CN114324329B - Nondestructive testing and evaluating method for strong laser damage characteristic of optical element - Google Patents
Nondestructive testing and evaluating method for strong laser damage characteristic of optical element Download PDFInfo
- Publication number
- CN114324329B CN114324329B CN202111588367.6A CN202111588367A CN114324329B CN 114324329 B CN114324329 B CN 114324329B CN 202111588367 A CN202111588367 A CN 202111588367A CN 114324329 B CN114324329 B CN 114324329B
- Authority
- CN
- China
- Prior art keywords
- optical element
- laser
- data
- damage
- matrix
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000003287 optical effect Effects 0.000 title claims abstract description 187
- 238000009659 non-destructive testing Methods 0.000 title claims abstract description 66
- 238000000034 method Methods 0.000 title claims abstract description 42
- 238000012360 testing method Methods 0.000 claims abstract description 67
- 238000001514 detection method Methods 0.000 claims abstract description 44
- 230000009471 action Effects 0.000 claims abstract description 11
- 239000011159 matrix material Substances 0.000 claims description 93
- 238000005259 measurement Methods 0.000 claims description 18
- 230000008569 process Effects 0.000 claims description 13
- 238000013519 translation Methods 0.000 claims description 13
- 238000004364 calculation method Methods 0.000 claims description 12
- 238000012545 processing Methods 0.000 claims description 11
- 230000005540 biological transmission Effects 0.000 claims description 9
- 230000007547 defect Effects 0.000 claims description 9
- 230000000007 visual effect Effects 0.000 claims description 9
- 238000012800 visualization Methods 0.000 claims description 9
- 238000006243 chemical reaction Methods 0.000 claims description 8
- 238000010521 absorption reaction Methods 0.000 claims description 7
- 238000013528 artificial neural network Methods 0.000 claims description 7
- 238000012549 training Methods 0.000 claims description 7
- 238000002834 transmittance Methods 0.000 claims description 5
- 230000010287 polarization Effects 0.000 claims description 4
- 238000012876 topography Methods 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 3
- 230000001131 transforming effect Effects 0.000 claims 1
- 238000011156 evaluation Methods 0.000 abstract description 11
- 238000013210 evaluation model Methods 0.000 abstract description 3
- 238000013135 deep learning Methods 0.000 abstract 1
- 238000002474 experimental method Methods 0.000 abstract 1
- 230000004907 flux Effects 0.000 description 10
- 238000003384 imaging method Methods 0.000 description 7
- 238000013527 convolutional neural network Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 238000007689 inspection Methods 0.000 description 6
- 238000013079 data visualisation Methods 0.000 description 5
- 230000001066 destructive effect Effects 0.000 description 5
- 238000003745 diagnosis Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 238000012821 model calculation Methods 0.000 description 4
- 238000005286 illumination Methods 0.000 description 3
- 101100045694 Caenorhabditis elegans art-1 gene Proteins 0.000 description 2
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N Silicium dioxide Chemical compound O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 description 2
- 230000004913 activation Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000009658 destructive testing Methods 0.000 description 2
- 239000005350 fused silica glass Substances 0.000 description 2
- 238000005305 interferometry Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000005192 partition Methods 0.000 description 2
- 230000001105 regulatory effect Effects 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000010998 test method Methods 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- 229910052779 Neodymium Inorganic materials 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 239000013078 crystal Substances 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 238000001917 fluorescence detection Methods 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 238000001499 laser induced fluorescence spectroscopy Methods 0.000 description 1
- 238000002536 laser-induced breakdown spectroscopy Methods 0.000 description 1
- QEFYFXOXNSNQGX-UHFFFAOYSA-N neodymium atom Chemical compound [Nd] QEFYFXOXNSNQGX-UHFFFAOYSA-N 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 238000010008 shearing Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
Landscapes
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention discloses a nondestructive testing and evaluating method for strong laser damage characteristics of an optical element, and aims to solve the problem that the laser damage characteristics of the optical element cannot be evaluated through nondestructive testing in the prior art. The invention comprises the following steps: obtaining nondestructive testing data of an optical element and laser parameters of damage under the action of strong laser through experiments; using nondestructive testing data and damage testing results as characteristic quantities, and using a deep learning method to obtain a detection evaluation model of the laser damage characteristics of the optical element according to the nondestructive testing data, the laser parameters and the damage testing results; and carrying out nondestructive testing on the optical element to be tested, taking the obtained nondestructive testing data and laser parameters as input quantity of a testing evaluation model, thereby obtaining a laser damage characteristic testing result of the optical element and realizing nondestructive testing and evaluation of strong laser damage characteristics.
Description
Technical Field
The invention relates to the field of detection of damage characteristics of optical elements, in particular to a nondestructive detection and evaluation method of strong laser damage characteristics of optical elements.
Background
The optical element can be damaged by laser induction under the action of laser, especially high-power/high-energy laser, and the existing direct evaluation method is a damage test method, and the representative technologies are prior art1, ISO 21254 (part 1-part 4), lasers and Lasers-related equipment-Test methods for laser-induced damage threshold and prior art 2,OPTICS EXPRESS Vol.22,No.23 (2014), damage modeling and statistical analysis of optics damage pertableance in MJ-class laser systems. Since the output capability of the laser is limited by the damage of the optical element, the prior art1 adopts a focusing method to improve the irradiation flux on the surface of the optical element, and obtains the laser flux which causes the damage of the optical element under the condition that the laser for testing does not generate the damage. The damage threshold of the surface or the body of the optical element is generally far higher than the output flux of the laser for testing, and under the condition of improving the laser flux by adopting a focusing mode, the test light spot is usually far smaller than the light passing area of the optical element, so that the damage performance of the whole optical element needs to be evaluated in a probability form, and the laser flux calculated value with zero damage probability is used as the laser damage threshold of the optical element. The damage test has the advantages that the damage flux or the damage threshold value of the optical element can be directly obtained, but the test period is long, the test cost is high (especially for large-caliber optical elements), and the test cannot be repeated, so that the problem is unavoidable.
