Metal damage monitoring method and equipment and storage medium
By combining fractal geometric theory and acoustic emission technology, a dynamic evaluation model of metal damage was constructed, which solved the problem of limited recognition ability of traditional detection methods for micro damage, real-time quantitative evaluation and early warning of metal damage, and improved detection efficiency and accuracy.
Patent Information
- Application Number
- CN202411092011.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-09
- Publication Date
- 2024-11-15
AI Technical Summary
Traditional metal damage detection methods have limited ability to recognize small damage and cannot achieve real-time monitoring and early warning, and are inefficient in detection and costly.
Combining fractal geometry theory and acoustic emission technology, through short-time Fourier transformation, fractal set calculation and fitting of acoustic emission signals, a mapping model and dynamic evolution model of fractal dimensions and metal damage are constructed to realize real-time damage detection and early warning.
Quantitative evaluation of the degree of metal damage is achieved, dynamic evaluation model is provided, monitoring sensitivity and accuracy is improved, false alarms and underreport risks are reduced, and is suitable for special metal materials in extreme environments.
Smart Images

Figure CN118961881A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of metal monitoring, and in particular to a metal damage monitoring method, equipment and storage medium. Background Art
[0002] With the acceleration of industrialization, metal materials and structures are increasingly used in various fields, such as construction, bridges, aerospace, nuclear power plants, etc. The health status of these metal structures is directly related to the safe operation of equipment and the safety of personnel. Therefore, effective damage monitoring and early warning of these metal structures have become an important demand for industrial safety.
[0003] Traditional metal damage monitoring methods often rely on regular visual inspections, ultrasonic testing or X-ray testing. Although these methods can detect damage to a certain extent, they have many limitations, such as low detection efficiency, high cost, and limited ability to identify minor damage and early damage. In addition, these methods are often offline detection and cannot achieve real-time monitoring and early warning. Summary of the invention
[0004] The purpose of the present invention is to propose a metal damage monitoring method, device and storage medium to solve the technical problems that the existing traditional metal damage detection methods are limited in identifying minor damages and cannot provide early warning.
[0005] The present invention provides a metal damage monitoring method, comprising the following steps:
[0006] S1: Use acoustic emission sensors to obtain acoustic emission signals of metals running in industrial systems;
[0007] S2: After preprocessing the acoustic emission signal, a discrete signal is obtained;
[0008] S3: Perform short-time Fourier transform on the discrete signal to obtain a time-frequency spectrum;
[0009] S4: Fractal set for calculating time-frequency spectra;
[0010] S5: Fit the fractal set to obtain the fractal dimension;
[0011] S6: Construct a mapping model between fractal dimension and metal damage;
[0012] S7: Construct a dynamic model of metal damage evolution;
[0013] S8: Input the fractal dimension of the metal to be monitored into the mapping model, and complete the metal damage detection based on the output result of the mapping model and the evolution dynamic model.
[0014] A storage medium stores instructions and data for implementing a metal damage monitoring method.
[0015] A metal damage monitoring device comprises: a processor and the storage medium; the processor loads and executes instructions and data in the storage medium to implement a metal damage monitoring method.
[0016] The beneficial effects provided by the present invention are:
[0017] 1. Innovation of theoretical framework: For the first time, fractal geometry theory is combined with acoustic emission technology and applied to metal damage monitoring. The quantitative evaluation of damage degree is achieved through parameters such as fractal dimension, providing a new perspective and evaluation tool for the health monitoring of metal materials.
[0018] 2. Construction of dynamic assessment model: A dynamic damage assessment model based on the fractal characteristics of acoustic emission signals was established. This model can track and reflect the evolution of material damage in real time. Compared with the traditional static "snapshot" assessment, this method is more flexible and forward-looking, which helps to deeply understand the damage mechanism and predict the damage development trend.
[0019] 3. Accuracy of feature extraction: Fractal features closely related to damage are successfully extracted from the acoustic emission signals of metal damage. These features provide an important basis for accurately evaluating the damage state of metal materials.
[0020] 4. Introduction of intelligent early warning system: Through adaptive threshold setting and intelligent early warning mechanism, this method can dynamically adjust the early warning threshold according to real-time monitoring data, so as to timely detect and warn of early signs of damage, significantly improving the sensitivity and accuracy of monitoring and reducing the risk of false alarms and missed alarms.
