Self-adaptive strain detection method, system and medium

Abstract

The invention relates to the field of strain detection, in particular to a self-adaptive strain detection method and system and a medium. The method comprises the following steps: inputting grid point coordinates, noisy displacement data and a maximum smooth half window; calculating the overall noise level sigma through local linear regression analysis; extracting local strain curvature characteristic quantity through cubic polynomial regression; establishing a noise and curvature balance relation based on the sigma and the curvature characteristic quantity, and calculating an adaptive smooth half window; obtaining the strain estimation value of each grid point through linear polynomial regression, and finally obtaining the strain distribution of the whole field. According to the method, the optimal smooth window can be automatically adjusted without manual intervention, balance is achieved between noise suppression and detail reservation, the problem that a traditional fixed window method is insufficient in precision in a non-uniform deformation area is solved, high-precision full-field strain detection of a noise-containing displacement field is achieved, and the detection precision is improved. The method is suitable for derivative calculation of digital image related systems and various noisy continuous signals.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of strain detection, in particular to a self-adaptive strain detection method, system and medium. BACKGROUND

[0002] In high-end equipment manufacturing, advanced material research and development, and service monitoring and safety evaluation of major engineering structures, stress and strain are the core physical quantities to characterize the mechanical state of the components, and the measurement accuracy directly determines the accuracy of the safety margin evaluation of the structure design, the fatigue life prediction of the material, and the reliability analysis of the equipment in service. With the development of high-end equipment towards lightweight, complex and extreme working conditions, the requirements for spatial resolution, accuracy and stability of strain measurement are increasingly stringent, and traditional measurement methods have been difficult to meet the actual needs. Traditional strain measurement mainly relies on point sensors such as resistance strain gauges, which, although the principle is mature and the response is sensitive, can only obtain local single-point strain data and cannot reflect the full-field strain distribution characteristics of complex structures under load, especially under multi-axial loading conditions, it is difficult to cover the key areas such as stress concentration, and there are significant limitations in the evaluation of the overall mechanical state.

[0003] In recent years, digital image correlation (DIC) technology has become the mainstream strain measurement method in civil engineering, mechanical manufacturing, aerospace and other fields due to its non-contact and full-field measurement advantages, and its accuracy and reliability in uniform strain measurement have been widely recognized. However, when applied to stress concentration areas, crack tips or material mesoscale areas, the strain measurement accuracy and stability of DIC technology are easily affected by noise interference, resulting in large errors and fluctuation of results. The strain extraction of DIC technology usually relies on local least squares method or Savitzky-Golay filtering method, and the selection of local smoothing window is a key factor affecting the strain calculation accuracy: when the window size is too small, the smoothing result is easily disturbed by random displacement noise and fluctuates sharply; when the window size is too large, the real local curvature change is hidden, resulting in excessive smoothing of the strain result and systematic deviation. Especially in areas with severe strain field spatial variation, the optimal smoothing scale at different positions differs significantly, and the traditional fixed window method needs to rely on manual experience or pre-judgment of strain distribution to set the window size. When the strain distribution of the measured area is unknown, it is difficult to determine the appropriate smoothing range, and even if the local area is set reasonably, the fixed window cannot consider the strain characteristics of different areas, ultimately leading to insufficient accuracy and distorted results of the global strain measurement.

[0004] In the prior art, although the prior art CN107255454B "ultra-high temperature multi-scale multi-functional strain measurement system and measurement method based on ultraviolet imaging DIC" realizes high-precision strain measurement in an ultra-high temperature environment, the technology focuses on image acquisition and multi-scale measurement in an ultra-high temperature working condition, does not involve adaptive adjustment of a smoothing window, still needs to rely on a fixed window or manually set a window size, and cannot solve the balance problem of random errors and system deviations under noisy and non-uniform deformation conditions. Therefore, how to adaptively determine the smoothing window under noisy and non-uniform deformation conditions, so as to achieve the optimal balance between random errors and system deviations, has become a key problem to be solved in the field of DIC and generalized strain measurement. SUMMARY

[0005] The purpose of the present application is to provide an adaptive strain detection method, system and medium to solve the problem that the traditional fixed window method cannot adaptively adjust the smoothing scale and balance random errors and system deviations.

[0006] To achieve the above-mentioned purpose, the following technical solutions are adopted.

[0007] An adaptive strain detection method comprises the following steps: Inputting grid step size d Grid point coordinates x i and corresponding noisy displacement data u i Meanwhile, inputting the maximum smoothing half-window n max for limiting the window search range, i =1,2,..., m , m is the total number of data points; For each internal grid point x i , linear regression analysis is performed in its local neighborhood interval, the local random noise estimate value of the internal grid point is calculated based on the regression result, and the local random noise estimate values of all internal grid points are statistically averaged to obtain the overall noise level σ reflecting the overall displacement data noise level; For each internal grid point x i , cubic polynomial regression analysis is performed in the neighborhood interval determined by the maximum smoothing half-window n max , and the cubic term coefficient is extracted from the regression result as the local strain curvature characteristic quantity a 3,i characterizing the strain field intensity of the point; Based on the overall noise levelσ and the local strain curvature characteristic quantity a 3,i The adaptive smoothing half-window corresponding to each internal grid point is calculated by establishing a balance relationship between the noise variance and the square of the curvature n i * ; For each internal grid point x i , a polynomial regression analysis is performed in the determined neighborhood interval n i * The first-order coefficient is extracted as the local strain estimate of the internal grid point ε i After processing all internal grid points in turn, the adaptive strain distribution of the entire field is obtained.

