Flame field parameter anti-artifact tomography method, device, system and storage medium

Abstract

The present invention discloses a flame field parameter anti-artifact tomography method and device, system, and storage medium, comprising: step S1, obtaining the center of gravity area of ​​the flame field of the area to be measured; step S2, obtaining the absorption rate distribution through the robust iterative reconstruction algorithm driven by the center of gravity momentum according to the center of gravity area of ​​the flame field; step S3, obtaining the temperature and concentration distribution of the flame according to the absorption rate distribution. The technical solution of the present invention is adopted to improve the measurement accuracy.

Description

Technical Field

[0001] The present invention belongs to the technical field of tomography, and in particular relates to a flame field parameter artifact-resistant tomography method and device, system, and storage medium. Background Art

[0002] Tomography is a technology that uses projection data values ​​at different angles to obtain cross-sectional images. As mentioned above, when early laser absorption spectroscopy technology was combined with tomography technology and used to measure stable combustion fields, the same structure as medical tomography technology was often used, that is, the transmission data at multiple angles were collected by rotating the measurement system components. When the imaging speed is not required to be high, a rich number of projection angles can be obtained by fine rotation angle design for high-resolution imaging reconstruction. However, as laser absorption spectroscopy tomography technology gradually moves towards industrial sites, driven by the measurement needs of a large number of dynamic combustion fields at industrial sites, the structures of most tomography systems are currently based on parallel detection structures. This result avoids the limitation of mechanical scanning on measurement speed, but also reduces the number of measurement projection angles used for tomographic reconstruction, resulting in discontinuous abnormal values ​​in imaging results, especially in high-resolution imaging results. Such abnormal values ​​are generally referred to as image artifacts in the images of imaging results. As discussed above, iterative reconstruction methods are often used in image reconstruction of sensors with fixed parallel structures with a small number of projections. Domestic and foreign researchers have also conducted extensive research on high-resolution image reconstruction artifact removal methods based on iterative reconstruction methods. At present, there are mainly two ideas: one is to use prior knowledge such as the continuity of the combustion field to design basis functions for fitting continuous distribution, eliminating discontinuous outliers, and achieving the effect of removing artifacts; the other is to add low-pass spatial filtering in the iterative process to achieve smooth filtering of outliers.

[0003] By designing basis functions based on prior knowledge such as the continuity of the combustion field, the imaging grid to be solved is represented as a linear combination of basis functions with specific expressions, and the problem of solving the high-resolution grid is converted into the problem of solving the coefficients of the basis functions, thereby reducing the pathological nature of the reconstruction equation. X. Gao et al. proposed a reconstruction method using Zernike orthogonal polynomials as basis functions, which converts the problem of fitting the grid absorption rate distribution into the problem of solving the coefficients of the Zernike polynomials, effectively reducing the number of unknown quantities to be solved, achieving an improvement in resolution, and effectively suppressing artifacts in high-resolution imaging. However, due to the symmetry of Zernike polynomials, the scope of application of fitting reconstruction as basis functions is limited to symmetrically distributed measurement objects, while the fitting error is large for complex asymmetric shapes. This also reflects the problem of the method based on basis functions, that is, the basis functions that can effectively reduce the number of reconstruction unknowns are often limited to reconstructions with simple distributions.