In terms of nondestructive testing, prior art4, UCRL-PROC-207944 (2004), correlation of Laser-Induced Damage to Phase Objects in Bulk Fused Silica uses correlation analysis to analyze phase defects. The relation between fluorescence and damage caused by impurities is analyzed in the prior art 5,OPTICS EXPRESS Vol.25,No.5 (2017) Spatially selective excitation in laser-induced breakdown spectroscopy combined with laser-induced fluorescence, and related researches are also carried out in the nondestructive fluorescence detection method of the sub-damage depth of the optical element in the prior art JOURNAL OF HARBIN INSTITUTE OF TECHNOLOGY Vol.50 No. 7. However, these prior art evaluations are only single nondestructive testing parameters and do not include the relationship between the damage caused by the laser parameter nondestructive testing parameters and the nondestructive testing parameters, and these correlation analyses are limited to the possibility between the specific parameters and the damage occurrence, and cannot solve the multi-parameter coupling evaluation problem in the optical element damage process. Since the damage of the optical element under the action of strong laser is a process of multi-nondestructive detection parameter action and determination but random, the certainty is shown in the damage test process, the damage result is determined when and where the damage occurs, and the damage is related to the defect of the optical element and the laser parameter. Therefore, the optical element damage multi-parameter coupling evaluation problem is a nonlinear, multi-dimensional and multi-nondestructive testing parameter problem, and the explicit evaluation cannot be performed by applying a pure mathematical theory.
In order to solve the problems that the damage test period of an optical element is long, the test cost is high and the damage of the optical element cannot be repeatedly evaluated, solve the problem that the damage of the optical element cannot be comprehensively represented by a single optical element nondestructive test result, and solve the problem of the relevance between the damage of the optical element and laser parameters and the damage test result of the optical element nondestructive test result, the invention provides a nondestructive test and evaluation method for the strong laser damage characteristic of the optical element.
Disclosure of Invention
The invention aims at: aiming at the problems that the laser damage test period of an optical element is long, the test cost is high, the test cannot be repeated and the like, and the damage problem of the optical element under specific laser parameters cannot be comprehensively reflected by a single nondestructive testing parameter, the invention provides a nondestructive testing and evaluating method for the strong laser damage characteristic of the optical element, and aims to realize the nondestructive testing and evaluating of the damage characteristic of the optical element, thereby reducing the use risk of the optical element, particularly a large-caliber optical element and improving the reliability of a laser system.
The technical scheme adopted by the invention is as follows:
the invention provides a nondestructive testing and evaluating method for strong laser damage characteristics of an optical element, which comprises the following steps:
step 1: obtaining optical element non-destructive inspection parameter data by non-destructive inspection and visualizing the data, including but not limited to the following parameters: optical element surface/subsurface defect distribution, surface topography height distribution, absorption distribution, stimulated fluorescence distribution, stress distribution, roughness distribution, transmittance distribution, refractive index distribution, surface curvature distribution, and the like. Converting the detection parameter data into a coordinate system according to a section perpendicular to the light transmission direction of the optical element, and normalizing each type of distribution data with the maximum value in the data set to form a visual digital matrix of the nondestructive detection parameter data of the optical element;
specifically, let the i-th lossless parameter distribution under the coordinate system of the optical element be f (i) The coordinate system of the vertical section of the light beam in the light passing direction is (u, v), and the conversion relation of the two coordinate systems is shown in the formula (1).
f (i) (u，v)＝t (i) ·f (i) (x，y)+b (i) (1)
Wherein t is (i) 、b (i) The rotation, expansion and translation matrixes are respectively arranged between the coordinate system (x, y) and the coordinate system (u, v).
Let f (i) (u, v) matrix maximum value f (i) max (u, v) normalizing the i-th lossless parameter distribution f 1 (i) The visualization matrix (u, v) is shown in formula (2).
f 1 (i) (u，v)＝f (i) (u，v)/f (i) max (u，v) (2)
In particular, when the optical element is a planar optical element and the optical element is perpendicular to the laser propagation direction, t (i) ＝1，b (i) ＝0。
Step 2: prior laser parameter data is obtained by laser parameter measurements and the detection data is visualized, including but not limited to the following parameters: laser intensity distribution, phase distribution, polarization distribution. Converting the coordinate system of the parameters by using projection of laser irradiation on the surface area of the optical element on the vertical section of the light passing direction of the optical element, and normalizing the distribution data of each type by using the maximum value in the data set to form a priori laser parameter detection data visualization digital matrix;
specifically, let the j-th laser parameter be distributed as g under its own coordinates (j) (X，Y)，T (j) 、B (j) The rotation, expansion and translation matrixes are respectively arranged between the coordinate system (X, Y) and the coordinate system (u, v) of the vertical section of the light beam in the light passing direction. And (3) calculating according to the formula (1) and the formula (2) to obtain a normalized visualization matrix of the j laser parameters.
In particular, when the first-pass laser parameter data is laser propagation direction perpendicular section data, T (j) ＝1，B (j) ＝0。
Step 3: the prior lossy detection data is obtained through offline lossy testing, and the data is visualized. Marking the damaged area corresponding to the optical element as 1, marking the undamaged area as 0, and performing coordinate system conversion on the damaged detection data according to the section perpendicular to the light transmission direction of the optical element to form an optical element priori damaged detection data visualization matrix;
specifically, let the prior lossy detection result be distributed as R (x, y) under the coordinate system of the optical element itself, wherein the value of the region where the damage occurs is 1, and the value of the region where the damage does not occur is 0. Let T, B be the rotation expansion and translation matrix between the laser parameter coordinate system (X, Y) and the vertical section coordinate system (u, v) of the beam light-passing direction. And (3) calculating a priori lossy detection result R (u, v) after translation and rotation expansion according to the formula (1). Setting a point greater than 0 in R (u, v) to 1 to obtain a priori lossy detection result visualization matrix R 1 (u，v)。
In particular, when the optical element is a planar optical element and the optical element is perpendicular to the direction of propagation of the test laser beam during the test, t=1, b=0.
