An adaptive optical reshaping lens design method
By using an adaptive design model based on neural networks, the problems of long design time and high cost of optical shaping lenses are solved, achieving efficient lens design and performance prediction, and reducing maintenance costs in practical applications.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIJING INST OF REMOTE SENSING EQUIP
- Filing Date
- 2025-12-23
- Publication Date
- 2026-06-23
AI Technical Summary
Existing optical reshaping lens design methods are time-consuming, costly, and cannot predict lens performance changes in actual applications, resulting in high trial-and-error costs.
An adaptive design model based on neural networks is adopted. Lens design data is systematically collected, standardized, and iteratively optimized using convolutional neural networks and recurrent neural networks. Simulation analysis is performed in conjunction with Maxwell's equations to generate lens design schemes and perform multiple rounds of optimization iterations until the design requirements are met.
It improves lens design efficiency, reduces verification costs, enables lens performance prediction, and reduces maintenance costs and risks in practical applications.
Smart Images
Figure CN122260632A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of optical lens design technology, and specifically relates to an adaptive optical shaping lens design method. Background Technology
[0002] Optical reshaping lenses, as an important component of optical elements, are widely used in optical imaging systems, fiber optic communications, and biomedicine. Through precise design, aberrations in optical systems can be corrected, improving image quality and resulting in more accurate and clearer images. Currently, most optical reshaping lens design methods are based on ray mapping or energy mapping, requiring extensive calculations for different light sources and reshaping needs, leading to lengthy design times. Furthermore, in traditional methods, the lens needs to be fabricated before the design can be verified, resulting in high trial-and-error costs. Therefore, there is an urgent need to find a design method for optical reshaping lenses that is efficient in design and low in verification costs.
[0003] With the emergence of the concept of neural networks and the continuous improvement of computer performance, more and more researchers hope to apply them to various fields such as medicine, communications, and materials processing. Neural network models possess powerful learning and adaptive capabilities, capable of simulating complex nonlinear relationships, learning and extracting features from sample data, providing reasonable predictions and classifications for unseen data, and resisting interference to a certain extent while maintaining stability. Therefore, by constructing a reasonable adaptive lens design model based on neural networks, and with input from a sample lens design library, lens design schemes can be provided and verified for different optical shaping needs, achieving the goal of cost reduction and efficiency improvement. Summary of the Invention
[0004] The purpose of this invention is to provide an adaptive optical reshaping lens design method to solve the problems of long time consumption, high cost, and inability to predict the performance changes of the lens in actual application that exist in the current traditional optical reshaping lens design methods.
[0005] To achieve the above objectives, the present invention adopts the following technical solution:
[0006] On one hand, the present invention provides an adaptive optical shaping lens design method, comprising:
[0007] S1. Systematically collect design data for optical lenses, including light source parameters, lens shape, size, material, optical performance, and other parameters, as well as the long-term performance of these lenses in real-world scenarios.
[0008] S2. Verify and standardize the collected data to ensure its accuracy and consistency;
[0009] S3. Input the processed data into the neural network-based adaptive design model, and fine-tune the model parameters through iterative optimization algorithms;
[0010] S4. Input the specific design requirements into the adaptive design model, including light source parameters, lens material, size, relative position, and output spot parameters;
[0011] S5. Generate a preliminary design scheme using an adaptive design model, and perform simulation analysis and verification of the designed lens based on Maxwell's equations;
[0012] S6. Based on the analysis results and design requirements, conduct multiple rounds of optimization and iteration until the design requirements are finally met;
[0013] S7. Predict the long-term performance changes of the designed lens to provide a basis for maintenance in the long-term application of the lens in actual optical systems.
[0014] On the other hand, the present invention discloses an electronic device including a processor, an internal bus, a network interface, memory, and non-volatile memory. The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it, forming an adaptive optical reshaping lens design method at the logical level.
