A method for rapidly optimizing glass materials in optical optimization design

Abstract

The present invention discloses a method for rapidly optimizing glass materials in optical optimization design, the steps of which are: 1) generating n values ​​of optical glass according to the parameters of each glass in the glass library; d ‑v d ‑dP g,F space and determine the boundary surface of the space; 2) record the evaluation function value mf(n) of the optical system before optimization; 3) determine the number of glass materials participating in the optimization through pseudo-random numbers; 4) read the parameters of the glass materials participating in the optimization as continuous variables and optimize them through damped least squares method; 5) determine the closest glass material based on the optimized glass parameters, and replace the optimized parameters with the actual parameters of the corresponding glass material in the glass library; 6) replace the glass parameters in the model with the corresponding specific glass parameters and optimize; record the updated function value mf(n+1); 7) when mf(n+1) is greater than mf(n), abandon the current model parameters; otherwise, retain the current model parameters.

Description

Technical Field

[0001] The present invention belongs to the technical field of optical system design, and in particular relates to a method for rapidly optimizing glass materials in optical optimization design. Background Art

[0002] The refractive index refers to the ratio of the sine of the incident angle γ to the sine of the refraction angle β when monochromatic light enters the glass from a vacuum as a function of wavelength.

[0003] n=sinγ / sinβ

[0004] The dispersion coefficient (Abbe number) is often used to characterize the "central dispersion" of glass. The larger the Abbe number, the smaller the dispersion.

[0005]

[0006] Relative partial dispersion: Relative partial dispersion is often considered in the design of apochromatic lenses. After the chromatic aberration of two colors of light is eliminated, the remaining chromatic aberration relative to the third color of light (secondary spectrum) becomes the main chromatic aberration.

[0007] According to Abbe's formula, most "normal glasses" satisfy the following linear relationship:

[0008] P x,y =m x,y ν d +b x,y

[0009] Where, ν d and P x,y are the horizontal and vertical coordinate variables m x,y is the slope, b x,y is the intercept.

[0010] For the correction of the secondary spectrum, at least one "abnormal glass" is required, that is, P' that deviates from the Abbe linear formula. x,y The deviation can be expressed as ΔP′ x,y For example, ΔP g,F The deviation characteristics of the special dispersion of "abnormal glass" from the Abbe line are quantitatively represented.

[0011]

[0012] ΔP g,F =P g,F -0.6457+0.001703·ν d

[0013] Traditional optical system design begins with ideal optical formulas to calculate the initial system structure. Aberration theory is then used to calculate the optical structure, and ray tracing is used to estimate and compensate for higher-order aberrations. In the optical lens optimization design process, the radius of curvature R and thickness T are used as continuous variables. The damped least squares method is used to find the minimum value of the evaluation function, i.e., the local optimal solution, within the constraints of the evaluation function. In addition to the radius of curvature and thickness, the glass material is also an important optimization variable. When the optical system aberrations do not meet the design requirements, it is often necessary to replace the glass or glass combination.

[0014] Over the years of development in optical design, optical designers have designed a wide variety of lens systems with varying parameters, putting them into production and use to verify that they meet the required specifications. In modern optical system design, optical designers typically search for lens design programs that closely match the design specifications and then perform local optimization adjustments based on these programs to achieve the optimal solution. These reference lenses often come from textbooks or optical lens patents. The recent development of extremely high refractive index, medium-to-low dispersion ZLaF glass has expanded the selection of glass materials in optical design. When using previous patents or case studies as reference designs, optimizing the system glass material can often further improve lens imaging quality.

[0015] The main parameters of optical glass refractive index n d and Abbe number v d The intervals are large, so they are discrete variables and are not suitable for damped least squares. The hammer optimization function in the optical design software ZEMAX can achieve the function of optimizing glass replacement, but the algorithm process is not public.

[0016] The selection of glass materials in the optical design process is a complex and difficult process. Optical glass is a discrete variable. What type of glass should be used in different types of optical systems? What type of glass should be replaced to compensate for different aberrations? Although the dispersion formula can provide preliminary estimates for these processes, the calculation process is complicated and requires a high level of optical design experience.

