Method for acquiring a full spectrum energy distribution

Abstract

This invention belongs to the field of optical measurement technology and discloses a method for obtaining the full-spectrum energy distribution. The method includes the following steps: Step 1, a broadband beam to be measured is perpendicularly incident onto a microstructure to obtain the near-field wavefront distribution U(x, y) of the broadband incident light; Step 2, the far-field diffracted light intensity distribution B(x, y) of the near-field wavefront distribution U(x, y) is obtained through Langhofer far-field diffraction, thereby obtaining a specific wavelength λ within the broadband band. c Step three involves solving for the spectral components w(λ) at each discrete wavelength using a broadband multivariate linear regression model. Step four involves smoothing and optimizing the solved spectral components w(λ) to obtain the full-spectrum energy distribution. This invention utilizes a partially coherent broadband diffraction intensity multivariate linear regression model to calculate and solve for the continuous spectral information to achieve rapid measurement of the spectrum.

Description

Technical Field

[0001] This invention belongs to the field of optical measurement technology, and more specifically, relates to a method for obtaining full-spectrum energy distribution. Background Technology

[0002] Spectroscopic measurement is one of the most commonly used and important optical measurement methods in scientific research and industrial applications. Existing precision spectroscopic measurement systems mainly consist of an illumination optical path, optical components, an amplification optical path, and a spectroscopic detector. These methods and devices are bulky, require stringent optical component manufacturing processes, are costly, and operate in harsh environments. Limited by factors such as price and size, most existing precision spectrometers are used as precision scientific instruments, hindering their portability and widespread adoption.

[0003] In recent years, spectral measurement technology has seen the emergence of novel miniature spectrometer methods (camera chips and imaging devices, CN114125196A). These methods utilize micro / nanomaterials as spectral sensors, directly converting spectral energy into analog electrical signals for sampling and detection, integrating the spectral measurement system into a single camera. However, due to the fluctuations in the spectral characteristics of these nanomaterials, frequent and precise spectral calibration is required, leading to significant losses in detection accuracy, short sensor lifespan, cumbersome calibration procedures, and limited application scenarios. Summary of the Invention

[0004] In view of the above-mentioned defects or improvement needs of the existing technology, the present invention provides a method for obtaining full-spectrum energy distribution. It uses a multivariate linear regression model algorithm for partially coherent broadband diffraction intensity to calculate and solve the continuous spectral information to achieve rapid measurement of the spectrum to be measured. This solves the technical problems of existing spectral measurement devices such as large size, high cost, complex system, and limited operating environment.

[0005] To achieve the above objectives, according to one aspect of the present invention, a method for obtaining full-spectrum energy distribution is provided, the method mainly comprising the following steps: Step 1: The broadband beam to be tested is incident perpendicularly onto the microstructure to obtain the near-field wavefront distribution of the broadband incident light. ; Step 2: Obtain the near-field wavefront distribution through Fraunhofer far-field diffraction. Far-field diffraction intensity distribution Through the known microstructure distribution function And the diffraction propagation model is used to calculate a specific wavelength within a wide spectral band. Coherent diffraction intensity distribution ; Step 3: Solve for the spectral components w(λ) at each discrete wavelength using a broadband multiple linear regression model. The mathematical expression of the broadband multiple linear regression model is as follows:

[0006] in,

[0007]

[0008] In the formula, For microstructures in the wavelength matrix The intensity distribution matrix of diffraction light under the condition of monochromatic coherent beam incidence. Represents the corresponding discrete wavelength Spectral weights of location This is system noise error; Step four: Smooth the solved spectral component w(λ) to obtain the full-spectrum energy distribution.

[0009] Furthermore,

[0010] in, Represents a broadband beam Incident on microstructure The resulting near-field wavefront distribution function.

[0011] Furthermore, the thickness of the microstructure is such that the transmission energy of the spectrum under test is less than 5%.

[0012] Furthermore, stainless steel micropores with a diameter of 20μm and a thickness of 100μm were used as the microstructure.

[0013] Furthermore, using Fresnel numbers Characterization of microstructure:

[0014] Where a is the size of the microstructure aperture; L is the distance between the microstructure and the detector.