[0015] Based on the above technical solution, the present invention can achieve the following technical effects:
[0016] This method employs an adaptive design model based on convolutional neural networks for the design of optical shaping lenses. This improves lens design efficiency while reducing the cost of lens performance verification through simulation, successfully achieving cost reduction and efficiency improvement. By using an adaptive design model based on recurrent neural networks to predict the long-term performance of the lens in practical applications, maintenance costs and risks during actual use can be reduced, effectively avoiding losses caused by sudden failures of optical lenses in practical applications. Attached Figure Description
[0017] Figure 1 This is a flowchart illustrating an adaptive optics shaping lens design method according to an embodiment of the present invention;
[0018] Figure 2 This is a schematic diagram of an electronic device in an embodiment of the present invention. Detailed Implementation
[0019] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The advantages and features of the present invention will become clearer from the following description and claims. It should be noted that the drawings are all in a very simplified form and are not to a precise scale, and are only used to facilitate and clarify the illustration of the embodiments of the present invention.
[0020] It should be noted that, in order to clearly illustrate the content of this invention, several embodiments are provided to further explain different implementations of the invention. These embodiments are enumerated rather than exhaustive. Furthermore, for the sake of brevity, content mentioned in the preceding embodiments is often omitted in the following embodiments. Therefore, content not mentioned in the later embodiments can be referred to in the preceding embodiments.
[0021] Example 1
[0022] Please refer to Figure 1 , Figure 1 The present invention provides an adaptive optics shaping lens design method according to this embodiment. In this embodiment, the method includes:
[0023] Step S1: Systematically collect design data for optical lenses, including light source parameters, lens shape, size, material, optical performance, and other parameters, shaped light spot parameters, as well as the long-term performance of these lenses in actual scenarios.
[0024] In this embodiment, one implementation of step S1 is as follows:
[0025] The collected data mainly includes two categories: time-independent initial lens design data and time-dependent performance change data of the lens during long-term operation.
[0026] The time-independent initial design data for the lens includes: light source parameters, namely the light intensity distribution function, light power, and spot size; lens parameters, namely the lens size, material, shape, surface coating parameters, and distance from the light source; and output spot parameters, namely the light intensity distribution, spot size, and distance between the spot and the lens.
[0027] The performance change data of the lens over a long period of time includes: environmental parameters, namely the changes of parameters such as temperature, humidity and air pressure in the actual optical system environment over time; and output spot parameters, namely the changes of parameters such as light intensity distribution and spot size of the output spot over time.
[0028] Step S2 involves verifying and standardizing the collected data to ensure its accuracy and consistency.
[0029] In this embodiment, one implementation of step S2 is as follows:
[0030] The standardization process is as follows: The actual input data is transformed into a d-dimensional vector x, and preferably normalized using the following formula:
[0031]
[0032] Where x i Let x be the i-th component of the vector in the input data.
[0033] Step S3: Input the processed data into the neural network-based adaptive design model, and finely adjust the model parameters through an iterative optimization algorithm.
[0034] In this embodiment, one way to implement step S3 is as follows:
[0035] The adaptive design model based on convolutional neural networks extracts and calculates light spot images from different planes in the data, and finally outputs the light spot image at the target plane. It mainly includes the following structure:
[0036] Convolutional layers primarily extract feature information by performing convolution processing on specific regions in a light spot image using convolution kernels.
[0037] Pooling layers primarily classify and merge the feature information output by different convolutional layers, reducing the overall computational cost of the model and preventing overfitting.
[0038] The activation function mainly determines whether a node in a neural network can be activated. The preferred activation function in this model is shown in the following formula, which has the advantages of fast convergence and less loss of features.
[0039]
[0040] The fully connected layer is mainly responsible for integrating the feature information output by the pooling layer and combining it with the activation function to calculate the final output result.
[0041] The adaptive design model based on recurrent neural networks is responsible for outputting the change of the light spot image on the target location plane over time, based on the input environmental parameters. The preferred model is shown in the following equation:
[0042]
[0043] Where x(k) is the state vector of the system at time k, φ(k) is the initial condition vector at time k, C is the state feedback coefficient matrix, A is the connection weight matrix, f(x(k)) is the activation function, and J is the external input vector.
