Precision optimization method and system of roll-to-roll high-precision projection exposure machine

Abstract

The application relates to the field of photoetching technology, and discloses a precision optimization method and system of a roll-to-roll high-precision projection exposure machine, the method comprising the following steps: acquiring microscopic depth data, material vibration data and an exposure overlapping area; performing energy distribution analysis according to the microscopic depth data to obtain exposure energy distribution; performing frequency domain analysis according to the material vibration data to obtain vibration frequency domain characteristics; performing correlation calculation according to the exposure energy distribution and the vibration frequency domain characteristics to obtain a correlation influence matrix; performing energy accumulation analysis on the exposure overlapping area to obtain an overlapping correction factor; and performing exposure parameter optimization according to the overlapping correction factor and the correlation influence matrix to update exposure machine control parameters. The method has the following effects: the method can improve the working precision of the roll-to-roll projection exposure machine.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of photolithography exposure technology, and in particular to a precision optimization method and system for a roll-to-roll high-precision projection exposure machine. BACKGROUND

[0002] With the rapid development of the electronic industry, especially the increasing demand for flexible electronic devices and high-density interconnection circuit boards, the application of roll-to-roll (R2R) high-precision projection exposure machines becomes particularly critical. Such devices are mainly used to accurately transfer fine patterns onto flexible substrates such as polyimide and other materials during the manufacturing process, which requires extremely high precision control at the nanometer level. Traditional batch processing methods have been unable to meet the needs of large-scale production, while R2R technology has become the mainstream choice in the industry due to its continuity and efficiency. In this context, how to optimize the precision of R2R high-precision projection exposure machines has become an important issue to improve product quality and production efficiency.

[0003] In one existing technology, in order to improve the precision of roll-to-roll high-precision projection exposure machines, a method combining multi-step calibration and real-time feedback control is adopted. First, during the installation stage of the device, a high-precision laser interferometer is used for preliminary calibration to ensure that the parallelism and perpendicularity errors between each motion axis are within the allowable range. Then, during actual operation, a sensor network integrated within the system is used to monitor the substrate position, speed, and distance between the exposure head and the substrate in real time, and these data are fed back to the control system. Based on the collected data, the control system can dynamically adjust the working parameters of each component to compensate for precision deviations caused by factors such as temperature changes and mechanical wear. In addition, modern R2R exposure machines are also equipped with advanced image processing software that can preprocess the design patterns before exposure, removing defects or noise that affect the final imaging quality, thereby further improving the accuracy of pattern transfer.

[0004] In traditional exposure processes, it is assumed that the substrate surface is completely flat and rigid, which allows the optical system to be designed based on a fixed focal length. However, in reality, flexible substrates not only have microscopic undulations due to uneven thickness, but also stretch or compress under tension. When the exposure head attempts to focus on such a non-ideal surface, its pre-set focal length parameter cannot adapt to this dynamic change, resulting in blurred or distorted images. Secondly, temperature changes during material processing can also cause thermal expansion or contraction effects, further exacerbating the degree of deformation of the material surface, ultimately leading to insufficient precision of the roll-to-roll projection exposure machine. SUMMARY

[0005] The application provides a precision optimization method and system of a roll-to-roll high-precision projection exposure machine, so as to improve the working precision of the roll-to-roll projection exposure machine.

[0006] In a first aspect, to solve the above technical problems, the application provides a precision optimization method of a roll-to-roll high-precision projection exposure machine, comprising:

[0007] Obtaining microscopic depth data, material vibration data and an exposure overlapping area;

[0008] Performing energy distribution analysis according to the microscopic depth data to obtain exposure energy distribution;

[0009] Performing frequency domain analysis according to the material vibration data to obtain vibration frequency domain characteristics;

[0010] Performing correlation calculation according to the exposure energy distribution and the vibration frequency domain characteristics to obtain a correlation influence matrix;

[0011] Performing energy accumulation analysis on the exposure overlapping area to obtain an overlapping correction factor;

[0012] Performing exposure parameter optimization according to the overlapping correction factor and the correlation influence matrix to update exposure machine control parameters.

[0013] In an optional embodiment, the energy distribution analysis according to the microscopic depth data comprises:

[0014] Dividing the microscopic depth data into grid units with a preset interval;

[0015] Performing numerical differential calculation according to the grid units and the microscopic depth data to obtain depth variation amplitude;

[0016] Performing second derivative calculation according to the grid units and the microscopic depth data to obtain depth variation frequency;

[0017] Performing weighted fusion according to the depth variation amplitude and the depth variation frequency to obtain a depth gradient value;

[0018] Inputting the depth gradient value and pre-stored exposure parameters into a preset energy compensation model to output an exposure energy result;

[0019] Performing distribution mapping according to the exposure energy result to obtain exposure energy distribution.

[0020] In an optional embodiment, the training process of the energy compensation model comprises:

[0021] The historical micro-depth data and the historical exposure parameters are taken as inputs, the historical exposure results are taken as outputs, and the training set and the verification set are divided through normalization processing;

[0022] Initialize the multi-layer perception model parameters;

[0023] Optimize the parameters through iterative training;

[0024] After the loss of the verification set converges or reaches the maximum number of iterations, if the error of the test set is lower than a preset threshold, the model parameters are saved as the final energy compensation model.

[0025] In an optional implementation, the frequency domain analysis based on the material vibration data to obtain vibration frequency domain features includes:

[0026] Wavelet denoising is performed on the material vibration data to obtain denoised vibration data;

[0027] Wavelet transform is performed on the denoised vibration data to obtain vibration frequency domain data;

[0028] Main frequency identification is performed based on the vibration frequency domain data to obtain vibration main frequency;

[0029] High-frequency interval identification is performed based on the vibration frequency domain data to obtain vibration high-frequency interval;

[0030] Normalization fusion is performed based on the vibration main frequency and the vibration high-frequency interval to obtain vibration frequency domain features.