[0017] When optimizing glass design using ZEMAX, after completing the initial system design based on the initial structure, the glass is typically set to "replacement." The global optimization process, known as Hammer Optimization, often takes a long time for the system to converge to a lower merit function result. However, the optimized result may not always meet the requirements, necessitating further optimization by replacing the initial structure with a different one. This requires repeating the merit function setting and Hammer glass replacement optimization process. Summary of the Invention

[0018] To address the problems existing in the prior art, the present invention aims to provide a method for rapidly optimizing glass materials in optical optimization design. The present invention primarily focuses on optimizing parameters such as the radius of curvature, thickness interval, and glass material. Key glass material parameters include refractive index, Abbe number, and relative partial dispersion deviation.

[0019] This invention provides a fast descent path method for simultaneously optimizing continuous and discrete variables. It optimizes discrete material variables as quasi-continuous variables through boundary constraints. Rather than always including all materials as variables during optimization, the optimization process uses pseudo-random numbers to determine which materials are optimized in each round.

[0020] The technical solution of the present invention is:

[0021] A method for rapidly optimizing glass materials in optical optimization design, comprising the following steps:

[0022] 1) According to the refractive index, Abbe number parameters and relative partial dispersion deviation value dP of each glass in the glass library g,F , generating n of optical glass d -v d -dP g,F space, and determine the n d -v d -dP g,F Boundary surfaces of space;

[0023] 2) Setting an evaluation function according to the index requirements of the optical system; recording the evaluation function value mf(n) of the optical system before optimization;

[0024] 3) Generate a set of 1 / 0 pseudo-random numbers based on the number of glass materials participating in the optimization in the optical system. The pseudo-random numbers correspond one-to-one to the materials to be optimized. A pseudo-random number of 1 indicates that the corresponding material is used as a variable in this round of optimization, and a pseudo-random number of 0 indicates that the corresponding material does not participate in this round of optimization.

[0025] 4) Read the refractive index n of the glass material involved in the optimization d , Abbe number v d and relative dispersion deviation dP g,F As a continuous variable, the continuous variable and other parameters in the optical system are optimized by a damped least squares method;

[0026] 5) reading the parameters of each glass in the glass library, determining the closest glass material based on the optimized glass parameters, and replacing the optimized parameters with the actual glass parameters of the corresponding glass material in the glass library;

[0027] 6) Replace the glass parameters in the system model corresponding to the optical system with the corresponding actual glass parameters in the material library one by one;

[0028] The replacement order is randomly generated; after each replacement, the updated system model is optimized using the damped least squares method; when all replacements and optimizations are completed, the evaluation function value mf(n+1) of the updated optical system is recorded;

[0029] 7) Compare mf(n) and mf(n+1). If the optimized mf(n+1) is greater than the value of mf(n) before the optimization, abandon the current system model parameters; otherwise, retain the current system model parameters, return to step 3) and generate a new set of 1 / 0 pseudo-random numbers.

[0030] 8) Repeat steps 3) to 7), each time step 3) generates a new set of 1 / 0 pseudo-random numbers and step 6) generates a new set of replacement orders; when the set number of times is reached, the currently optimal system model parameters are selected from the retained system model parameters and output.

[0031] Furthermore, the method for determining the boundary surface is as follows: reading the refractive index, Abbe number and relative partial dispersion deviation value parameters of each glass in the glass library participating in the optimization; according to n d -v d -dP g,F The scattered point distribution in the three-dimensional coordinate system of space is fitted with several boundary function surfaces.

[0032] Further, the closest glass determination method: calculate the distance between the unmatched glass and each glass in the glass library Determine the replacement for each unmatched glass based on the distance d; where N dm 、V dm , ΔP g,Fm is the refractive index, Abbe number and relative partial dispersion deviation of the unmatched glass in the system model; N dg 、V dg , ΔP g,Fg is the refractive index, Abbe number and relative partial dispersion deviation value of the glass in the glass library; W n 、W v 、W p are the weights corresponding to the refractive index, Abbe number and relative partial dispersion deviation value.

[0033] Furthermore, the information used to calculate the evaluation function includes focal length, constraint on lens thickness, RMS value of the spot diagram or wavefront, and PTV value.

[0034] A server, characterized in that it includes a memory and a processor, the memory stores a computer program, the computer program is configured to be executed by the processor, and the computer program includes instructions for executing each step in the above method.

[0035] A computer-readable storage medium stores a computer program thereon, wherein the computer program implements the steps of the above method when executed by a processor.

[0036] The advantages of the present invention are as follows:

[0037] The designer has low design experience requirements and does not need to understand the performance of various optical glass materials. The optical system design indicators can be improved by replacing the glass;

[0038] Discrete variables can be optimized by taking the derivative of the damped least squares method to obtain a more suitable glass combination.