[0044] Step S31: Based on the input time-independent initial lens design data, perform iterative parameter optimization on the adaptive design model part based on convolutional neural network.
[0045] Step S32: Based on the input time-related performance change data of the lens during long-term operation, perform parameter iterative optimization on the adaptive design model part based on recurrent neural network.
[0046] In this embodiment, one implementation of steps S31 and S32 is as follows:
[0047] S301, Calculate the loss function, which is mainly used to measure the error between the model design results and the actual results.
[0048] S302, backpropagation, mainly involves backpropagating the error obtained in the previous step in the model to calculate the gradient of the parameters in the model;
[0049] S303, parameter update, mainly involves updating the model's parameters based on the gradient and learning rate to improve the model's design accuracy.
[0050] That is, the above three sub-steps are included in the parameter iterative optimization of both the adaptive design model part based on convolutional neural networks and the adaptive design model part based on recurrent neural networks.
[0051] Step S4: Input the specific design requirements into the adaptive design model, including light source parameters, lens material, size, relative position, and output spot parameters.
[0052] Step S5: Generate a preliminary design scheme using the adaptive design model, and perform simulation analysis and verification on the designed lens based on Maxwell's equations.
[0053] In this embodiment, one implementation of step S5 is as follows:
[0054] The planar distribution change of the light spot at each interface in the optical system is calculated based on Maxwell's equations, and the light spot image at the target position is finally obtained. The performance of the designed lens is then verified by simulation analysis. The calculation expression is shown in the following formula:
[0055]
[0056] Where E is the electric field strength, H is the magnetic field strength, B is the magnetic flux density, D is the electric displacement vector, and J is the magnetic flux density. m Where J is the equivalent magnetic flux density and J is the current density. is the Hamiltonian operator in Maxwell's equations, representing the differential with respect to the x, y, z coordinate system.
[0057] Step S6: Based on the analysis results and design requirements, perform multiple rounds of optimization and iteration until the design requirements are finally met.
[0058] In this embodiment, one implementation of step S6 is as follows:
[0059] The target location spot image evaluation function needs to be used as the standard for iterative optimization. The preferred evaluation function is shown in the following formula:
[0060]
[0061] Where n is the number of grids in the target location spot image, w i y represents the weighting coefficients of the grid. i r represents the light intensity value of the target location spot image under the grid. i This represents the light intensity value of the target location spot image calculated by the designed lens under the grid. The iteration is considered complete when the error value calculated using this evaluation function for the designed lens is within the set allowable range.
[0062] Step S7: Predict the long-term performance changes of the designed lens to provide a maintenance basis for the long-term application of the lens in a practical optical system.
[0063] Example 2
[0064] Please refer to Figure 2 This embodiment provides an electronic device including a processor, an internal bus, a network interface, memory, and non-volatile memory, and may also include other hardware required for business operations. The processor reads the corresponding computer program from the non-volatile memory into memory and then runs it, forming an adaptive optics reshaping lens design method at the logical level. Of course, in addition to software implementation, this specification does not exclude other implementation methods, such as logic devices or a combination of hardware and software, etc. That is, the execution subject of the following processing flow is not limited to individual logic units, but can also be hardware or logic devices.
[0065] Network interfaces, processors, and memory can be interconnected via a bus system. These buses can be categorized as address buses, data buses, control buses, etc.
[0066] Memory is used to store programs. Specifically, programs may include program code, which includes computer operation instructions. Memory may include read-only memory and random access memory, and provides instructions and data to the processor.
[0067] The processor is used to execute the program stored in the aforementioned memory, and specifically perform the following:
[0068] S2. Verify and standardize the data to ensure its accuracy and consistency;
[0069] S3. Input the processed data into the neural network-based adaptive design model, and fine-tune the model parameters through iterative optimization algorithms;
[0070] S4. Input the specific design requirements into the adaptive design model, including light source parameters, lens material, size, relative position, and output spot parameters;
[0071] S5. Generate a preliminary design scheme using an adaptive design model, and perform simulation analysis and verification of the designed lens based on Maxwell's equations;
[0072] S6. Based on the analysis results and design requirements, conduct multiple rounds of optimization and iteration until the design requirements are finally met;
[0073] S7. Predict the long-term performance changes of the designed lens to provide a basis for maintenance in the long-term application of the lens in actual optical systems.
[0074] A processor may be an integrated circuit chip with signal processing capabilities. In implementation, the steps of the above method can be completed through the processor's integrated hardware logic circuits or software instructions.
[0075] Based on the same invention, embodiments of this specification also provide a computer-readable storage medium that stores one or more programs, which, when executed by an electronic device including multiple applications, cause the electronic device to perform... Figure 1 A corresponding embodiment provides an adaptive optics shaping lens design method.
[0076] Those skilled in the art will understand that the embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, this specification may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this specification may take the form of a computer program product embodied on one or more computer-readable storage media containing computer-usable program code.
[0077] Furthermore, the specific implementation of the above system is basically similar to the method implementation, so the description is relatively simple. For relevant details, please refer to the description of the method implementation. Moreover, it should be noted that in the various modules of the system of this application, the components are logically divided according to the functions they are to perform. However, this application is not limited to this and can re-divide or combine the components as needed.
[0078] This specification discloses an adaptive optical shaping lens design method, belonging to the field of optical lens design technology. Its core steps include: systematically collecting design data for optical lenses, covering key parameters such as lens shape, size, material, and optical performance, as well as the long-term performance of these lenses in real-world scenarios; verifying and standardizing the collected data to ensure its accuracy and consistency; inputting the processed data into a neural network-based adaptive design model and fine-tuning the model parameters through iterative optimization algorithms; inputting specific design requirements into the adaptive design model to quickly generate corresponding lens design schemes; performing simulation analysis on the designed lens based on Maxwell's equations and conducting multiple rounds of optimization iterations according to the design requirements until the final design requirements are met; after design completion, further predicting the long-term performance of the designed lens to provide a basis for the overall performance changes of the lens during actual use, facilitating the maintenance of the optical system during actual use. This design method effectively solves the problems of long time consumption, high cost, and difficult long-term maintenance associated with traditional optical lens design methods due to its advantages of speed, effectiveness, low cost, and ability to predict the long-term performance of the designed lens.
[0079] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences between it and other embodiments.
[0080] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps described in the claims may be performed in a different order than those shown in the embodiments and still achieve the desired result. Furthermore, the specific order or sequential order shown in the drawings is not necessarily required to achieve the desired result; in some embodiments, multitasking and parallel processing are possible or may be advantageous.
[0081] The above description is merely an embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of this application should be included within the scope of the claims of this application.
Claims
1. A method for designing an adaptive optical shaping lens, characterized in that, include: S1. Systematically collect design data for optical lenses, as well as the long-term performance of these lenses in real-world scenarios; S2. Verify and standardize the collected data to ensure its accuracy and consistency; S3. Input the processed data into the neural network-based adaptive design model, and fine-tune the model parameters through iterative optimization algorithms; S4. Input the specific design requirements into the adaptive design model, including light source parameters, lens material, size, relative position, and output spot parameters; S5. Generate a preliminary design scheme using an adaptive design model, and perform simulation analysis and verification of the designed lens based on Maxwell's equations; S6. Based on the analysis results and design requirements, conduct multiple rounds of optimization and iteration until the design requirements are finally met; S7. Based on the lens's operating environment, predict the long-term performance changes of the designed lens to provide a maintenance basis for the long-term application of the lens in actual optical systems.
2. The method according to claim 1, characterized in that, In step S1, the collected data mainly includes two types: time-independent initial lens design data and time-dependent performance change data of the lens during long-term operation. The time-independent initial lens design data includes: Light source parameters, namely, light intensity distribution function, light power, spot size, etc. Lens parameters, including lens size, material, shape, surface coating parameters, and distance from the light source; Output spot parameters, namely, the light intensity distribution of the output spot, the spot size, and the distance between the spot and the lens; The time-related performance change data of the lens during long-term operation includes: Environmental parameters, namely the changes in parameters such as temperature, humidity, and air pressure of the actual optical system's environment over time; Output spot parameters, namely the changes in parameters such as light intensity distribution and spot size of the output spot over time.
3. The method according to claim 2, characterized in that, In step S2, the standardization process is as follows: the actual input data is converted into a d-dimensional vector x, and preferably normalized using the following formula: Where x i Let x be the i-th component of the vector x in the input data, and softmax be the normalization exponential function.
4. The method according to claim 3, characterized in that, Step S3 specifically includes: The adaptive model adjusts its parameters based on the input data, including the following sub-steps: S31. Based on the input time-independent initial lens design data, perform iterative parameter optimization on the adaptive design model part based on convolutional neural network; S32. Based on the input time-dependent performance change data of the lens during long-term operation, perform parameter iterative optimization on the adaptive design model based on recurrent neural network.
5. The method according to claim 4, characterized in that, The adaptive design model based on convolutional neural networks extracts and calculates light spot images from different planes in the data, and finally outputs the light spot image at the target plane. It mainly includes the following structure: Convolutional layers primarily extract feature information by performing convolution processing on specific regions in a light spot image using convolution kernels. Pooling layers primarily classify and merge the feature information output by different convolutional layers, reducing the overall computational cost of the model and preventing overfitting. The activation function mainly determines whether a node in a neural network can be activated. The preferred activation function in this model is shown in the following formula, which has the advantages of fast convergence and less loss of features. The fully connected layer is mainly responsible for integrating the feature information output by the pooling layer and combining it with the activation function to calculate the final output result.
6. The method according to claim 5, characterized in that, The adaptive design model based on recurrent neural networks is responsible for outputting the change of the light spot image on the target location plane over time, based on the input environmental parameters. The model is shown in the following equation: Where x(k) is the state vector of the system at time k, φ(k) is the initial condition vector at time k, C is the state feedback coefficient matrix, A is the connection weight matrix, f(x(k)) is the activation function, and J is the external input vector.
7. The method according to claim 6, characterized in that, The parameter iterative optimization of both the convolutional neural network-based adaptive design model and the recurrent neural network-based adaptive design model includes the following sub-steps: S301. Calculate the loss function, which is mainly used to measure the error between the model design results and the actual results. S302. Backpropagation mainly involves backpropagating the error obtained in the previous step in the model to calculate the gradient of the parameters in the model. S303. Parameter Update: This mainly involves updating the model's parameters based on the gradient and learning rate to improve the model's design accuracy.
8. The method according to claim 7, characterized in that, Step S5 specifically includes: generating a preliminary design scheme using an adaptive design model, calculating the planar distribution change of the light spot through each interface in the optical system based on Maxwell's equations, and finally obtaining the light spot image at the target position for simulation analysis to verify the performance of the designed lens.
9. The method according to claim 8, characterized in that, Step S6 specifically includes: In the multi-round optimization iteration based on the analysis results and design requirements, it is necessary to use the target position spot image evaluation function as the standard for iterative optimization; the evaluation function is shown in the following formula: Where n is the number of grids in the target location spot image, w i y represents the weighting coefficients of the grid. i r represents the light intensity value of the target location spot image under the grid. i The light intensity value of the target position spot image calculated by the designed lens under the grid is considered to be the light intensity value. When the error value calculated by the designed lens using the evaluation function is within the set allowable range, the iteration is considered to be complete.
10. An electronic device, characterized in that, The electronic device includes a processor, an internal bus, a network interface, memory, and non-volatile memory. The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it, forming an adaptive optical reshaping lens design method at the logical level.