Step 4: and processing nondestructive testing parameters of the optical element and laser parameter testing data. Processing the normalized prior optical element detection Data matrix and the laser parameter matrix to obtain a multidimensional multi-parameter prior nondestructive detection Data matrix Data (u, v);
specifically, let the normalized matrix of optical element parameters be f 1 (i) (u, v) matrix size M f ×N f Laser parameter normalization matrix is g (j) (X, Y) matrix size M g ×N g When M f ≠M g Or N f ≠N g When using the difference method to divide f 1 (i) (u, v) and g (j) (X, Y) are adjusted to the same size, e.g. if M f >M g And N is f <N g Then, data (u, v) has a size M f ×N g 。
Step 5: taking the multi-parameter prior nondestructive testing data matrix obtained in the step 4 as input quantity, taking the prior lossy testing data obtained in the step 3 as to-be-fitted quantity or label value, and substituting the to-be-fitted quantity or label value into a neural network for training to obtain a relation model LIDRM (Laser Induced Damage Relationship Model) between the prior nondestructive testing data and the prior lossy testing data;
specifically, taking convolutional neural networks as an example, but not limited to, a convolutional neural network, a LIDAM model calculation method is exemplified.
Data (u, v) is used as input layer to input Data, R 1 (u, v) is used as output layer standard test data, y (u, v) is fitting data after model calculation, and Loss function loss= ||y (u, v) -R1 (u, v) |is used 2 And obtaining the number of convolution kernels, the optimal convolution kernel weight and a corresponding full-connection matrix and offset matrix through iterative fitting by taking the least square residual as a loss function. The LIDAM model includes: input layer, hidden layer and number, output layer, convolution kernel number and convolution kernel weight, activation function, full connection matrix and weight, bias matrix and weight, etc. See fig. 3 for a specific example.
To reduce the computer performance requirements, the fit Data size is increased by obtaining Data (u, v) and R 1 (u, v) matrix and then forming (Data) by dividing the matrix j (u，v)，R 1j (u, v)) numberData set, by Data j Matrix and R 1j Fitting of the matrix yields the LIDAM model at that size. When the Data matrix is taken as input for integral calculation, the Data matrix is taken as input for integral calculation j Matrix size partitions the Data matrix, R 1 The matrix is composed of R 1j And (5) splicing the matrixes to obtain the matrix.
Step 6: and (2) processing nondestructive testing data and laser parameter data of the optical element which are not subjected to damage testing according to the steps 1-2, obtaining the input quantity of an optimal LIDAM model as training, and obtaining a damage estimation result of the optical element under the action of the laser parameters through LIDAM calculation.
Step 7: under the action of a specific laser parameter, the spatial position of the optical element is regulated, and the area with the highest damage estimation result is used as a working area, so that the optical element can be optimally used under the laser parameter.
In summary, due to the adoption of the technical scheme, the beneficial effects of the invention are as follows:
1. according to the invention, the damage probability of the optical element under the action of a specific light spot can be estimated under the condition that the damage detection is not carried out, and the laser damage test cost of the optical element is reduced;
2. in the process of improving the optical element technology, the invention can evaluate the relation between the technological characteristic parameters and the laser damage, can realize damage performance evaluation when the laser light field distribution is constant, avoids fluctuation of test results caused by uneven test laser light field distribution, can pertinently improve the processing technology in the processing process of the optical element, and shortens the research and development economic cost and the time cost of the processing technology of the optical element;
3. the invention can evaluate the damage possibility of the optical element under the action of a specific laser light field, reduce the damage of the optical element in the actual application process, reduce the development cost of the laser and improve the reliability and usability of the system by adjusting the application mode of the optical element;
4. the invention can extrapolate the test results of the light spots with small caliber and medium caliber to the actual light spot action results with large caliber, which is beneficial to evaluating the reliability of the device in the design stage of the laser and reducing the design risk of the laser;
5. the invention is used as an expandable and reducible nondestructive evaluation method for damage of the optical element, and is suitable for LIDAM model fitting of the undamaged state and the damaged state of the surface of the same optical element; the method is suitable for fitting known nondestructive testing parameters and laser parameters, is also suitable for model fitting of new nondestructive testing parameters obtained by a new technology and a new method, and has good model expansibility; when only partial key lossless parameters are concerned, other non-concerned parameter matrixes can be set as identity matrixes, and the key parameters are subjected to reduction fitting by using a model.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art. The drawings are not intended to be drawn to scale on actual dimensions, emphasis instead being placed upon illustrating the principles of the invention.
FIG. 1 is a schematic diagram of a nondestructive evaluation flow of damage to an optical element in accordance with the present invention;
FIG. 2 is a schematic diagram of a point-by-point scan during acquisition of nondestructive inspection data of an optical element in accordance with the present invention;
FIG. 3 is a schematic diagram of a lossy test data acquisition process for an optical element according to the present invention;
fig. 4 is a schematic diagram of the convolutional neural network calculation lidam of the present invention.