[0015] Furthermore, the far-field diffraction intensity distribution The calculation formula is:

[0016] in, For illumination wavelength, This is the distance between the detector and the microstructure. To be at wavelength Spectral weights at the location.

[0017] Furthermore, for the coherent diffraction matrix distribution at other wavelengths, the monochromatic diffraction intensity distribution at any wavelength can be obtained through calibration. Obtained by linear transformation:

[0018] in, It is the transfer function for linear transformation.

[0019] Furthermore, since the multiple linear regression model is a singular matrix, it can be iteratively fitted using gradient descent or least squares methods to obtain the globally optimal approximate solution.

[0020] In summary, compared with the prior art, the method for obtaining the full-spectrum energy distribution provided by this invention has the following advantages: 1. This invention utilizes a partially coherent broadband diffraction intensity multiple linear regression model to calculate and solve for the continuous spectral information, thereby achieving rapid measurement of the spectrum to be measured. This invention provides a simple and fast solution for the spectrum to be measured.

[0021] 2. The computational spectral measurement method provided by this invention has the advantages of system simplification, algorithm robustness, fast solution and high measurement accuracy, and can realize simple and fast measurement of spectral information in multiple scenarios and wide wavelengths. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating a method for obtaining full-spectrum energy distribution provided by the present invention; Figure 2 yes Figure 1 A schematic diagram of the structure involved in the method for obtaining the full-spectrum energy distribution; Figure 3 This is a detailed flowchart of the invention for obtaining the full-spectrum energy distribution based on a broadband multivariate linear regression model.

[0023] In all the accompanying drawings, the same reference numerals are used to denote the same elements or structures, wherein: 201-the broadband beam to be measured, 202-microstructure, 203-detector. Detailed Implementation

[0024] 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.

[0025] Please see Figure 1 , Figure 2 and Figure 3 The method for obtaining the full-spectrum energy distribution provided by this invention mainly includes the following steps: Step 1: The broadband beam to be tested is incident perpendicularly onto the microstructure to obtain the near-field wavefront distribution of the broadband incident light. .

[0026] There are three incident forms of the broadband beam 201 to be tested: using optical fiber or other light guiding devices to guide the light path to be incident on the microstructure 202; designing an imaging optical path, such as a reflector or focusing lens, to converge the broadband beam 201 to be tested onto the microstructure 202; and directly incident the broadband beam 201 to be tested onto the microstructure.

[0027] The spectral bandwidth of the broadband beam to be tested ranges from the extreme ultraviolet (EUV) band to the far-infrared band. Specifically, the bandwidth applicable to this embodiment is 12 nm to 15 nm in the EUV band.

[0028] Step one also includes determining broadband measurement parameters, given the microstructure size and morphology characteristics, to obtain the near-field wavefront distribution of broadband incident light. .

[0029]

[0030] in, Represents a broadband beam Incident on microstructure The resulting near-field wavefront distribution function.

[0031] The microstructure is generally a simple, thin dielectric microstructure such as a perforated micropore or microslit. The dielectric material is required to have low transmittance to the broadband spectrum being measured. Generally, the thickness of the microstructure is sufficient to ensure that the transmitted energy of the spectrum being measured is less than 5%. In this embodiment, a stainless steel micropore with a diameter of 20 μm and a thickness of 100 μm is used as the microstructure. No requirements are placed on its size or positional accuracy. The partially coherent broadband diffraction field only needs to satisfy the Fraunhofer diffraction approximation condition, generally using the Fresnel number. Characterization:

[0032] Where a is the size of the microstructure aperture, such as radius, and L is the distance between the microstructure and the detector 203.

[0033] Step 2: Obtain the near-field wavefront distribution using Langhefeld far-field diffraction. Far-field diffraction intensity distribution Thus, a specific wavelength within a wide spectral band can be obtained. Coherent diffraction intensity distribution , as a calibration parameter for the single-wavelength coherent diffraction matrix.