[0039] The damped least squares method in this model replacement method can be replaced by other optimization algorithms, such as the orthogonal descent method, the adaptation method, as well as simulated annealing, global search, variable scaling method and damped least squares method with escape function in the global optimization algorithm.

[0040] Glass boundary determination method: In addition to fitting several boundary surfaces, you can also divide the three-dimensional space into grids with appropriate density. By judging whether there is glass in the cubic grid, you can determine whether the three-dimensional grid in the space belongs to the optimization range. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0042] The present invention will be described in further detail below with reference to the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0043] Figure 1 It is the main flow chart of the technical solution of the present invention. The processing flow of the optimization method of the present invention includes the following steps.

[0044] Step 1: Describe the n of optical glass based on the refractive index, Abbe number parameters and relative partial dispersion deviation values ​​in the glass library d -v d -dP g,F The space is then mapped to the glass boundary surface. By fitting the glass boundary, the optimization range of the refractive index and Abbe number can be reduced, thus preventing the optimization parameters from failing to find a suitable glass.

[0045] Method for determining the boundary surface: read the refractive index, Abbe number and relative partial dispersion deviation parameters of each glass in the glass library involved in the optimization; according to the scattered point distribution in the three-dimensional coordinate system, fit several boundary function surfaces. The surface restriction range is slightly larger than the actual glass range. The expansion of the optimization range is conducive to searching for more suitable glass. The three-dimensional coordinate system here is the n d -vd -dP g,F The three-dimensional coordinates of the space and the scattered point distribution are drawn according to the parameters of each glass in the glass library.

[0046] Step 2: According to the actual optical system index requirements, write a suitable evaluation function, including focal length, constraint on lens thickness, RMS value of point diagram or wavefront, PTV value, etc. The curvature, thickness and material in the optical system are respectively used as optimization variables, where the independent variables of the material variable are refractive index, Abbe number and relative partial dispersion deviation value. Record the system evaluation function value mf(n) before optimization. The setting of the evaluation function varies from program to program. The constraints of focal length and lens thickness are constrained according to the design index requirements. The constraint method can provide an upper boundary, a lower boundary or fix a certain quantity to the target value; the RMS value and PTV value of the point diagram and wavefront are the results obtained on the image plane by tracing rays through the optical system.

[0047] Step 3: Generate a set of 1 / 0 pseudo-random numbers based on the number of materials participating in the optimization. The pseudo-random numbers correspond one-to-one to the materials to be optimized. A pseudo-random number of 1 indicates that the corresponding material is selected as a variable for this round of optimization, while a pseudo-random number of 0 indicates that the corresponding material is not selected for this round of optimization.

[0048] Step 4: Read the refractive index n of the material involved in the optimization d , Abbe number v d , relative partial dispersion deviation coefficient dP g,F The parameter is treated as a continuous variable and optimized with other system parameters through the damped least squares method. If the initial design is appropriate and the evaluation function is set reasonably, the evaluation function will drop to the minimum value quickly.

[0049] Step 5: The refractive index and Abbe number calculated by the damped least squares method may not have a completely corresponding glass, so the distance is determined by judging the model material n d ,v d , dP g,F The closest glass material is used for replacement. Random sorting is performed based on the number of glass models that need to be replaced, and the material models are replaced with specific glass one by one.

[0050] The closest glass determination method:

[0051]

[0052] Where d represents the distance between the model material and the actual glass; N dm 、V dm , ΔP g,Fm is the refractive index, Abbe number and relative partial dispersion of the model material; N dg 、V dg , ΔPg,Fg are the refractive index, Abbe number, and relative partial dispersion of the actual glass material; W n , W v , W p are weights. By comparing the distances between the model material and each glass material, the model glass is replaced with the "nearest" actual glass.

[0053] Step Six: When replacing to the actual glass from the three parameters, due to the discrete characteristics of the glass, the system is no longer at the optimized minimum value, and the evaluation function often becomes larger. Therefore, after each replacement, the system is optimized for the curvature radius, thickness, and the parameters of the glass material that has not been replaced with a specific glass by the damped least squares method. After completing all glass replacements and optimizations, record m.f.(n + 1) at this time.

[0054] Step Seven: Compare m.f.(n) and m.f.(n + 1). When the optimized m.f.(n + 1) is larger than the value of m.f.(n) before the start of the optimization, abandon this model optimization and replacement. Repeat the operations in Step Three - Step Six.