Reference numeral 3:
1-testing a laser light source; 2-an energy attenuation system; a 3-focus system; 4-a laser parameter measurement system; 5-optical element samples; 6-a damage diagnosis system; 7-an assisted mobile system; 8-an energy absorbing device; 9-computer control system.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention provides a nondestructive testing and evaluating method for strong laser damage characteristics of an optical element, which comprises the following steps:
step 1: obtaining prior optical element non-destructive inspection parameter data by non-destructive inspection and visualizing the data, including but not limited to the following parameters: optical element surface/subsurface defect distribution, surface topography height distribution, absorption distribution, stimulated fluorescence distribution, stress distribution, roughness distribution, transmittance distribution, refractive index distribution, surface curvature distribution, and the like. Converting the detection parameter data into the parameter coordinate system according to the section perpendicular to the light transmission direction of the optical element, and normalizing each type of distribution data with the maximum value in the data set to form a priori nondestructive detection parameter data visualization digital matrix of the optical element;
in particular, in terms of obtaining the nondestructive testing parameter data of the prior optical element, absorption, stimulated fluorescence, roughness, transmittance, refractive index, surface curvature distribution and the like can be obtained in a point-by-point scanning mode, the surface defect distribution of the optical element is obtained in a bright field or dark field imaging mode, the subsurface defect distribution of the optical element is obtained in an internal reflection illumination and bright field imaging mode, the surface topography height distribution is obtained in an interferometry mode, and the stress distribution is obtained in a stress birefringence measurement mode. In the point-by-point scanning process, the coordinate position of the scanning area relative to the optical element should be recorded and the measured value is recorded in the corresponding position, and when the measuring area is large, the parameter value in the area is replaced by the measured value.
Specifically, the aspect of detecting the data coordinate system conversion is provided withThe ith lossless parameter distribution under the self-coordinate system of the optical element is f (i) The coordinate system of the vertical section of the light beam in the light passing direction is (u, v), and the conversion relation of the two coordinate systems is shown in the formula (1).
f (i) (u，v)＝t (i) ·f (i) (x，y)+b (i) (1)
Wherein t is (i) 、b (i) The rotation, expansion and translation matrixes are respectively arranged between the coordinate system (x, y) and the coordinate system (u, v).
Let f (i) (u, v) matrix maximum value f (i) max (u, v) normalizing the i-th lossless parameter distribution f 1 (i) The visualization matrix (u, v) is shown in formula (2).
f 1 (i) (u，v)＝f (i) (u，v)/f (i) max (x 0 ，y 0 ) (2)
In particular, when the optical element is a planar optical element and the optical element is perpendicular to the laser propagation direction, t (i) ＝1，b (i) ＝0。
Step 2: prior laser parameter data is obtained by laser parameter measurements and the detection data is visualized, including but not limited to the following parameters: laser intensity distribution, phase distribution, polarization distribution. Converting the coordinate system of the parameters by using projection of laser irradiation on the surface area of the optical element on the vertical section of the light passing direction of the optical element, and normalizing the distribution data of each type by using the maximum value in the data set to form a priori laser parameter detection data visualization digital matrix;
specifically, in the aspect of laser parameter acquisition, laser intensity distribution can be obtained in a conjugated imaging mode, laser beam phase distribution can be obtained in a Hartmann method or a shearing interference method, and laser beam polarization distribution can be obtained in a conjugated imaging and polarization-detecting mode.
Specifically, in terms of laser parameter data coordinate system conversion, the relation between the laser beam coordinate system and the optical element coordinate system should be calibrated before the damage test, and the coordinate system of the vertical interface of the light beam is taken as an intermediate coordinate system, so that the two coordinate systems are connected together.
Let j-th laser parameters be distributed as g under own coordinates (j) (X，Y)，T (j) 、B (j) The rotation, expansion and translation matrixes are respectively arranged between the coordinate system (X, Y) and the coordinate system (u, v) of the vertical section of the light beam in the light passing direction. And (3) calculating according to the formula (1) and the formula (2) to obtain a normalized visualization matrix of the j laser parameters.
In particular, when the first-pass laser parameter data is laser propagation direction perpendicular section data, T (j) ＝1，B (j) ＝0。
Step 3: the prior lossy detection data is obtained through offline lossy testing, and the data is visualized. Marking the damaged area corresponding to the optical element as 1, marking the undamaged area as 0, and performing coordinate system conversion on the damaged detection data according to the section perpendicular to the light transmission direction of the optical element to form an optical element priori damaged detection data visualization matrix;
specifically, let the prior lossy detection result be distributed as R (x, y) under the coordinate system of the optical element itself, wherein the value of the region where the damage occurs is 1, and the value of the region where the damage does not occur is 0. Let T, B be the rotation expansion and translation matrix between the laser parameter coordinate system (X, Y) and the vertical section coordinate system (u, v) of the beam light-passing direction. And (3) calculating a priori lossy detection result R (u, v) after translation and rotation expansion according to the formula (1). Setting a point greater than 0 in R (u, v) to 1 to obtain a priori lossy detection result visualization matrix R 1 (u，v)。
In particular, when the optical element is a planar optical element and the optical element is perpendicular to the direction of propagation of the test laser beam during the test, t=1, b=0.
Step 4: and processing nondestructive testing parameters of the optical element and laser parameter testing data. Processing the normalized prior optical element detection Data matrix and the laser parameter matrix to obtain a multidimensional multi-parameter prior nondestructive detection Data matrix Data (u, v);
specifically, let the normalized matrix of optical element parameters be f 1 (i) (u, v) matrix size M f ×N f Laser parameter normalization matrix is g (j) (x, y) matrix size M g ×N g When M f ≠M g Or N f ≠N g When using the difference method to divide f 1 (i) (u, v) and g (j) (x, y) are adjusted to the same size, e.g. if M f >M g And N is f <N g Then, data (u, v) has a size M f ×N g 。
Step 5: taking the multi-parameter prior nondestructive testing data matrix obtained in the step 4 as input quantity, taking the prior lossy testing data obtained in the step 3 as to-be-fitted quantity or label value, and substituting the to-be-fitted quantity or label value into a neural network for training to obtain a relation model LIDRM (Laser Induced Damage Relationship Model) between the prior nondestructive testing data and the prior lossy testing data;
specifically, taking convolutional neural networks as an example, but not limited to, a convolutional neural network, a LIDAM model calculation method is exemplified.
