A method for monitoring and controlling a capacitor production process

Abstract

The embodiment of the application provides a capacitor production process monitoring control method, the method comprises the following steps: deploying a detection unit; the detection unit identifies candidate pinholes by analyzing the distortion characteristics of interference fringes generated by laser irradiation on the surface of the film, verifies and screens target pinholes from the candidate pinholes, and synchronously records the three-dimensional coordinates of the target pinholes; a traceability analysis library is constructed, the traceability analysis library comprises a traceability code; based on the traceability code, the target process parameters of the base film preparation, vacuum coating and slitting processes are bound, a space-time correlation model is used, and a machine learning algorithm is combined to perform traceability analysis to obtain a traceability result; based on the traceability result, a process closed-loop regulation mechanism is started, and a parameter regulation instruction is sent to the production equipment of the corresponding process. The application can effectively avoid the missed detection and false detection problems caused by a single detection method; can identify pinhole defects in the early stage of each process, reduce material waste and rework cost caused by defect flow to subsequent processes, and finally improve the yield of capacitor products.

Description

Technical Field

[0001] This manual relates to the field of capacitor production monitoring and control, and in particular to a method for monitoring and controlling capacitor production processes. Background Technology

[0002] Metallized film capacitors, with their high voltage withstand capability, low loss, and long lifespan, have become core components in high-end equipment such as on-board chargers for new energy vehicles, photovoltaic inverters, and traction converters for rail transit. Their performance is highly dependent on the surface integrity of the metallized film, especially pinhole defects with a diameter of less than 1μm. These pinholes can create localized field concentrations under an electric field, leading to dielectric breakdown or localized overheating, causing sudden capacitor failure under high-frequency, high-power conditions. As the voltage platform of new energy vehicles upgrades to 800V and above, the film thickness continues to decrease, further amplifying the impact of pinholes on reliability. Data from a leading automaker shows that early capacitor failures due to film pinholes account for more than 35%.

[0003] Current pinhole detection technologies for metallized thin films have significant limitations and struggle to meet the quality requirements of high-end applications. Offline inspection, represented by spark testing, identifies defects by generating breakdown discharge through contact between a high-voltage electrode and the thin film. However, limited by detection speed and voltage threshold settings, it suffers from a 10%-20% false negative rate for tiny pinholes <1μm in diameter. Furthermore, it can only determine the presence or absence of defects, failing to record crucial information such as pinhole location and density, making it difficult to pinpoint specific production stages during subsequent traceability. While online inspection can adapt to production line speeds, the mainstream corona discharge method relies on localized corona signal changes caused by pinholes. Its minimum detectable diameter is 2μm, rendering it completely ineffective for pinholes smaller than 1μm. These tiny pinholes, after the thin film is wound into a capacitor, expand with temperature cycling and mechanical stress, becoming potential failure hazards.

[0004] Tiny impurities in the base film raw materials, target particle detachment during vacuum coating, and micro-damage to the cutting blades during slitting can all generate pinholes smaller than 1μm. However, existing detection technologies cannot capture these defects in real time throughout the entire production process, resulting in a lack of accurate data support for process optimization. For example, on one production line, the pinhole density increased by 30% after the base film supplier changed batches, but the detection system failed to issue a timely warning, ultimately leading to a 12% decrease in the pass rate of the withstand voltage test for the entire batch of 500,000 capacitors. Furthermore, as film widths expand to over 1.2m, the scanning coverage and signal-to-noise ratio of traditional detection equipment further deteriorate, making the detection of pinholes smaller than 1μm a critical bottleneck restricting the upgrade of metallized film capacitors to higher reliability and lower loss. Summary of the Invention

[0005] This specification provides one or more embodiments of a capacitor manufacturing process monitoring and control method, the method comprising:

[0006] A detection unit is deployed, which identifies candidate pinholes by analyzing the distortion characteristics of interference fringes generated by laser irradiation on the thin film surface, verifies and screens target pinholes from the candidate pinholes, and simultaneously records the three-dimensional coordinates of the target pinholes.

[0007] A traceability analysis library is constructed, which includes traceability codes. Based on the traceability codes, the target process parameters of the base film preparation, vacuum coating and slitting processes are bound together. The traceability results are obtained by performing traceability analysis through a spatiotemporal correlation model and machine learning algorithms.

