A method for measuring beam divergence angle of high-current particle beam based on infrared diagnostic target

Abstract

The application discloses a kind of based on infrared diagnosis target's strong current particle beam beam divergence angle measurement method, it is related to strong current particle beam diagnostic field.Processing to background frame and experimental frame;Each candidate beam divergence angle is generated respectively each corresponding single-hole two-dimensional Gaussian beam spot of extraction hole, and N single-hole Gaussian components are superimposed according to actual spatial position and weight, obtain the multiple Gaussian simulation beam field capable of characterizing multi-peak, peak valley and shoulder structure;In the same transverse and longitudinal strip region, using the same group of weights respectively extracts experimental temperature rise curve and simulation beam curve, completes field alignment, and with the shape error, gradient error and half-height width error between experimental curve and simulation curve constructs composite objective function.For different candidate beam divergence angles, multiple Gaussian simulation beam field and simulation curve are repeatedly generated, by coarse step search and local fine step search, with the beam divergence angle corresponding to the minimum composite error as the inversion result, and output root mean square beam divergence angle and its evolution with time.

Description

Technical Field

[0001] This invention relates to the field of high-current particle beam diagnostic technology, and specifically to a method for measuring the beam divergence angle of a high-current particle beam based on an infrared diagnostic target. Background Technology

[0002] Beam divergence is a critical beam parameter in neutral beam injection devices, ion beam devices, and other high-energy particle beam systems. It directly affects beam transmission efficiency, target power density distribution, and device safety margin. Acquiring the temperature field formed after a particle beam bombards a diagnostic target using an infrared camera, and then calculating the beam divergence angle from the hotspot morphology, is a non-contact diagnostic method suitable for high-power, short-pulse beam conditions.

[0003] In actual experiments, infrared cameras are often not directly aligned with the diagnostic target, resulting in significant perspective distortion in the original temperature field. Furthermore, different experimental frames exhibit background radiation, boundary noise, and hotspot position shifts. Traditional methods often involve manually selecting a single row or column from the experimental temperature rise field and directly performing a single Gaussian fit on the selected one-dimensional curve, then calculating the beam divergence angle from the fitted width. This type of method does not establish a simulated beam curve based on candidate beam divergence angles and repeatedly compare it with the actual curve. It represents a one-way calculation from the experimental curve to the beam divergence angle, making it difficult to utilize the overall hotspot peak shape, gradient, and horizontal / vertical differences to form a joint constraint on the beam divergence angle.

[0004] Particle beams are typically extracted from multiple electrode apertures, each forming a single-aperture beam with its own central position and propagation direction. The individual aperture beams are influenced by the aperture array arrangement, aperture offset focusing, inter-plate angle, and zone deflection parameters. When superimposed on the diagnostic target surface, the resulting overall beam often exhibits multiple Gaussian characteristics, including multi-peak, peak-valley, shoulder, and asymmetry. The diagnostic target is usually positioned downstream of the extraction electrodes. As the single-aperture beams propagate over a certain distance, they gradually expand and undergo multiple spatial superpositions on the target surface. Simultaneously, the lateral installation space within the beam source and the conditions for avoiding adjacent components are limited. If two aperture array plates are arranged coplanarly side-by-side, the lateral occupancy will significantly increase, potentially causing structural interference. Therefore, a folded arrangement where two aperture array plates are connected at a common edge, slightly tilted in opposite directions relative to the vertical direction, and converged along the beam propagation direction allows for the retention of the required number of apertures and effective extraction area within the limited installation space, while ensuring the two sets of aperture beams achieve the desired coverage and superposition on the distant diagnostic target surface. If the overall beam is described by a single Gaussian function or an array of equally spaced parallel apertures, the local topographic information caused by the actual electrode structure will be lost, thus affecting the physical accuracy of beam divergence angle inversion.

[0005] Therefore, a forward model is needed that represents each aperture as a single aperture two-dimensional Gaussian component and performs multiple superpositions based on the actual aperture position and beam direction. At the same time, it is necessary to extract the simulated curve and the actual curve under the conditions of unified coordinates, unified strip region and unified weight, and repeatedly generate the simulated curve by changing the candidate beam divergence angle and minimizing the error between the two, so as to complete the beam divergence angle inversion with clear physical meaning. Summary of the Invention

[0006] This invention provides a method for measuring the beam divergence angle of high-current particle beams based on infrared diagnostic targets, in order to solve the problems of existing methods, such as the use of a single Gaussian approximation for the overall beam, insufficient utilization of real porous structures, lack of closed-loop inversion between simulated and actual curves, insufficient two-dimensional morphology constraints, and inconsistent evaluation criteria for beam divergence angle.

[0007] To achieve the above objectives, the present invention provides a method for measuring the beam divergence angle of a high-current particle beam based on an infrared diagnostic target, comprising the following steps:

[0008] Step 1: Select one frame of background data and multiple experimental data from different times, sort the selected data files naturally according to the numbers in the filenames, and identify the background frame;

[0009] Step 2: Subtract the temperature matrix of each experimental frame from the temperature matrix of the background frame pixel by pixel, and set the negative values ​​in the difference to zero to obtain the temperature rise field at each time.