[0031] In an optional implementation, the correlation calculation based on the exposure energy distribution and the vibration frequency domain features to obtain an association influence matrix includes:

[0032] The vibration main frequency and the vibration high-frequency interval extracted from the vibration frequency domain features are subjected to block processing to form a frequency vector;

[0033] The exposure energy result of each grid unit is extracted based on the exposure energy distribution;

[0034] Correlation coefficient calculation is performed based on the frequency vector and the exposure energy result to obtain vibration exposure correlation degree;

[0035] Normalization processing is performed on the vibration exposure correlation degree to obtain normalized correlation degree;

[0036] The normalized correlation degree is used to construct an association influence matrix according to the correspondence between the grid unit and the frequency vector.

[0037] In an optional implementation, the energy accumulation analysis on the exposure overlapping area to obtain an overlapping correction factor includes:

[0038] integrating according to the exposure overlap area and the exposure energy distribution, to obtain cumulative exposure energy;

[0039] performing correction factor calculation according to the cumulative exposure energy, to obtain attenuation factor and enhancement factor;

[0040] when the cumulative exposure energy is less than a preset energy threshold, taking the enhancement factor as the overlap correction factor;

[0041] when the cumulative exposure energy is greater than or equal to the energy threshold, taking the attenuation factor as the overlap correction factor.

[0042] In an optional implementation, the exposure parameter optimization according to the overlap correction factor and the correlation influence matrix, to update the exposure machine control parameter, comprises:

[0043] performing weighted fusion according to the overlap correction factor and the correlation influence matrix, to obtain a modified influence matrix;

[0044] performing particle swarm parameter initialization according to the modified influence matrix;

[0045] when the number of iterations reaches a preset upper limit or the fitness function of the particle meets a constraint condition, completing the iteration, and outputting the exposure machine control parameter.

[0046] In a second aspect, the present application provides a precision optimization system of a roll-to-roll high-precision projection exposure machine, comprising:

[0047] a data acquisition module, configured to acquire micro-depth data, material vibration data and exposure overlap area;

[0048] an energy distribution module, configured to perform energy distribution analysis according to the micro-depth data, to obtain exposure energy distribution;

[0049] a vibration analysis module, configured to perform frequency domain analysis according to the material vibration data, to obtain vibration frequency domain characteristics;

[0050] a correlation calculation module, configured to perform correlation calculation according to the exposure energy distribution and the vibration frequency domain characteristics, to obtain a correlation influence matrix;

[0051] an overlap correction module, configured to perform energy accumulation analysis on the exposure overlap area, to obtain an overlap correction factor;

[0052] a parameter update module, configured to perform exposure parameter optimization according to the overlap correction factor and the correlation influence matrix, to update the exposure machine control parameter.

[0053] In a third aspect, the present application also provides an electronic device comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor implements the precision optimization method of the roll-to-roll high-precision projection exposure machine according to any one of the above aspects when executing the computer program.

[0054] In a fourth aspect, the present application also provides a computer-readable storage medium comprising a stored computer program, wherein the computer-readable storage medium controls the device where it is located to execute the precision optimization method of the roll-to-roll high-precision projection exposure machine according to any one of the above aspects when the computer program runs.

[0055] Compared with the prior art, the present application has the following beneficial effects:

[0056] (1) The process of obtaining micro-depth data, material vibration data, and exposure overlap area ensures a comprehensive understanding of the material surface characteristics and processing environment. Through precise data acquisition means, including high-resolution sensors and image processing technology, subtle changes in the material surface and vibration conditions can be effectively captured, and the overlap area in the exposure process can be identified. These high-quality data provide a solid foundation for subsequent analysis, improving the reliability and accuracy of the entire system.

[0057] (2) Energy distribution analysis is performed according to the micro-depth data to obtain the exposure energy distribution. This step uses advanced algorithm models to extract energy demand information at different positions from the micro-depth data, achieving precise calculation of the exposure energy distribution. This method not only adapts to complex surface morphology, but also effectively avoids quality problems caused by uneven energy, improving the consistency and quality of the final pattern.

[0058] (3) Frequency domain analysis is performed according to the material vibration data to obtain vibration frequency domain characteristics. Through frequency domain analysis of material vibration data, the system can identify different frequency components of vibration and their impact on the exposure process. This frequency spectrum-based analysis method helps to optimize exposure parameters, reduce vibration interference, and improve pattern precision and stability.

[0059] (4) Correlation calculation is performed according to the exposure energy distribution and the vibration frequency domain characteristics to obtain the correlation influence matrix. Combining the exposure energy distribution and the vibration frequency domain characteristics, the correlation analysis method generates the correlation influence matrix, revealing the interaction between the two. This quantitative analysis provides a scientific basis for adjusting exposure parameters, enhancing the adaptability and flexibility of the system.

[0060] (5) Energy accumulation analysis is performed on the exposure overlap area to obtain an overlap correction factor. By performing detailed analysis on the energy accumulation of the exposure overlap area, the system can determine an overlap correction factor to compensate for the energy accumulation effect caused by multiple exposures. This method ensures the uniformity of exposure in the overlap area, prevents overexposure or underexposure, and ensures the consistency and integrity of the image.