[0008] Based on the traceability results, the process closed-loop control mechanism is activated, and parameter control instructions are sent to the production equipment of the corresponding process.

[0009] In some embodiments, the detection unit includes a first detection unit, a second detection unit, and a third detection unit:

[0010] The first detection unit corresponds to the base film preparation process. The first detection unit uses a continuous laser of a preset wavelength to irradiate the surface of the base film and extracts the first distortion feature by Fourier transform through interference fringes.

[0011] The first distortion feature includes stripe offset, rate of curvature change, and fracture length;

[0012] The second detection unit corresponds to the vacuum coating process. It uses a tunable laser to irradiate the surface of the coated film and extracts the second distortion feature through the phase demodulation algorithm of the interference fringes.

[0013] The second distortion feature includes the stripe contrast attenuation rate, the phase shift of the discontinuous region of the metal layer, and the diffusion radius of the annular stripe;

[0014] The third detection unit corresponds to the slitting process. It uses linearly polarized laser to irradiate along the parallel direction of the slitting edge and extracts the third distortion feature through a stripe edge detection algorithm.

[0015] The third distortion feature includes the abrupt change in stripe density near the slit edge, the length of the serrated distortion segment, and the stripe breakage spacing across the slit line.

[0016] In some embodiments, the target pinhole includes a first target pinhole, a second target pinhole, and a third target pinhole, and the step of verifying and screening the target pinhole from the candidate pinholes includes:

[0017] For the first candidate pinhole identified by the first detection unit, its first distortion feature and the base film thickness data of the corresponding area are obtained; in response to the stripe offset reaching the preset offset threshold, the curvature change rate exceeding the preset curvature threshold, the fracture length meeting the preset fracture threshold, and the base film thickness data deviating from the preset thickness threshold of the normal area, it is determined to be the first target pinhole.

[0018] For the second candidate pinhole identified by the second detection unit, its second distortion feature and the coating layer thickness uniformity data of the corresponding area are obtained; in response to the contrast attenuation rate reaching the preset attenuation threshold, the phase offset exceeding the preset phase threshold, the annular stripe diffusion radius meeting the preset diffusion threshold, and the coating layer thickness uniformity deviating from the preset uniformity threshold, it is determined to be the second target pinhole.

[0019] For the third candidate pinhole identified by the third detection unit, its third distortion feature and the flatness data of the cutting edge of the corresponding region are obtained; in response to the stripe density mutation value reaching the preset mutation threshold, the length of the serrated distortion segment exceeding the preset distortion length threshold, the cross-cutting line breakage distance meeting the preset distance threshold, and the flatness of the cutting edge deviating from the preset flatness threshold, it is determined to be the third target pinhole.

[0020] In some embodiments, the traceability analysis library also includes the following associated data:

[0021] The target pinhole data for each process includes the three-dimensional coordinates, number and distribution density of the first target pinhole, the damaged area and location information of the coating layer of the second target pinhole, and the cutting edge correlation data of the third target pinhole.

[0022] The equipment status data for the entire process includes the wear of the die head of the base film preparation equipment, the stability of the evaporation source power of the vacuum coating machine, and the tool wear parameters of the slitting machine;

[0023] Environmental monitoring data, including the cleanliness level of the base film preparation workshop, the residual gas composition in the vacuum coating chamber, and the temperature and humidity fluctuation curves of the slitting workshop;

[0024] The traceability code uniquely binds the aforementioned target process parameters, target pinhole data, equipment status data, and environmental monitoring data through a mapping relationship between timestamps and spatial coordinates, forming a traceable full-chain data archive.

[0025] In some embodiments, the source tracing results obtained by combining a spatiotemporal correlation model with machine learning algorithms include:

[0026] The spatiotemporal correlation model establishes time axis correlation information based on the timestamps of each process. The three-dimensional coordinates of the target pinhole in the base film preparation process are mapped to the corresponding spatial positions in the vacuum coating process and the slitting process, forming the spatiotemporal trajectory of the pinhole's entire life cycle.