[0010] Step 3: Select a rough quadrilateral region of interest on the first experimental frame, perform edge detection and line detection within the region, filter the main lines corresponding to the four sides of the rough quadrilateral, and obtain the refined four corner points from the intersection of adjacent main lines;

[0011] Step 4: Establish homography transformation based on the four corner points of the refinement, correct the perspective of the temperature rise field into a rectangular area, and determine the horizontal and vertical pixel dimensions based on the actual physical width and actual physical height of the rectangular area;

[0012] Step 5: Establish a real multi-hole beam forward model based on the hole array position of the particle beam extraction electrode, the beam direction of each hole, the inter-plate angle, and the deflection parameters of the partitioned holes; use the candidate beam divergence angle as the single-hole beam spot width parameter to generate a single-hole two-dimensional Gaussian beam spot that propagates from each extraction hole to the diagnostic target surface, and weight and superimpose the N single-hole Gaussian components to form a multi-Gaussian simulated beam field;

[0013] Step 6: Determine the alignment offset of the simulated field relative to the experimental field using the normalized cross-correlation between the multiple Gaussian simulated beam field and the experimental temperature rise field, and reuse the alignment offset in the inversion calculation of each candidate beam divergence angle.

[0014] Step 7: Determine the transverse and longitudinal banded regions based on the weighted centroid of the experimental temperature rise field. Using the same set of weights determined by the experimental temperature rise intensity, extract the transverse and longitudinal normalized curves from the experimental temperature rise field and the multiple Gaussian simulated beam field, respectively.

[0015] Step 8: Compare the experimental curve and the simulated curve point by point under the same physical coordinates, calculate the shape error, gradient error and half width at half maximum (WHM) error of the transverse and longitudinal curves respectively, and combine them to obtain the composite objective function for beam divergence angle inversion;

[0016] Step 9: Change the candidate beam divergence angle one by one within the preset beam divergence angle range. Each time the candidate beam divergence angle is changed, regenerate each single-hole Gaussian beam spot, the multi-Gaussian simulated beam field and its transverse and longitudinal simulated curves. First, perform a coarse step search, and then perform a local fine step search around the optimal value of the coarse search. The beam divergence angle corresponding to the minimum composite error of the experimental curve and the simulated curve is used as the inversion value.

[0017] Step 10: According to the correspondence between the Gaussian width parameter and the root mean square width parameter, convert the inverted model beam divergence angle into root mean square beam divergence angle as the final measurement result, and output the beam divergence angle time evolution results corresponding to multiple experimental frames.

[0018] The present invention has the following beneficial effects:

[0019] First, each real aperture is treated as an independent single-aperture two-dimensional Gaussian component, and N-fold superposition is performed based on the aperture array position, aperture beam direction, inter-plate angle, and partition deflection. The resulting multi-gaussian simulated beam field can naturally retain the multi-peak, peak-valley, shoulder, and two-dimensional non-uniformity of the overall beam, avoiding the substitution of a single Gaussian curve for the real multi-aperture beam. Second, experimental and simulated curves are extracted synchronously under the same spatial band and weight. For each candidate beam divergence angle, a multi-gaussian simulated beam field is regenerated. Closed-loop inversion is performed through the joint error of shape, gradient, and half-width at half-maximum, so that the beam divergence angle is automatically determined by the best match between the simulated curve and the actual curve. Third, the region of interest is selected only once in the first frame, and the perspective transformation and physical scale are reused in subsequent frames. A joint search of coarse and fine scans is used to stably output the root mean square beam divergence angle and its temporal evolution for multiple frames. Fourth, it adopts a compact folded surface arrangement with a double-plate common edge connection and reverse slight tilt, which can accommodate two sets of aperture arrays without significantly increasing the lateral installation size, and utilizes a longer propagation distance to enable the two sets of aperture beams to form the required multiple coverage on the diagnostic target surface. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the experimental system and data processing relationship of the present invention;

[0021] Figure 2This is a flowchart illustrating the overall process of this invention, from single-hole Gaussian modeling and multiple Gaussian superposition to experimental-simulation curve inversion.

[0022] Figure 3 This is a flowchart of the multi-file reading, natural sorting, and background frame recognition process of the present invention;

[0023] Figure 4 This is a schematic diagram of the fine-tuning of corner points and perspective correction in the region of interest of this invention;

[0024] Figure 5 This is a schematic diagram showing the three-dimensional structure of the double-plate folded electrode of the present invention, the actual hole array arrangement, and the side view relationship of the hole beam direction;

[0025] Figure 6 This is a schematic diagram of the propagation of the two-dimensional Gaussian beam spot in each single aperture and the superposition of the Gaussian components of N single apertures to form a multiple Gaussian simulated beam field according to the present invention.