[0004] In contrast, the spatial filtering method places fewer restrictions on the reconstructed object. Spatial filtering eliminates outliers by averaging or weighted averaging the values ​​of the grid points to be determined and the values ​​of the grid points around it. Therefore, for general continuous distributions, good reconstruction results can be obtained in principle. In the process of designing spatial filters, early researchers usually introduced fixed filters to smooth the reconstruction results or the intermediate reconstruction process, such as adding median filtering to the iterative process of the Landweber algorithm, or introducing discrete Laplace operators into the Tikhonov regularization reconstruction process by simulating the molecular diffusion motion that leads to the smooth distribution of component concentrations. These methods have improved the effect of suppressing artifacts. However, the fixed filter structure often leads to reduced reconstruction accuracy due to the lack of consideration of the adaptive adjustment of the filter template coefficients. N. Li et al. proposed an improved adaptive algebraic reconstruction algorithm. This method adds an adaptive weighted mean filtering process to the basic ART algorithm. The filter coefficients are adjusted by setting a smoothing factor during the iterative process. While suppressing artifacts, the accuracy and convergence speed of the reconstruction results are effectively improved. X. Yang et al. introduced weighted median filtering results into the SART algorithm, firstly performed median filtering on the results of the previous iteration, and then added the filtered results to the results of this iteration according to the weights, where the median filter is fixed, but the added weights are determined by simulation. N. Terzija et al. smoothed the results of the Landweber algorithm based on wavelet transform, which improved the iteration speed and reconstruction accuracy, but this method still relies on simulation calculation to determine the threshold of the approximate subband in the wavelet decomposition process. At present, whether it is directly using a fixed filter or using an adaptive filter under the influence of coefficients determined by simulation, it depends on empirical selection to determine, and adaptive filter selection cannot be achieved. This leads to the contradiction between detail resolution and anti-artifact performance in the specific imaging process due to the size selection of the filter. The main reason is that the spatial filtering algorithm is an algorithm that balances the reconstruction error and reconstruction resolution, and does not add new constraints or new test data to the reconstruction process. Therefore, in order to fundamentally achieve de-artifacting imaging with high resolution, it is necessary to consider studying adaptive algorithms based on physical field characteristics to supplement information in the iterative reconstruction process. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a flame field parameter artifact-resistant tomography method and device, system and storage medium to improve measurement accuracy.

[0006] To achieve the above object, the present invention adopts the following technical solution:

[0007] A flame field parameter anti-artifact tomography method, comprising:

[0008] Step S1, obtaining the center of gravity area of ​​the flame field of the area to be measured based on the flame convex shape assumption and the optical path layout structure;

[0009] Step S2, according to the center of gravity area of ​​the flame field, the absorption rate distribution is obtained by using a robust iterative reconstruction algorithm driven by the momentum of the center of gravity;

[0010] Step S3: Obtain the temperature and concentration distribution of the flame according to the absorption rate distribution.

[0011] In step S2, the absorption rate distribution is obtained by a robust iterative reconstruction algorithm driven by the momentum of the center of gravity, as follows:

[0012]

[0013] in, Indicates the serial number of the laser beam, Indicates Grid pairs The contribution of the laser path measurements, is the integrated absorption rate vector of all input paths, is the number of iterations, For the The local integrated absorption rate distribution of the spectral line of the iteration, For the The relaxation factor for the iteration, for The updated value of

[0014] When considering using the error to update the value of this iteration, the weight of the robust estimation distribution will be introduced , iterative update is:

[0015]

[0016] in, is the result of comparing the normalized update with the estimated robust distribution, The update is:

[0017]

[0018] in, and They respectively represent the average values ​​of the calculated absorption rates in all pixel grids used to normalize the update amount; is a weighting factor used to adjust the proportion of the correction value according to the noise level of the projected test signal;

[0019] When the number of iterations reaches the set maximum number of iterations, the iteration is terminated and the last updated absorption rate distribution result is output as the final absorption rate distribution result.

[0020] The present invention also provides a flame field parameter anti-artifact tomography device, comprising:

[0021] An acquisition module is used to acquire the center of gravity area of ​​the flame field of the area to be measured based on the flame convex shape assumption and the optical path layout structure;

[0022] A reconstruction module is used to obtain the absorption rate distribution according to the center of gravity area of ​​the flame field through a robust iterative reconstruction algorithm driven by the momentum of the center of gravity;

[0023] An imaging module is used to obtain the temperature and concentration distribution of the flame according to the absorption rate distribution;

[0024] Among them, the absorption rate distribution is obtained by the robust iterative reconstruction algorithm driven by the center of gravity momentum, as follows:

[0025]

[0026] in, Indicates the serial number of the laser beam, Indicates Grid pairs The contribution of the laser path measurements, is the integrated absorption rate vector of all input paths, is the number of iterations, For the The local integrated absorption rate distribution of the spectral line of the iteration, For the The relaxation factor for the iteration, for The updated value of

[0027] When considering using the error to update the value of this iteration, the weight of the robust estimation distribution will be introduced , iterative update is:

[0028]

[0029] in, is the result of comparing the normalized update with the estimated robust distribution, The update is:

[0030]

[0031] in, and They respectively represent the average values ​​of the calculated absorption rates in all pixel grids used to normalize the update amount; is a weighting factor used to adjust the proportion of the correction value according to the noise level of the projected test signal;

[0032] When the number of iterations reaches the set maximum number of iterations, the iteration is terminated and the last updated absorption rate distribution result is output as the final absorption rate distribution result.

