyze the frequency components of the signal at different time points through short-time Fourier transform, and generate a signal time-frequency diagram; According to the signal time-frequency diagram, perform peak detection, identify the frequency points where the key energy is concentrated, and generate a frequency candidate list; Based on the frequency candidate list, perform energy threshold screening and statistical significance test, using the formula: ； Determine the key vibration frequencies and generate a list of key vibration frequencies, where represents a frequency point, represents a frequency point is the energy value of represents a frequency point is the statistical significance index value of is a constant, is the threshold of the energy-to-significance ratio; The acquisition steps of the vibration feature data set are as follows: Based on the key vibration frequency list, using the formula: ； Calculate the amplitude value at each frequency point , generate a list of amplitude values, where represents the power spectral density of the frequency , and is the normalization constant; Perform pattern recognition on the amplitude value list, identify the vibration modes, including the normal operation mode and the potential fault mode, and output the classified vibration mode set; Integrate the classified vibration mode set, record the key frequencies, amplitudes and classification labels of all vibration modes, and construct a vibration feature data set.

2. The engineering vibration detection system based on machine vision according to claim 1, characterized in that, The acquisition steps of the vibration fluctuation signal data are as follows: Configure a camera at the key parts of the engineering structure to capture the surface changes in real time and generate real-time video stream data; Perform image analysis on the real-time video stream data, identify the motion information and changes in each frame, extract the time series, and obtain a time series data set; Based on the time series data set, using the formula: ； Calculate the change amount between frames , and obtain the vibration fluctuation signal data, where and respectively represent the -th and the -th key frames, represents the total number of key frames.

3. The engineering vibration detection system based on machine vision according to claim 1, characterized in that The construction steps of the mapping relationship between the frequency and the amplitude are as follows: Extract the time series of each data point from the vibration feature data set, using the formula: ； Calculate the frequency value of each data point , generate spectral data, where is the signal value at the th time point, is the total number of points in the time series, is the frequency point; Analyze the spectrum data, identify the key frequency components, and record the peaks and the corresponding amplitudes to obtain the frequency peak and amplitude pairing data; Utilize the frequency peak and amplitude pairing data to capture the relationship between the frequency and the amplitude, and generate a mapping relationship table between the frequency and the amplitude.

4. The engineering vibration detection system based on machine vision according to claim 3, characterized in that, The acquisition steps of the frequency response diagram are as follows: According to the frequency and amplitude mapping relationship table, monitor the new vibration data in real time. For each newly monitored frequency data point, query the mapping table to extract the corresponding amplitude value and generate real-time pairing data; Based on the real-time pairing data, perform digital signal processing to remove the noise and outliers in the data, and obtain the processed frequency and amplitude data; For the processed frequency and amplitude data, dynamically display the data, reveal the changes of the vibration characteristics in a graphical way, and draw the frequency response diagram.

5. The engineering vibration detection system based on machine vision according to claim 4, characterized in that The execution steps of the data comparison and historical trend analysis are as follows: Collect vibration sensor data in real time, fuse it with the frequency response diagram, integrate the two data forms, and obtain a comprehensive data set; Perform time series analysis on the comprehensive data set, extract the historical trend of the data, and obtain the historical trend analysis result; Based on the historical trend analysis result, perform pattern matching and data comparison to identify abnormal fluctuations in the data, and obtain the data comparison and trend analysis result.

6. The engineering vibration detection system based on machine vision according to claim 5, characterized in that, The steps for obtaining the engineering structure health analysis result are as follows: Based on the data comparison and trend analysis result, perform deviation analysis to determine whether the deviation exceeds a preset threshold, identify abnormal points in the data, and generate a preliminary deviation identification result; Perform cluster analysis on the preliminary deviation identification result, group the deviation data points according to vibration characteristics, identify common abnormal patterns, including periodic faults and random faults, and generate a cluster analysis result; Using the cluster analysis result, record all abnormal vibration patterns and their potential impacts, formulate maintenance suggestions, and generate the engineering structure health analysis result.

Images