[0026] Figure 7 This is a schematic diagram illustrating how the present invention extracts experimental and simulated curves within the same strip region and inverts the beam divergence angle based on the closed-loop composite error.

[0027] Figure 8 This is a schematic diagram showing the output of the coarse scan, fine scan, and single-beam divergence angle inversion results of the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other. To achieve the above objectives, this invention adopts the following technical solution.

[0029] like Figure 1 As shown, this invention provides a high-current particle beam divergence angle measurement system based on an infrared diagnostic target. The system is housed in a vacuum chamber and includes an ion source, a real double-plate porous electrode, a diagnostic target, an infrared camera, and a data acquisition and processing module. The ion source generates a particle beam, which is extracted through the real double-plate porous electrode to form a high-current particle beam that propagates to the surface of the diagnostic target, forming beam thermal deposition. The diagnostic target withstands the bombardment of the high-current particle beam and forms a corresponding temperature distribution. The infrared camera is positioned towards the diagnostic target to collect infrared temperature information of the target before and after particle beam bombardment. The data acquisition and processing module is connected to the infrared camera to receive the temperature data collected by the infrared camera and process the temperature data to obtain experimental data for subsequent beam distribution analysis and divergence angle inversion.

[0030] This invention provides a method for measuring the beam divergence angle of high-current particle beams based on an infrared diagnostic target, such as... Figure 2 As shown, it includes the following steps:

[0031] Step 1: Select one frame of background data and multiple experimental data from different times, sort the selected data files naturally according to the numbers in the filenames, and identify the background frame;

[0032] Step 2: Subtract the temperature matrix of each experimental frame from the temperature matrix of the background frame pixel by pixel, and set the negative values ​​in the difference to zero to obtain the temperature rise field at each time point; further, Step 2 includes: first reading the background frame and multiple experimental frames, subtracting the temperature matrix of the k-th experimental frame from the temperature matrix of the background frame pixel by pixel, and setting the negative values ​​to zero to obtain the non-negative temperature rise field:

[0033] ;

[0034] This represents the coordinates of the diagnostic target surface at the k-th experimental moment. Temperature rise at that location This represents the temperature value at the location corresponding to the k-th experimental time. This represents the temperature value at the corresponding location in the background frame or at the initial time, where k represents the experimental frame or experimental time number. and These represent the lateral and longitudinal coordinates on the diagnostic target surface, respectively. This means that when the temperature difference is less than 0, the value is 0; otherwise, the actual temperature difference is used.

[0035] Step 3: Select a rough quadrilateral region of interest on the first experimental frame, perform edge detection and Hough line detection within the region, filter the main lines corresponding to the four sides of the rough quadrilateral, and obtain the refined four corner points from the intersection of adjacent main lines.

[0036] Step 4: Establish homography transformation based on the refined four corner points, correct the temperature rise field perspective to a rectangular region, and determine the horizontal and vertical pixel dimensions based on the actual physical width and height of the rectangular region; further, Step 4 includes: establishing a homography transformation matrix from the homography coordinates of the original image and the homography coordinates of the target rectangle:

[0037] ;

[0038] The non-zero scale factor representing homogeneous coordinates, Represents the homogeneous coordinates of a point in the original image. This represents the homogeneous coordinates of the corresponding point after perspective correction. This represents the homography transformation matrix mapping from the original image plane to the target rectangular plane. The horizontal and vertical physical resolutions of the corrected image are:

[0039] ;

[0040] in, and These represent the actual physical size resolution of the image after perspective correction in the horizontal and vertical directions, respectively. and These represent the actual physical width and actual physical height of the selected diagnostic target region, respectively. and These represent the number of pixels in the horizontal and vertical directions of the image after perspective correction, respectively.

[0041] Step 5: Based on the aperture array position of the particle beam extraction electrodes, the beam direction of each aperture, the inter-plate angle, and the deflection parameters of the partitioned apertures, establish a realistic multi-aperture beam forward model; using the candidate beam divergence angle as the single-aperture beam width parameter, generate a single-aperture two-dimensional Gaussian beam spot for each extraction aperture propagating to the diagnostic target surface, and weight and superimpose the N single-aperture Gaussian components to form a multi-Gaussian simulated beam field; further, Step 5 includes: calculating the center of the single-aperture beam spot formed on the diagnostic target surface after the i-th extraction aperture has a propagation distance D based on its local position and unit direction vector. Under paraxial conditions, this can be written as:

[0042] ;

[0043] in, and These represent the x and y coordinates of the center of the beam after propagation from the i-th aperture to the diagnostic target surface, respectively. and Let x and y represent the initial x and y coordinates of the i-th extraction hole, respectively, and D represent the propagation distance from the extraction hole to the diagnostic target surface. , and These represent the unit direction vector of the beam at the i-th aperture. , and The component in the direction, i represents the hole number.

