Cosmic ray muon distribution three-dimensional reconstruction method and system

Abstract

The invention relates to the technical field of muon imaging, in particular to a cosmic ray muon distribution three-dimensional reconstruction method and system.The method comprises the steps that a monitoring environment is built, an object to be detected is placed at the circle center of an annular guide rail, two lifting platforms are fixed to the annular guide rail, supporting frames are installed on the lifting platforms respectively, and detector panels are arranged on the supporting frames; detecting position information left when muons penetrate through a detector panel, calculating muon tracks and muon flux to obtain angular distribution, and establishing point-surface images of the angular distribution and flux distribution of the muons based on the muon tracks; performing weighted summation to obtain a face-to-face image; and compressing Monte Carlo output muon position information by using an auto-encoder, obtaining momentum, time and position information of muons, and extracting angular distribution in a preset area for classifying the angular distribution and the position information according to the interval. The method is suitable for various use scenes, belongs to passive detection, does not generate radioactive contamination, and can replace industrial CT with radioactive contamination in a specific scene.

Description

Technical Field

[0001] The present invention relates to the field of muon detection technology, and in particular to a method and system for three-dimensional reconstruction of cosmic ray muon distribution. Background Art

[0002] In many industrial monitoring and inspection fields, accurately acquiring internal structural information of the object being tested and achieving real-time and safe monitoring are crucial. Currently, there are various technical means for internal inspection and monitoring of objects in different application scenarios, but all have certain limitations.

[0003] In the field of industrial inspection, industrial CT, as a commonly used inspection method, can obtain three-dimensional internal structural information of objects and plays a vital role in product quality control and defect detection. However, industrial CT uses an active detection method, emitting radioactive radiation such as X-rays to scan and image the object being inspected. This process generates radioactive contamination, posing potential hazards to operators and the environment. Therefore, it requires complex shielding and protection facilities, increasing equipment cost and operational difficulty. Furthermore, it can only be used in specific, relatively closed environments, limiting its scope of application and flexibility. Summary of the Invention

[0004] In order to solve the technical problems existing in the above background technology, the present invention provides a.

[0005] The technical solutions of the present invention are as follows:

[0006] A method for three-dimensional reconstruction of cosmic ray muon distribution comprises the following steps:

[0007] S1. Set up the monitoring environment. Place the object to be measured at the center of the circular guide rail. Fix the lifting platform on the circular guide rail. Install a support frame on each lifting platform. Each support frame is equipped with a set of detector panels. Each set of detector panels has three layers, numbered from bottom to top. Establish a coordinate system with the lower left corner of the bottom detector panel as the origin.

[0008] S2. Each layer of detector panels consists of two layers of sub-panels, each with embedded scintillator elements perpendicular to each other. The two sub-panels detect the position information left by muons as they pass through the detector panels, retain the position information recorded simultaneously on the three layers of detector panels, and calculate the muon trajectory on each set of detector panels.

[0009] S3. Calculate the muon flux and the angle between the muon trajectory and the plane of each two coordinate axes on each set of detector panels to obtain the muon angular distribution on each set of detector panels. Based on the muon trajectory, establish a point-to-surface image of the muon angular distribution and flux distribution.

[0010] S4, multiplying all point-surface images by weight coefficients and then superimposing them to obtain line-surface images;

[0011] S5, multiplying all line-surface images by weight coefficients and then superimposing them to obtain surface images;

[0012] S6. Complete the three-dimensional reconstruction of the object to be measured based on the surface images and position information on each set of detector panels.

[0013] The specific method of establishing the coordinate system is: the horizontal direction is the y-axis, the direction perpendicular to the bottom detection panel is the z-axis, and the right-hand rule is used. Determine the x-axis.

[0014] The position information on the first layer detector panel is (x1, y1, z1) = (x1 ， y1, 0), the position information on the second layer detector panel is (x2, y2, z2) = (x2, y2, D), and the position information on the third layer detector panel is (x3, y3, z3) = (x3, y3, 2D), where D is the distance between two adjacent layers of detector panels.