[0027] The spatiotemporal trajectory and the real-time process parameter sequence of each process are input into a machine learning algorithm. The model learns the correlation between historical pinholes and process parameter anomalies and outputs the source tracing results.

[0028] In some embodiments, the process closed-loop control mechanism includes a three-level response:

[0029] When the number of target pinholes in a single batch is less than or equal to a first preset threshold, a fine-tuning command is sent.

[0030] When the number of target pinholes in a single batch is greater than the first preset threshold and less than or equal to the second preset threshold, a mid-level adjustment command is sent.

[0031] When the number of pinholes in a single batch exceeds a second preset threshold, a shutdown inspection command is sent and an alarm is triggered.

[0032] Beneficial effects: By configuring differentiated detection units for the three processes of base film preparation, vacuum coating, and slitting, and designing exclusive distortion features for pinhole defects in each process, and combining multi-dimensional verification parameters to screen target pinholes, the problems of missed detection and false detection caused by a single detection method can be effectively avoided. Moreover, the detection covers the entire process from base film formation to final slitting, and can identify pinhole defects at an early stage in each process, reducing material waste and rework costs caused by defects flowing to subsequent processes, and ultimately significantly improving the yield of capacitor products. Attached Figure Description

[0033] This specification will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting; in these embodiments, the same reference numerals denote the same structures, wherein:

[0034] Figure 1 This is an exemplary flowchart of a capacitor production monitoring and control method according to some embodiments of this specification. Detailed Implementation

[0035] To more clearly illustrate the technical solutions of the embodiments in this specification, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are merely some examples or embodiments of this specification. For those skilled in the art, these drawings can be applied to other similar scenarios without creative effort. Unless obvious from the context or otherwise specified, the same reference numerals in the drawings represent the same structures or operations.

[0036] It should be understood that the terms "system," "unit," and / or "module" used herein are a method of distinguishing different components, elements, parts, sections, or assemblies at different levels. However, if other terms can achieve the same purpose, they may be replaced by other expressions.

[0037] Unless the context clearly indicates an exception, words such as "a," "an," "a kind," and / or "the" do not specifically refer to the singular and may also include the plural. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of explicitly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0038] Flowcharts are used in this specification to illustrate the operations performed by the system according to embodiments of this specification. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, the steps can be processed in reverse order or simultaneously. Furthermore, other operations can be added to these processes, or one or more steps can be removed from them.

[0039] Example 1: Figure 1 This is an exemplary flowchart of a capacitor manufacturing process monitoring and control method according to some embodiments of this specification. Figure 1 As shown, a method for monitoring and controlling a capacitor manufacturing process is provided, the method comprising:

[0040] A detection unit is deployed, which identifies candidate pinholes by analyzing the distortion characteristics of interference fringes generated by laser irradiation on the thin film surface, verifies and screens target pinholes from the candidate pinholes, and simultaneously records the three-dimensional coordinates of the target pinholes.

[0041] The detection unit includes a first detection unit, a second detection unit, and a third detection unit:

[0042] The first detection unit corresponds to the base film preparation process. The first detection unit uses a continuous laser of a preset wavelength to irradiate the surface of the base film and extracts the first distortion feature by Fourier transform through interference fringes.

[0043] The first distortion feature includes stripe offset, rate of curvature change, and fracture length;

[0044] For example, high-precision identification of early pinhole defects on the substrate surface is achieved through the principle of laser interference: a preset band continuous laser matching the optical properties of the substrate material is selected, and the laser is split into reference light and probe light by a beam splitter to form stable parallel interference fringes; a high-speed industrial camera is simultaneously started to continuously acquire interference fringe images, and the acquired images are preprocessed to eliminate interference from ambient light and slight undulations of the substrate film, and then the fringe signal in the spatial domain is converted to the frequency domain by a two-dimensional Fourier transform. The distortion frequency component caused by pinholes is separated by a designed bandpass filter, and the clear contour of the distorted region is reconstructed by an inverse Fourier transform; based on the reconstructed contour, the first distortion feature is further extracted: the fringe offset is obtained by calculating the pixel difference between the fringe center of the distorted region and the fringe center of the normal region and combining it with the imaging magnification conversion; the rate of curvature change is determined by calculating the change in the radius of curvature after fitting the distorted fringe segment with an arc; and the break length is located by locating the start and end points of the fringe interruption by an edge detection algorithm and converted into the actual physical length by combining pixel calibration parameters, thereby realizing the early feature capture of potential pinhole defects in the substrate preparation stage.