[0044] The two aperture array plates are continuously connected at a common connection edge and slightly tilted to opposite sides relative to the global vertical direction. Due to the certain propagation distance between the diagnostic target and the extraction electrode, the beams from each aperture expand during propagation and form multiple spatial superpositions on the diagnostic target surface. At the same time, due to the limitations of the lateral installation space inside the beam source and the avoidance conditions of adjacent components, the folded surface arrangement allows the two sets of aperture arrays to converge back and forth along the beam propagation direction, retaining the effective extraction area of ​​the two sets of aperture arrays without increasing the lateral installation size, and enabling the two sets of beams to form the required coverage and superposition on the far-end diagnostic target surface.

[0045] For the i-th hole, first determine its local lateral coordinates. and the line number Determine the horizontal deflection angle and vertical deflection angle The central region is not deflected in the corresponding direction; holes on both sides of the plate width are deflected towards the center of the plate width, and the same horizontal deflection parameter is used in the same column; holes at the top and bottom of the plate height are deflected towards the center of the plate height, and the same vertical deflection parameter is used in the same row. The three components of its local direction vector are respectively... , and Determine and normalize, where Represents a symbolic function. Represents the tangent function. Indicates the first The horizontal deflection angle of each lead-out hole Indicates the first The vertical deflection angle of each lead-out hole Indicates the first The vertical deflection direction sign of each lead-out hole; in the specified... With the shaft pointing upwards, the upper hole is taken as follows: Take the lower end hole The central non-deflection region passed the order accomplish; The sign of the hole is opposite depending on whether it is located at the upper or lower end of the plate height; the direction vector of the second hole array plate is first mirrored with respect to the vertical component, and then a rotation transformation corresponding to the plate tilt angle is applied together with the hole coordinates.

[0046] In one implementation, the lateral center non-focus area covers |x i For diameters ≤32 mm, the outer holes are divided into six levels based on their distance from the center of the plate width, with horizontal deflection angles of 0.182973522 degrees, 0.219567711 degrees, 0.256161619 degrees, 0.292755199 degrees, 0.329348405 degrees, and 0.365941189 degrees respectively. The upper three rows and lower three rows of the plate height are each divided into three levels, with vertical deflection angles of 0.222240985 degrees, 0.277815183 degrees, and 0.333388561 degrees respectively, all pointing towards the center of the plate surface. The above values ​​illustrate the implementation method of the partitioned hole offset. In practical applications, the values ​​can be set according to the electrode design dimensions, propagation distance, and target focusing position.

[0047] For any candidate beam divergence angle θ, a single-aperture two-dimensional Gaussian model is established for each aperture based on its corresponding beam spread width. The equivalent width of the single aperture corresponding to the candidate beam divergence angle is:

[0048] ;

[0049] in, Indicates the candidate beam divergence angle as The transverse equivalent width of a single-aperture beam spot. Indicates the candidate beam divergence angle as The longitudinal equivalent width of a single-aperture beam spot. This represents the initial equivalent width of the single-aperture beam at the extraction position, and D represents the propagation distance from the extraction aperture to the diagnostic target surface. Indicates the candidate beam divergence angle. This represents the ratio between the vertical width and the horizontal width.

[0050] The multi-Gaussian simulated beam distributions obtained by superimposing the i-th single-aperture Gaussian beam spot with the N single-aperture Gaussian components are as follows:

[0051] ;

[0052] and The calculation formula is:

[0053] ;

[0054] ;

[0055] in, and , respectively, represent the standard deviations of the beam spot in the transverse and longitudinal directions, where, This represents the coordinates of the single-hole beam corresponding to the i-th extraction hole on the diagnostic target surface. Simulated intensity at the location, This indicates the coordinates of the diagnostic target surface after the superposition of all the outflow beams. The total simulated beam intensity at the location, and Let i be the center coordinates of the beam propagating from the i-th aperture to the diagnostic target surface. The weight is the weight of the i-th single-aperture Gaussian component, and the single-aperture beam spot is a two-dimensional Gaussian distribution; and The equivalent width parameters are lateral and longitudinal, and N is the total number of extraction holes participating in the superposition. The simulated beam distribution is the sum of the Gaussian components of N single holes with different spatial positions and propagation directions, rather than using a single Gaussian function to describe the overall beam. Exponent 2 corresponds to a normal two-dimensional Gaussian beam spot, N is the total number of extraction holes, and w i The weight is the i-th single aperture. Since the center position and propagation direction of each single aperture Gaussian component are determined by the real aperture array, the superposition result can retain the multi-peak, peak-valley, shoulder and transverse and longitudinal differences generated by the porous structure. Therefore, the multi-Gaussian simulated beam field is not a single Gaussian fit to the overall hot spot.

[0056] Step 6: Determine the alignment offset of the simulated field relative to the experimental field using the normalized cross-correlation between the reference multi-Gaussian simulated beam field and the experimental temperature rise field, and reuse the alignment offset in the inversion calculation of each candidate beam divergence angle.