[0015] The position information on the three-layer detector panel is retained and recorded simultaneously, specifically:

[0016] If the following conditions are met, the position information on the three-layer detector panel will be retained and recorded simultaneously:

[0017] x1*y1*x2*y2*x3*y3≠0.

[0018] The specific method for calculating muon trajectories is:

[0019]

[0020] The position information on the first layer detector panel is (x1, y1, z1), and the position information on the second layer detector panel is (x2, y2, z2).

[0021] The specific steps to obtain the angular distribution are:

[0022]

[0023] in, is the vector representation of the muon trajectory, is the vector representation of the plane where each two coordinate axes are located, α is the angle between the side of the support frame and the horizontal plane, that is, the zenith angle. Since the lifting platform is fixed on the circular guide rail, the azimuth angle is 0-360°. Each zenith angle and azimuth angle determines a muon trajectory, and then the muon angle distribution with respect to the zenith angle and azimuth angle in the spherical coordinate system is obtained.

[0024] The specific operation of S6 can be: based on the surface images on each set of detector panels, use the autoencoder to compress the Monte Carlo simulation of the muon angular distribution, obtain the momentum, time, and position information of the muon, output the surface images and position information of the muon before and after passing through the object to be measured, and complete the three-dimensional reconstruction of the object to be measured based on the surface images and position information of the muon before and after passing through the object to be measured.

[0025] A cosmic ray muon distribution three-dimensional reconstruction system, comprising:

[0026] Annular guide rails for carrying the lifting platform;

[0027] Lifting platform, two lifting platforms are fixed on the circular guide rail;

[0028] The muon detection equipment includes a support frame and detector panels. A support frame is installed on each lifting platform. Each support frame is equipped with a set of detector panels. A set of detector panels is arranged in three layers and numbered from bottom to top. The coordinate system is established with the lower left corner of the bottom detector panel as the origin. Each layer of detector panels contains two levels of panels, each of which is embedded with plastic scintillator strips or plastic scintillator optical fibers in a vertical direction. The two levels of panels detect the position information left by muons when they pass through the detector panels, and retain and record the position information on the three layers of detector panels at the same time.

[0029] A trajectory calculation module calculates the muon trajectory on each set of detector panels based on the position information on each set of detector panels;

[0030] The point-surface image acquisition module calculates the muon flux on each set of detector panels and the angle between the muon trajectory and the plane containing each two coordinate axes, obtains the muon angular distribution on each set of detector panels, and builds point-surface images of the muon angular distribution and flux distribution based on the muon trajectory.

[0031] The line-surface image acquisition module multiplies all point-surface images by weight coefficients and then superimposes them to obtain line-surface images;

[0032] The surface image acquisition module multiplies all line-surface images by weight coefficients and then superimposes them to obtain the surface image;

[0033] The three-dimensional simulation module uses an autoencoder to compress Monte Carlo simulation of the muon angular distribution based on the surface images on each set of detector panels, obtains the momentum, time, and position information of the muons, and outputs the surface images and position information of the muons before and after passing through the object to be tested. Based on the surface images and position information of the muons before and after passing through the object to be tested, it completes the three-dimensional reconstruction of the object to be tested, extracts the angular distribution within the preset area, classifies the angular distribution and position information according to the interval they are in, and infers the information of the material to be tested.

[0034] The support bracket is used to adjust the zenith angle direction of the detector panel.

[0035] A photomultiplier tube is installed at the front end of the detector panel to read the position information of the muon, convert the light signal into an electrical signal and record it.

[0036] The beneficial effects of the present invention are:

[0037] The present invention can replace industrial CT in specific scenarios, because industrial CT is an active detection method that will produce radioactive contamination and require shielding protection. The present invention, on the other hand, is a passive detection method that utilizes natural muons, does not actively produce radioactive rays, does not require shielding protection, and can be used in an open environment.

[0038] The method of the present invention is applicable to a variety of usage scenarios, including but not limited to real-time monitoring of nuclear waste, as well as monitoring of columnar bridge piers and industrial pipelines.