As shown in FIG. 1, data (u, v) is input as an input layer, R 1 (u, v) is used as output layer standard test data, y (u, v) is fitting data after model calculation, and Loss function loss= ||y (u, v) -R1 (u, v) |is used 2 And obtaining the number of convolution kernels, the optimal convolution kernel weight and a corresponding full-connection matrix and offset matrix through iterative fitting by taking the least square residual as a loss function. The LIDAM model includes: input layer, hidden layer and number, output layer, convolution kernel number and convolution kernel weight, activation function, full connection matrix and weight, bias matrix and weight, etc.
To reduce the computer performance requirements, the fit Data size is increased by obtaining Data (u, v) and R 1 (u, v) matrix and then forming (Data) by dividing the matrix j (u，v)，R 1j (u, v)) dataset by applying to Data j Matrix and R 1j Fitting of the matrix yields the LIDAM model at that size. When the Data matrix is taken as input for integral calculation, the Data matrix is taken as input for integral calculation j Matrix size partitions the Data matrix, R 1 The matrix is composed of R 1j And (5) splicing the matrixes to obtain the matrix.
Step 6: and (2) processing nondestructive testing data and laser parameter data of the optical element which are not subjected to damage testing according to the steps 1-2, obtaining the input quantity of an optimal LIDAM model as training, and obtaining a damage estimation result of the optical element under the action of the laser parameters through LIDAM calculation.
Embodiment a priori optical element non-destructive inspection parameter data acquisition
As shown in fig. 2, the transmittance, absorption, stimulated fluorescence, roughness, refractive index, surface curvature distribution, and the like are obtained in a point-by-point scanning manner in the manner shown in fig. 2 by using a spectrophotometer, an absorption coefficient tester, a stimulated fluorescence tester, a roughness tester, a refractive index tester, a surface curvature tester, and the like. In the point-by-point scanning process, the coordinate position of the scanning area relative to the optical element should be recorded and the measured value is recorded in the corresponding position. When the measurement area is large and the measurement value is stable in the measurement area, as shown in fig. 2, all the values of the parameter in the measurement area can be replaced by the measurement value, for example, in fig. 2, the black dots represent the measurement values, and the black dots in the dotted line represent the measurement area which can be represented by the black dots in the discrete measurement process, so that the values in the shadow area are all the measurement values of the black dots.
And obtaining the surface defect distribution image of the optical element by adopting an imaging mode under bright field or dark field illumination through a microscopic system.
And in the vertical section of the light passing direction of the optical element, obtaining a subsurface defect distribution image of the optical element by internal reflection illumination and adopting a microscopic imaging mode.
And obtaining the surface morphology height distribution by adopting a laser interferometer through an interferometry method.
Stress distribution is obtained by stress birefringence measurement by adopting a stress meter.
By adopting the measurement, the prior optical element nondestructive testing parameter data are obtained.
Example two priori optical element lossy detection parameter data acquisition
Schematic diagram of the process of acquiring destructive testing data of an optical element according to the present invention referring to fig. 3, the destructive testing data testing system comprises: a test laser light source 1, an energy attenuation system 2, a focusing system 3, a laser parameter measurement system 4, an optical element sample 5, a damage diagnosis system 6, an auxiliary moving system 7, an energy absorption device 8 and a computer control system 9. The auxiliary moving system performs position adjustment of the optical element sample 5. Besides obtaining laser parameter distribution data, the laser parameter measurement system 4 also has the function of calibrating the position relationship of the laser beam and the optical element coordinate system, calibrating the characteristic position of the laser beam, and calculating the rotation and translation matrixes between the laser beam and the optical element corresponding to the coordinate system. The damage diagnosis system 6 needs to record the coordinate positions of the damage points relative to the optical element in addition to the damage profile of the optical element.
Based on a schematic diagram of the test system, the detailed description of the process of acquiring the optical element damage test data in this embodiment is given below:
the laser beam emitted by the test laser source 1 passes through the energy attenuation system 2, and the energy of the laser beam is regulated by the energy attenuation system to meet the test requirement. The focusing system 3 focuses and irradiates the laser beam onto the surface of the optical element 7. The damage diagnosis system 6 respectively performs surface imaging on the laser beam irradiation area before and after irradiation, compares the surface irradiation change of the optical element, judges whether damage occurs through a computer, and if damage occurs, records the irradiation point and the damage point position and marks the damage area in addition to recording the damage morphology.
After the damage has occurred, the auxiliary movement system 7 moves the optical element 5 to the next position to be irradiated under the control of the computer control system 9.
During irradiation, the residual laser light is absorbed by the energy absorbing device 8. The computer control system 9 controls the laser output, energy attenuation, auxiliary movement, and records and saves laser parameters, damage diagnosis results.
In order to evaluate the damage condition of the optical element under different fluxes, the optical element should be scanned with a certain flux, and after the scanning is completed, the damage image of the optical element is recorded. The laser flux is changed, the optical element is scanned again, and the damaged area is skipped and is not used as damage data under the flux. Until the number of undamaged areas of the optical element does not meet the requirement of iterative fitting (such as not less than 50).
Through the flow, the prior optical element lossy detection parameter data are obtained.
Wherein the connection in the drawings represents a possibly indirect connection and does not necessarily refer to a mechanical or an electrical direct connection, and the relationship should be understood in terms of text;
wherein the optical element test sample is usually tested in the form of a test specimen;
wherein the optical element mainly refers to a laser type optical element;
the laser under the above-mentioned set laser parameters mainly refers to laser meeting the requirements, but what specific requirements are determined according to the data actually needed.