[0057] Step 7: Determine the transverse and longitudinal banded regions based on the weighted centroid of the experimental temperature rise field. Using the same set of weights determined by the experimental temperature rise intensity, extract the transverse and longitudinal normalized curves from the experimental temperature rise field and the multiple Gaussian simulated beam field, respectively. Further, Step 7 includes: to reduce the sensitivity of single-row or single-column selection to noise and local anomalies, this invention determines the hotspot center based on the weighted centroid of the smoothed temperature rise field, forming a transverse profile band within a preset half-width range above and below the center, and a longitudinal profile band within a preset half-width range to the left and right of the center. The experimental temperature rise field and the multiple Gaussian simulated beam field use the same set of weights determined by the experimental temperature rise intensity to extract the actual transverse curve, actual longitudinal curve, simulated transverse curve, and simulated longitudinal curve within the same spatial range, respectively. The transverse and longitudinal weighted banded curves are respectively represented as:

[0058] ;

[0059] in, This represents the horizontal weighted profile value corresponding to column j. This represents the vertical weighted profile value corresponding to the q-th row; and These represent the image intensity values ​​at the corresponding pixel locations; and These represent the weights of the corresponding pixels when extracting the horizontal and vertical profiles, respectively; r represents the row index, c represents the column index, and q and j represent the current row number and column number, respectively.

[0060] Step 8: Compare the experimental curve and the simulated curve point by point under the same physical coordinate system, calculate the shape error, gradient error, and half-width at half-maximum (WHM) error of the transverse and longitudinal curves respectively, and combine them to obtain the composite objective function for beam divergence angle inversion; further, Step 8 includes: normalizing the actual curve and the simulated curve to zero to one respectively, and comparing them point by point under the same physical coordinate system. Construct curve shape error, gradient error, and relative WHM error for the transverse and longitudinal directions respectively, so that the inversion not only constrains the width near the peak, but also simultaneously constrains the attenuation on both sides of the curve, local peaks and valleys, and the overall trend of change. The three types of errors are dimensionless before combination, and non-negative, normalized coefficients are used for combination; the coefficients are determined by constrained optimization from known beam divergence angle calibration samples or repeated experimental data to minimize the inversion deviation and repeated measurement discreteness of the calibration conditions, and the weights in the transverse and longitudinal directions are determined according to the same principle:

[0061] ;

[0062] in, Represents the simulated curve. Indicates the experimental curve; and These represent the gradients of the simulated curve and the experimental curve, respectively. and These represent the half-width and full height of the simulated curve and the experimental curve, respectively. It is a preset minimum positive number with the same dimensions as the half-width to avoid the denominator being zero; This indicates that the average value is taken for each data point within the parentheses. The composite error of the actual-simulated curve in one direction and the joint fitting error in the horizontal and vertical directions are:

[0063] ;

[0064] in, This represents the composite error in a single direction. , and These represent shape error, gradient error, and half-width at half-maximum (WHM) error, respectively. , and These are the weighting coefficients corresponding to the three types of errors; This represents the fitting error of the horizontal and vertical curves. and These represent the combined errors in the horizontal and vertical directions, respectively. and These are the weighting coefficients for the horizontal and vertical errors, respectively;

[0065] Based on the joint fitting error between the actual curve and the simulated curve, a quadratic penalty term for the divergence angle deviating from the prior value is added:

[0066] ;

[0067] Indicates the total error. This represents the weighting coefficient of the prior constraint term for the bundle divergence angle. Indicates the candidate beam divergence angle. This represents the prior value of the beam divergence angle.

[0068] The candidate beam divergence angle is determined based on the criterion of minimizing the total error between the experimental curve and the simulated curve:

[0069] ;

[0070] in, This represents the final determined beam divergence angle. This indicates the preset candidate beam divergence angle search range. This represents the total error corresponding to the candidate beam divergence angle being . This indicates the search for the beam divergence angle that minimizes the total error.

[0071] Step 9: Within the preset beam divergence angle range, change the candidate beam divergence angles one by one. Each time the candidate beam divergence angle is changed, regenerate each single-aperture Gaussian beam spot, the multi-Gaussian simulated beam field, and its transverse and longitudinal simulated curves. First, perform a coarse step search, and then perform a local fine step search around the optimal value of the coarse search. The beam divergence angle corresponding to the minimum composite error between the experimental curve and the simulated curve is used as the inversion value. Further, Step 9 includes: within the candidate angle range, recalculate each single-aperture Gaussian beam spot for each candidate beam divergence angle, and superimpose all single-aperture components to obtain the corresponding multi-Gaussian simulated beam field. Then, use the same strip region and weight as the experimental field to extract the simulated curve, and calculate the total error between the simulated curve and the actual curve. First, perform a coarse step search to obtain the optimal coarse search angle, and then perform a fine step search in its vicinity, thus forming a closed-loop inversion process of "candidate beam divergence angle - multi-Gaussian simulated beam field - simulated curve - actual curve error - updated beam divergence angle".

[0072] Step 10: According to the correspondence between the Gaussian width parameter and the root mean square width parameter, convert the inverted model beam divergence angle into root mean square beam divergence angle as the final measurement result, and output the beam divergence angle time evolution results corresponding to multiple experimental frames.