[0008] Optionally, the steps of calculating the local random noise estimate and the overall noise level σ specifically include: For each internal grid point x i , five displacement data points including the center point are selected in the neighborhood interval x i -2 d , x i +2 d , a local linear approximation function is established by performing a linear least squares regression analysis on the five displacement data points, and the fitting values of each data point in the neighborhood interval are calculated using the local linear approximation function. The difference between the actual displacement data u i+j and the fitting value of each data point is taken as the residual sequence r i+j , wherein j =-2,-1,0,1,2; The residual sequence r i+j is subjected to second-order difference processing to obtain a second-order difference sequence Δ 2 r i+k , wherein k =-1,0,1, based on the variance characteristics of the second-order difference sequence, the local random noise estimate of the internal grid point is calculated, and after the local random noise estimate calculation of all internal grid points is completed, the arithmetic mean of all local random noise estimates is taken to obtain the overall noise level σ .

[0009] Optionally, the local random noise estimate is calculated by the following formula:

[0010] wherein Var(·) represents variance operation, Δ 2 r is the second-order difference sequence, is the local noise estimate, and the overall noise level σ is calculated by the following formula:

[0011] The formula averages the local random noise estimates of all internal grid points to obtain the overall noise level representing the noise level of the entire displacement field σ .

[0012] Optionally, the step of extracting the local strain curvature feature quantity a 3,i specifically includes: For each internal grid point x i , 2 n max +1 displacement data points including the center point are selected within the neighborhood interval determined by the maximum smoothing half-window x i n max d , x i + n max d 2 n max +1 displacement data points including the center point are selected within the neighborhood interval determined by the maximum smoothing half-window n max +1 displacement data points, a cubic polynomial regression analysis is performed on the 2 a 3,i , the coefficient of the cubic term in the cubic polynomial regression function is taken as the local strain curvature feature quantity a 3,i of the internal grid point, and the local strain curvature feature quantity

[0013] Optionally, the adaptive smoothing half-window n i * is calculated by the following piecewise function: the formula averages the local random noise estimates of all internal grid points to obtain the overall noise level representing the noise level of the entire displacement field​σ .

[0014] Optionally, the local strain curvature feature quantity is extracted a 3,i The step specifically comprises: for each internal grid point x i , a neighborhood interval n max determined by the maximum smoothing half-window x i - n max d , x i + n max d 2 n max +1 displacement data points including the center point, cubic polynomial regression analysis is performed on the 2 n max +1 displacement data points, a cubic polynomial regression function including a constant term, a first-order term, a second-order term and a third-order term is established, and the coefficient of the third-order term in the cubic polynomial regression function is taken as the local strain curvature feature quantity a 3,i of the internal grid point a 3,i , which is used to quantify the non-uniform intensity of the strain field at the position.

[0015] Optionally, the adaptive smoothing half-window n i * is calculated by the following piecewise function:

[0016] The piecewise function establishes a balance relationship between the overall noise level σ and the local strain curvature feature quantity a 3,i , the minimum half-window 1 is taken when the calculated value is less than 1, the calculated value itself is taken when the calculated value is within the effective range, and the maximum smoothing half-window n max is taken when the calculated value exceeds the maximum smoothing half-window n max .

[0017] Optionally, the correction coefficient c in the piecewise function is used to adjust the sensitivity of the noise and curvature balance relationship, and the correction coefficient cthe value range of the correction coefficient c is adjusted to make the adaptive smoothing half-window n i * The optimal balance between random noise suppression and system bias control can be achieved, ensuring that the window size is appropriately increased to smooth noise in areas with high noise levels, and the window size is appropriately reduced to retain local strain characteristics in areas with sharp curvature changes.

[0018] Optionally, the maximum smoothing half-window n max According to the grid step d and the characteristic length when there is a significant change in strain gradient L The two satisfy the approximate relationship ,

[0019] This relationship determines the upper limit of the window search range based on the spatial scale characteristics of the strain gradient change, and is used to constrain the adaptive smoothing half-window n i * The value range of the correction coefficient

[0020] Optionally, the polynomial regression analysis step specifically includes: For each internal grid point x i , the adaptive smoothing half-window n i * The neighborhood interval x i n i * d , x i + n i * d Within the neighborhood interval n i * +1 displacement data points containing the center point, polynomial regression analysis is performed on the 2 n i * +1 displacement data points, a polynomial regression function containing a constant term and a linear term is established, and the linear term coefficient of the polynomial regression function is taken as the local strain estimate of the internal grid point​ε i , the local strain estimation value ε i is the final strain detection result of the internal grid point.

[0021] An adaptive strain detection system, comprising, a data input module configured to receive grid point coordinates d with a grid step size of x i and corresponding noisy displacement data u i and receive a set value of a maximum smoothing half-window n max , wherein i = 1, 2, …, m , m is the total number of data points; a noise level calculation module connected with the data input module and configured to perform linear regression analysis on a neighborhood interval x i of each internal grid point, calculate a local random noise estimation value based on a regression residual, and statistically obtain an overall noise level σ ; a curvature feature extraction module connected with the data input module and configured to perform cubic polynomial regression analysis on a neighborhood interval x i of each internal grid point in a range of a maximum smoothing half-window n max , extract a cubic term coefficient as a local strain curvature feature quantity a 3,i ; a window calculation module connected with the noise level calculation module and the curvature feature extraction module and configured to calculate an adaptive smoothing half-window σ corresponding to each internal grid point based on the overall noise level a 3,i and the local strain curvature feature quantity n i * ; a strain estimation module connected with the window calculation module and configured to perform linear polynomial regression analysis on a neighborhood interval x i of each internal grid point in a range of the adaptive smoothing half-window n i * , extract a linear term coefficient as a local strain estimation value ε iThe modules are sequentially executed to complete the whole-field adaptive strain distribution calculation.

[0022] A computer-readable storage medium stores at least one instruction, which is executed by a processor to implement the adaptive strain detection method.