[0039] The present invention can perform a three-dimensional angle adjustment function when monitoring an object to be measured, and can realize imaging of the object to be measured at multiple angles, thereby achieving the purpose of three-dimensional imaging. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In the attached figure:

[0041] Figure 1 Schematic diagram of the imaging effect and point-surface image of the object to be measured;

[0042] Figure 2 Schematic diagram of the imaging effect and line-surface image of the object to be measured;

[0043] Figure 3 It is a schematic diagram of the imaging effect and surface image of the object to be measured;

[0044] Figure 4 This is a structural diagram of the muon detection equipment.

[0045] The components represented by the reference numerals in the figure are:

[0046] 1. Detector panel; 2. Support frame. DETAILED DESCRIPTION

[0047] The technical solutions of the present invention are as follows:

[0048] A method for three-dimensional reconstruction of cosmic ray muon distribution comprises the following steps:

[0049] S1. Build a monitoring environment, place the object to be measured at the center of the circular guide rail, fix the lifting platform on the circular guide rail, install a support frame 2 on each lifting platform, and each support frame 2 is provided with a group of detector panels 1. Each group of detector panels 1 is provided with three layers, which are numbered from bottom to top. The coordinate system is established with the lower left corner of the bottom detector panel 1 as the origin.

[0050] The coordinate system takes the horizontal direction as the y-axis and the direction perpendicular to the bottom detection panel as the z-axis. According to the right-hand rule Compared with the traditional coordinate system establishment method, the present invention uses the right-hand rule to determine the x-axis, which has a wider application range and is more accurate.

[0051] The lifting platform can rotate on the circular guide rail and is fixed by the pulley. Compared with the starting position, it can rotate 360° around the center of the circular guide rail. The central angle formed is the azimuth angle φ. The lifting platform e can be lifted and lowered in the z-axis direction.

[0052] The detector panel 1 is a square plate, each layer is arranged in parallel, and the support frame 2 includes four telescopic rods, which are sequentially connected to the four corners of each layer of the detector panel 1. By adjusting the support frame 2, the zenith angle θ of the detector panel 1 can be changed, and the adjustable range is greater than 60°.

[0053] During use, the circular guide rail's placement should be determined based on the monitoring site's operating conditions to ensure the detector panel 1's field of view fully covers the object being monitored. The detector panel 1 should not be placed too close to the object being measured, as this reduces the field of view and affects detection accuracy. The circular guide rail stabilizes the distance between the detector panel 1 and the object being measured, ensuring a fixed relative position.

[0054] After the annular guide rail is arranged, the lifting platform is fixed on the annular guide rail, and the height of the lifting platform is adjusted to ensure that the detector panel 1 can observe the monitored object from the top and bottom directions.

[0055] The muon support frame 2 is installed on the lifting platform, and the relative height of the detector panel 1 is adjusted to ensure that the field of view of the detector panel 1 is obliquely upward and the zenith angle is not greater than 50°, so as to ensure that the muon flux is large enough and improve the detection efficiency.

[0056] The above installation steps should be coordinated to ensure that the object to be detected is within the field of view of the detector panel 1. If adjustment is required, use the circular guide rail to adjust the φ direction angle, use the lifting platform to adjust the z-axis height, and use the support frame 2 to adjust the θ direction angle.

[0057] The instrument and equipment of the present invention can be used for real-time monitoring of nuclear waste before, during and after transportation. The lifting platform, the detector panel 1 and the support frame 2 constitute a set of detection equipment. The present invention uses two sets of detection equipment to complete the monitoring of the object to be measured. The field of view directions of the two sets of detection equipment must be consistent. Muon scattering imaging is used to determine the internal structure of the object to be measured to achieve the purpose of real-time monitoring.

[0058] S2. Each layer of detector panels 1 contains two layers of level panels, each of which has scintillator elements embedded in it. The two level panels are embedded in perpendicular directions to each other. The position information left by muons when they pass through the detector panels 1 is detected, the position information recorded simultaneously on the three layers of detector panels 1 is retained, and the muon trajectory on each group of detector panels 1 is calculated.

[0059] The scintillator element is a plastic scintillator bar or a plastic scintillator fiber.

[0060] The general process of muon imaging is that muons pass through one detector panel (1), leaving behind a set of muon position information, then pass through the object being measured, and continue through another detector panel (1), leaving behind another set of muon position information. The two sets of muon position information are different, depending on the object being measured. Therefore, by comparing the muon information measured by the two detectors, information about the object being measured can be inferred, which is how the image is formed.