[0073] like Figure 3 As shown, the present invention provides a method for reading and sorting experimental data files, including multiple file selection, file type filtering, natural number sorting, background frame identification, and experimental frame sequence generation. The processing steps are connected in sequence to obtain background frames and experimental data arranged in experimental order.

[0074] like Figure 4 As shown, the present invention provides an infrared temperature rise field perspective correction and physical scale calibration method. By establishing homography transformation between the four corner points of the original oblique region of interest and the four corner points of the target rectangular region, the original region is corrected into a rectangular region, and the horizontal and vertical physical resolutions are determined according to the actual physical width, height and image pixel count.

[0075] like Figure 5 As shown, the present invention provides a real dual-plate multi-hole electrode and a hole beam propagation model, including an upper electrode plate and a lower electrode plate arranged along a common connection edge. The two electrode plates are inclined in opposite directions relative to the global vertical direction, and each lead-out hole has a corresponding beam propagation direction, so that the beam of each hole propagates to the far-end diagnostic target surface.

[0076] like Figure 6As shown, this invention provides a method for real porous beam multi-Gaussian forward modeling. Based on the aperture positions and beam propagation directions of a real porous electrode, the position of each single-aperture beam on the diagnostic target surface is determined, and two-dimensional Gaussian beam spots are established for each aperture. The Gaussian components of each aperture are then superimposed to form a multi-Gaussian simulated beam field. Wherein, Indicates the candidate beam divergence angle as At that time, the first A single-aperture Gaussian beam spot in the coordinates of the diagnostic target surface The corresponding normalized quadratic exponent term; Figure 6 In It has the same physical meaning as the aforementioned expression for a single-aperture two-dimensional Gaussian beam spot.

[0077] like Figure 7 As shown, this invention provides a closed-loop inversion method for beam divergence angle based on the comparison of experimental curves and simulated curves. The method extracts transverse and longitudinal curves from the infrared temperature rise field and the multiple Gaussian simulated beam field corresponding to the candidate beam divergence angle, respectively. By comparing the experimental curves and the simulated curves and calculating the composite error, the candidate beam divergence angle is continuously updated according to the error results until the beam divergence angle corresponding to the minimum error is obtained.

[0078] like Figure 8 As shown, the present invention provides a method for determining the beam divergence angle based on coarse step size search and local fine step size search. First, a coarse step size search is performed within a preset beam divergence angle range. Then, a local fine step size search is performed around the region with smaller error obtained from the coarse search. The beam divergence angle corresponding to the point with the minimum total error is used as the inversion result at the current experimental moment, and the root mean square beam divergence angle is calculated.

[0079] The invention will be further described below with reference to the accompanying drawings. In this embodiment, the user selects one frame of background data and multiple experimental frames at a time, supporting csv, xlsx, xls, txt, dat, and common image formats. The program sorts the data naturally by the numbers in the filenames, prioritizes identifying background identifiers, and then reads the first experimental frame to establish a unified region of interest.

[0080] In this embodiment, the single-hole array is arranged in 13 rows, with the number of holes in each row being 5, 10, 11, 10, 11, 10, 11, 10, 11, 10, 11, 10, 11 respectively; the horizontal hole spacing is 16 mm, and the staggered adjacent row spacing is 17 mm. The second hole array is obtained by mirroring the first hole array relative to the common connection edge, and the two plates form a 262-hole structure; the two plates are continuously connected at the common connection edge and are tilted 0.94 degrees to opposite sides relative to the global vertical direction, forming a symmetrical folded surface of 1.88 degrees. The arrangement is used to retain the effective area of ​​the two sets of hole arrays within the limited horizontal installation space, and to allow the single-hole beam after propagation over a certain distance to form multiple coverages on the diagnostic target surface. The hole offset is determined according to the above-mentioned horizontal and vertical partitioning rules, with no deflection applied in the corresponding direction of the central region, the holes on the outer side of the plate width deflected towards the horizontal center, and the holes at the upper and lower ends of the plate height deflected towards the vertical center. Thus, each aperture has a defined initial coordinate and unit direction vector, forming a single-aperture Gaussian component on the diagnostic target surface corresponding to its spatial position and propagation direction.

[0081] In this embodiment, the propagation distance from the aperture array to the diagnostic target is 2677.5 mm. The single-aperture beam spot directly adopts the ordinary two-dimensional Gaussian form without introducing thermal diffusion smoothing and lateral scale correction. For any beam divergence angle to be measured, the program sequentially calculates the beam spot center and width of 262 apertures on the diagnostic target surface, generating a single-aperture two-dimensional Gaussian beam spot for each aperture. Then, the 262 single-aperture Gaussian components are added point by point to form a multi-Gaussian simulated beam field. The double peaks, local peaks and valleys, and shoulders in this simulated field are naturally generated by the superposition of the actual aperture positions and beam directions, and are not obtained by setting additional peaks or artificially modifying the peak shape of the overall experimental curve. The candidate beam divergence angle coarse scan range is 0.75 degrees to 1.30 degrees, with a step size of 0.05 degrees; fine scan is performed around the optimal coarse scan value within a range of ±0.08 degrees, with a fine scan step size of 0.015 degrees.