[0023] Compared with the prior art, the adaptive strain detection method has the following beneficial effects: The adaptive strain detection method of the present application, by inputting the grid point coordinates, noisy displacement data and maximum smoothing half-window, calculating the overall noise level through local linear regression analysis, extracting the local strain curvature characteristic quantity through cubic polynomial regression, determining the adaptive smoothing half-window by establishing the noise and curvature balance relationship, and finally obtaining the local strain estimate value through a polynomial regression, effectively solves the problem that the traditional fixed window method cannot adaptively adjust the smoothing scale and balance the random error and system deviation, and realizes high-precision full-field strain detection under the condition of noisy and non-uniform deformation. This method does not need to rely on artificial experience or multiple trial calculations, and can automatically adjust the optimal smoothing window size according to the noise level of the displacement field and the local strain curvature characteristics, increase the window size in the area with high noise level to smooth the noise, and reduce the window size in the area with sharp curvature change to retain the local strain characteristics. At the same time, the maximum allowed window is automatically used in the uniform strain area, so that the method naturally degenerates into the traditional fixed window form, and the measurement stability and reliability under different deformation characteristics are considered.

[0024] The local random noise estimate value and the overall noise level are calculated through 5-point linear least squares regression, residual second-order difference and variance operation, which further improves the accuracy of noise level evaluation and provides reliable data support for accurate calculation of adaptive window; the cubic term coefficient is extracted as the local strain curvature characteristic quantity through cubic polynomial regression, which realizes the accurate quantification of the non-uniformity and severity of the strain field, and provides a scientific basis for the adaptive adjustment of the window; the balance relationship between noise and curvature is optimized through the piecewise function, the correction coefficient adjustment and the maximum smoothing half-window estimation method, which ensures the flexible adjustment of the adaptive window within the effective range and further improves the accuracy of strain measurement; the strain estimate value is extracted through a polynomial regression within the adaptive window, which simplifies the calculation process while ensuring the reliability of the strain results; the adaptive strain detection system realizes the automation of data input, noise calculation, curvature extraction, window adjustment and strain estimation through modular design, which improves the engineering practicability of the technology; the computer-readable storage medium provides a convenient carrier for the landing application of the method, which can be flexibly embedded in various strain measurement devices or systems, and expands the application scenarios of the technology. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1This is a schematic flowchart of an embodiment of the adaptive strain detection method of the present invention; Figure 2 This is a schematic diagram of a module of an embodiment of the adaptive strain detection system of the present invention; Figure 3 yes T =300, σ Comparison of the true sinusoidal displacement field and the noisy displacement data under the condition of 0.02; Figure 4 yes T =300, σ Comparison of strain detection results of various methods under the condition of 0.02; Figure 5 yes T =300, σ Comparison of strain detection errors of various methods under the condition of 0.02; Figure 6 Different periods T Comparison of root mean square error (RMSE) of strain under the given conditions; Figure 7 Different noise standard deviations σ Comparison of root mean square error (RMSE) of strain under the given conditions. Detailed Implementation

[0026] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0027] The following detailed description is exemplary and intended to provide further detailed explanation of the invention. Unless otherwise specified, all technical terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. The terminology used in this invention is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention.

[0028] Example 1 like Figure 1 As shown, this embodiment details the specific implementation process of an adaptive strain detection method. This method is suitable for extracting high-precision strain fields from noisy displacement data, and is particularly applicable to displacement field data obtained in digital image correlation techniques. It can also be extended to the calculation of derivatives of other types of noisy continuous signals.

[0029] This embodiment details a window-adaptive strain detection method based on the principle of noise and curvature balance. This method analyzes the dynamic relationship between noise characteristics and strain field curvature characteristics in displacement data to automatically optimize and adjust the size of the smooth window, thereby improving strain calculation accuracy while maintaining spatial resolution.

[0030] The method starts by first receiving necessary input parameters. These parameters include the position information of grid points distributed at fixed intervals and their corresponding displacement data containing measurement noise, as well as a parameter for defining the maximum window range. The maximum window parameter is determined according to the density of the measurement grid and the characteristic scale of the strain field variation, ensuring that the subsequent calculation is performed within a reasonable window range.

[0031] Next, the noise level evaluation stage is entered. In this stage, local noise characteristics are analyzed for each valid internal data point. Specifically, in a small neighborhood around each data point, several adjacent data points including the point are selected, and linear fitting is performed on the displacement values of these points. By comparing the differences between the actual displacement values and the predicted values of the fitted straight line, a set of residual data is obtained. Then, specific mathematical processing is performed on this set of residual data to separate the random fluctuation component, thereby estimating the noise intensity of the local region. After completing the local noise estimation of all data points, the local estimation values are comprehensively averaged to obtain the global noise evaluation value representing the noise level of the entire displacement field.

[0032] After completing the noise evaluation, the strain field curvature characteristic analysis stage is entered. This stage aims to quantify the degree of strain variation at each location. For each valid internal data point, in a larger neighborhood around it determined by the maximum window parameter, multiple adjacent data points including the point are selected, and cubic curve fitting is performed on the displacement distribution of these points. From the fitting result, a parameter representing the degree of curve bending is extracted, which can effectively reflect the local variation characteristics of the strain field at this location. The larger the curvature characteristic parameter value, the more dramatic the strain variation at the location; the smaller the value, the more gradual the strain variation.

[0033] Based on the global noise evaluation value and the local curvature characteristic parameter obtained from the previous steps, the core adaptive window calculation stage is entered. In this stage, a specific calculation rule is used to determine the most suitable smoothing window size for each data point. The design of the calculation rule ensures that in areas with high noise, a larger window is automatically selected to effectively suppress random errors, while in areas with dramatic strain variation, a smaller window is automatically selected to preserve local detail characteristics. In addition, the calculation rule ensures that the window size is within a reasonable preset range, avoiding extreme cases of too small or too large. This balancing mechanism also includes an adjustable parameter for fine-tuning the weight relationship between noise and curvature, allowing the method to adapt to different application scenarios.