[0061] Furthermore, the secondary panel is composed of plastic scintillator strips or plastic scintillator optical fibers wrapped in a reflective layer in the most compact arrangement. The scintillators in the two layers of the secondary panel are perpendicular to each other.

[0062] The muon signal is introduced by an electronic circuit and connected to the detector panel 1. The front end of the plastic scintillator strip or plastic scintillator optical fiber is a photomultiplier tube. The two are coupled to convert the optical signal into an electrical signal to perform data selection, that is, the data in the signal that is less than the data that can be recorded simultaneously on two layers of detection panels is discarded, and the position information on the three layers of detector panels 1 is retained to save storage space and improve data processing efficiency. The interaction between muons and plastic scintillator strips or plastic scintillator optical fibers is used to emit light, and the optical signal is collected by the photomultiplier tube and converted into an electrical signal to record the position information of the muons. The whole process is automatically realized.

[0063] The distance between each layer of detector panels 1 is known, so the position information of the original data only records the values of the x and y axes. The original data of the three layers of detector panels 1 are (x1, y1), (x2, y2) and (x3, y3). Combined with the layer of the detector panel 1, the position information on the first layer of detector panel 1 is (x1, y1, z1) = (x1, y1, 0), the position information on the second layer of detector panel 1 is (x2, y2, z2) = (x2, y2, D), and the position information on the third layer of detector panel 1 is (x3, y3, z3) = (x3, y3, 2D), where D is the distance between two adjacent layers of detector panels. (x1, y1), (x2, y2) and (x3, y3) are taken as a row and used Determine whether there are 6 complete data lines by line, and retain and record the position information on the three-layer detector panel 1, specifically:

[0064] Eligible if:

[0065] x1*y1*x2*y2*x3*y3≠0.

[0066] To limit the saving conditions, the data is stored in rows, which makes it easier to use for other operations in the future and saves storage space.

[0067] If you do not meet the conditions, that is:

[0068] x1*y1*x2*y2*x3*y3=0,

[0069] This indicates that the position information is not fully recorded, and is judged as a "false" signal. The data in this row is discarded, and then the judgment statement is returned to continue judging the data in the next row. The above steps are repeated until all muon position information that meets the conditions is selected.

[0070] Using two points in the mathematical thought space to determine a straight line, we can select the muon signal that satisfies the straight line equation. The specific method for calculating the muon trajectory is as follows:

[0071]

[0072] The position information on the first layer detector panel 1 is (x1, y1, z1), and the position information on the second layer detector panel 1 is (x2, y2, z2).

[0073] Only the position information on the two-layer detector panel 1 is used to obtain the muon trajectory, and the position information of the third layer is substituted back into the muon trajectory to determine whether the position information of the third layer is on the muon trajectory. If the error exceeds the error threshold, it is not, and the position information is reselected to obtain the muon trajectory.

[0074] Because muons move in straight lines in space, muon trajectories that meet the error requirements are determined to be muon trajectories.

[0075] S3. Calculate the muon flux on each set of detector panels and the angle between the muon trajectory and the plane of each two coordinate axes to obtain the muon angular distribution on each set of detector panels. Based on the muon trajectory, establish a point-surface image of the muon angular distribution and flux distribution.

[0076] The muon flux is calculated as:

[0077]

[0078] Among them E μ represents the muon energy, cosθ * represents the muon zenith angle distribution function. The first term in the brackets represents the decay contribution of pions in the atmosphere, and the second term represents the decay contribution of kaons in the atmosphere.

[0079] The specific steps to obtain the angular distribution are:

[0080]

[0081] in, is the vector representation of the muon trajectory, is the vector representation of the plane where each two coordinate axes are located, α is the angle between the side of the support frame 2 and the horizontal plane, that is, the zenith angle. Since the lifting platform is fixed on the circular guide rail, the azimuth angle is 0-360°. Each zenith angle and azimuth angle determines a muon trajectory, and then the muon angle distribution with respect to the zenith angle and azimuth angle in the spherical coordinate system is obtained.