[0082] In this embodiment, the half-width of both the horizontal and vertical profile bands is set to 10 pixels, the low-temperature weight threshold ratio is set to 0.25, and the weight power is set to 1.2. After determining the hot spot center and pixel weights from the experimental temperature rise field, the experimental field and the multiple Gaussian simulated beam field both use the same horizontal band, the same vertical band, and the same weights to obtain the actual X curve, the actual Y curve, the simulated X curve, and the simulated Y curve, respectively. This embodiment does not limit the specific values ​​of the weights for shape error, gradient error, half-width error, and horizontal and vertical errors; after normalizing each error term, optimization is performed based on the known beam divergence angle calibration conditions or repeated experimental data, under the constraint that the weights are non-negative and their sum is 1, so as to minimize the beam divergence angle inversion deviation and the discreteness of repeated measurements. The prior value of the beam divergence angle and its penalty coefficient are determined based on the equipment design parameters and historical experimental statistical results.

[0083] During the inversion process, for each candidate beam divergence angle tested, 262 single-aperture Gaussian beam spots and their multiple superposition fields were regenerated. The corresponding simulated X and Y curves were extracted, and the composite error was calculated by comparing them with the actual X and Y curves of the same frame. The candidate beam divergence angle corresponding to the lowest point of the error curve is the model inversion value for that frame. The model inversion value is converted to the root mean square beam divergence angle by dividing by √2 according to the definition of Gaussian width. The root mean square beam divergence angle, fitting error, and quality label are output for each experimental frame and written into the results table. The time axis is set according to the actual time interval between adjacent experimental frames to obtain the variation of beam divergence angle with time after bombardment.

[0084] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection. Any equivalent substitutions made by those skilled in the art, without departing from the inventive principle of "single-hole Gaussian modeling—multiple-hole superposition—closed-loop inversion of simulated curves and actual curves," regarding the number of hole arrays, hole spacing, deflection parameters, single-hole weights, search range, error weights, or image processing parameters, should fall within the scope of protection of the present invention.

Claims

1. A method for measuring the beam divergence angle of high-current particle beams based on an infrared diagnostic target, characterized in that, Includes the following steps: Step 1: Select one frame of background data and multiple experimental data from different times, sort the selected data files naturally according to the numbers in the filenames, and identify the background frame; Step 2: Subtract the temperature matrix of each experimental frame from the temperature matrix of the background frame pixel by pixel, and set the negative values ​​in the difference to zero to obtain the temperature rise field at each time. Step 3: Select a rough quadrilateral region of interest on the first experimental frame, perform edge detection and line detection within the region, filter the main lines corresponding to the four sides of the rough quadrilateral, and obtain the refined four corner points from the intersection of adjacent main lines; Step 4: Establish homography transformation based on the four corner points of the refinement, correct the perspective of the temperature rise field into a rectangular area, and determine the horizontal and vertical pixel dimensions based on the actual physical width and actual physical height of the rectangular area; Step 5: Establish a real multi-hole beam forward model based on the hole array position of the particle beam extraction electrode, the beam direction of each hole, the inter-plate angle, and the deflection parameters of the partitioned holes; use the candidate beam divergence angle as the single-hole beam spot width parameter to generate a single-hole two-dimensional Gaussian beam spot that propagates from each extraction hole to the diagnostic target surface, and weight and superimpose the N single-hole Gaussian components to form a multi-Gaussian simulated beam field; Step 6: Determine the alignment offset of the simulated field relative to the experimental field using the normalized cross-correlation between the multiple Gaussian simulated beam field and the temperature rise field, and reuse the alignment offset in the inversion calculation of each candidate beam divergence angle. Step 7: Determine the transverse and longitudinal banded regions based on the weighted centroid of the temperature rise field. Using the same set of weights determined by the experimental temperature rise intensity, extract the transverse and longitudinal normalized experimental and simulated curves from the experimental temperature rise field and the multiple Gaussian simulated beam field, respectively. Step 8: Compare the experimental curve and the simulated curve point by point under the same physical coordinates, calculate the shape error, gradient error and half width at half maximum (WHM) error of the transverse and longitudinal curves respectively, and combine them to obtain the composite objective function for beam divergence angle inversion; Step 9: Change the candidate beam divergence angle one by one within the preset beam divergence angle range. Each time the candidate beam divergence angle is changed, regenerate each single-hole Gaussian beam spot, the multi-Gaussian simulated beam field and its transverse and longitudinal simulated curves. First, perform a coarse step search, and then perform a local fine step search around the optimal value of the coarse search. The beam divergence angle corresponding to the minimum composite error of the experimental curve and the simulated curve is used as the inversion value. Step 10: According to the correspondence between the Gaussian width parameter and the root mean square width parameter, convert the inverted model beam divergence angle into root mean square beam divergence angle as the final measurement result, and output the beam divergence angle time evolution results corresponding to multiple experimental frames.