[0034] Finally, the strain value calculation stage is entered. According to the optimization window size determined for each data point, an appropriate number of adjacent data points are selected within its corresponding neighborhood range, and the displacement distribution of these points is linearly fitted. The slope parameter is extracted from the fitting result as the strain estimate of the point. After sequentially performing this operation on all valid data points, the complete strain field distribution result is obtained.

[0035] The method exhibits significant advantages in practical application. It can automatically adjust the calculation parameters according to the actual characteristics of the strain field, using a larger smoothing window in areas with a gentle strain change to improve the signal-to-noise ratio, and using a smaller window in areas with a sharp strain change to maintain spatial resolution. This adaptability allows the method to maintain high calculation accuracy in scenarios with different noise levels and different strain distribution characteristics, making it particularly suitable for precise measurement of complex strain fields such as stress concentration and crack tip.

[0036] The entire processing process forms a complete technical closed loop, from data input to result output, with each link designed to achieve optimal strain calculation accuracy. This method is not only suitable for one-dimensional strain analysis, but can also be extended to two-dimensional and three-dimensional cases, providing reliable technical support for material performance testing, structural health monitoring, and other engineering applications.

[0037] Example 2 The method of the present application includes the following main steps:

[0038] 1) Input the grid point coordinates d with a grid step size of x i and the corresponding noisy displacement data u i , where i =1,2,..., m , m is the total number of data points. At the same time, input the maximum smoothing half-window size for strain detection n max This parameter can be estimated according to the grid step size d and the characteristic length L when there is a significant strain gradient change, and is generally taken as an integer n max ≈ L / (2 d ). It should be noted that n max here is only an empirical upper limit used to limit the window search range, and its value is an estimate, not too precise.

[0039] 2) For each internal data point x i(Excluding 4 invalid data points at both ends), calculate its local random noise estimate And the overall noise level is obtained by: , Where, is the local noise estimate, which is calculated as follows: For each valid data point x i , in its neighborhood interval[ x i - 2 d , x i + 2 d ] Select a total of 5 displacement data, linear least squares regression on the local data, establish local approximation function: , Where a 0, a 1 is the regression coefficient. Calculate the residual sequence of each data point in the neighborhood interval: .

[0040] Further, the residual sequence is differentiated twice to get: , According to the second-order difference sequence obtained from the above formula, the noise standard deviation estimate value of the local area is calculated by: , 3) For each internal grid point x i (Excluding 2 n max invalid data at both ends), in its neighborhood interval[ x i - n max d , x i + n max d Select a total of 2 n max +1 displacement data, cubic polynomial regression on the local data, the regression function is: , Where a 0, a 1, a 2,a 3 is the regression coefficient. Take the cubic term coefficient a 3 as the grid point x i Local strain curvature feature quantity a 3,i That is: .

[0041] 4) For each internal grid point x i , according to the above two steps to get the noise level σ And the strain curvature feature quantity a 3,i , the size of its smoothing half window is determined by the following formula n i * : , Where the constant c is a correction coefficient, generally 1.05~1.2, preferably 1.18.

[0042] 5) For each effective grid point x i , according to its corresponding n i * Adaptive selection of smoothing window interval[ x i - n i * d , x i + n i * d ], 2 n i * +1 data in the smoothing window is executed once polynomial regression, and the regression function is established: , Where, b 0, b 1 is the regression coefficient. Take the linear term coefficient b 1 as the local strain estimate value of this position: .

[0043] By sequentially executing the above operation on all data points excluding boundary invalid data points, the adaptive strain distribution of the whole field εi .

[0044] As a specific example of the above examples, specifically includes: A one-dimensional sinusoidal displacement field is selected for testing, and its displacement function is defined as follows: , Wherein, η (x) ~ N (0, σ ²) is a random error obeying zero mean and variance σ ². By adjusting the period parameter T , the strain curvature size (i.e. the degree of strain non-uniformity) can be changed, and by adjusting the noise standard deviation σ , different levels of random noise can be simulated.

[0045] In this embodiment, the grid step is taken as d =2, the number of grid points m =400, and the grid point coordinates xᵢ = 2, 4, 6,…, 800. The grid point displacement is: , The noise standard deviation is taken as σ = 0.005, 0.02, 0.04, respectively, the period is taken as T = 150, 300, 450, respectively, the characteristic length L is taken as T / 2, and the maximum smoothing half-window is taken as n max = round( T / 8). Further, the adaptive window method based on noise-curvature balance proposed by the present application is used to perform adaptive smoothing and strain extraction on the above noisy displacement data. The specific steps are as follows: 1) For each internal data point xᵢ ∈[6, 796], calculate its local random noise estimate , and obtain the overall noise level by the following formula: , Wherein, the calculation method of the local noise estimate is as follows: For each valid data point x i , in its neighborhood interval[ x i – 2 d , x i + 2 dLinear regression is performed, and the regression function is: , where a 0, a 1 are regression coefficients. The residual sequence of each sampling point in the neighborhood interval is calculated: .

[0046] Further, the residual sequence r i+j is subjected to second-order difference: , According to the second-order difference sequence obtained from the above formula, the noise standard deviation estimate value of the local area is calculated using the following formula: , where Var(·) represents variance operation.

[0047] 2) For each internal grid point x i (excluding 2 n max invalid data at both ends), a cubic polynomial regression is performed in its neighborhood interval[ x i - n max d , x i + n max d The regression function is: , where a 0, a 1, a 2, a 3 are regression coefficients. Take the cubic term coefficient a 3 of them as the local strain curvature characteristic quantity x i of the grid point a 3,i , that is: .