[0082] The angle between the calculated muon trajectory and the plane is used to establish the muon angular distribution and flux distribution images. The two are usually expressed as The function is denoted as P i , where i represents a pixel marker, is called a point-to-surface perspective map, consistent with industrial CT imaging. For muon imaging, this solution has a single perspective, insufficient data statistics, and low image clarity. Therefore, the following processing is required for the point-to-surface image.

[0083] S4. Multiply all point-surface images by weight coefficients and then superimpose them to obtain line-surface images.

[0084] S5. Multiply all line-surface images by weight coefficients and then superimpose them to obtain surface images.

[0085] Point refers to the pixel point in the detector panel 1, which is actually a small area, similar to the small squares on the Go board, such as Figure 1 As shown, the right side is the point surface image, and the left side is the imaging effect of the object to be measured; the line refers to the line formed by connecting the pixel points, which is similar to the strip surface formed by connecting the small squares in the same direction on the Go board, such as Figure 2 As shown, the right side is the line surface image, and the left side is the imaging effect of the object to be measured; the surface is the surface composed of all the points on the plane, similar to the combination of small squares on the Go board, which has nothing to do with energy, such as Figure 3 As shown in the figure, the right side is the surface image, and the left side is the imaging effect of the object to be measured. What is measured at the pixel point is the number of muons passing through the pixel, or flux, which is equal to the number per unit area per unit time.

[0086] The points in the point-surface image are the pixels left by muons passing through the bottom detector panel 1, because muons move from top to bottom.

[0087] The point-surface image P i Expanded to line-surface image P L =∑ i ∈ i *P i , specifically, for all point-surface images Pi Weighted summation, the weight factor is ∈ i , determined by the pixel coordinates,∈ i ∝x i *y i The summation method is superposition, the point surface image is multiplied by the weight coefficient and then superimposed, and the line surface image is recorded as P L .

[0088] Repeat the above operation to convert the line-surface image P L Expanded to face image P s =∑ L∈L *P L , represents all line-surface images P L Weighted summation, the weight factor is ∈ L , determined by the pixel coordinates,∈ L ∝x L *y L The summation method is superposition, the line-surface image is multiplied by the weight coefficient and then superimposed, and the surface image is recorded as P S .

[0089] S6. Complete the three-dimensional reconstruction of the object to be measured based on the surface images and position information on each group of detector panels 1.

[0090] During the execution of the above method, there may be a problem of insufficient number of muons collected in S2, which will affect the accuracy of three-dimensional reconstruction. The above method can be further optimized by increasing the amount of data. For example, Geant4 software can be used to perform Monte Carlo simulation of the interaction process between muons and matter in nature, and an autoencoder can be used to compress the Monte Carlo simulation of the muon angular distribution to obtain the momentum, time, and position information of the muons. The surface images and position information of the muons before and after passing through the object to be measured are output, and the three-dimensional reconstruction of the object to be measured is completed based on the surface images and position information of the muons before and after passing through the object to be measured.

[0091] It is also possible to extract the angular distribution within a preset area, classify the angular distribution and position information according to the interval, and infer the information of the substance to be tested.

[0092] The operation can be achieved through the following steps: Using Geant4 software, a Monte Carlo simulation of the interaction between muons and matter in nature is performed, combining surface images and position information. First, an autoencoder (AE) is used to compress the Monte Carlo output position information, retaining the key features acquired: momentum, time, and position information. Then, a principal component analysis (PCA) algorithm coupled with a convolutional neural network (CNN) is used to extract the main patterns in the angular distribution. This is used to classify the angular distribution and coordinates, allowing for rapid classification based on intervals.

[0093] The most important goal of shortening experimental time and improving efficiency is to increase statistics, or proportionally increase the muon flux per pixel, rather than process refinement. Deep learning is used to optimize data, for example, when actual detectors are inefficient or exhibit excessive errors. Deep learning can improve efficiency and reduce errors.