The structure of the test system is not limited to that shown in the embodiment, and those skilled in the art can implement the present invention by using equivalent substitutions, etc., and the present invention is also within the scope of the present invention.
In order to avoid damage of the laser for testing, the laser induced damage test of the optical device (such as fused quartz material, K9 material, crystal, neodymium glass, etc.) is usually a beam shrinking test, i.e. the area to be tested is far smaller than the aperture of the optical device, so that multiple areas need to be tested or sampled to represent the whole optical device, and these areas are all on the same optical device.
According to the neural network fitting law, when the sample is large enough, the established model is more reliable, so that more prior nondestructive testing data means that a more reliable nondestructive evaluation model of the damage of the optical element is obtained.
Example three convolutional neural network calculation LIDRM
Based on the first embodiment and the second embodiment, the relationship model LIDRM (Laser Induced Damage Relationship Model) between the obtained priori nondestructive testing data and the obtained priori lossy testing data is obtained by analyzing the surface testing data of the priori optical element. See fig. 4 for a specific flow.
The convolution network is randomly generated and at least comprises an input layer, a hidden layer and an output layer. In this embodiment, a multi-layer neural network is used.
N original Data matrixes data_n (u, v) are randomly selected in the Data set, convolved by a plurality of convolution kernels W which are randomly generated and added with a random bias matrix B, and a plurality of FeatureMap, featureMap activated functions (such as sigmaio functions or Relu functions) are generated and enter the next network layer. The FeatureMap of the next layer is obtained by overlapping the FeatureMap of the previous layer after convolution of convolution kernels, and finally the FeatureMap of the next layer is tidied into one-dimensional vectors at an output layer, and the one-dimensional vectors are connected with a unidimensional R1 matrix through a randomly generated full-connection matrix and a bias matrix.
And obtaining a least square deviation matrix (Loss function) of an output result and an R1 matrix by first iterative calculation, and reversely correcting the full-connection matrix weight, the bias matrix weight, the convolution kernel weight and the bias matrix weight by using a gradient descent algorithm according to a Back Propagation (BP) algorithm of a neural network and a Loss function error matrix.
Repeating the steps until the model meets the error requirement.
And when the calculation result of the LIDAM model meets the error requirement, the obtained LIDAM model is the calculation result.
The foregoing is merely illustrative embodiments of the present invention, and the present invention is not limited thereto, and any changes or substitutions that may be easily contemplated by those skilled in the art within the scope of the present invention should be included in the scope of the present invention.
Claims (10)
1. The nondestructive testing and evaluating method for the strong laser damage characteristic of the optical element is characterized by comprising the following steps:
step 1: obtaining nondestructive testing parameter data of the optical element in a nondestructive testing mode, and visualizing the nondestructive testing parameter data of the optical element:
converting a coordinate system of the nondestructive testing parameter data of the optical element according to a section perpendicular to the light transmission direction of the optical element, and normalizing the nondestructive testing parameter data of the optical element according to the maximum value in the data set of each type of distribution data to form a visual digital matrix of the nondestructive testing parameter data of the optical element;
step 2: obtaining laser parameter data through laser parameter measurement, and visualizing the laser parameter data:
performing coordinate system conversion on the laser parameter data by using projection of laser irradiation on the surface area of the optical element on the vertical section of the light passing direction of the optical element, and normalizing each type of distribution data by using the maximum value in the data set to form a visual digital matrix of laser parameter detection data;
step 3: obtaining optical element damage detection parameter data through off-line damage test, and visualizing the optical element damage detection parameter data:
marking a damaged area corresponding to the optical element as 1, and marking an undamaged area as 0;
performing coordinate system conversion on the damage detection data according to the section perpendicular to the light transmission direction of the optical element to form a visual digital matrix of the damage detection parameter data of the optical element;
step 4: processing the visual digital matrix of the nondestructive testing parameter data of the optical element obtained in the step 1 and the visual digital matrix of the laser parameter testing data obtained in the step 2 to obtain a multidimensional and multiparameter nondestructive testing data matrix;
specifically, the step 4 is to process the normalized prior optical element detection Data matrix and the laser parameters to obtain a multidimensional multi-parameter prior nondestructive detection Data matrix Data (u, v), specifically:
let the normalized matrix of the optical element parameters be f 1 (i) (u, v) matrix size M f ×N f Laser parameter normalization matrix is g (j) (x, y) matrix size M g ×N g ；
When M f ≠M g Or N f ≠N g When using the difference method to divide f 1 (i) (u, v) and g (j) (X, Y) is adjusted to the same size;
if M is f >M g And N is f <N g Then, data (u, v) has a size M f ×N g ；
Step 5: taking the multi-parameter non-destructive testing data matrix obtained in the step 4 as an input quantity, taking the optical element damage testing parameter data obtained in the step 3 as a quantity to be fitted or a label value, substituting the quantity to be fitted or the label value into a neural network for training, and obtaining a relation model LIDRM between the non-destructive testing data and the damage testing data;
step 6: processing nondestructive testing parameter data and laser parameter data of an optical element which is not subjected to damage testing according to the steps 1-2, obtaining input quantity of an optimal LIDAM model as training, and obtaining nondestructive testing results of damage characteristics of the optical element under the action of the laser parameters through LIDAM calculation;
step 7: under a specific laser parameter, adjusting the spatial position of the optical element, and taking the area with the highest damage estimation result as a working area to realize the optimal use of the optical element under the laser parameter;
step 8: and (3) aiming at the influence of single or multiple nondestructive testing parameters on damage characteristics, after the LIDAM model is obtained, setting other nondestructive testing parameter matrixes in the step (6) as all 1 identity matrixes, and substituting the single or multiple nondestructive testing parameter matrixes to be considered into the model, so that nondestructive testing results of the damage characteristics of the optical element under the condition of the single or multiple nondestructive testing parameters are obtained.