[0061] (6) Exposure parameter optimization is performed according to the overlap correction factor and the correlation influence matrix to update the exposure machine control parameters. The last step integrates the results of all previous steps, and optimizes and adjusts the exposure parameters by intelligent algorithm to generate updated exposure machine control parameters. This process not only considers the influence of the overlap correction factor and the correlation influence matrix, but also ensures the optimization of the overall exposure strategy, significantly improves the accuracy and efficiency of the exposure process, and ensures high-quality manufacturing results. BRIEF DESCRIPTION OF DRAWINGS

[0062] Figure 1 is a precision optimization method flowchart of a roll-to-roll high-precision projection exposure machine provided by the first embodiment of the present application;

[0063] Figure 2 is a precision optimization system structure schematic diagram of a roll-to-roll high-precision projection exposure machine provided by the second embodiment of the present application. DETAILED DESCRIPTION

[0064] The technical solutions in the embodiments of the present application will be described in detail below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor are within the scope of protection of the present application.

[0065] Referring to Figure 1 , the first embodiment of the present application provides a precision optimization method for a roll-to-roll high-precision projection exposure machine, including the following steps:

[0066] S11, obtaining micro-depth data, material vibration data and exposure overlap area;

[0067] S12, performing energy distribution analysis according to the micro-depth data to obtain exposure energy distribution;

[0068] S13, performing frequency domain analysis according to the material vibration data to obtain vibration frequency domain characteristics;

[0069] S14, performing correlation calculation according to the exposure energy distribution and the vibration frequency domain characteristics to obtain a correlation influence matrix;

[0070] S15, performing energy accumulation analysis on the exposure overlap area to obtain an overlap correction factor;

[0071] S16, performing exposure parameter optimization according to the overlap correction factor and the correlation influence matrix to update the exposure machine control parameter.

[0072] In step S11, micro-depth data, material vibration data and exposure overlap area are obtained.

[0073] In an embodiment, the synchronous acquisition of the three types of key data is realized by integrating a multi-source sensing system: the micro-depth data is measured by a white light interference scanner with 5 nm step precision for three-dimensional topography measurement, 4096 sampling points of depth information per square centimeter are generated, and the data is stored as a 16-bit floating-point three-dimensional matrix; the material vibration data is captured by a three-axis MEMS acceleration sensor distributed at the transmission roller at a sampling rate of 100 kHz, the vibration transmission path is verified by a laser Doppler vibration tester, and the original time domain data is stored as a structured array containing frequency-amplitude-phase parameters after FFT conversion; the exposure overlap area is captured by a high-speed linear array CCD (5000 fps) triggered by an encoder to capture the real-time position of the moving substrate, and the step error predicted by the feedforward motion control model is combined to map the overlap area coordinates to a two-dimensional probability distribution graph containing a ±15 μm deviation range, the three types of data are stored in an industrial-level time sequence database through a unified timestamp, and a spatial coordinate conversion matrix is established to realize cross-sensor data fusion.

[0074] In a specific embodiment, taking flexible circuit board exposure as an example: when the substrate passes through the exposure station at a speed of 0.2 m / s, the white light interference scanner scans the PI film surface at an interval of 5 μm, and it is found that there is a 0.32 μm concave topography in a certain area (stored as a three-dimensional matrix element X

[128]

[128] = 0.32); at the same time, the three-axis MEMS sensor at the transmission roller detects that there is an abnormal vibration of 85 Hz / 0.15 g in the X-axis direction (stored as a structure {frequency: 85.3 Hz, amplitude: 0.147 g, phase: 32°} after FFT conversion), and the vibration is verified by a laser Doppler vibration tester to find that it causes an amplitude amplification effect of 0.8 μm; the high-speed linear array CCD captures the exposure area of the N+1 frame in real time, which produces a +0.125 μm offset compared with the theoretical position, and the three types of data are unified in coordinates through the spatial conversion matrix to form a correlated data set with microsecond-level synchronization accuracy in the time sequence database, which lays a foundation for subsequent vibration-topography coupling analysis.

[0075] In step S12, energy distribution analysis is performed according to the micro-depth data to obtain an exposure energy distribution.

[0076] In an embodiment, the micro-depth data is divided into grid units of a preset interval; a numerical differentiation calculation is performed according to the grid units and the micro-depth data to obtain a depth variation amplitude; a second derivative calculation is performed according to the grid units and the micro-depth data to obtain a depth variation frequency; a weighted fusion is performed according to the depth variation amplitude and the depth variation frequency to obtain a depth gradient value; the depth gradient value and a pre-stored exposure parameter are input into a preset energy compensation model to output an exposure energy result; and a distribution mapping is performed according to the exposure energy result to obtain an exposure energy distribution.

[0077] It is worth noting that, before the second derivative calculation, high-speed smoothing processing is performed on the micro-depth data to reduce noise interference.

[0078] It is worth noting that, in the energy distribution analysis, the three-dimensional micro-depth data collected by the high-precision white light interferometer is first intelligently grid-divided according to process characteristics, for example, a 50 μm x 50 μm square grid unit (about covering 20 sampling points) is used for the flexible circuit board substrate, when it is detected that there is an abrupt edge similar to a micro-crack in a certain region, the system will automatically dynamically encrypt the grid to 25 μm x 25 μm to improve the analysis precision.

[0079] It is worth noting that, the fluctuation amplitude of each grid unit is quantified by calculating the maximum depth difference in the adjacent four directions (up, down, left and right), for example, when the east-west depth difference of a certain polyimide film region is measured to be 3.2 μm, the gradient value of this direction will be marked as a red warning level; at the same time, the degree of depth variation is captured by the second derivative operation, for example, when it is detected that there is a periodic fluctuation of 0.05 μm per μm interval in a certain 0.5 mm 2 region, the system will identify it as a high-frequency tremor feature region.

[0080] It is worth noting that, after the gradient amplitude and the fluctuation frequency are fused with a weight ratio of 6:4, a quantized heat map containing topography complexity classification is generated (for example, a flat area is marked as level 1 and a wrinkle area is marked as level 5). This index is input into the energy compensation model based on 100,000 sets of experimental data calibration together with the parameters such as the spot diameter of the exposure machine and the focal depth limit.