[0045] The second detection unit corresponds to the vacuum coating process. It uses a tunable laser to irradiate the surface of the coated film and extracts the second distortion feature through the phase demodulation algorithm of the interference fringes.

[0046] The second distortion feature includes the stripe contrast attenuation rate, the phase shift of the discontinuous region of the metal layer, and the diffusion radius of the annular stripe;

[0047] For example, the second detection unit uses a tunable laser, which is split into a measurement beam and a reference beam via fiber coupling. The measurement beam illuminates the continuously transmitted thin film surface after coating at a perpendicular incident angle. After being reflected by the metal layer, it merges with the reference beam in the interferometric imaging module, forming interference fringes containing information on the coating thickness and integrity. Simultaneously, a high-resolution camera is activated to acquire the fringes image. First, adaptive threshold segmentation is used to eliminate background noise caused by minor undulations in the coating layer. Then, a phase demodulation algorithm is used to extract the phase of the fringes. By analyzing the phase difference between adjacent pixels, the brightness changes of the fringes are converted into quantized phase distribution data, thereby accurately capturing abrupt changes in the optical properties of discontinuous areas of the metal layer. Based on this... The second distortion feature is extracted: the fringe contrast attenuation rate is obtained by calculating the ratio of the fringe contrast of the defect area to that of the normal area, reflecting the degree of decrease in the reflectivity of the metal layer; the phase shift of the discontinuous area of ​​the metal layer is determined by comparing the phase value difference between the defect center and the surrounding normal area, quantifying the sudden change in optical path difference caused by the absence of coating; the annular fringe diffusion radius is for the annular interference fringes formed by the gradual change in coating thickness at the edge of the pinhole. The outline of the outermost annular fringe is located by the edge detection algorithm, and converted into the actual physical radius by combining the magnification of the imaging system, thereby characterizing the influence range of the pinhole and realizing the accurate identification and feature quantification of pinhole defects related to the metal layer in the vacuum coating process.

[0048] The third detection unit corresponds to the slitting process. It uses linearly polarized laser to irradiate along the parallel direction of the slitting edge and extracts the third distortion feature through a stripe edge detection algorithm.

[0049] The third distortion feature includes the abrupt change in stripe density near the slit edge, the length of the serrated distortion segment, and the stripe breakage spacing across the slit line.

[0050] For example, the third detection unit uses a linearly polarized laser. A laser collimation system compresses the beam into a line beam with a width ≤0.5mm, illuminating the edge region of the slit film along a direction parallel to the slit edge. After reflection from the edge surface, the laser interferes with a preset reference optical path beam, forming continuous interference fringes distributed along the slit edge. Simultaneously, a linear array camera acquires fringe images. Morphological filtering first eliminates scattering noise from slit debris, then a fringe edge detection algorithm extracts the continuous contours of the fringes. By locating the light-dark boundary lines of the fringes, abrupt changes in fringe morphology caused by edge defects are identified. Based on this, a third distortion feature is extracted: the abrupt change in fringe density near the slit edge is calculated per unit length of the defect area. The difference between the number of stripes within the normal edge area is obtained; the length of the serrated distortion segment is obtained by tracking the continuous interval of the periodic serrated bending of the stripes, and the physical length is obtained by combining the conversion relationship between image pixels and actual size, which corresponds to the irregular cutting marks formed by the wear of the cutting tool edge; the stripe breakage spacing across the cutting line is determined by measuring the straight distance between the stripe endpoints on both sides of the breakage point to address the stripe interruption caused by tearing or misalignment on both sides of the cutting line, reflecting the tension fluctuation or tool positioning error during cutting, thereby achieving accurate identification and quantitative characterization of pinholes related to edge defects unique to the cutting process.