2. The method for non-destructive testing and evaluating the strong laser damage characteristic of an optical element according to claim 1, wherein step 1 is to convert the non-destructive testing parameter data of the optical element according to the cross section perpendicular to the light transmission direction of the optical element, normalize the distribution data of each type with the maximum value in the data set, and form the visual digital matrix of the non-destructive testing parameter data of the optical element, and comprises the following specific steps:
let the ith lossless parameter distribution in the optical element's own coordinate system (x, y) be f (i) (x, y) coordinates of vertical section of beam passing directionAnd (u, v), the two coordinate systems are converted as follows:
f (i) (u，v) ＝ t (i) ·f (i) (x，y)+b (i) (1)
wherein t is (i) 、b (i) The optical element coordinate system (x, y) and the cross-section coordinate system (u, v) are respectively a rotation expansion matrix and a translation matrix;
let f (i) (u, v) matrix maximum value f (i) max (u, v) normalizing the i-th lossless parameter distribution f 1 (i) (u, v) the visualization matrix is as follows:
f 1 (i) (u，v) ＝ f (i) (u，v)/ f (i) max (u，v) (2)。
3. the method for non-destructive testing and evaluating of a strong laser damage characteristic of an optical element according to claim 2, wherein when the optical element is a planar optical element and the optical element is perpendicular to the propagation direction of the laser beam, t (i) ＝1，b (i) ＝0。
4. The method for nondestructive testing and evaluating of optical element strong laser damage characteristics according to claim 1 or 2, wherein the optical element nondestructive testing parameter data comprises: optical element surface/subsurface defect distribution, surface topography height distribution, absorption distribution, stimulated fluorescence distribution, stress distribution, roughness distribution, transmittance distribution, refractive index distribution, and surface curvature distribution.
5. The method for non-destructive testing and evaluating the strong laser damage characteristic of an optical element according to claim 1, wherein the step 2 is characterized in that the laser parameter data is transformed by the projection of the laser irradiation on the surface area of the optical element on the vertical section of the optical element in the light transmission direction, and the distribution data of each type is normalized by the maximum value in the data set, so as to form a visual digital matrix of the laser parameter detection data, specifically:
let j-th laser parameters be distributed as under the own coordinatesg (j) (X，Y)，T (j) 、B (j) The rotation, expansion and translation matrixes are respectively arranged between the laser parameter coordinate system (X, Y) and the vertical section coordinate system (u, v) of the light beam passing direction;
and (3) calculating according to the formula (1) and the formula (2) to obtain a normalized visualization matrix of the j laser parameters.
6. The method for non-destructive testing and evaluating of a strong laser damage characteristic of an optical element according to claim 5, wherein when said first-pass laser parameter data is laser propagation direction perpendicular cross-section data, T (j) ＝1，B (j) ＝0。
7. The method for non-destructive testing and evaluating of the strong laser damage characteristics of an optical component according to claim 5 or 6, wherein the laser parameter data includes, but is not limited to: laser intensity distribution, phase distribution, and polarization distribution.
8. The method for non-destructive testing and evaluating of optical element strong laser damage characteristics according to claim 1, wherein said step 3 comprises transforming said damage detection data according to a cross section perpendicular to the light transmission direction of the optical element to form a visual digital matrix of the damage detection parameter data of the optical element, specifically:
the prior lossy detection results are distributed as R (x, y) under the coordinate system of the optical element, wherein the value of a region where damage occurs is 1, and the value of a region where no damage occurs is 0;
setting T, B as a rotation expansion and translation matrix between a laser parameter coordinate system (X, Y) and a beam light-passing direction vertical section coordinate system (u, v);
calculating the priori lossy detection result R (u, v) after translation and rotation expansion according to the formula (1), and setting the point greater than 0 in R (u, v) to 1 to obtain the priori lossy detection result visualization matrix R 1 (u，v)。
9. The method for non-destructive testing and evaluating of the strong laser damage characteristics of an optical element according to claim 8, wherein t=1, b=0 when the optical element is a planar optical element and the optical element is perpendicular to the propagation direction of the test laser beam during the test.
10. The method for non-destructive testing and evaluating of the strong laser damage characteristic of an optical element according to claim 1, wherein in step 5, a plurality of non-destructive testing parameter data are used as input values, damage testing data are used as label values, and a neural network training is performed to obtain a relation model of the non-destructive testing data and the damage testing data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111588367.6A CN114324329B (en) | 2021-12-23 | 2021-12-23 | Nondestructive testing and evaluating method for strong laser damage characteristic of optical element |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111588367.6A CN114324329B (en) | 2021-12-23 | 2021-12-23 | Nondestructive testing and evaluating method for strong laser damage characteristic of optical element |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114324329A CN114324329A (en) | 2022-04-12 |
| CN114324329B true CN114324329B (en) | 2023-10-31 |
Family
ID=81054845
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111588367.6A Active CN114324329B (en) | 2021-12-23 | 2021-12-23 | Nondestructive testing and evaluating method for strong laser damage characteristic of optical element |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114324329B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115014718B (en) * | 2022-05-30 | 2023-05-02 | 西安交通大学 | Research method for damage of laser irradiation optical element |