[0048] 3) For each internal grid point x i , the smooth half-window size n i * , where the correction factor in the piecewise function c is used to adjust the sensitivity of the noise-curvature balance c The value of the correction factor c is in the range of 1.05-1.2 c The value of the correction factor n is 1.18, which is used to adjust the adaptive smoothing half-window i * The optimal balance between random noise suppression and system bias control is achieved, which ensures the window size is properly increased to smooth the noise in the area with high noise level, and properly decreased to preserve the local strain features in the area with sharp curvature change.

[0049] 4) For each valid grid point x i , the curvature n i * is calculated according to the result of step 3), and the smoothing window interval x i is adaptively selected n i * d , x i + n i * d , and a polynomial regression is performed on the 2 n i * +1 data points within the smoothing window to establish a regression function , where b 0, b 1 are the regression coefficients. The linear coefficient b 1 is taken as the local strain estimate at this position: .

[0050] By sequentially performing the above operations on all data points excluding the invalid data points at the boundaries, the adaptive strain distribution of the whole field is obtained ε i .

[0051] Referring to Figures 3 to 7Compared to traditional fixed-window methods, this invention can automatically adjust the smoothing scale according to the non-uniformity of the strain field, achieving higher accuracy strain results under different noise levels and strain curvature conditions. This method exhibits good noise robustness and curvature adaptability, achieving an optimal balance between random errors and systematic biases, thereby significantly improving the global accuracy and stability of strain calculations.

[0052] Example 3 like Figure 2 As shown, this embodiment details the specific implementation architecture of an adaptive strain detection system. The system adopts a modular design, with each functional module working collaboratively to achieve a complete detection process. It can be implemented based on hardware circuits or software programs.

[0053] The data input module is responsible for receiving external data and configuring system parameters, and includes a displacement data receiving unit and a parameter configuration unit. The displacement data receiving unit connects to a digital image correlation system or external measurement equipment via a data interface, including but not limited to USB, Ethernet, CameraLink, and wireless transmission modules. The displacement data receiving unit has a built-in data buffer and protocol parser, capable of handling displacement data in different formats, including text, binary, and standard image correlation software output formats. The data buffer uses a circular queue structure, and its capacity is dynamically adjusted according to the data volume to ensure the continuity and integrity of data transmission. The parameter configuration unit provides a human-machine interface, which can be a graphical user interface or a command-line interface, allowing users to input the grid step size. d Maximum smooth half window n max Correction coefficient c Key parameters, etc. The parameter configuration unit has a built-in parameter validation function to check the reasonableness of input parameters and ensure... d Greater than zero n max positive integers c Within the range of 1.05 to 1.2. The parameter configuration unit supports the import and export of parameter files, facilitating batch processing and repeated experiments.

[0054] The noise level calculation module is connected to the data input module and includes a linear regression processor and a noise estimation processor. The linear regression processor has a built-in linear least squares algorithm unit, using matrix operations to achieve fast regression calculations. For each internal grid point, the linear regression processor obtains the neighborhood interval from the data input module. x i – 2 d to x i +2 d5 displacement data points within the range of [x a 0 and a 1. The linear regression processor supports single-precision floating-point and double-precision floating-point operations, which can be selected according to the accuracy requirement. The noise estimation processor is built-in with a second-order difference operation unit and a variance statistical unit, which receives the regression coefficients and the original displacement data output by the linear regression processor. First, the residual sequence r i+j is calculated, then the second-order difference operation is performed to obtain 2 r i+k , and finally the variance is calculated and the square root is taken to obtain the local random noise estimation value. The noise estimation processor uses a sliding window technique to avoid repeated calculation and improve processing efficiency. The noise estimation processor is built-in with an accumulator that accumulates the local random noise estimation values of all internal grid points, and finally obtains the overall noise level σ through division operation.

[0055] The curvature feature extraction module is connected with the data input module and contains a cubic polynomial regression processor. The cubic polynomial regression processor is built-in with a cubic polynomial least squares algorithm unit and uses the generalized inverse matrix method to solve the regression coefficients. For each internal grid point x i , the cubic polynomial regression processor obtains 2 x i - n max d , x i + n max d displacement data points within the range of [x n max +1, constructs a design matrix containing constant terms, first-order terms, second-order terms, and third-order terms, and obtains the regression coefficients by least squares method a 0, a 1, a 2, a 3. The cubic polynomial regression processor extracts the third-order term coefficient a 3 as the local strain curvature feature quantity a 3,i output. The cubic polynomial regression processor supports polynomial regression of different orders, and the user can select the regression order through the parameter configuration unit, with the default being cubic. The cubic polynomial regression processor is built-in with a condition number checking function, which automatically uses a regularization method to improve numerical stability when the design matrix is ill-conditioned.

[0056] The window calculation module is connected with the noise level calculation module and the curvature feature extraction module, and contains a balance relationship operator and a comparison and judgment device. The balance relationship operator is internally provided with an exponential operation unit and a multiplication operation unit, receives the overall noise level σ and the local strain curvature feature quantity a 3,i As input, the following formula is executed .

[0057] The exponential operation unit is implemented by using a fast power algorithm or a lookup table method, so as to ensure the calculation accuracy and speed. The comparison and judgment device is internally provided with a threshold storage register, stores the preset threshold 1 and n max receives the intermediate calculation value output by the balance relationship operator, and executes numerical comparison and judgment. The comparison and judgment device adopts a pipeline structure, and sequentially judges whether the intermediate calculation value is less than 1, whether it is between 1 and n max , and whether it is greater than n max According to the judgment result, the corresponding output value is selected as the adaptive smoothing half-window n i * The comparison and judgment device supports vectorized comparison, and can simultaneously process the judgment tasks of multiple grid points.