[0094] A cosmic ray muon distribution three-dimensional reconstruction system, comprising:

[0095] Annular guide rails for carrying the lifting platform;

[0096] Lifting platform, two lifting platforms are fixed on the circular guide rail;

[0097] Muon detection equipment, including a support frame 2 and a detector panel 1, refer to Figure 4 , a support frame 2 is installed on each lifting platform, and a group of detector panels 1 is set on each support frame 2. A group of detector panels 1 is provided with three layers, which are numbered from bottom to top. The coordinate system is established with the lower left corner of the bottom detector panel 1 as the origin; each layer of detector panel 1 includes two levels of panels, and the two levels of panels are respectively embedded with plastic scintillator strips or plastic scintillator optical fibers, and the embedding direction is vertical. The position information left by the muon when it passes through the detector panel 1 is detected, and the position information on the three layers of detector panels 1 is retained and recorded simultaneously;

[0098] A trajectory calculation module calculates the muon trajectory on each set of detector panels 1 according to the position information on each set of detector panels 1;

[0099] The point-surface image acquisition module calculates the muon flux and the angle between the muon trajectory and the plane of each two coordinate axes on each set of detector panels 1, obtains the muon angular distribution on each set of detector panels 1, and establishes point-surface images of the muon angular distribution and flux distribution based on the muon trajectory;

[0100] The line-surface image acquisition module multiplies all point-surface images by weight coefficients and then superimposes them to obtain line-surface images;

[0101] The surface image acquisition module multiplies all line-surface images by weight coefficients and then superimposes them to obtain the surface image;

[0102] The three-dimensional simulation module uses an autoencoder to compress Monte Carlo simulation of the muon angular distribution based on the surface images on each group of detector panels 1, obtains the momentum, time, and position information of the muons, and outputs the surface images and position information of the muons before and after passing through the object to be tested. Based on the surface images and position information of the muons before and after passing through the object to be tested, the three-dimensional reconstruction of the object to be tested is completed, the angular distribution within the preset area is extracted, the angular distribution and position information are classified according to the interval in which they are located, and the information of the substance to be tested is inferred.

[0103] The support frame 2 is used to adjust the zenith angle direction of the detector panel 1.

[0104] A photomultiplier tube is provided at the front end of the detector panel 1, which is used to read the position information of the muons, convert the light signal into an electrical signal and record it.

Claims

1. A three-dimensional reconstruction method of cosmic ray muon distribution, characterized in that: The following steps are involved: S1. Build a monitoring environment, place the object to be measured at the center of the circular guide rail, fix the lifting platform on the circular guide rail, install a support frame (2) on each lifting platform, and each support frame (2) is provided with a group of detector panels (1). Each group of detector panels (1) is provided with three layers, which are numbered from bottom to top, and a coordinate system is established with the lower left corner of the lowest detector panel (1) as the origin; S2. Each layer of detector panels (1) includes two layers of level panels, each of which has scintillator elements embedded therein, and the embedding directions are perpendicular to each other, detecting the position information left by muons when they pass through the detector panels (1), retaining the position information recorded simultaneously on the three layers of detector panels (1), and calculating the muon trajectory on each set of detector panels (1); S3, calculating the muon flux on each set of detector panels (1) and the angle between the muon trajectory and the plane where each two coordinate axes are located, obtaining the muon angular distribution on each set of detector panels (1), and establishing a point-surface image of the muon angular distribution and flux distribution based on the muon trajectory; S4, multiplying all point-surface images by weight coefficients and then superimposing them to obtain line-surface images; S5, multiplying all line-surface images by weight coefficients and then superimposing them to obtain surface images; S6. Complete the three-dimensional reconstruction of the object to be measured based on the surface images and position information on each set of detector panels (1).

2. The method for three-dimensional reconstruction of cosmic ray muon distribution according to claim 1, characterized in that: The specific method of establishing the coordinate system is: the horizontal direction is the y-axis, the direction perpendicular to the bottom detection panel is the z-axis, and the right-hand rule is used. Determine the x-axis.

3. The method for three-dimensional reconstruction of cosmic ray muon distribution according to claim 1, characterized in that: The position information on the first layer of detector panels is (x1, y1, z1) = (x1, y1, 0), the position information on the second layer of detector panels is (x2, y2, z2) = (x2, y2, D), and the position information on the third layer of detector panels is (x3, y3, z3) = (x3, y3, 2D), where D is the distance between two adjacent layers of detector panels.