[0033] An embodiment of the present invention further provides a flame field parameter anti-artifact tomography system, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes a flame field parameter anti-artifact tomography method when executed by the processor.

[0034] An embodiment of the present invention further provides a storage medium, on which a computer program is stored, and when the computer program is run, a flame field parameter anti-artifact tomography method is executed.

[0035] Based on the diffusion property of the combustion process, the present invention presets the area to be measured as a continuous convex distribution area. In the iterative process, the initial value is considered as a diffusion area close to a circle. By analyzing the geometric relationship between the optical path layout of the sensor and the radius of the assumed area to be measured, and further combining the signals of multiple optical path layouts, the area and value of the imaging area are preliminarily estimated, the accuracy of the initial value of the iteration is improved, the constraints in the iterative process are increased, the final reconstruction quality is improved, and the appearance of image artifacts is avoided; in addition, the present invention can effectively supplement the number of constraints in the iterative process of tomographic imaging, thereby improving the underdetermination of the overdetermined equation to be solved in the image reconstruction process, reducing its pathological nature, and thus improving the measurement accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0037] Figure 1 This is a flow chart of the flame field parameter anti-artifact tomography method according to an embodiment of the present invention;

[0038] Figure 2 Schematic diagram of the optical path of the five-angle tomography sensor;

[0039] Figure 3 Schematic diagram of the geometric relationship between the area to be tested and the path absorption rate. DETAILED DESCRIPTION

[0040] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0042] Embodiment 1:

[0043] like Figure 1 As shown, an embodiment of the present invention provides a flame field parameter anti-artifact tomography method, comprising:

[0044] Step S1, obtaining the center of gravity area of ​​the flame field of the area to be tested;

[0045] Step S2, according to the center of gravity area of ​​the flame field, the absorption rate distribution is obtained by using a robust iterative reconstruction algorithm driven by the momentum of the center of gravity;

[0046] Step S3: Obtain the temperature and concentration distribution of the flame according to the absorption rate distribution.

[0047] As an implementation of the embodiment of the present invention, in step S1, the center of gravity area of ​​the flame field of the area to be measured is obtained based on the flame convex shape assumption and the optical path layout structure.

[0048] Furthermore, the area to be measured is assumed to be a circular distribution, and the principle of using geometric relationships to obtain the initial value of the iterative area is demonstrated based on the model of the five-angle imaging sensor. Figure 2 As shown, two fan-shaped luminous angles among the five angles are selected as examples, and the position where the light passing through the area to be tested intersects with the area to be tested, that is, its chord length, is the position where absorption occurs. Since the product concentration and gas temperature in the combustion area are much higher than those in the non-combustion area, there is a linear relationship between the overall absorbance detected by this path and the chord length. According to the Beer-Lambert law:

[0049]

[0050] Where A is the integrated absorption area, P is the ambient pressure, is the length of the area to be measured, T is the temperature at a point on the laser transmission path, and S(T) is the line intensity of the absorption line

[0051] When making a preliminary estimate, it is assumed that the concentration and temperature distribution in the measured area is uniform. Figure 3The path shown has simple geometric constraints according to the Pythagorean theorem as follows:

[0052]

[0053] Therefore, the following equation can be constructed:

[0054]

[0055] in, is the absorption density,

[0056]

[0057] make H (n,m) The proposed burning area center and n The laser irradiates m The vertical distance between the optical paths of the detectors can be expressed as:

[0058]

[0059] in, is the coordinate of the center point of the proposed burning area, It is n The coordinates of the laser emission point, yes n The laser irradiates m The coordinates of the detector are the two coordinate points that determine the direction of the straight line and the position of the optical path. The coordinates of the laser emission point and the detector are known quantities, and the path absorption rate is also a known quantity. Therefore, according to the test results and coordinates of lasers with different paths, the center point of the combustion field can be fitted, and the range of the combustion field can be estimated to obtain the center of gravity of the flame field in the area to be tested.