| CN116399850B (en) * | 2023-05-31 | 2023-09-15 | 无锡光煜晞科技有限责任公司 | Spectrum detection and identification system for optical signal processing and detection method thereof |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101226148A (en) * | 2008-02-19 | 2008-07-23 | 中国原子能科学研究院 | Method and device for detecting laser damage threshold of optical element |
| CN106770128A (en) * | 2017-01-11 | 2017-05-31 | 中国工程物理研究院激光聚变研究中心 | Quick three-dimensional detects the detection means and detection method of subsurface defect of optical element |
| CN206348270U (en) * | 2017-01-11 | 2017-07-21 | 中国工程物理研究院激光聚变研究中心 | Quick three-dimensional detects the detection means of subsurface defect of optical element |
| CN107063641A (en) * | 2017-01-16 | 2017-08-18 | 中国工程物理研究院激光聚变研究中心 | A kind of method of Nondestructive Evaluation optic element damage performance |
| CN107907086A (en) * | 2017-11-15 | 2018-04-13 | 西安交通大学 | Chemical etching measuring method, auxiliary experiment device and the test method of optical element sub-surface damage |
| CN108287061A (en) * | 2018-01-19 | 2018-07-17 | 中国工程物理研究院激光聚变研究中心 | A kind of laser optical element life tests and service life probabilistic testing method and system |
| CN110736600A (en) * | 2019-12-06 | 2020-01-31 | 中国科学院长春光学精密机械与物理研究所 | wind tunnel experiment system and test method for strong laser irradiation effect |
| CN111678919A (en) * | 2020-06-17 | 2020-09-18 | 中国人民解放军国防科技大学 | Strong laser element laser damage threshold prediction method, system and medium based on multi-mode in-situ test |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109668838B (en) * | 2018-11-13 | 2020-07-03 | 浙江大学 | Device and method capable of simultaneously detecting surface and subsurface defects of optical element |
-
2021
- 2021-12-23 CN CN202111588367.6A patent/CN114324329B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101226148A (en) * | 2008-02-19 | 2008-07-23 | 中国原子能科学研究院 | Method and device for detecting laser damage threshold of optical element |
| CN106770128A (en) * | 2017-01-11 | 2017-05-31 | 中国工程物理研究院激光聚变研究中心 | Quick three-dimensional detects the detection means and detection method of subsurface defect of optical element |
| CN206348270U (en) * | 2017-01-11 | 2017-07-21 | 中国工程物理研究院激光聚变研究中心 | Quick three-dimensional detects the detection means of subsurface defect of optical element |
| CN107063641A (en) * | 2017-01-16 | 2017-08-18 | 中国工程物理研究院激光聚变研究中心 | A kind of method of Nondestructive Evaluation optic element damage performance |
| CN107907086A (en) * | 2017-11-15 | 2018-04-13 | 西安交通大学 | Chemical etching measuring method, auxiliary experiment device and the test method of optical element sub-surface damage |
| CN108287061A (en) * | 2018-01-19 | 2018-07-17 | 中国工程物理研究院激光聚变研究中心 | A kind of laser optical element life tests and service life probabilistic testing method and system |
| CN110736600A (en) * | 2019-12-06 | 2020-01-31 | 中国科学院长春光学精密机械与物理研究所 | wind tunnel experiment system and test method for strong laser irradiation effect |
| CN111678919A (en) * | 2020-06-17 | 2020-09-18 | 中国人民解放军国防科技大学 | Strong laser element laser damage threshold prediction method, system and medium based on multi-mode in-situ test |
Non-Patent Citations (2)
| Title |
|---|
| 光学元件亚表面损伤深度的无损荧光检测方法;侯晶;王洪祥;王储;王景贺;朱本温;;哈尔滨工业大学学报(07);全文 * |
| 基于高斯脉冲激光空间分辨测量光学元件表面激光损伤阈值研究;单;赵元安;张喜和;胡国行;王岳亮;彭小聪;李成;;中国激光(01);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114324329A (en) | 2022-04-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114324329B (en) | Nondestructive testing and evaluating method for strong laser damage characteristic of optical element | |
| CN109060822B (en) | Long-pulse infrared nondestructive testing sequence image processing method and system | |
| AU2017245369B2 (en) | Wrinkle characterization and performance prediction for composite structures | |
| JP5328319B2 (en) | Inspection apparatus and inspection method using terahertz waves | |
| US6874932B2 (en) | Methods for determining the depth of defects | |
| US10094794B2 (en) | Characterization of wrinkles and periodic variations in material using infrared thermography | |
| US7724925B2 (en) | System for generating thermographic images using thermographic signal reconstruction | |
| US6751342B2 (en) | System for generating thermographic images using thermographic signal reconstruction | |
| RU2431823C2 (en) | Method and device for nondestructive test of isolation coat | |
| Holland | Thermographic signal reconstruction for vibrothermography | |
| Zhao et al. | A novel defect depth measurement method based on Nonlinear System Identification for pulsed thermographic inspection | |
| JP2015512526A (en) | Probabilistic fatigue life prediction using ultrasonography data considering EIFS uncertainty | |
| Zhao et al. | A coefficient clustering analysis for damage assessment of composites based on pulsed thermographic inspection | |
| Patil et al. | Surface roughness measurement based on singular value decomposition of objective speckle pattern | |
| Greene et al. | Crack detection in rail using infrared methods | |
| US20240013044A1 (en) | Method for characterizing a part through non-destructive inspection | |
| KR102619601B1 (en) | Apparatus, system and method for analyzing thin films with improved precision | |
| Ramirez-Granados et al. | Reconstruction and analysis of pulsed thermographic sequences for nondestructive testing of layered materials | |
| KR102492065B1 (en) | Apparatus and method for optical measuring | |
| Sasanpour et al. | Determination of probabilistic distribution function of background and defect optical densities for X-ray radiography images of a steel plate | |
| Tian et al. | A statistical framework for improved automatic flaw detection in nondestructive evaluation images | |
| Szymanik et al. | Qualitative evaluation of 3D printed materials’ structure using active infrared thermography and signal analysis based on LSTM neural networks | |
| Hurley et al. | FY17 Status Report for the Development of Infrared Thermography for In-Pile Fuel Behavior Applications | |
| US20230221247A1 (en) | Optically super-resolved infrared impulse spectroscopy (osiris) - a technique for high contrast chemical imaging of chemical composition | |
| JP2014130124A (en) | Inner surface spatial resolution improvement method, improvement device, and improvement program |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |