Lens mold pressing equipment control method and system based on AI
Through the AI-based lens molding equipment control method, temperature monitoring network, variational autoencoder and Peltier element array are used to detect and control the temperature field abnormality, which solves the problem of temperature gradient in the mold, and improves the molding accuracy and optical performance of the lens.
Patent Information
- Application Number
- CN202510552726.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-29
- Publication Date
- 2025-07-11
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing control methods of lens molding equipment cannot effectively perceive and eliminate the internal temperature gradient unevenness of the mold, resulting in the deviation of the lens geometric accuracy and uneven internal stress distribution, affecting the imaging quality and service life of the optical components.
Using the AI-based lens molding equipment control method, temperature field data is collected through the temperature monitoring network, space-time fusion and feature extraction are performed, abnormal detection is performed using a variational autoencoder, thermal field abnormality correction instructions are generated, Peltier element array is adjusted for differential multi-rate control, and multi-objective optimization is performed in combination with optical performance prediction model to achieve closed-loop control of the temperature field.
Real-time perception and active compensation of the temperature gradient during lens molding process is achieved, and the molding accuracy and optical performance of optical components are improved.
Smart Images
Figure CN120289070A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of equipment control, and particularly to an AI-based control method and system for lens molding equipment. Background Art
[0002] Lens molding technology is an important method for manufacturing high-precision optical components. By precisely forming glass materials in a high-temperature environment, complex aspherical and free-form optical components can be produced. Existing lens molding equipment usually adopts an upper and lower symmetric die set structure, combined with a multi-zone heating system, a pressure control unit, and a cooling system to achieve heating, forming, and cooling of glass materials. However, the current control method mainly relies on fixed PID parameter adjustment and a preset temperature curve, and performs simple feedback adjustment based on limited temperature sensor data.
[0003] During the molding process of complex aspherical or free-form lenses, due to the irregularity of the die geometry and the non-linear characteristics of material heat conduction, unpredictable temperature gradients often form inside the die. These small but critical temperature non-uniformities cannot be effectively detected by traditional sparse sensing networks, ultimately leading to geometric accuracy deviations and uneven internal stress distribution of the lens, seriously affecting the imaging quality and service life of optical components. Especially in high-precision optical system applications, the small deviations caused by temperature gradients will be amplified into significant degradation of optical performance. Summary of the Invention
[0004] The main purpose of the present invention is to solve the technical problem that the existing control method for lens molding equipment cannot effectively sense and eliminate the non-uniformity of temperature gradients inside the die; The first aspect of the present invention provides an AI-based control method for lens molding equipment. The lens molding equipment includes an upper die, a lower die, a heating system, and a control system. The AI-based control method for lens molding equipment includes: Collecting temperature field data of the molding cavity by using a temperature monitoring network arranged on the surfaces of the upper die and the lower die, and performing spatio-temporal fusion and feature extraction processing on the temperature field data to obtain temperature field feature data; Performing variational autoencoder anomaly detection processing on the temperature field of the molding cavity according to the temperature field feature data to obtain a thermal field anomaly correction instruction; According to the thermal field anomaly correction instruction, adjusting the temperature field of the molding cavity by using a Peltier element array arranged on the edges of the upper die and the lower die through a control system for differential multi-rate control to obtain glass material phase change state data and temperature gradient history data; Using a preset optical property prediction model, perform multi-objective optimization analysis on the temperature field according to the phase change state data and temperature gradient history data of the glass material, and adjust the heating power of the heating system through the control system to achieve closed-loop control of the temperature field.
[0005] Optionally, in the first implementation manner of the first aspect of the present invention, collecting temperature field data of the molding cavity by using a temperature monitoring network arranged on the surfaces of the upper mold and the lower mold, and performing spatio-temporal fusion and feature extraction processing on the temperature field data to obtain temperature field feature data, including: Collecting and preprocessing temperature field data of the molding cavity by using a temperature monitoring network arranged on the surfaces of the upper mold and the lower mold; Performing time-domain interpolation calculation and spatial-domain variational approximation calculation on the temperature field data to obtain regular grid temperature data; Calculating the radial temperature gradient along the curvature direction of the mold surface, the tangential temperature gradient along the isocurvature line direction of the mold surface, and the normal temperature gradient along the normal direction of the mold surface according to the regular grid temperature data to obtain temperature field feature data.
[0006] Optionally, in the second implementation manner of the first aspect of the present invention, performing variational autoencoder anomaly detection processing on the temperature field of the molding cavity according to the temperature field feature data to obtain a thermal field anomaly correction instruction, including: Inputting the temperature field feature data into the encoding network of the variational autoencoder for feature dimensionality reduction processing to obtain a latent space feature vector; Inputting the latent space feature vector into the decoding network of the variational autoencoder for feature reconstruction processing to obtain the reconstructed temperature field feature data; Calculating the error between the temperature field feature data before and after reconstruction to obtain a radial temperature anomaly region, a tangential temperature anomaly region, and a normal temperature anomaly region; Performing spatial position correlation analysis on the radial temperature anomaly region, the tangential temperature anomaly region, and the normal temperature anomaly region to obtain a temperature field anomaly distribution map of the molding cavity, and generating the thermal field anomaly correction instruction according to the temperature field anomaly distribution map.
[0007] Optionally, in the third implementation manner of the first aspect of the present invention, performing spatial position correlation analysis on the radial temperature anomaly region, the tangential temperature anomaly region, and the normal temperature anomaly region to obtain a temperature field anomaly distribution map of the molding cavity, and generating the thermal field anomaly correction instruction according to the temperature field anomaly distribution map, including: Performing spatial coordinate transformation on the radial temperature anomaly region, the tangential temperature anomaly region, and the normal temperature anomaly region in the cylindrical coordinate system to obtain a three-dimensional distribution matrix of the anomaly region; Detect the overlapping regions in the three-dimensional distribution matrix, and obtain the abnormal overlap degree coefficient by calculating the intersection of the abnormal regions in each direction; Perform weighted superposition calculation on the abnormal regions according to the abnormal overlap degree coefficient to obtain the abnormal distribution map of the temperature field; Extract the morphological features of the abnormal regions in the abnormal distribution map of the temperature field, and combine the position and distribution form of the abnormal regions to obtain abnormal characteristic parameters including the central overheating index, the edge cooling rate, the radial temperature gradient value, and the tangential temperature non-uniformity; Quantify and classify the abnormal characteristic parameters using a fuzzy rule set, and map each abnormal characteristic parameter to the power adjustment interval of the Peltier element to obtain the abnormal correction instruction for the thermal field.
[0008] Optionally, in the fourth implementation manner of the first aspect of the present invention, the differential multi-rate control of the temperature field of the molding cavity is performed by adjusting the Peltier element array arranged at the edges of the upper mold and the lower mold through a control system according to the abnormal correction instruction for the thermal field, and the phase change state data and the temperature gradient history data of the glass material are obtained, including: Perform dynamic region division on the molding cavity according to the abnormal correction instruction for the thermal field, and allocate different control response frequencies to different regions to obtain regional differential control parameters; Calculate the power output values of the Peltier element array arranged at the edges of the upper mold and the lower mold according to the regional differential control parameters using fuzzy control rules to obtain a heat flow redirection control signal; Adjust the working state of the Peltier element array according to the heat flow redirection control signal to obtain the temperature-pressure coupling data of each region of the molding cavity; Analyze the rheological state of the glass material according to the temperature-pressure coupling data to obtain the phase change state data and the temperature gradient history data of the glass material.
[0009] Optionally, in the fifth implementation manner of the first aspect of the present invention, the analysis of the rheological state of the glass material according to the temperature-pressure coupling data to obtain the phase change state data and the temperature gradient history data of the glass material includes: Segment and intercept the temperature-pressure coupling data at a sampling time interval, and use the sliding window method to obtain a temperature change rate sequence and a pressure change rate sequence; Perform cross-correlation analysis on the temperature change rate sequence and the pressure change rate sequence, calculate the temperature-pressure response time delay and the coupling coefficient, and obtain the viscoelastic characteristic vector of the glass material; Establish a viscoelastic constitutive equation according to the viscoelastic characteristic vector, and fit the material parameters by the least squares method to obtain the rheological state equation set of the glass material. Numerically integrate and solve the rheological state equations to obtain the shear stress distribution, strain rate distribution, and apparent viscosity distribution of the glass material in each region; Store the shear stress distribution, strain rate distribution, and apparent viscosity distribution data in a circular buffer, establish a temperature field state sequence according to the timestamp marking, and obtain the phase change state data and temperature gradient history data of the glass material.
[0010] Optionally, in the sixth implementation manner of the first aspect of the present invention, the multi-objective optimization analysis of the temperature field is performed by using a preset optical performance prediction model based on the phase change state data and temperature gradient history data of the glass material, and the closed-loop control of the temperature field is realized by adjusting the heating power of the heating system through the control system, including: Input the phase change state data and temperature gradient history data of the glass material into a preset optical performance prediction model, calculate the wavefront error, aberration coefficient, and stress birefringence according to the curvature-temperature-refractive index coupling relationship, and obtain the predicted data of the optical performance index; Establish a reinforcement learning controller according to the predicted data of the optical performance index, optimize the temperature control trajectory with the predicted data of the optical performance index as the reward signal, and obtain the optimized temperature control trajectory; Perform optical performance sensitivity analysis on each region in the temperature field to obtain the control accuracy parameters of the optically sensitive region; Perform differential control calculation on the temperature control trajectory according to the control accuracy parameters to obtain the heating power compensation instructions for each region; Input the heating power compensation instructions into the process parameter-optical performance mapping table for adaptive optimization calculation, and adjust the heating power of the heating system through the control system to realize the closed-loop control of the temperature field.
[0011] The second aspect of the present invention provides an AI-based control system for a lens molding device. The lens molding device includes an upper mold, a lower mold, a heating system, and a control system. The AI-based control system for a lens molding device includes: A data acquisition module, configured to collect temperature field data of the molding cavity by using a temperature monitoring network arranged on the surfaces of the upper mold and the lower mold, and perform spatio-temporal fusion and feature extraction processing on the temperature field data to obtain temperature field feature data; An anomaly detection module, configured to perform variational autoencoder anomaly detection processing on the temperature field of the molding cavity according to the temperature field feature data to obtain a thermal field anomaly correction instruction; A differential control module, configured to perform differential multi-rate control on the temperature field of the molding cavity by adjusting, according to the thermal field anomaly correction instruction, a Peltier element array arranged at the edges of the upper mold and the lower mold through a control system, so as to obtain glass material phase change state data and temperature gradient history data; A closed-loop optimization module, configured to perform multi-objective optimization analysis on the temperature field by using a preset optical performance prediction model according to the glass material phase change state data and the temperature gradient history data, and adjust the heating power of the heating system through the control system to achieve closed-loop control of the temperature field.
[0012] The above AI-based lens molding equipment control method and system first collect temperature field data of the molding cavity by using a temperature monitoring network arranged on the mold surface, perform spatio-temporal fusion and feature extraction; then perform anomaly detection on the temperature field through a variational autoencoder to generate a thermal field anomaly correction instruction; then use the control system to adjust the Peltier element array at the edge of the mold to implement differential multi-rate control to obtain glass material phase change state data and temperature gradient history data; finally, perform multi-objective optimization on the temperature field based on the optical performance prediction model, and achieve closed-loop control of the temperature field by adjusting the power of the heating system. The present invention can realize real-time perception and active compensation of the temperature gradient during the lens molding process, and effectively improve the molding accuracy and optical performance of optical elements.
[0013] Other features and advantages of the present invention will be described in the following specification, and, in part, will be obvious from the specification, or will be understood by practicing the present invention. The objectives and other advantages of the present invention are realized and obtained by the structures specifically pointed out in the specification, the claims, and the drawings.
[0014] To make the above objectives, features, and advantages of the present invention more obvious and understandable, the following specific preferred embodiments are given, and in conjunction with the accompanying drawings, the detailed description is as follows. Description of the Drawings
[0015] Figure 1 It is a schematic diagram of the first embodiment of the AI-based lens molding equipment control method in an embodiment of the present invention; Figure 2 It is a schematic diagram of an embodiment of the AI-based lens molding equipment control system in an embodiment of the present invention. Detailed Embodiments
[0016] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.
[0017] As used in the embodiments of the present invention, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally further include other unlisted steps or units, or may optionally further include other steps or units inherent to these processes, methods, products, or devices.
[0018] To facilitate the understanding of this embodiment, first, a control method for an AI-based lens molding device disclosed in the embodiments of the present invention will be introduced in detail. As Figure 1 shown, this method includes the following steps: 101. Use the temperature monitoring network arranged on the surfaces of the upper mold and the lower mold to collect the temperature field data of the molding cavity, and perform spatio-temporal fusion and feature extraction processing on the temperature field data to obtain temperature field feature data; In an embodiment of the present invention, the step of using the temperature monitoring network arranged on the surfaces of the upper mold and the lower mold to collect the temperature field data of the molding cavity, and performing spatio-temporal fusion and feature extraction processing on the temperature field data to obtain temperature field feature data includes: using the temperature monitoring network arranged on the surfaces of the upper mold and the lower mold to collect and preprocess the temperature field data of the molding cavity; performing time-domain interpolation calculation and spatial-domain variational approximation calculation on the temperature field data to obtain regular grid temperature data; calculating the radial temperature gradient along the curvature direction of the mold surface, the tangential temperature gradient along the isocurvature line direction of the mold surface, and the normal temperature gradient along the normal direction of the mold surface according to the regular grid temperature data to obtain temperature field feature data.
[0019] Specifically, 102. According to the temperature field feature data, perform variational autoencoder anomaly detection processing on the temperature field of the molding cavity to obtain a thermal field anomaly correction instruction; In an embodiment of the present invention, the step of performing variational autoencoder anomaly detection processing on the temperature field of the molding cavity according to the temperature field feature data to obtain a thermal field anomaly correction instruction includes: inputting the temperature field feature data into the encoding network of the variational autoencoder for feature dimensionality reduction processing to obtain a latent space feature vector; inputting the latent space feature vector into the decoding network of the variational autoencoder for feature reconstruction processing to obtain the reconstructed temperature field feature data; calculating the error between the temperature field feature data before and after reconstruction to obtain the radial temperature anomaly region, the tangential temperature anomaly region, and the normal temperature anomaly region; performing spatial position correlation analysis on the radial temperature anomaly region, the tangential temperature anomaly region, and the normal temperature anomaly region to obtain the temperature field anomaly distribution map of the molding cavity, and generating the thermal field anomaly correction instruction according to the temperature field anomaly distribution map.
[0020] Specifically, Further, the spatial position correlation analysis of the radial temperature anomaly region, the tangential temperature anomaly region, and the normal temperature anomaly region is performed to obtain the temperature field anomaly distribution map of the molding cavity, and generating the thermal field anomaly correction instruction according to the temperature field anomaly distribution map includes: performing spatial coordinate transformation on the radial temperature anomaly region, the tangential temperature anomaly region, and the normal temperature anomaly region in the cylindrical coordinate system to obtain the three-dimensional distribution matrix of the anomaly regions; detecting the overlapping regions of the anomaly regions in the three-dimensional distribution matrix, and calculating the intersection of the anomaly regions in each direction to obtain the anomaly overlap coefficient; performing weighted superposition calculation on the anomaly regions according to the anomaly overlap coefficient to obtain the temperature field anomaly distribution map; extracting the morphological features of the anomaly regions in the temperature field anomaly distribution map, and combining the position and distribution form of the anomaly regions to obtain the anomaly characteristic parameters including the central overheating index, the edge cooling rate, the radial temperature gradient value, and the tangential temperature non-uniformity; quantifying and grading the anomaly characteristic parameters by using the fuzzy rule set, and mapping each anomaly characteristic parameter to the power adjustment interval of the Peltier element to obtain the thermal field anomaly correction instruction.
[0021] Specifically, 103. According to the thermal field anomaly correction instruction, the temperature field of the molding cavity is differentially multi-rate controlled by adjusting the Peltier element array arranged on the edges of the upper mold and the lower mold through the control system, and the phase change state data and the temperature gradient history data of the glass material are obtained. In an embodiment of the present invention, the differentially multi-rate control of the temperature field of the molding cavity by adjusting the Peltier element array arranged on the edges of the upper mold and the lower mold according to the thermal field anomaly correction instruction to obtain the phase change state data and the temperature gradient history data of the glass material includes: dynamically dividing the molding cavity according to the thermal field anomaly correction instruction, and assigning different control response frequencies to different regions to obtain the regional differential control parameters; calculating the power output values of the Peltier element array arranged on the edges of the upper mold and the lower mold according to the regional differential control parameters by using the fuzzy control rule to obtain the heat flow redirection control signal; adjusting the working state of the Peltier element array according to the heat flow redirection control signal to obtain the temperature-pressure coupling data of each region of the molding cavity; analyzing the rheological state of the glass material according to the temperature-pressure coupling data to obtain the phase change state data and the temperature gradient history data of the glass material.
[0022] Specifically, Further, analyzing the rheological state of the glass material based on the temperature-pressure coupling data to obtain the phase change state data and temperature gradient history data of the glass material includes: segmenting and intercepting the temperature-pressure coupling data at a sampling time interval, and using a sliding window method to obtain a temperature change rate sequence and a pressure change rate sequence; performing cross-correlation analysis on the temperature change rate sequence and the pressure change rate sequence, calculating the temperature-pressure response time delay and coupling coefficient, and obtaining the viscoelastic characteristic vector of the glass material; establishing a viscoelastic constitutive equation based on the viscoelastic characteristic vector, and fitting the material parameters by the least square method to obtain the rheological state equation set of the glass material; performing numerical integration on the rheological state equation set to solve, and obtaining the shear stress distribution, strain rate distribution and apparent viscosity distribution of the glass material in each region; storing the shear stress distribution, strain rate distribution and apparent viscosity distribution data in a circular buffer, and establishing a temperature field state sequence according to the time stamp mark to obtain the phase change state data and temperature gradient history data of the glass material.
[0023] Specifically, 104. Using a preset optical performance prediction model, based on the phase change state data and temperature gradient history data of the glass material, perform multi-objective optimization analysis on the temperature field, and adjust the heating power of the heating system through a control system to achieve closed-loop control of the temperature field.
[0024] In an embodiment of the present invention, the using a preset optical performance prediction model, based on the phase change state data and temperature gradient history data of the glass material, perform multi-objective optimization analysis on the temperature field, and adjust the heating power of the heating system through the control system to achieve closed-loop control of the temperature field includes: inputting the phase change state data and temperature gradient history data of the glass material into a preset optical performance prediction model, calculating the wavefront error, aberration coefficient and stress birefringence according to the curvature-temperature-refractive index coupling relationship, and obtaining the predicted data of the optical performance index; establishing a reinforcement learning controller based on the predicted data of the optical performance index, and optimizing the temperature control trajectory with the predicted data of the optical performance index as the reward signal to obtain the optimized temperature control trajectory; performing optical performance sensitivity analysis on each region in the temperature field to obtain the control precision parameters of the optical sensitive region; performing differential control calculation on the temperature control trajectory according to the control precision parameters to obtain the heating power compensation instruction for each region; inputting the heating power compensation instruction into a process parameter-optical performance mapping table for adaptive optimization calculation, and adjusting the heating power of the heating system through the control system to achieve closed-loop control of the temperature field.
[0025] Specifically, In this embodiment, first, the temperature field data of the molding cavity is collected by using the temperature monitoring network arranged on the surface of the mold, and spatio-temporal fusion and feature extraction are performed; then, the variational autoencoder is used to detect anomalies in the temperature field to generate a thermal field anomaly correction instruction; next, the control system is used to adjust the Peltier element array at the edge of the mold to implement differential multi-rate control, and the phase change state data of the glass material and the temperature gradient history data are obtained; finally, based on the optical performance prediction model, multi-objective optimization of the temperature field is performed, and the closed-loop control of the temperature field is achieved by adjusting the power of the heating system. The present invention can realize the real-time perception and active compensation of the temperature gradient during the lens molding process, and effectively improve the molding accuracy and optical performance of the optical element.
[0026] The control method of the AI-based lens molding equipment in the embodiment of the present invention is described above. Next, the control system of the AI-based lens molding equipment in the embodiment of the present invention is described. The lens molding equipment includes an upper mold, a lower mold, a heating system, and a control system. Please refer to Figure 2 , an embodiment of the control system of the AI-based lens molding equipment in the embodiment of the present invention includes: A data acquisition module 201, configured to collect temperature field data of the molding cavity by using a temperature monitoring network arranged on the surfaces of the upper mold and the lower mold, and perform spatio-temporal fusion and feature extraction processing on the temperature field data to obtain temperature field feature data; An anomaly detection module 202, configured to perform variational autoencoder anomaly detection processing on the temperature field of the molding cavity according to the temperature field feature data to obtain a thermal field anomaly correction instruction; A differential control module 203, configured to perform differential multi-rate control on the temperature field of the molding cavity by adjusting the Peltier element array arranged at the edges of the upper mold and the lower mold through the control system according to the thermal field anomaly correction instruction to obtain phase change state data of the glass material and temperature gradient history data; A closed-loop optimization module 204, configured to perform multi-objective optimization analysis on the temperature field by using a preset optical performance prediction model according to the phase change state data of the glass material and the temperature gradient history data, and realize closed-loop control of the temperature field by adjusting the heating power of the heating system through the control system.
[0027] In the embodiments of the present invention, the AI-based lens molding equipment control system operates the above-mentioned AI-based lens molding equipment control method. The AI-based lens molding equipment control system first collects the temperature field data of the molding cavity by using the temperature monitoring network arranged on the surface of the mold, and performs spatio-temporal fusion and feature extraction; then performs anomaly detection on the temperature field through a variational autoencoder to generate a thermal field anomaly correction instruction; then uses the control system to adjust the Peltier element array at the edge of the mold to implement differential multi-rate control to obtain the phase change state data of the glass material and the temperature gradient history data; finally, performs multi-objective optimization on the temperature field based on the optical performance prediction model, and realizes the closed-loop control of the temperature field by adjusting the power of the heating system. The present invention can realize the real-time perception and active compensation of the temperature gradient in the lens molding process, and effectively improve the molding accuracy and optical performance of optical elements.
[0028] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described system, device, or unit can refer to the corresponding processes in the foregoing method embodiments, and will not be described herein again.
[0029] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.
[0030] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.
Claims
1. A control method for an AI-based lens molding device, characterized in that, The lens molding equipment includes an upper mold, a lower mold, a heating system, and a control system. The AI-based control method for the lens molding equipment includes: Collecting temperature field data of the molding cavity using a temperature monitoring network arranged on the surfaces of the upper mold and the lower mold, and performing spatio-temporal fusion and feature extraction processing on the temperature field data to obtain temperature field feature data; Performing variational autoencoder anomaly detection processing on the temperature field of the molding cavity according to the temperature field feature data to obtain a thermal field anomaly correction instruction; According to the thermal field anomaly correction instruction, adjusting the Peltier element array arranged on the edges of the upper mold and the lower mold through the control system to perform differential multi-rate control on the temperature field of the molding cavity, and obtaining glass material phase change state data and temperature gradient history data; Using a preset optical property prediction model to perform multi-objective optimization analysis on the temperature field according to the glass material phase change state data and the temperature gradient history data, and adjusting the heating power of the heating system through the control system to achieve closed-loop control of the temperature field.
2. The control method of the lens molding equipment according to claim 1, characterized in that The step of collecting temperature field data of the molding cavity using a temperature monitoring network arranged on the surfaces of the upper mold and the lower mold, and performing spatio-temporal fusion and feature extraction processing on the temperature field data to obtain temperature field feature data includes: Collecting and preprocessing the temperature field data of the molding cavity using a temperature monitoring network arranged on the surfaces of the upper mold and the lower mold; Performing time-domain interpolation calculation and spatial-domain variational approximation calculation on the temperature field data to obtain regular grid temperature data; Calculating the radial temperature gradient along the curvature direction of the mold surface, the tangential temperature gradient along the isocurvature line direction of the mold surface, and the normal temperature gradient along the normal direction of the mold surface according to the regular grid temperature data to obtain temperature field feature data.
3. The control method of the lens molding equipment according to claim 1, characterized in that, The step of performing variational autoencoder anomaly detection processing on the temperature field of the molding cavity according to the temperature field feature data to obtain a thermal field anomaly correction instruction includes: Inputting the temperature field feature data into the encoding network of the variational autoencoder for feature dimensionality reduction processing to obtain a latent space feature vector; Inputting the latent space feature vector into the decoding network of the variational autoencoder for feature reconstruction processing to obtain the reconstructed temperature field feature data; Calculating the error between the temperature field feature data before and after reconstruction to obtain a radial temperature anomaly region, a tangential temperature anomaly region, and a normal temperature anomaly region; Performing spatial position correlation analysis on the radial temperature anomaly region, the tangential temperature anomaly region, and the normal temperature anomaly region to obtain a temperature field anomaly distribution map of the molding cavity, and generating the thermal field anomaly correction instruction according to the temperature field anomaly distribution map.
4. The method for controlling a lens molding device according to claim 3, characterized in that, The step of performing spatial position correlation analysis on the radial temperature anomaly region, the tangential temperature anomaly region, and the normal temperature anomaly region to obtain a temperature field anomaly distribution map of the molding cavity, and generating the thermal field anomaly correction instruction according to the temperature field anomaly distribution map includes: Performing spatial coordinate transformation on the radial temperature anomaly region, the tangential temperature anomaly region, and the normal temperature anomaly region in the cylindrical coordinate system to obtain a three-dimensional distribution matrix of the anomaly regions; Detect the overlapping regions of the abnormal regions in the three-dimensional distribution matrix, and obtain the abnormal overlap degree coefficient by calculating the intersection of the abnormal regions in each direction; Perform weighted superposition calculation on the abnormal regions according to the abnormal overlap degree coefficient to obtain the abnormal distribution map of the temperature field; Extract the morphological features of the abnormal regions in the abnormal distribution map of the temperature field, and combine the positions and distribution morphologies of the abnormal regions to obtain abnormal characteristic parameters including the central overheating index, the edge cooling rate, the radial temperature gradient value, and the tangential temperature non-uniformity; Quantify and classify the abnormal characteristic parameters using a fuzzy rule set, and map each abnormal characteristic parameter to the power adjustment range of the Peltier element to obtain a thermal field abnormal correction instruction; 5. The control method of the lens molding equipment according to claim 1, wherein According to the thermal field abnormal correction instruction, adjust the temperature field of the molding cavity by a differential multi-rate control through a control system for the Peltier element array arranged at the edges of the upper mold and the lower mold, and obtain the phase change state data and the temperature gradient history data of the glass material, including: Dynamically divide the molding cavity according to the thermal field abnormal correction instruction, and assign different control response frequencies to different regions to obtain regional differential control parameters; Calculate the power output values of the Peltier element array arranged at the edges of the upper mold and the lower mold according to the regional differential control parameters using fuzzy control rules to obtain a heat flow redirection control signal; Adjust the working state of the Peltier element array according to the heat flow redirection control signal to obtain the temperature-pressure coupling data of each region of the molding cavity; Analyze the rheological state of the glass material according to the temperature-pressure coupling data to obtain the phase change state data and the temperature gradient history data of the glass material.
6. The control method of the lens molding equipment according to claim 5, characterized in that The analysis of the rheological state of the glass material according to the temperature-pressure coupling data to obtain the phase change state data and the temperature gradient history data of the glass material includes: Segment and intercept the temperature-pressure coupling data at a sampling time interval, and use the sliding window method to obtain the temperature change rate sequence and the pressure change rate sequence; Perform cross-correlation analysis on the temperature change rate sequence and the pressure change rate sequence, calculate the temperature-pressure response time delay and the coupling coefficient, and obtain the viscoelastic characteristic vector of the glass material; Establish a viscoelastic constitutive equation according to the viscoelastic characteristic vector, and fit the material parameters by the least squares method to obtain the rheological state equation set of the glass material; Perform numerical integration and solution on the rheological state equation set to obtain the shear stress distribution, the strain rate distribution, and the apparent viscosity distribution of the glass material in each region; Store the shear stress distribution, the strain rate distribution, and the apparent viscosity distribution data in a circular buffer, and establish a temperature field state sequence according to the time stamp mark to obtain the phase change state data and the temperature gradient history data of the glass material.
7. The control method of the lens molding equipment according to claim 1, wherein Perform multi-objective optimization analysis on the temperature field using a preset optical property prediction model according to the phase change state data and the temperature gradient history data of the glass material, and adjust the heating power of the heating system through the control system to achieve closed-loop control of the temperature field, including: Input the phase change state data and temperature gradient history data of the glass material into a preset optical performance prediction model, and calculate the wavefront error, aberration coefficient, and stress birefringence according to the curvature-temperature-refractive index coupling relationship to obtain the predicted data of optical performance indicators; Establish a reinforcement learning controller based on the predicted data of the optical performance indicators, and optimize the temperature control trajectory with the predicted data of the optical performance indicators as the reward signal to obtain the optimized temperature control trajectory; Conduct an optical performance sensitivity analysis on each region in the temperature field to obtain the control accuracy parameters of the optically sensitive regions; Perform differential control calculations on the temperature control trajectory according to the control accuracy parameters to obtain the heating power compensation instructions for each region; Input the heating power compensation instructions into the process parameter-optical performance mapping table for adaptive optimization calculations, and adjust the heating power of the heating system through the control system to achieve closed-loop control of the temperature field.
8. An AI-based control system for a lens molding device, characterized in that, The lens molding equipment includes an upper mold, a lower mold, a heating system, and a control system. The AI-based control system for the lens molding equipment includes: A data acquisition module for collecting the temperature field data of the molding cavity by using the temperature monitoring network arranged on the surfaces of the upper mold and the lower mold, and performing spatio-temporal fusion and feature extraction processing on the temperature field data to obtain the temperature field feature data; An anomaly detection module for performing variational autoencoder anomaly detection processing on the temperature field of the molding cavity according to the temperature field feature data to obtain a thermal field anomaly correction instruction; A differential control module for performing differential multi-rate control on the temperature field of the molding cavity by adjusting the Peltier element array arranged on the edges of the upper mold and the lower mold through the control system according to the thermal field anomaly correction instruction to obtain the phase change state data and temperature gradient history data of the glass material; A closed-loop optimization module for performing multi-objective optimization analysis on the temperature field by using a preset optical performance prediction model according to the phase change state data and temperature gradient history data of the glass material, and adjusting the heating power of the heating system through the control system to achieve closed-loop control of the temperature field.