[0060] As an implementation method of an embodiment of the present invention, in step S2, the SART algorithm is an iterative imaging algorithm commonly used in laser absorption spectral tomography calculations. The center of gravity region estimated in step S1 is fused with the SART algorithm to achieve robust SART tomography with artifact resistance. In the SART iteration process, the error is calculated by comparing the projection value of the reconstructed area with the test value, and the update amount of the iteration process is designed based on the error value. This error-driven iterative algorithm is overly dependent on the accuracy of the test data. When there is high noise in the projected test data, this iterative mechanism will cause the iteration result to deviate from the true value of the actual image in order to cater to the noisy projection data, which is specifically manifested as lower fitting efficiency and the appearance of local artifacts or noise. Considering the combination of the optical path layout structure and the convex shape assumption, the iterative variables corrected by simple error correction are corrected according to the expected imaging area and position calculated by the overall projection value.

[0061] Prior to this, the center of gravity region estimation algorithm can calculate the center of the combustion product absorption rate and the high absorption rate range based on the total absorption rate of the boundary. The distribution of the absorption rate corresponds to the concentration distribution and temperature distribution of the combustion products. Therefore, the preliminary distribution calculated from the boundary information can be used as the starting quantity of the SART iterative algorithm to speed up the iterative convergence process. The calculated estimated absorption rate distribution is :

[0062]

[0063] in, is the distance from the pixel to be solved to the center of the absorption rate distribution, that is, the coordinates are The distance between the points, is the average absorption rate of the high temperature and high concentration product area corresponding to the combustion, is the absorption rate of the non-combustion area, which is calculated based on the environmental state constant of the non-combustion state.

[0064] The initial value of the iteration Let be the estimated absorptivity distribution:

[0065]

[0066] During the iteration process, the update amount calculated directly from the error information is recorded as , which uses an update method similar to the SART update formula as follows:

[0067]

[0068] in, Indicates the serial number of the laser beam, Indicates Grid pairs The contribution of the laser path measurements, is the integrated absorption rate vector of all input paths, is the number of iterations, For the The local integrated absorption rate distribution of the spectral line of the iteration, For the The relaxation factor for the iteration, for The updated value of

[0069] When considering using the error to update the value of this iteration, the weight of the robust estimation distribution will be introduced , iterative update is:

[0070]

[0071] in, is the result of comparing the normalized update with the estimated robust distribution, The update is:

[0072]

[0073] in, and They respectively represent the average values ​​of the calculated absorption rates in all pixel grids used to normalize the update amount; is a weighting factor used to adjust the proportion of the correction value according to the noise level of the projection test signal. When the signal-to-noise ratio of the projection test signal is high, It can be a small value. When the signal-to-noise ratio of the projected test signal decreases, it should be increased. To reduce the over-driving of the update value by the projection test error. When the number of iterations reaches the set maximum number of iterations, the iteration is terminated and the last updated absorption rate distribution result is output as the final absorption rate distribution result.

[0074] As an implementation of the embodiment of the present invention, in step S3,

[0075] Since the integral area of ​​the absorption spectrum It is a quantity related to pressure, concentration, path length and temperature at the same time. Therefore, in order to measure temperature, it is necessary to add new information to eliminate parameters. Select two independent frequency lasers, and the center frequencies are recorded as , , so that it passes through the area to be tested synchronously in time and space, and the line intensities are and And extract the integrated absorption rate of the two absorption lines and , and then define the ratio of the twoR T for:

[0076]

[0077] in, R T is a variable uniquely related to temperature. R T - The lookup table of T can be obtained by R T The temperature T is obtained by solving the problem. Then, by using the obtained temperature value, any one of the two absorption lines can be selected to calculate the partial pressure corresponding to the measured area, which is the product of pressure and mole fraction:

[0078]

[0079] Then, under the condition of constant pressure, the gas concentration value can be determined.

[0080] Embodiment 2:

[0081] The embodiment of the present invention further provides a flame field parameter anti-artifact tomography device, comprising:

[0082] An acquisition module is used to acquire the center of gravity area of ​​the flame field of the area to be measured based on the flame convex shape assumption and the optical path layout structure;

[0083] A reconstruction module is used to obtain the absorption rate distribution according to the center of gravity area of ​​the flame field through a robust iterative reconstruction algorithm driven by the momentum of the center of gravity;

[0084] An imaging module is used to obtain the temperature and concentration distribution of the flame according to the absorption rate distribution;

[0085] Among them, the absorption rate distribution is obtained by the robust iterative reconstruction algorithm driven by the center of gravity momentum, as follows:

[0086]