[0058] The strain estimation module is connected with the window calculation module, and contains a first-order polynomial regression processor. The first-order polynomial regression processor is internally provided with a first-order polynomial least square algorithm unit, and adopts a matrix operation method similar to the linear regression processor. For each internal grid point x i , the first-order polynomial regression processor obtains the adaptive smoothing half-window n i * from the data input module, determines 2 x i – n i * d , x i + n i * d displacement data points within the neighborhood interval n i * +1, constructs a design matrix and an observation vector, and solves the regression coefficients b 0 and b 1 by using the least square method. The first-order polynomial regression processor extracts the first-order coefficient b 1 as the local strain estimation value εi Output. The polynomial regression processor has a built-in error estimation unit that calculates the standard deviation of the strain estimate, providing the user with a confidence assessment of the result.

[0059] The internal data bus of the system uses a high-speed serial bus or parallel bus architecture to connect the various functional modules, ensuring that the data transmission rate meets the real-time processing requirements. The data bus width is selected according to the data precision, supporting 32-bit or 64-bit data transmission. The data bus has a built-in arbitration mechanism that prioritizes the data requirements of the noise level calculation module and the curvature feature extraction module when multiple modules request data simultaneously. For software implementation, the data bus can be replaced by a memory sharing mechanism, and each module exchanges data by accessing shared memory, with mutual exclusion and semaphores to achieve synchronization.

[0060] The system clock management unit provides a unified clock signal for each module, supporting both synchronous and asynchronous operation modes. In synchronous mode, all modules execute according to a fixed timing cycle, suitable for high real-time requirements. In asynchronous mode, each module triggers execution when data is ready, suitable for non-real-time or event-driven scenarios. The clock frequency is selected according to the processing speed requirement, ranging from tens of megahertz to several gigahertz.

[0061] The system power management unit uses low-power design, supporting dynamic voltage and frequency adjustment. When the system load is low, the operating voltage and clock frequency are automatically reduced to save power. Each functional module supports independent power control, and unused modules can be turned off individually. The power management unit has a built-in temperature sensor that automatically reduces the frequency when the chip temperature exceeds the threshold.

[0062] The system can be implemented using an application-specific integrated circuit, a field-programmable gate array, or a digital signal processor. The application-specific integrated circuit solution has high integration, low power consumption, and fast processing speed, making it suitable for mass production. The field-programmable gate array solution has high flexibility, short development cycle, and reconfigurability, making it suitable for algorithm iteration and customization needs. The digital signal processor solution is easy to program, mature in ecosystem, and has strong floating-point operation capabilities, making it suitable for complex algorithm implementation and research applications.

[0063] For software implementation, the system can be deployed on a personal computer, server, or embedded platform. The programming languages can be C / C++, Python, or MATLAB. C / C++ implementation has high execution efficiency, making it suitable for real-time processing. Python implementation has high development efficiency, making it suitable for rapid prototyping. MATLAB implementation has convenient mathematical operations, making it suitable for algorithm research and simulation. The software architecture uses modular design, with each functional module encapsulated as an independent function or class, and data exchange is achieved through parameter passing and return values.

[0064] The system supports distributed deployment. For large-scale data, the data domain can be divided into multiple sub-regions, allocated to different computing nodes for parallel processing. Each node independently performs noise level calculation, curvature feature extraction, window calculation, and strain estimation, and the main node finally aggregates the results. Nodes are interconnected through high-speed networks and communicate using MPI or TCP / IP protocols.

[0065] The system has built-in self-checking and diagnostic functions. When starting up, it performs functional tests on each module, including memory tests, operation unit tests, and interface tests. During operation, it monitors the status of each module in real time and records error logs and triggers alarms when errors are detected. Error logs include error type, occurrence time, related parameters, and other information to facilitate troubleshooting.

[0066] In terms of user interface, the system provides a graphical user interface and an application programming interface. The graphical user interface displays input data, intermediate results, and final strain fields, supporting various visualization methods such as two-dimensional curves, three-dimensional surfaces, and color cloud maps. Users can adjust parameters, start calculations, and save results through the interface. The application programming interface provides standard function call interfaces, supports integration with other software systems, and includes initialization, parameter setting, data input, execution calculation, and result acquisition functions.

[0067] In terms of performance optimization, the system uses a multi-level cache mechanism to store frequently accessed data in high-speed cache, reducing memory access latency. Linear regression and polynomial regression operations use SIMD instruction sets for acceleration, processing multiple data with a single instruction. For large-scale matrix operations, GPU parallel computing can be used, utilizing CUDA or OpenCL programming models to divide the computing task into a large number of threads for parallel execution.

[0068] In terms of precision assurance, the system supports double-precision floating-point operations and uses high-precision algorithms for key calculation steps. Linear regression and polynomial regression use singular value decomposition methods to avoid numerical errors introduced by matrix inversion. Noise estimation and variance calculation use compensated summation algorithms to reduce floating-point operation cumulative errors. The system provides result verification functions, such as residual analysis and consistency checks, to evaluate the reliability of strain results.

[0069] In practical engineering applications, the system can seamlessly interface with digital image correlation software. After the digital image correlation software completes displacement field calculation, it outputs displacement data in a standard format, which is read by the data input module of this system to automatically execute subsequent strain detection processes. The detection results can be imported into finite element analysis software for verification or compared with theoretical solutions to evaluate accuracy. The system supports batch processing, automatically processing data for multiple samples or various working conditions, and generating statistical reports.

[0070] For two-dimensional digital image correlation systems, the system is expanded to a dual-channel architecture, processingx direction and y direction displacement. The data input module receives two-dimensional grid coordinates and displacement data, the noise level calculation module and the curvature feature extraction module process two directions in parallel, the window calculation module can share noise and curvature information or calculate separately, and the strain estimation module outputs x direction normal strain, y direction normal strain and shear strain. The system supports color image display, superimposes the strain field on the image of the measured object in the form of a cloud chart, and intuitively displays the strain distribution.

[0071] For a three-dimensional digital image correlation system, the system is expanded to a three-channel architecture, processing x , y , z three direction displacements. The data input module receives three-dimensional grid coordinates and displacement data, the noise level calculation module and the curvature feature extraction module process three directions in parallel, the strain estimation module outputs three normal strains and three shear strains, forming a complete strain tensor. The system supports three-dimensional visualization, displaying the strain field in the form of a three-dimensional cloud chart or a slice.

[0072] In the application of material mechanical property testing, the system is used for strain measurement in tensile, compression, bending and other experiments. The speckle pattern is made on the surface of the sample, the displacement field is obtained by the digital image correlation system, the strain field is calculated by the system, and then the stress-strain curve, elastic modulus, Poisson's ratio and other material parameters are obtained. The system supports real-time processing, and can display the strain distribution in real time during the experiment to monitor the deformation process of the sample.

[0073] In the application of structural health monitoring, the system is used for strain monitoring of structures such as bridges, buildings and mechanical equipment. By periodically shooting images of the structure surface, the strain field change is calculated to identify the damage location and development trend. The system supports long-term data storage and trend analysis, and can set a strain threshold to trigger an alarm when the strain exceeds the threshold.

[0074] In the application of mesoscopic mechanics research, the system is used for strain measurement of material micro areas. The material surface image is obtained by a high-resolution microscope, the mesoscopic strain field is calculated, and the local deformation behaviors such as crack propagation, interface slip and phase change are studied. The system supports fine analysis of small areas, and can extract strain singularity information at the crack tip.

[0075] In the application of impact dynamics, the system is used for strain measurement in high-speed deformation process. The impact process is shot by a high-speed camera to obtain the transient displacement field, and the strain field is quickly calculated by the system to analyze stress wave propagation, localized deformation and other dynamic phenomena. The system optimizes the efficiency of the algorithm, and supports high-frame-rate data processing.

[0076] In fatigue experiment applications, the system is used for strain measurement under cyclic loading. By continuously recording the fatigue process, the strain field of each cycle is calculated, and the fatigue behaviors such as cyclic hardening, softening, and crack initiation are studied. The system supports phase-locked averaging, which improves the signal-to-noise ratio of cyclic strain measurement.

[0077] In creep experiment applications, the system is used for strain measurement under long-term loading. By periodically recording the creep process, the creep strain field is calculated, and the creep deformation mechanism and damage evolution are studied. The system supports time series analysis, which can extract the variation law of the creep strain rate over time.

[0078] In thermal-mechanical fatigue applications, the system considers the influence of temperature changes. By synchronously recording the temperature field and displacement field, the thermal strain and mechanical strain are separated, and the strain caused by true stress is calculated. The system supports multi-physical field coupling analysis, which can study the influence of temperature on strain distribution.

[0079] In composite material mechanical property characterization, the system is used for multi-axial strain measurement of anisotropic materials. By measuring strain components in different directions, the anisotropic parameters of the material are calculated, and the constitutive model is verified. The system supports multi-directional strain synchronous measurement, which can comprehensively characterize the deformation behavior of composite materials.

[0080] In welding joint performance evaluation, the system is used for strain measurement in the weld and heat-affected zone. By high-resolution measurement, the weld residual strain distribution is captured, and the welding quality and joint strength are evaluated. The system supports fine measurement in high gradient areas, which can accurately capture the strain concentration near the weld.

[0081] Example 4 A computer-readable storage medium stores at least one instruction, which is executed by a processor to implement an adaptive strain detection method.

[0082] The modules / units integrated in the electronic device 100, if implemented in the form of software function units and sold or used as independent products, can be stored in a computer readable storage medium. Based on such understanding, all or part of the processes in the above-mentioned embodiments of the method of the present application can also be instructed by a computer program to relevant hardware, and the computer program can be stored in a computer readable storage medium. When the processor executes the computer program, the steps of each method embodiment described above can be implemented. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate forms. The computer readable medium can include any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, and read-only memory (ROM). Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can be in the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can be in the form of a computer program product implemented on one or more computer readable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) containing computer usable program code. The present application is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to produce a machine, so that the instructions executed by the computer or other programmable data processing devices produce a device that implements the functions specified in the flow Figure 1 One or more flows and / or blocks Figure 1 One or more flows and / or blocks Figure 1 One or more flows and / or blocks Figure 1 One or more flows and / or blocks Figure 1 One or more flows and / or blocksFigure 1 the steps of the functions specified in the one or more blocks.

[0083] It is apparent that the application can be carried out in series of ways, and utilizing equivalent and / or similar elements and / or functions without departing from the spirit or essential characteristics of the application. Thus, all modifications the scope of the application are included within the scope of the application and will be protected by the scope of the appended claims as construed according to the doctrine of equivalents.

Claims

1. An adaptive strain detection method, characterized in that, Includes the following steps: Input grid step size is d Grid point coordinates x i and the corresponding noisy displacement data u i Simultaneously input the maximum smooth half-window used to limit the window search range. n max ,in i =1,2,..., m , m This represents the total number of data points. For each internal grid point x i Linear regression analysis is performed within its local neighborhood interval. Based on the regression results, the local random noise estimate of the internal grid point is calculated. The local random noise estimates of all internal grid points are then statistically averaged to obtain the overall noise level, which reflects the overall noise level of the displacement data. σ ; For each internal grid point x i In the maximum smooth half window n max Perform cubic polynomial regression analysis within the defined neighborhood interval, and extract the coefficients of the cubic term from the regression results as a local strain curvature characteristic quantity characterizing the intensity of the strain field at that point. a 3,i ; Based on the overall noise level σ and the local strain curvature characteristic quantity a 3,i By establishing a balance between noise variance and curvature square, the adaptive smoothing half-window corresponding to each internal grid point is calculated. n i * ; For each internal grid point x i In adaptive smooth half-window n i * Perform a first-order polynomial regression analysis within the defined neighborhood interval, and extract the coefficients of the first-order term as the local strain estimate for that internal grid point. ε i After processing all internal grid points in sequence, the adaptive strain distribution of the entire field is obtained.

2. The adaptive strain detection method according to claim 1, characterized in that, Calculate the local random noise estimate and the overall noise level. σ The specific steps include: For each internal grid point x i In the neighborhood interval [ x i – 2 d , x i +2 d Five displacement data points, including the center point, were selected. Linear least squares regression analysis was performed on these five data points to establish a local linear approximation function. This function was then used to calculate the fitted value for each data point within its neighborhood interval. The actual displacement data of each data point was then... u i+j The difference between the fitted value and the actual value is used as the residual sequence. r i+j ,in j =-2,-1,0,1,2; For the residual sequence r i+j Perform second-order difference processing to obtain the second-order difference sequence Δ 2 r i+k ,in k =-1,0,1. Based on the variance characteristics of the second-order difference sequence, the local random noise estimate of the internal grid point is calculated. After calculating the local random noise estimate of all internal grid points, the arithmetic mean of all local random noise estimates is taken to obtain the overall noise level. σ .

3. The adaptive strain detection method according to claim 2, characterized in that, The estimated value of the local random noise is calculated using the following formula: ; Where Var(·) represents variance operation, Δ 2 r For the second-order difference sequence, The overall noise level is a local noise estimate. σ It is calculated using the following formula: ; This formula averages the local random noise estimates of all internal grid points to obtain the overall noise level, which represents the noise level of the entire displacement field. σ .

4. The adaptive strain detection method according to claim 1, characterized in that, Extracting local strain curvature characteristics a 3,i The specific steps include: For each internal grid point x i In the case of the maximum smooth half window n max Determined neighborhood interval [ x i – n max d , x i + n max d [Select 2 points including the center point] n max +1 displacement data points, for the 2 n max A cubic polynomial regression analysis was performed on +1 displacement data points to establish a cubic polynomial regression function containing constant, linear, quadratic, and cubic terms. The coefficient of the cubic term in this function was taken as the local strain curvature characteristic of the internal grid point. a 3,i The local strain curvature characteristic quantity a 3,i Used to quantify the severity of non-uniformity in the strain field at that location.

5. The adaptive strain detection method according to claim 1, characterized in that, The adaptive smooth half-window n i * It is obtained through the following piecewise function: ； The piecewise function establishes the overall noise level. σ With local strain curvature characteristic quantity a 3,i The balance relationship is as follows: when the calculated value is less than 1, the minimum half-window size of 1 is used; when the calculated value is within the valid range, the calculated value itself is used; when the calculated value exceeds the maximum smoothing half-window size... n max Time to take n max .

6. The adaptive strain detection method according to claim 5, characterized in that, The correction coefficient in the piecewise function c The correction coefficient is used to adjust the sensitivity of the noise-curvature balance relationship. c The value range is 1.05 to 1.2, and this correction coefficient is used. c The adjustment enables adaptive smooth half-window n i * It can achieve an optimal balance between random noise suppression and system deviation control, ensuring that the window size is appropriately increased in areas with high noise levels to smooth out noise, and that the window size is appropriately reduced in areas with drastic curvature changes to preserve local strain characteristics.

7. The adaptive strain detection method according to claim 1, characterized in that, The maximum smooth half window n max Based on grid step size d and the characteristic length corresponding to the presence of a significant strain gradient change. L The predictions determine that the two satisfy an approximate relationship. ； This relationship determines the upper limit of the window search range based on the spatial scale characteristics of strain gradient changes, and is used to constrain the adaptive smoothing half-window. n i * The range of values ​​ensures that a larger window size can be used in regions with gentle strain changes to improve the signal-to-noise ratio, while limiting the window size in regions with drastic strain changes to avoid over-smoothing.

8. The adaptive strain detection method according to claim 1, characterized in that, The specific steps of a first-order polynomial regression analysis include: For each internal grid point x i In adaptive smooth half-window n i * Determined neighborhood interval [ x i – n i * d , x i + n i * d [Select 2 points including the center point] n i * +1 displacement data point, for this 2 n i * A first-order polynomial regression analysis was performed on +1 displacement data points to establish a first-order polynomial regression function containing a constant term and a first-order term. The coefficient of the first-order term in this function was taken as the local strain estimate for that internal grid point. ε i The estimated local strain value ε i This is the final strain detection result for that internal grid point.

9. An adaptive strain detection system, based on the adaptive strain detection method according to any one of claims 1-8, characterized in that, include, The data input module is used to receive data with a grid step size of [value missing]. d Grid point coordinates x i and the corresponding noisy displacement data u i and receive the maximum smooth half window n max The set value, of which i =1,2,..., m , m This represents the total number of data points. The noise level calculation module, connected to the data input module, is used to calculate the noise level for each internal grid point. x i Linear regression analysis is performed within the neighborhood interval. Based on the regression residuals, the local random noise estimate is calculated, and the overall noise level is statistically obtained. σ ; The curvature feature extraction module, connected to the data input module, is used to extract features from each internal grid point. x i In the maximum smooth half window n max A cubic polynomial regression analysis was performed within the neighborhood of the range, and the coefficients of the cubic term were extracted as local strain curvature characteristics. a 3,i ; The window calculation module, connected to the noise level calculation module and the curvature feature extraction module, is used to calculate based on the overall noise level. σ and the local strain curvature characteristic quantity a 3,i The adaptive smoothing half-window corresponding to each internal grid point is calculated by balancing the noise and curvature. n i * ; The strain estimation module, connected to the window calculation module, is used to perform calculations on each internal grid point. x i In the adaptive smooth half window n i * Perform a first-order polynomial regression analysis within the neighborhood of the range, and extract the coefficients of the first-order term as the local strain estimate. ε i Each module executes sequentially to complete the adaptive strain distribution calculation for the entire field.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one instruction, which, when executed by a processor, implements an adaptive strain detection method as described in any one of claims 1 to 8.

Images