[0034]

[0035] in, For illumination wavelength, This is the distance between detector 203 and diffraction microstructure 202. To be at wavelength The spectral weights at a given location are used to solve for unknown quantities.

[0036] Generally, it can be obtained in two ways: one is by using a given specific wavelength. The single-wavelength coherent diffraction intensity matrix is ​​obtained through the methods described above; secondly, it is obtained through the known microstructure distribution function. The results were obtained from calculations using a diffraction propagation model.

[0037] For the coherent diffraction matrix distribution at other wavelengths, according to wave optics theory, the monochromatic diffraction intensity distribution at any wavelength can be determined by calibration. Obtained by linear transformation:

[0038] in, It is the transfer function for linear transformation.

[0039] Step 3: Solve for the spectral components w(λ) at each discrete wavelength using a broadband multiple linear regression model. The mathematical expression of the broadband multiple linear regression model is as follows:

[0040] in,

[0041]

[0042] In the formula, For microstructures in the wavelength matrix The intensity distribution matrix of diffraction light under the condition of monochromatic coherent beam incidence. Represents the corresponding discrete wavelength Spectral weights of location This represents system noise error.

[0043] Multiple linear regression models are usually singular matrices and generally do not have a single particular solution. They can be iteratively fitted using gradient descent or least squares methods to obtain a globally optimal approximate solution.

[0044] Step four: Smooth the solved spectral component w(λ) to obtain the full-spectrum energy distribution.

[0045] In this embodiment, the optimization method for this step generally consists of two steps: the first step is to use a uniform sampling method to adjust the spectral weight matrix. Uniform sampling is performed; the second step involves data optimization using convolutional filtering.

[0046] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for obtaining full-spectrum energy distribution, characterized in that, The method includes the following steps: Step 1: The broadband beam to be tested is incident perpendicularly onto the microstructure to obtain the near-field wavefront distribution of the broadband incident light. ; Step 2: Obtain the near-field wavefront distribution through Fraunhofer far-field diffraction. Far-field diffraction intensity distribution and through the known microstructure distribution function And the diffraction propagation model calculates a specific wavelength within a wide spectral band. Coherent diffraction intensity distribution ; Step 3: Solve for the spectral components w(λ) at each discrete wavelength using a broadband multiple linear regression model. The mathematical expression of the broadband multiple linear regression model is as follows: in, In the formula, For microstructures in the wavelength matrix The intensity distribution matrix of diffraction light under the condition of monochromatic coherent beam incidence. Represents the corresponding discrete wavelength Spectral weights of location This is system noise error; Step 4: Perform smoothing optimization on the solved spectral component w(λ) to obtain the full-spectrum energy distribution; For the coherent diffraction matrix distribution at other wavelengths, the monochromatic diffraction intensity distribution at any wavelength can be obtained through calibration. Obtained by linear transformation: in, It is the transfer function for linear transformation.

2. The method for obtaining the full-spectrum energy distribution as described in claim 1, characterized in that: in, Represents a broadband beam Incident on microstructure The resulting near-field wavefront distribution function.

3. The method for obtaining the full-spectrum energy distribution as described in claim 1, characterized in that: The thickness of the microstructure is such that the transmission energy of the spectrum to be measured is less than 5%.

4. The method for obtaining the full-spectrum energy distribution as described in claim 3, characterized in that: Stainless steel micropores with a diameter of 20μm and a thickness of 100μm were used as the microstructure.

5. The method for obtaining the full-spectrum energy distribution as described in claim 3, characterized in that: Using Fresnel numbers Characterization of microstructure: Where a is the size of the microstructure aperture; L is the distance between the microstructure and the detector.

6. The method for obtaining the full-spectrum energy distribution as described in claim 2, characterized in that: Far-field diffraction intensity distribution The calculation formula is: in, For illumination wavelength, This is the distance between the detector and the microstructure. To be at wavelength Spectral weights at the location.

7. The method for obtaining the full-spectrum energy distribution as described in any one of claims 1-6, characterized in that: The multiple linear regression model is a singular matrix, which can be iteratively fitted using gradient descent or least squares methods to obtain the globally optimal approximate solution.

Images