[0055] When the optimized m.f.(n + 1) is less than or equal to the value of m.f.(n) before the start of the optimization, retain this model optimization and glass replacement, update the system parameters, and repeat the operations in Step Three - Step Six.

[0056] By using the material optimization algorithm proposed in this patent, good output results can usually be obtained after several groups of optimization replacements.

[0057] Once a lower rating function value m.f.(n + 1) < m.f.(n) is obtained, the entire system will be replaced with a new glass combination, and subsequent m.f.(n + 2) will be compared with the previous group's m.f.(n + 1). The evaluation function value can be updated in real time. Every time a lower evaluation function value is calculated, the system is updated to the latest state.

[0058] Although specific embodiments of the present invention are disclosed for illustrative purposes, which are intended to help understand the content of the present invention and implement it accordingly, those skilled in the art can understand that: without departing from the spirit and scope of the present invention and the appended claims, various substitutions, changes, and modifications are possible. Therefore, the present invention should not be limited to the content disclosed in the best embodiments, and the scope of protection claimed by the present invention is defined by the scope of the claims.

Claims

1. A method for rapidly optimizing glass materials in optical optimization design, comprising the following steps: 1) According to the refractive index, Abbe number and relative partial dispersion deviation value dP of each glass in the glass library g,F , generating n of optical glass d -v d -dP g,F space, and determine the n d -v d -dP g,F Boundary surfaces of space; 2) Setting an evaluation function according to the index requirements of the optical system; Record the evaluation function value mf(n) of the optical system before optimization; 3) Generate a set of 1 / 0 pseudo-random numbers based on the number of glass materials participating in the optimization in the optical system. The pseudo-random numbers correspond one-to-one to the materials to be optimized. A pseudo-random number of 1 indicates that the corresponding material is used as a variable in this round of optimization, and a pseudo-random number of 0 indicates that the corresponding material does not participate in this round of optimization. 4) Read the refractive index n of the glass material involved in the optimization d , Abbe number v d and relative dispersion deviation dP g,F As a continuous variable, the continuous variable and other parameters in the optical system are optimized by a damped least squares method; 5) reading the parameters of each glass in the glass library, determining the closest glass material based on the optimized glass parameters, and replacing the optimized parameters with the actual glass parameters of the corresponding glass material in the glass library; 6) Replace the glass parameters in the system model corresponding to the optical system with the corresponding actual glass parameters in the material library one by one; the replacement order is randomly generated; after each replacement, perform a damped least squares optimization on the updated system model; when all replacements and optimizations are completed, record the evaluation function value mf(n+1) of the updated optical system at this time; 7) Compare mf(n) and mf(n+1). If the optimized mf(n+1) is greater than the value of mf(n) before the optimization, abandon the current system model parameters; otherwise, retain the current system model parameters, return to step 3) and generate a new set of 1 / 0 pseudo-random numbers. 8) Repeat steps 3) to 7), each time step 3) generates a new set of 1 / 0 pseudo-random numbers and step 6) generates a new set of replacement orders; when the set number of times is reached, the currently optimal system model parameters are selected from the retained system model parameters and output.

2. The method according to claim 1, characterized in that The method for determining the boundary surface is as follows: reading the refractive index, Abbe number and relative partial dispersion deviation value parameters of each glass in the glass library involved in the optimization; d -v d -dP g,F The scattered point distribution in the three-dimensional coordinate system of space is fitted with several boundary function surfaces.

3. The method according to claim 1, characterized in that The closest glass determination method: Calculate the distance d between the unmatched glass and each glass in the glass library = [W n (N dm -N dg ) 2 +W v (V dm -V dg ) 2 +W p (ΔP g,Fm -ΔP g,Fg ) 2 ] 1 / 2 ; Determine the replacement of each unmatched glass according to the distance d; Wherein, N dm 、V dm , ΔP g,Fm is the refractive index, Abbe number and relative partial dispersion deviation of the unmatched glass in the system model; N dg 、V dg , ΔP g,Fg is the refractive index, Abbe number and relative partial dispersion deviation value of the glass in the glass library; W n 、W v 、W p are the weights corresponding to the refractive index, Abbe number and relative partial dispersion deviation value.

4. The method according to claim 1, 2 or 3, characterized in that: The information used to calculate the evaluation function includes focal length, constraint on lens thickness, RMS value of the spot diagram or wavefront, and PTV value.

5. A server, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, the computer program is configured to be executed by the processor, and the computer program comprises instructions for executing each step of the method according to any one of claims 1 to 4.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 4 are implemented.

Images