[0081] In an embodiment, the training process of the energy compensation model includes:

[0082] The historical micro-depth data and the historical exposure parameters are taken as inputs and the historical exposure results are taken as outputs, and the training set and the validation set are divided by normalization processing; the multi-layer perception model parameters are initialized; the parameters are optimized by iterative training; after the validation set loss converges or the maximum iteration number is reached, if the test set error is lower than a preset threshold, the model parameters are saved as the final energy compensation model.

[0083] In an embodiment, the training process of the energy compensation model comprises: firstly, multi-dimensional preprocessing of the collected historical data, spatiotemporal alignment of three-dimensional microscopic depth data (containing surface gradient, curvature and other features) and corresponding exposure parameters (such as light source wavelength, mask transmittance, scanning speed), elimination of dimensional differences after Z-score standardization, and division of training set, validation set and test set according to a ratio of 7:2:1; then, a multilayer perception model containing 3 hidden layers is constructed, wherein the input layer is designed as 1024 nodes to accommodate the topological structure parameters extracted by feature engineering, a LeakyReLU activation function is used to prevent gradient disappearance, and the network weight is set through the Xavier initialization method; in the iterative training stage, an adaptive moment estimation (Adam) optimizer is used, the initial learning rate is set to 0.001, and a cosine annealing strategy is used for dynamic adjustment; in each training period, the mean square error loss of the predicted exposure dose and the actual development line width is calculated by forward propagation, and an L2 regularization term (λ=0.01) is added to suppress the overfitting phenomenon; when the validation set loss decreases by less than 1e-4 for 5 consecutive epochs or reaches the preset maximum iteration number of 500 times, the independent test set is used to evaluate the model performance, if the key indicators (such as energy compensation error rate ≤3%, line width control precision ±1.5μm standard rate ≥98%) meet the preset threshold, the optimal model parameters are solidified and stored as a compensation engine that can be called in the production environment, and an online updating mechanism is established to continuously incorporate newly collected process data to realize model iterative optimization.

[0084] In step S13, frequency domain analysis is performed according to the material vibration data to obtain vibration frequency domain features.

[0085] In an embodiment, the material vibration data is wavelet denoised to obtain denoised vibration data.

[0086] The denoised vibration data is wavelet transformed to obtain vibration frequency domain data.

[0087] The vibration frequency domain data is used to identify a vibration main frequency.

[0088] The vibration frequency domain data is used to identify a vibration high frequency interval.

[0089] The vibration main frequency and the vibration high frequency interval are normalized and fused to obtain vibration frequency domain features.

[0090] It is worth mentioning that the vibration frequency domain feature extraction process is realized by a multi-stage signal processing method: first, the original material vibration data is denoised by wavelet, the multi-scale characteristics of the wavelet basis function are used to separate the effective vibration signal and the environmental noise, the high-frequency noise component is suppressed, and the low-frequency main body vibration waveform reflecting the material characteristics is retained; then the denoised vibration data is subjected to wavelet packet transform, the time domain waveform is decomposed into a plurality of sub-frequency bands containing different frequency components, and the frequency spectrum distribution map is drawn by calculating the energy proportion of each sub-frequency band; in the main frequency identification stage, the sliding window energy integral method is used to select the frequency band with energy concentration degree exceeding the set threshold (such as more than 60% of the total frequency band energy) as the dominant vibration frequency interval; for high frequency feature extraction, the variation coefficient of the instantaneous amplitude of each sub-frequency band is calculated to identify the high frequency sensitive area with violent amplitude fluctuation and frequency higher than 3 times the basic resonance frequency; finally, the main frequency energy proportion and the high frequency region fluctuation intensity are normalized and weighted to form a three-dimensional frequency domain feature vector representing the dynamic response of the material, which provides multi-dimensional vibration modal parameters for subsequent fatigue life prediction.

[0091] In one embodiment, an example of identifying a high-frequency sensitive area by calculating the variation coefficient includes: a composite material blade generates a vibration signal with a basic resonance frequency of 500 Hz when running. The denoised vibration data is decomposed into 16 sub-frequency bands by wavelet packet transform, and the center frequency of the 12th sub-frequency band is 1800 Hz (more than 3 times the basic frequency). Segment analysis is performed on the sub-frequency band: take a 10-second vibration waveform, divide it into 100 time segments, and calculate the fluctuation degree of the amplitude value in each segment (i.e. the difference between the maximum and minimum values accounts for the average value). It is found that the amplitude fluctuation rate of this frequency band increases from the normal 15% to 65% when the blade rotates to a certain angle, and the amplitude difference between adjacent segments reaches more than 3 times the normal state. Combined with the energy distribution characteristics of this frequency band (accounting for 28% of the total high frequency energy), the system determines that the 1800 Hz region is a high frequency sensitive area, and actual inspection finds that there is a hidden crack damage about 3 cm long at the corresponding position, verifying the effectiveness of the method.

[0092] It is worth mentioning that the coefficient of variation is a relative index in statistics for measuring the degree of dispersion of data, and its core function is to eliminate dimensional differences and compare the volatility of data sets of different scales or units. Specifically, this index represents the stability of data distribution by calculating the ratio of standard deviation to mean value - for example, the average charging time of a certain type of battery is 2 hours, with a standard deviation of 0.3 hours, and the coefficient of variation is 15%, indicating that there is a 15% relative fluctuation in charging time; while another battery has an average charging time of 8 hours, with a standard deviation of 0.8 hours, and its coefficient of variation is 10%, indicating that the latter is more stable in relative terms, although the absolute deviation is larger.

[0093] In step S14, a correlation calculation is performed according to the exposure energy distribution and the vibration frequency domain feature to obtain a correlation influence matrix.

[0094] In one embodiment, the vibration main frequency and the vibration high frequency interval extracted from the vibration frequency domain feature are subjected to block processing to form a frequency vector.

[0095] Based on the exposure energy distribution, the exposure energy result of each grid cell is extracted.

[0096] According to the frequency vector and the exposure energy result, a correlation coefficient is calculated to obtain a vibration exposure correlation degree.

[0097] The vibration exposure correlation degree is subjected to normalization processing to obtain a normalized correlation degree.

[0098] According to the normalized correlation degree, a correlation influence matrix is constructed according to the corresponding relationship between the grid cell and the frequency vector.

[0099] It is worth noting that the construction of the correlation influence matrix is achieved through multi-dimensional data fusion: first, the vibration main frequency interval and the high-frequency sensitive area are divided into several frequency blocks according to the preset bandwidth (such as 50Hz per analysis unit), forming a multi-dimensional frequency vector containing parameters such as energy proportion and fluctuation intensity; At the same time, the material surface is divided according to the millimeter grid, and the cumulative exposure energy value of each grid point in the laser processing process is extracted. Subsequently, a sliding window matching strategy is adopted to time-align the vibration feature sequence corresponding to each frequency block with the exposure energy distribution of the spatial grid, and the degree of synchronization of the change trend of the two (for example, whether the vibration energy suddenly increases is accompanied by a sudden drop in exposure energy) is calculated to obtain the preliminary correlation value. Then use the range normalization method to map the correlation to the 0 to 1 interval, eliminate the influence of different dimensions, and set a dynamic correction coefficient according to the material type (such as increasing the high-frequency area weight of metal materials by 0.3). Finally, the pairwise correlation of 256 grid units and 32 frequency blocks is arranged as a matrix according to the space-frequency correspondence, where the row represents the grid position, the list represents the frequency component, and the matrix element value greater than 0.8 is identified as a high-correlation risk area, providing a quantitative basis for optimizing laser processing parameters.

[0100] It is worth noting that the sliding window matching strategy achieves time alignment by dynamically intercepting data segments, and its core lies in setting an analysis window of fixed time length and moving it step by step. For example, in laser welding quality monitoring, suppose we need to correlate 1000 frames of infrared thermal imager data (recording material temperature changes) per second with 200Hz acceleration signals collected by vibration sensors. The system sets a window width of 500 milliseconds, and slides with a step of 100 milliseconds. When the window covers the interval from 3.2 seconds to 3.7 seconds, the synchronization of the temperature rise slope (such as from 850℃ to 920℃) and the vibration energy increase (such as from 0.5 to 1.2) in the window is calculated. If the consistency of the change trend exceeds the threshold (such as 80%), it is determined that there is abnormal coupling of process parameters in this period. This time-segment comparison method can effectively capture transient correlation characteristics and avoid masking local anomalies by overall correlation calculation.

[0101] It is worth mentioning that the range normalization method eliminates the dimensional difference by data range compression, and the specific operation includes two steps of determining the extreme value boundary and linear scaling. For example, in the detection of lithography materials, the range normalization method realizes the comparison of multiple parameters by unifying the dimension. Taking surface roughness (0.8-2.4 nanometers), assembly pressure (50-200 micro-newtons), and laser welding time (2-8 femtoseconds) as examples: first, calculate the range of each parameter, the surface roughness range is 1.6 nanometers (2.4-0.8), the assembly pressure range is 150 micro-newtons (200-50), and the welding time range is 6 femtoseconds (8-2). Then subtract the minimum value from the measured value and divide by the range, for example, when the assembly pressure is detected to be 125 micro-newtons, the normalized value is (125-50) / 150=0.5; if the welding time is 5 femtoseconds, the normalized value is (5-2) / 6≈0.5. Through this method, parameters of different physical dimensions can be converted into dimensionless values in the range of 0-1, so that each parameter is in the same order of magnitude range. This standardization is crucial in the multi-sensor data fusion scene, for example, by synchronously inputting the normalized vibration displacement (nanometer level), temperature (degree Celsius), and driving current (milliampere) parameters into a neural network model, the accuracy of lithography equipment fault diagnosis can be effectively improved.

[0102] In step S15, energy accumulation analysis is performed on the exposure overlap area to obtain an overlap correction factor.

[0103] In one embodiment, the cumulative exposure energy is obtained by integral calculation according to the exposure overlap area and the exposure energy distribution;

[0104] The attenuation factor and the enhancement factor are obtained by correction factor calculation according to the cumulative exposure energy;

[0105] When the cumulative exposure energy is less than a preset energy threshold, the enhancement factor is taken as the overlap correction factor;

[0106] When the cumulative exposure energy is greater than or equal to the energy threshold, the attenuation factor is taken as the overlap correction factor.

[0107] It is worth mentioning that the calculation of the overlap correction factor is realized by piecewise linear model, and the parameters are calibrated by experiment: first, the energy density integration is performed on the overlapping area of adjacent spots (for example, the crescent-shaped area formed by 40% overlapping of circular spots with a diameter of 0.1 mm), and the specific method is to divide the area into a micron-level grid, and the single-point energy is calculated according to the laser power (for example, 200 mW) and the residence time (for example, 50 microseconds) at each grid point (200 mW x 0.00005 seconds = 0.00001 joules), and then the total cumulative exposure energy is obtained by accumulating the energy values of all 5000 grid points (for example, 0.01 x 5000 = 0.050 joules). For example, in the photolithography process, when calculating the correction factor, a piecewise function is designed according to different parts of the overlapping area. For example, when two circular spots with a diameter of 0.2 mm are exposed with 50% overlap, it is found by actual measurement that the central area reaches 1.5 times the theoretical value due to the superposition of double-spot energy, and the edge area only reaches 0.8 times the theoretical value due to energy attenuation. The piecewise model refers to the use of two linear functions with different slopes for the part greater than one and the part less than one. Specifically, 1 / 1.5 is multiplied by the experimentally calibrated attenuation slope to obtain 0.7, and 1 / 0.8 is multiplied by the experimentally calibrated enhancement slope and then subtracted by 1 to obtain 1.2. The specific value of the attenuation slope is 1.05, and the specific value of the enhancement slope is 1.76.

[0108] It is worth mentioning that when the cumulative energy exceeds the threshold value (for example, the critical value set in the metal mask plate processing), the use of the attenuation factor (for example, 0.8 times energy compensation) can effectively suppress heat accumulation. For example, on a continuously wound polyimide film, if the energy of a certain area reaches 450 mJ / cm 2 (super-threshold 50 mJ) due to spot overlap, attenuation can prevent micron-level deformation caused by thermal expansion of the substrate, which is crucial for maintaining the accuracy of 10 μm line width circuits. When the energy is below the threshold value, the enhancement factor (for example, 1.2 times energy compensation) can ensure that the reaction proceeds sufficiently. For example, in the manufacture of flexible OLED display panels, if the energy of a certain area is only 350 mJ / cm 2 (below the 400 mJ threshold), enhancement compensation can prevent nanoscale residual defects after development, which is decisive for achieving sub-pixel level evaporation alignment accuracy.

[0109] In step S16, exposure parameter optimization is performed according to the overlap correction factor and the correlation influence matrix, and the exposure machine control parameters are updated.

[0110] In one embodiment, weighted fusion is performed according to the overlap correction factor and the correlation influence matrix to obtain a corrected influence matrix;

[0111] Particle swarm parameter initialization is performed according to the corrected influence matrix;

[0112] When the number of iterations reaches the preset upper limit or the fitness function of the particle meets the constraint condition, the iteration is completed, and the exposure machine control parameters are output.

[0113] It is worth noting that in the optimization of exposure parameters, the system first dynamically weights and fuses the overlap correction factor (such as the 0.85 attenuation coefficient of metal parts) and the associated influence matrix (containing the mutual influence weight of parameters such as scanning speed, laser power, etc.): For heat-sensitive materials (such as polyimide film), a correction factor of 60% weight is given, and for motion accuracy sensitive parameters (such as galvanometer acceleration), a matrix data of 70% weight is allocated, and a correction influence matrix with process adaptability is generated. Based on the guidance range of the matrix (such as the scanning speed limited to 0.01-0.5m / s, and the laser power window set to 10-100mW), 200 groups of parameter combinations are initialized to generate a particle group, each particle containing parameter values randomly distributed in the limited interval (such as particle A setting speed 0.23m / s, power 180mW, overlap compensation rate 92%). In the iteration process, the exposure effect score corresponding to each particle is calculated every round - by real-time simulation calculation of the positioning accuracy (such as ±1.2μm deviation) and thermal influence score (such as substrate temperature rise 28℃) under the parameters, and then the comprehensive score is synthesized according to the weight of 6:4. When it is detected that a certain particle reaches the preset standard (such as the score of the optimal particle exceeding 95 points for 5 consecutive generations) or completes 50 generations of iteration, the historical optimal parameter combination (such as the combination of speed 3.1m / s + power 220mW + compensation 88%) is selected as the final control parameter. In the optimization process, the simulated annealing mechanism is particularly introduced, when it is detected that the optimal solution is not updated for 10 consecutive generations, the parameter search range is automatically expanded by 10%, avoiding local optimal trap, such as in the OLED panel manufacturing scene, this strategy successfully reduces the exposure positioning error from ±2.5μm to ±1.1μm, and reduces the unit area heat accumulation by 40%.

[0114] In one embodiment, the weighted fusion of the overlap correction factor and the associated influence matrix needs to dynamically adjust the weight proportion according to the material characteristics and process requirements. For example, in the flexible circuit board manufacturing scene, if the system detects that the cumulative energy of a certain polyimide substrate area is 380mJ / cm 2When the threshold value is lower than 400 mJ, a 1.2 times enhancement factor is enabled. At this time, the enhancement factor is hierarchically fused with the parameter weight in the correlation influence matrix: for the photosensitive reaction key parameters (such as laser power), an adjustment weight of 70% is given to the enhancement factor, and the 40% weight of the parameter in the original matrix is correspondingly increased to 68% (i.e. 40% x 70% superimposed 1.2 times enhancement effect); for the motion stability sensitive parameters (such as the coiled material traction speed), the original 60% weight of the matrix is maintained at 80%, and only 20% of the enhancement effect is applied. After normalization processing, the weight of the laser power in the generated new matrix is increased to 52%, the scanning speed weight is adjusted to 38%, and a 10% dynamic compensation weight is introduced to offset the influence of the coiled material jitter. This intelligent weight allocation mechanism enables the manufacture of 5 μm line width circuit, which can not only ensure the sufficient reaction of photoresist through power enhancement, but also maintain a stable walking speed of 0.8 m / s, and finally realize an exposure alignment accuracy of ±1.3 μm.

[0115] In summary, the present application discloses a precision optimization method and system for a roll-to-roll high-precision projection exposure machine, aiming to improve the accuracy in the exposure process through a series of precise data collection and analysis steps. The method first involves obtaining micro-depth data, material vibration data and exposure overlap area, which ensures a comprehensive understanding of the substrate surface characteristics and its processing environment. Specifically, a white light interferometer is used for three-dimensional topography measurement to obtain micro-depth data, a three-axis MEMS acceleration sensor is used in conjunction with a laser Doppler vibration meter to capture material vibration data, and a high-speed linear array CCD is responsible for determining the position of the exposure overlap area. Based on these data, the next step is to analyze the energy distribution according to the micro-depth data, thereby obtaining the exposure energy distribution. In this process, the micro-depth data is divided into grid units with a predetermined interval, and the depth variation amplitude and frequency are calculated by numerical differentiation, finally forming the depth gradient value. Then, the pre-trained energy compensation model is used to output the exposure energy result combined with the depth gradient value and historical exposure parameters, and then mapped into an exposure energy distribution map.

[0116] In terms of processing material vibration, the method uses frequency domain analysis technology to extract vibration frequency domain features from the original vibration data. This process includes steps such as wavelet denoising, wavelet transform and main frequency identification, and finally generates vibration frequency domain features by normalizing fusion of vibration main frequency and vibration high frequency interval. Next, the construction of the correlation influence matrix depends on the correlation calculation of the exposure energy distribution and the vibration frequency domain features, which not only considers the influence of vibration on exposure effect, but also reveals the interaction between the two. By block processing of the vibration frequency domain features to form a frequency vector, and based on the exposure energy distribution to extract the exposure energy result of each grid unit, the vibration exposure correlation degree can be calculated, and then the correlation influence matrix is constructed.

[0117] In addition, in order to cope with the energy accumulation effect caused by multiple exposures, energy accumulation analysis needs to be performed on the exposure overlap area to determine the overlap correction factor. This involves integrating the cumulative exposure energy and selecting an appropriate attenuation or enhancement factor as the overlap correction factor according to its relationship with the preset energy threshold. The last step is to update the exposure parameters based on all the above information, i.e. the overlap correction factor and the correlation influence matrix. By weighting and fusing these two key factors, a corrected influence matrix is generated, and the particle swarm algorithm is used to initialize the parameters until a certain condition is met, thereby completing the update process of the exposure machine control parameters and improving the working precision.

[0118] Referring to Figure 2 The second embodiment of the present application provides a precision optimization system of a roll-to-roll high-precision projection exposure machine, comprising:

[0119] a data acquisition module configured to acquire microscopic depth data, material vibration data, and an exposure overlap area;

[0120] an energy distribution module configured to perform energy distribution analysis according to the microscopic depth data to obtain exposure energy distribution;

[0121] a vibration analysis module configured to perform frequency domain analysis according to the material vibration data to obtain vibration frequency domain characteristics;

[0122] a correlation calculation module configured to perform correlation calculation according to the exposure energy distribution and the vibration frequency domain characteristics to obtain a correlation influence matrix;

[0123] an overlap correction module configured to perform energy accumulation analysis on the exposure overlap area to obtain an overlap correction factor;

[0124] a parameter update module configured to perform exposure parameter optimization according to the overlap correction factor and the correlation influence matrix to update exposure machine control parameters.

[0125] It should be noted that the precision optimization system of a roll-to-roll high-precision projection exposure machine provided by the embodiments of the present application is used to perform all process steps of the precision optimization method of a roll-to-roll high-precision projection exposure machine provided by the above embodiments, and the working principles and beneficial effects of the two are one-to-one corresponding, so they will not be repeated here.

[0126] The embodiments of the present application also provide an electronic device. The electronic device includes a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a data acquisition program. The processor implements the steps in the above-mentioned various precision optimization methods of a roll-to-roll high-precision projection exposure machine when executing the computer program, such as Figure 1The step S11 is shown. Alternatively, the processor implements the functions of each module / unit in the above-mentioned device embodiments when executing the computer program, such as the data acquisition module.

[0127] For example, the computer program can be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present application. The one or more modules / units can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program in the electronic device.

[0128] The electronic device can be a desktop computer, a notebook computer, a palm computer, a smart tablet and the like. The electronic device can include, but is not limited to, a processor, a memory. Those skilled in the art can understand that the above components are only examples of the electronic device and do not constitute a limitation on the electronic device, and can include more or less components than the above, or combine certain components, or different components, for example, the electronic device can also include an input / output device, a network access device, a bus, etc.

[0129] The processor can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The processor is the control center of the electronic device, which connects all parts of the electronic device through various interfaces and lines.

[0130] The memory can be used to store the computer program and / or modules, and the processor realizes various functions of the electronic device by running or executing the computer program and / or modules stored in the memory, and calling data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application program required by a function (such as a sound playing function, an image playing function, etc.), and the like; and the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), and the like. In addition, the memory can include a high-speed random access memory, and can also include a nonvolatile memory, for example, a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state memory devices.

[0131] The modules / units integrated in the electronic device, if realized in the form of software function units and sold or used as independent products, can be stored in a computer readable storage medium. Based on this understanding, all or part of the processes in the above-mentioned embodiment methods can also be completed by a computer program instructing related hardware, and the computer program can be stored in a computer readable storage medium. The computer program can realize the steps of the above-mentioned various method embodiments when executed by a processor. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate forms, etc. The computer readable medium can include any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the contents included in the computer readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, the computer readable medium does not include electrical carrier signals and telecommunication signals.

[0132] It should be noted that the apparatus embodiments described above are merely illustrative, and the units described as separate units can or can not be physically separate, and the units displayed as units can or can not be physical units, i.e. can be located in one place, or can be distributed to multiple network units. Part or all of the modules can be selected to achieve the purpose of the embodiment scheme according to actual needs. In addition, the connection relationship between the modules in the apparatus embodiment provided by the present application indicates that there is a communication connection between them, which can be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement it without creative labor.

[0133] The above specific embodiments further illustrate the purpose, technical scheme and beneficial effects of the present application. It should be understood that the above description is only a specific embodiment of the present application and is not intended to limit the protection scope of the present application. It is particularly pointed out that any modification, equivalent replacement, improvement, etc. made by those skilled in the art within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A precision optimization method of a roll-to-roll high precision projection exposure machine, characterized by, The method comprises the following steps: acquiring microscopic depth data, material vibration data and exposure overlapping area; performing energy distribution analysis according to the microscopic depth data to obtain exposure energy distribution; performing frequency domain analysis according to the material vibration data to obtain vibration frequency domain characteristics; performing correlation calculation according to the exposure energy distribution and the vibration frequency domain characteristics to obtain an associated influence matrix; performing energy accumulation analysis on the exposure overlapping area to obtain an overlapping correction factor; performing exposure parameter optimization according to the overlapping correction factor and the associated influence matrix to update exposure machine control parameters.

2. The precision optimization method of the roll-to-roll high precision projection exposure machine according to claim 1, wherein, The energy distribution analysis according to the microscopic depth data comprises the following steps: dividing the microscopic depth data into grid units with a preset interval; performing numerical differentiation calculation on the grid units and the microscopic depth data to obtain depth variation amplitude; performing second derivative calculation on the grid units and the microscopic depth data to obtain depth variation frequency; performing weighted fusion on the depth variation amplitude and the depth variation frequency to obtain depth gradient value; inputting the depth gradient value and the pre-stored exposure parameter into a preset energy compensation model to output exposure energy result; performing distribution mapping on the exposure energy result to obtain exposure energy distribution.

3. The precision optimization method of the roll-to-roll high precision projection exposure machine according to claim 2, characterized in that, The training process of the energy compensation model comprises the following steps: taking historical microscopic depth data and historical exposure parameters as input and taking historical exposure result as output, and dividing training set and verification set through normalization processing; initializing multi-layer perception model parameters; optimizing parameters through iterative training; after the loss of the verification set converges or reaches the maximum number of iterations, if the error of the test set is lower than a preset threshold, saving the model parameters as the final energy compensation model.

4. The precision optimization method of the roll-to-roll high precision projection exposure machine according to claim 1, wherein The frequency domain analysis according to the material vibration data comprises the following steps: performing wavelet denoising on the material vibration data to obtain denoised vibration data; performing wavelet transform on the denoised vibration data to obtain vibration frequency domain data; performing main frequency identification on the vibration frequency domain data to obtain vibration main frequency; performing high frequency interval identification on the vibration frequency domain data to obtain vibration high frequency interval; performing normalization fusion on the vibration main frequency and the vibration high frequency interval to obtain vibration frequency domain characteristics.

5. The precision optimization method of the roll-to-roll high precision projection exposure machine according to claim 1, wherein, The correlation calculation according to the exposure energy distribution and the vibration frequency domain characteristics comprises the following steps: performing block processing on the vibration main frequency and the vibration high frequency interval extracted from the vibration frequency domain characteristics to form a frequency vector; extracting exposure energy result of each grid unit based on the exposure energy distribution; performing correlation coefficient calculation on the frequency vector and the exposure energy result to obtain vibration exposure correlation degree; performing normalization processing on the vibration exposure correlation degree to obtain normalized correlation degree; constructing an associated influence matrix according to the corresponding relationship between the grid unit and the frequency vector based on the normalized correlation degree.

6. The precision optimization method of the roll-to-roll high precision projection exposure machine according to claim 1, wherein, The energy accumulation analysis on the exposure overlapping area comprises the following steps: performing integral calculation on the exposure overlapping area and the exposure energy distribution to obtain accumulated exposure energy; According to the cumulative exposure energy, a correction factor calculation is performed to obtain an attenuation factor and an enhancement factor; When the cumulative exposure energy is less than a preset energy threshold, the enhancement factor is taken as an overlap correction factor; When the cumulative exposure energy is greater than or equal to the energy threshold, the attenuation factor is taken as an overlap correction factor.

7. The precision optimization method of the roll-to-roll high precision projection exposure machine according to claim 1, wherein According to the overlap correction factor and the correlation influence matrix, exposure parameter optimization is performed to update the exposure machine control parameters, including: According to the overlap correction factor and the correlation influence matrix, weighted fusion is performed to obtain a corrected influence matrix; According to the corrected influence matrix, particle swarm parameter initialization is performed; When the number of iterations reaches a preset upper limit or the fitness function of the particles meets the constraint condition, the iteration is completed, and the exposure machine control parameters are output.

8. A precision optimization system for a roll-to-roll high precision projection exposure machine, characterized by, It includes: A data acquisition module is configured to acquire micro-depth data, material vibration data, and an exposure overlap area; An energy distribution module is configured to perform energy distribution analysis according to the micro-depth data to obtain an exposure energy distribution; A vibration analysis module is configured to perform frequency domain analysis according to the material vibration data to obtain vibration frequency domain characteristics; A correlation calculation module is configured to perform correlation calculation according to the exposure energy distribution and the vibration frequency domain characteristics to obtain a correlation influence matrix; An overlap correction module is configured to perform energy accumulation analysis on the exposure overlap area to obtain an overlap correction factor; A parameter update module is configured to perform exposure parameter optimization according to the overlap correction factor and the correlation influence matrix to update the exposure machine control parameters.

9. An electronic device, comprising: The computer readable storage medium includes a stored computer program, wherein when the computer program runs, the device where the computer readable storage medium is located performs the precision optimization method of the roll-to-roll high-precision projection exposure machine as claimed in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium includes a stored computer program, wherein when the computer program runs, the device where the computer readable storage medium is located performs the precision optimization method of the roll-to-roll high-precision projection exposure machine as claimed in any one of claims 1 to 7.

Images