4. The method for three-dimensional reconstruction of cosmic ray muon distribution according to claim 3, characterized in that: The position information on the three-layer detector panel (1) is retained and recorded simultaneously, specifically: If the following conditions are met, the position information on the three-layer detector panel (1) is retained and recorded simultaneously: x1*y1*x2*y2*x3*y3≠0.

5. The method for three-dimensional reconstruction of cosmic ray muon distribution according to claim 1, characterized in that: The specific method for calculating muon trajectories is: The position information on the first layer detector panel is (x1, y1, z1), and the position information on the second layer detector panel is (x2, y2, z2).

6. The method for three-dimensional reconstruction of cosmic ray muon distribution according to claim 1, characterized in that: The specific steps to obtain the angular distribution are: in, is the vector representation of the muon trajectory, is the vector representation of the plane where each two coordinate axes are located, α is the angle between the side of the support frame (2) and the horizontal plane, that is, the zenith angle. Since the lifting platform is fixed on the circular guide rail, the azimuth angle is 0-360°. Each zenith angle and azimuth angle determines a muon trajectory, and then the muon angle distribution with respect to the zenith angle and azimuth angle in the spherical coordinate system is obtained.

7. The method for three-dimensional reconstruction of cosmic ray muon distribution according to claim 1, characterized in that: The specific operation of S6 can be as follows: according to the surface images on each set of detector panels (1), the autoencoder is used to compress the Monte Carlo simulation of the muon angular distribution to obtain the momentum, time, and position information of the muon, and the surface images and position information of the muon before and after passing through the object to be measured are output; based on the surface images and position information of the muon before and after passing through the object to be measured, the three-dimensional reconstruction of the object to be measured is completed.

8. A 3D reconstruction system for cosmic ray muon distribution, characterized in that: include: Annular guide rails for carrying the lifting platform; Lifting platform, two lifting platforms are fixed on the circular guide rail; A muon detection device comprises a support frame (2) and a detector panel (1). The support frame (2) is respectively installed on each lifting platform. A group of detector panels (1) is provided on each support frame (2). The group of detector panels (1) is provided with three layers, which are numbered from bottom to top. A coordinate system is established with the lower left corner of the bottom detector panel (1) as the origin. Each layer of the detector panel (1) comprises two levels of panels, and the two levels of panels are respectively embedded with plastic scintillator strips or plastic scintillator optical fibers, and the embedding direction is vertical. The position information left by the muon when passing through the detector panel (1) is detected, and the position information on the three layers of detector panels (1) is retained and recorded simultaneously. A trajectory calculation module calculates the muon trajectory on each set of detector panels (1) based on the position information on each set of detector panels (1); A point-surface image acquisition module calculates the muon flux on each set of detector panels (1) and the angle between the muon trajectory and the plane where each two coordinate axes are located, obtains the muon angular distribution on each set of detector panels (1), and establishes a point-surface image of the muon angular distribution and flux distribution based on the muon trajectory; The line-surface image acquisition module multiplies all point-surface images by weight coefficients and then superimposes them to obtain line-surface images; The surface image acquisition module multiplies all line-surface images by weight coefficients and then superimposes them to obtain the surface image; The three-dimensional simulation module uses an autoencoder to compress Monte Carlo simulation of the muon angular distribution based on the surface images on each set of detector panels (1), obtains the momentum, time, and position information of the muon, outputs the surface images and position information of the muon before and after passing through the object to be tested, completes the three-dimensional reconstruction of the object to be tested based on the surface images and position information of the muon before and after passing through the object to be tested, extracts the angular distribution in the preset area, classifies the angular distribution and position information according to the interval, and infers the information of the substance to be tested.

9. The method and system for three-dimensional reconstruction of cosmic ray muon distribution according to claim 8, characterized in that: The support frame (2) is used to adjust the zenith angle direction of the detector panel (1).

10. The method for three-dimensional reconstruction of cosmic ray muon distribution according to claim 8, characterized in that: A photomultiplier tube is provided at the front end of the detector panel (1) for reading the position information of the muons, converting the light signal into an electrical signal and recording the electrical signal.

Images