[0021] 5. Wide range of application scenarios: This method is not only suitable for damage monitoring of conventional metal materials, but can also be extended to special metal materials used in nuclear power plants, aerospace, marine engineering and other fields, showing high reliability and wide adaptability in extreme and complex environments. BRIEF DESCRIPTION OF THE DRAWINGS
[0022] Figure 1 It is a schematic flow chart of the method of the present invention;
[0023] Figure 2 It is a working schematic diagram of the hardware device of the present invention. DETAILED DESCRIPTION
[0024] To make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be further described below with reference to the accompanying drawings.
[0025] Before formally describing the present invention, the scheme of the present invention is first generally described for easy understanding.
[0026] Please refer to Figure 1 , Figure 1 It is a schematic flow diagram of the method of the present invention;
[0027] The present invention provides a metal damage monitoring method, comprising the following steps:
[0028] S1: Use acoustic emission sensors to obtain acoustic emission signals of metals running in industrial systems;
[0029] S2: After preprocessing the acoustic emission signal, a discrete signal is obtained;
[0030] It should be noted that the acoustic emission signal can be expressed as a time function x(t), which is preprocessed in the present invention, such as noise removal, smoothing, etc., and then discretized to obtain its corresponding discrete signal x[n].
[0031] S3: Perform short-time Fourier transform on the discrete signal to obtain a time-frequency spectrum;
[0032] It should be noted that the time-frequency spectrum of a time-series acoustic emission signal is a graphical representation that shows the characteristics of a signal in two dimensions: time and frequency. A time-series acoustic emission signal is a non-stationary signal whose characteristics and intensity may change over time. The time-frequency spectrum can simultaneously show the energy distribution of a signal at different times and frequencies.
[0033] The time-frequency spectrum is usually generated by short-time Fourier transform (STFT), wavelet transform, etc. In the time-frequency spectrum, the intensity or energy of the signal is usually represented by color or brightness.
[0034] For example, darker colors or higher brightness may indicate higher signal strength at a specific time and frequency. By observing the distribution and changes of colors in the spectrum, the characteristics of the signal at different times and frequencies can be analyzed. The time-frequency spectrum is very useful for analyzing and processing non-stationary signals, especially complex signals such as acoustic emission signals. It can reveal the periodic components in the signal, the instantaneous frequency changes, and the time-frequency distribution of the signal energy.
[0035] Short-Time Fourier Transform (STFT) is a commonly used signal processing technique used to analyze the time-frequency characteristics of non-stationary signals. The basic idea of STFT is to divide the signal into many small time periods and perform Fourier transform on each time period to obtain the representation of the signal at different times and frequencies.
[0036] As an example, in practical applications, STFT is usually implemented in a discrete form, that is, the discrete signal x[n] is transformed:
[0037]
[0038] Where: ω is the discrete window function; R is the number of samples of the frame shift; N is the number of points in the FFT, which determines the frequency resolution; k is the frequency index; x(n) is the discrete signal; m is an integer representing the index of the time period; STFT stands for short-time Fourier transform.
[0039] By calculating the above formula, we can get a complex matrix whose rows represent different time periods (or frames) and columns represent different frequency components. The modulus of this matrix represents the amplitude of the signal at different times and frequencies, while the phase angle represents the phase information. Usually, we only care about the amplitude to form a time-frequency spectrum.
[0040] S4: Fractal set for calculating time-frequency spectra;
[0041] As an embodiment, step S4 is specifically as follows:
[0042] S41, using the time-frequency spectrum as the input of the fractal box, and selecting an initial box size ε;
[0043] It should be noted that the size ε can also be understood as the grid size or resolution.
[0044] S42, for each size ε, calculating the minimum number of boxes N(ε) that can cover all phase space points;
[0045] S43, if the box contains at least one point, it is considered occupied;
[0046] S44, using boxes of size ε to cover the entire signal; counting the minimum number of boxes N(ε) required to completely cover the signal;
[0047] S45. Gradually reduce the box size ε and repeat S44 to obtain a series of box sizes and corresponding box numbers to form a fractal set.
[0048] S5: Fit the fractal set to obtain the fractal dimension;
[0049] As an example, for a fractal set, plotting the relationship between N(ε) and ε in a double logarithmic coordinate system usually results in a straight line. The slope of the straight line is the fractal dimension D. The mathematical relationship can be expressed as:
[0050] log(N(ε))=-Dlog(ε)+C
[0051] Among them, C is a constant term.
[0052] The fractal dimension D can be obtained from the slope of the fitted line. That is, D = -slope. The fractal dimension D can be obtained by linearly fitting the above equation using the least squares method.
[0053] As an example, in Python, libraries such as numpy and scipy can be used to complete the above calculations. For example, numpy is used to calculate the number of boxes required for signal coverage, and then the linear regression function in scipy.stats is used to estimate the fractal dimension.
[0054] S6: Construct a mapping model between fractal dimension and metal damage;
[0055] It should be noted that the mapping model in step S6 includes: a linear model and a nonlinear model.
[0056] As an example, in practical applications, the relationship between the fractal dimension D and metal damage (such as stress corrosion cracking SCC) may present linear or nonlinear characteristics. In order to determine this relationship, parameter fitting can be performed through experimental data to verify whether it is a linear relationship or a nonlinear relationship.
[0057] If it is assumed that there is a linear relationship between the fractal dimension D and the damage degree S, the linear regression method can be used to fit the experimental data. If the experimental data show that the relationship between D and S is not linear, a nonlinear model can be considered for fitting.
[0058] In order to perform parameter fitting, a series of experimental data points (D i ,S i ), where D i is the measured fractal dimension, S i is the corresponding damage degree indicator (such as crack length, area, etc.). Then, the model parameters are estimated using the fitting function in the statistical software or programming language.
[0059] After fitting is completed, the goodness of fit of the model needs to be evaluated. Commonly used evaluation indicators include the coefficient of determination (R-squared), mean squared error (MSE), etc. These indicators can help us determine whether the model can better describe the experimental data.
[0060] The linear correlation model between fractal dimension and damage degree is as follows:
[0061] There is a linear relationship between the fractal dimension D and the severity S of metal damage, which can be mathematically expressed as:
[0062] S=aD+b
[0063] Where: S represents the quantitative index of damage, which can be the length, area, volume or other physical quantities that can reflect the degree of damage. a and b are model parameters determined by fitting experimental data.
[0064] The nonlinear correlation model between fractal dimension and damage degree is as follows:
[0065] In some cases, the relationship between fractal dimension and damage degree may be nonlinear. In this case, a power law relationship can be used to describe this nonlinear association:
[0066] S=cD d
[0067] Where: c and d are parameters obtained by fitting experimental data. d reflects the rate at which the damage degree increases with the fractal dimension.
[0068] S7: Construct a dynamic model of metal damage evolution;
[0069] As an example, considering the evolution of metal damage over time t, the rate of change of the fractal dimension can be described by the following differential equation:
[0070]
[0071] in, represents the rate of change of fractal dimension over time, k is the rate constant of damage expansion; D max is the maximum fractal dimension.
[0072] S8: Input the fractal dimension of the metal to be monitored into the mapping model, and complete the metal damage detection based on the output result of the mapping model and the evolution dynamic model.
[0073] It should be noted that by comparing the output results of the mapping model with the preset threshold and combining them with the output results of the dynamic model of damage evolution, a comprehensive judgment is made to issue a metal damage warning.
[0074] Through the output of nonlinear model or linear model, and based on the threshold of the severity S of metal damage, the fractal dimension change rate can be combined To alarm and corresponding operation and maintenance strategies of equipment components.
[0075] The adaptive threshold update strategy is as follows:
[0076] Set a threshold update cycle. In each update cycle, update the threshold (T) based on the monitoring data. The update strategy can be based on the damage indicator change trend in the recent period, or based on the prediction results of the dynamic model of damage evolution. For example, if the damage indicator is detected to continue to rise, the threshold can be appropriately increased, otherwise the threshold can be lowered.
[0077] The adaptive threshold update algorithm is as follows:
[0078] Initial threshold t=0:S0
[0079] Adaptive Threshold Update: Among them, k is a proportional coefficient, which can be determined by experiment. to adjust the threshold value. Figure 2 , Figure 2 It is a schematic diagram of the working of the hardware device of an embodiment of the present invention, and the hardware device specifically includes: a metal damage monitoring device 401, a processor 402 and a storage medium 403.
[0080] A metal damage monitoring device 401: The metal damage monitoring device 401 implements the metal damage monitoring method.
[0081] Processor 402: The processor 402 loads and executes the instructions and data in the storage medium 403 to implement the metal damage monitoring method.
[0082] Storage medium 403: The storage medium 403 stores instructions and data; the storage medium 403 is used to implement the metal damage monitoring method.
[0083] The beneficial effects of the present invention are:
[0084] 1. Innovation of theoretical framework: For the first time, fractal geometry theory is combined with acoustic emission technology and applied to metal damage monitoring. The quantitative evaluation of damage degree is achieved through parameters such as fractal dimension, providing a new perspective and evaluation tool for the health monitoring of metal materials.
[0085] 2. Construction of dynamic assessment model: A dynamic damage assessment model based on the fractal characteristics of acoustic emission signals was established. This model can track and reflect the evolution of material damage in real time. Compared with the traditional static "snapshot" assessment, this method is more flexible and forward-looking, which helps to deeply understand the damage mechanism and predict the damage development trend.
[0086] 3. Accuracy of feature extraction: Fractal features closely related to damage are successfully extracted from the acoustic emission signals of metal damage. These features provide an important basis for accurately evaluating the damage state of metal materials.
[0087] 4. Introduction of intelligent early warning system: Through adaptive threshold setting and intelligent early warning mechanism, this method can dynamically adjust the early warning threshold according to real-time monitoring data, so as to timely detect and warn of early signs of damage, significantly improving the sensitivity and accuracy of monitoring and reducing the risk of false alarms and missed alarms.
[0088] 5. Wide range of application scenarios: This method is not only suitable for damage monitoring of conventional metal materials, but can also be extended to special metal materials used in nuclear power plants, aerospace, marine engineering and other fields, showing high reliability and wide adaptability in extreme and complex environments.
[0089] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims
1. A metal damage monitoring method, Features: The method comprises the following steps: S1: Use acoustic emission sensors to obtain acoustic emission signals of metals running in industrial systems; S2: After preprocessing the acoustic emission signal, a discrete signal is obtained; S3: Perform short-time Fourier transform on the discrete signal to obtain a time-frequency spectrum; S4: Fractal set for calculating time-frequency spectra; S5: Fit the fractal set to obtain the fractal dimension; S6: Construct a mapping model between fractal dimension and metal damage; S7: Construct a dynamic model of metal damage evolution; S8: Input the fractal dimension of the metal to be monitored into the mapping model, and complete the metal damage detection based on the output result of the mapping model and the evolution dynamic model.
2. A metal damage monitoring method as claimed in claim 1, Features: The preprocessing in step S2 includes: denoising, smoothing and discretization.
3. A metal damage monitoring method as claimed in claim 1, Features: Step S3 is as follows: S31. Perform short-time Fourier transform on the discrete signal, as shown in the following formula: Where: ω is the discrete window function; R is the number of samples of the frame shift; N is the number of points in the FFT, which determines the frequency resolution; k is the frequency index; x(n) is the discrete signal; m is an integer representing the index of the time period; STFT stands for short-time Fourier transform; S32. After short-time Fourier transform, a complex matrix is obtained. The information contained in the time-frequency spectrum is composed of this complex matrix, in which the rows of the matrix represent different time periods and the columns represent different frequency components; the module of the matrix represents the amplitude of the signal at different times and frequencies, and the phase angle represents the phase information.
4. A metal damage monitoring method as claimed in claim 1, Features: Step S4 is specifically as follows: S41, using the time-frequency spectrum as the input of the fractal box, and selecting an initial box size ε; S42, for each size ε, calculating the minimum number of boxes N(ε) that can cover all phase space points; S43, if the box contains at least one point, it is considered occupied; S44, using boxes of size ε to cover the entire signal; counting the minimum number of boxes N(ε) required to completely cover the signal; S45. Gradually reduce the box size ε and repeat S44 to obtain a series of box sizes and corresponding box numbers to form a fractal set.
5. A metal damage monitoring method as claimed in claim 4, Features: Step S5 is as follows: for the fractal set, a relationship graph between N(ε) and ε is plotted in a double logarithmic coordinate system to obtain a straight line; the slope of the straight line is the fractal dimension D; the mathematical relationship can be expressed as: log(N(ε))=-Dlog(ε)+C, Where C is a constant term.
6. A metal damage monitoring method as claimed in claim 1, Features: The mapping model in step S6 includes: a linear model and a nonlinear model.
7. A metal damage monitoring method as claimed in claim 1, Features: The dynamic model of damage evolution is as follows: in, represents the rate of change of fractal dimension over time, k is the rate constant of damage expansion; D max is the maximum fractal dimension.
8. A metal damage monitoring method as claimed in claim 1, Features: Step S8 is specifically as follows: By comparing the output results of the mapping model with the preset threshold and combining them with the output results of the dynamic model of damage evolution, a metal damage warning is issued after a comprehensive judgment.
9. A storage medium, Features: The storage medium stores instructions and data for implementing a metal damage monitoring method as described in any one of claims 1 to 8.
10. A metal damage monitoring device, Features: include: Processor and storage medium; the processor loads and executes instructions and data in the storage medium to implement a metal damage monitoring method as described in any one of claims 1 to 8.
Citation Information
Cited By
Natural resource intelligent monitoring integrated system and method
CN119721451A