2. The method according to claim 1, characterized in that, Files whose filenames contain the 0-frame identifier are prioritized for identification.

3. The method according to claim 1, characterized in that, The process of obtaining the refined four corner points includes: performing Gaussian smoothing and edge detection on the image within the region of interest, performing lateral and longitudinal morphological dilation on the edge image, and performing Hough transform to obtain candidate lines; constructing a screening score based on the angle difference, midpoint distance, and line segment length between the candidate lines and the corresponding sides of the rough quadrilateral, selecting the main lines on the top, right, bottom, and left sides respectively, and calculating the intersection points of adjacent main lines.

4. The method according to claim 1, characterized in that, The perspective correction establishes a projection transformation matrix between the four corner points of the original image and the four corner points of the target rectangle, and uses bilinear interpolation to generate the corrected temperature rise field; the horizontal pixel size is equal to the actual physical width divided by the number of columns in the corrected image, and the vertical pixel size is equal to the actual physical height divided by the number of rows in the corrected image.

5. The method according to claim 1, characterized in that, The multi-aperture beam forward model consists of two aperture array plates connected along a common edge and tilted in opposite directions. The second aperture array plate is obtained by mirroring the first aperture array plate. The two aperture array plates are arranged to adapt to the limited installation space and to make the two sets of aperture beams cover and overlap on the far-end diagnostic target surface. The beam direction of each lead-out aperture is determined according to the position of the aperture and the tilt angle of the plate surface, so that the outer aperture beams are deflected towards the array center, thereby obtaining the propagation center of each single aperture beam on the diagnostic target surface.

6. The method according to claim 5, characterized in that, The i-th exit aperture forms an independent single-aperture two-dimensional Gaussian beam spot. The weighted superposition of all single-aperture beam spots forms a multiple Gaussian simulated beam distribution, which respectively satisfy the following equations: ； in, This represents the coordinates of the single-hole beam corresponding to the i-th extraction hole on the diagnostic target surface. Simulated intensity at the location, This indicates the coordinates of the diagnostic target surface after the superposition of all the outflow beams. The total simulated beam intensity at the location, and Let i be the center coordinates of the beam propagating from the i-th aperture to the diagnostic target surface. The weight is the weight of the i-th single-aperture Gaussian component, and the single-aperture beam spot is a two-dimensional Gaussian distribution; and The equivalent width parameters are lateral and longitudinal, and N is the total number of extraction holes participating in the superposition. The simulated beam distribution is the sum of the Gaussian components of N single holes with different spatial positions and propagation directions, rather than using a single Gaussian function to describe the overall beam.

7. The method according to claim 1, characterized in that, The experimental curves and simulated curves are extracted using the same horizontal and vertical strip regions and pixel weights. The weights are determined by the normalized experimental temperature rise field. Pixel weights that are lower than the maximum temperature rise of the strip region by a preset ratio are reset to zero, and the weights of the remaining pixels are calculated according to a preset power, so that the two types of curves have the same spatial sampling conditions.

8. The method according to claim 1, characterized in that, The single-direction curve error is formed by the combination of shape error, gradient error, and full width at half maximum (FWHM) error between the experimental and simulated curves. Each error term is dimensionless before combination, and the weighting coefficients for the three types of errors are non-negative and normalized, as are the weights for the lateral and longitudinal directions. These weighting coefficients are determined based on calibration data or repeated measurement data of the known beam divergence angle, with the goal of minimizing inversion bias and result dispersion. The total error is the sum of the fitting error and the penalty term for the beam divergence angle deviating from the prior value. ； ； in, This represents the composite error in a single direction. , and These represent shape error, gradient error, and half-width at half-maximum (WHM) error, respectively. , and These are the weighting coefficients corresponding to the three types of errors; This represents the fitting error of the horizontal and vertical curves. and These represent the combined errors in the horizontal and vertical directions, respectively. and These are the weighting coefficients for the horizontal and vertical errors, respectively; Indicates the total error. This represents the weighting coefficient of the prior constraint term for the bundle divergence angle. Indicates the candidate beam divergence angle. This represents the prior value of the beam divergence angle.

9. The method according to claim 6, characterized in that, The single-aperture two-dimensional Gaussian beam spot is represented by a Gaussian width parameter, and its intensity distribution satisfies the following equation: ； and The calculation formula is: ； ； in, and The standard deviations of the beam spot in the transverse and longitudinal directions are respectively used. For each candidate beam divergence angle, all Gaussian components of the single aperture are recalculated and multiple superpositions are performed to extract the corresponding simulation curve. The candidate beam divergence angle is determined by minimizing the total error between the experimental curve and the simulation curve. ； in, This represents the final determined beam divergence angle. This indicates the preset candidate beam divergence angle search range. This represents the total error corresponding to the candidate beam divergence angle being . This indicates the search for the beam divergence angle that minimizes the total error.

Images