[0087] in, Indicates the serial number of the laser beam, Indicates Grid pairs The contribution of the laser path measurements, is the integrated absorption rate vector of all input paths, is the number of iterations, For the The local integrated absorption rate distribution of the spectral line of the iteration, For the The relaxation factor for the iteration, for The updated value of

[0088] When considering using the error to update the value of this iteration, the weight of the robust estimation distribution will be introduced , iterative update is:

[0089]

[0090] in, is the result of comparing the normalized update with the estimated robust distribution, The update is:

[0091]

[0092] in, and They respectively represent the average values ​​of the calculated absorption rates in all pixel grids used to normalize the update amount; is a weighting factor used to adjust the proportion of the correction value according to the noise level of the projected test signal;

[0093] When the number of iterations reaches the set maximum number of iterations, the iteration is terminated and the last updated absorption rate distribution result is output as the final absorption rate distribution result.

[0094] Embodiment 3:

[0095] An embodiment of the present invention further provides a flame field parameter anti-artifact tomography system, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes a flame field parameter anti-artifact tomography method when executed by the processor.

[0096] Embodiment 4:

[0097] An embodiment of the present invention further provides a storage medium, on which a computer program is stored, and when the computer program is run, a flame field parameter anti-artifact tomography method is executed.

[0098] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A flame field parameter anti-artifact tomography method, characterized in that: include: Step S1, obtaining the center of gravity area of ​​the flame field of the area to be measured based on the flame convex shape assumption and the optical path layout structure; Step S2, according to the center of gravity area of ​​the flame field, the absorption rate distribution is obtained by using a robust iterative reconstruction algorithm driven by the momentum of the center of gravity; Step S3, obtaining the temperature and concentration distribution of the flame according to the absorption rate distribution; In step S2, the absorption rate distribution is obtained by a robust iterative reconstruction algorithm driven by the momentum of the center of gravity, as follows: in, Indicates the serial number of the laser beam, Indicates Grid pairs The contribution of the laser path measurements, is the integrated absorption rate vector of all input paths, is the number of iterations, For the The local integrated absorption rate distribution of the spectral line of the iteration, For the The relaxation factor for the iteration, for The updated value of When considering using the error to update the value of this iteration, the weight of the robust estimation distribution will be introduced , iterative update is: in, is the result of comparing the normalized update with the estimated robust distribution, The update is: in, and They respectively represent the average values ​​of the calculated absorption rates in all pixel grids used to normalize the update amount; is a weighting factor used to adjust the proportion of the correction value according to the noise level of the projected test signal; When the number of iterations reaches the set maximum number of iterations, the iteration is terminated and the last updated absorption rate distribution result is output as the final absorption rate distribution result.

2. A flame field parameter anti-artifact tomography device, characterized in that: include: An acquisition module is used to acquire the center of gravity area of ​​the flame field of the area to be measured based on the flame convex shape assumption and the optical path layout structure; A reconstruction module is used to obtain the absorption rate distribution according to the center of gravity area of ​​the flame field through a robust iterative reconstruction algorithm driven by the momentum of the center of gravity; An imaging module is used to obtain the temperature and concentration distribution of the flame according to the absorption rate distribution; Among them, the absorption rate distribution is obtained by the robust iterative reconstruction algorithm driven by the center of gravity momentum, as follows: in, Indicates the serial number of the laser beam, Indicates Grid pairs The contribution of the laser path measurements, is the integrated absorption rate vector of all input paths, is the number of iterations, For the The local integrated absorption rate distribution of the spectral line of the iteration, For the The relaxation factor for the iteration, for The updated value of When considering using the error to update the value of this iteration, the weight of the robust estimation distribution will be introduced , iterative update is: in, is the result of comparing the normalized update with the estimated robust distribution, The update is: in, and They respectively represent the average values ​​of the calculated absorption rates in all pixel grids used to normalize the update amount; is a weighting factor used to adjust the proportion of the correction value according to the noise level of the projected test signal; When the number of iterations reaches the set maximum number of iterations, the iteration is terminated and the last updated absorption rate distribution result is output as the final absorption rate distribution result.

3. A flame field parameter anti-artifact tomography system, characterized in that: include: A memory and a processor, wherein the memory stores a computer program executed by the processor, and when the computer program is executed by the processor, the flame field parameter anti-artifact tomography method according to claim 1 is executed.

4. A storage medium, characterized in that: The storage medium stores a computer program, which executes the flame field parameter anti-artifact tomography method according to claim 1 when running.

Images