[0051] The target pinholes include a first target pinhole, a second target pinhole, and a third target pinhole. The step of verifying and screening target pinholes from candidate pinholes includes:

[0052] For the first candidate pinhole identified by the first detection unit, its first distortion feature and the base film thickness data of the corresponding area are obtained; in response to the stripe offset reaching the preset offset threshold, the curvature change rate exceeding the preset curvature threshold, the fracture length meeting the preset fracture threshold, and the base film thickness data deviating from the preset thickness threshold of the normal area, it is determined to be the first target pinhole.

[0053] For the second candidate pinhole identified by the second detection unit, its second distortion feature and the coating layer thickness uniformity data of the corresponding area are obtained; in response to the contrast attenuation rate reaching the preset attenuation threshold, the phase offset exceeding the preset phase threshold, the annular stripe diffusion radius meeting the preset diffusion threshold, and the coating layer thickness uniformity deviating from the preset uniformity threshold, it is determined to be the second target pinhole.

[0054] For the third candidate pinhole identified by the third detection unit, its third distortion feature and the flatness data of the cutting edge of the corresponding region are obtained; in response to the stripe density mutation value reaching the preset mutation threshold, the length of the serrated distortion segment exceeding the preset distortion length threshold, the cross-cutting line breakage distance meeting the preset distance threshold, and the flatness of the cutting edge deviating from the preset flatness threshold, it is determined to be the third target pinhole.

[0055] A traceability analysis library is constructed, which includes traceability codes. Based on the traceability codes, the target process parameters of the base film preparation, vacuum coating and slitting processes are bound together. The traceability results are obtained by performing traceability analysis through a spatiotemporal correlation model and machine learning algorithms.

[0056] The source tracing analysis library also includes the following related data:

[0057] The target pinhole data for each process includes the three-dimensional coordinates, number and distribution density of the first target pinhole, the damaged area and location information of the coating layer of the second target pinhole, and the cutting edge correlation data of the third target pinhole.

[0058] For example, for the first target pinhole, an absolute coordinate system is established with the starting end of the base film roll as the origin. The X-axis is along the base film running direction, the Y-axis is along the width direction, and the Z-axis is along the thickness direction. The quantity statistics include the total number of pinholes within a single meter length. The distribution density is calculated by the sliding window algorithm to determine the frequency of pinholes per unit area in order to identify areas with concentrated defects.

[0059] For the second target pinhole, the damaged area of ​​the coating layer is calculated from the contour of the distorted region through image pixel calibration. The position association information is mapped to the coordinates of the first target pinhole in the base film preparation stage through a coordinate transformation matrix to determine whether the coating layer damage is caused by the original defect of the base film.

[0060] For the third target pinhole, the correlation data of the cutting edge is used to quantify the correlation between defects and cutting operations by calculating the vertical distance between the pinhole center and the cutting line and the overlap between the pinhole edge and the wear trajectory of the cutting edge.

[0061] The equipment status data for the entire process includes the wear of the die head of the base film preparation equipment, the stability of the evaporation source power of the vacuum coating machine, and the tool wear parameters of the slitting machine;

[0062] The wear of the die head in the base film preparation equipment is detected in real time by a laser profilometer, which measures the linear wear depth of the die head lip and correlates it with the base film thickness fluctuation range. The power stability of the evaporation source in the vacuum coating machine is measured by a power sensor that collects instantaneous power and calculates the standard deviation of a preset time interval. The standard deviation reflects the consistency of the metal evaporation rate. The tool wear parameters of the slitting machine include the blade fillet radius and the tool vibration frequency, which are directly related to the integrity of the slitting edge.

[0063] Environmental monitoring data, including the cleanliness level of the base film preparation workshop, the residual gas composition in the vacuum coating chamber, and the temperature and humidity fluctuation curve of the slitting workshop;

[0064] For example, the cleanliness level of the base film preparation workshop is recorded according to the ISO 14644-1 standard, including the concentration per cubic meter of 0.5μm and 5μm particles, which is related to pinholes caused by surface contamination of the base film; the residual gas components in the vacuum coating chamber are detected by mass spectrometry, with a focus on recording the partial pressure of oxygen and water vapor, reflecting the risk of coating oxidation; the temperature and humidity fluctuation curves of the slitting workshop are recorded with a sampling interval of 1 minute, recording the deviation values ​​of temperature and relative humidity, which are related to the impact of film tension stability on slitting quality.

[0065] The traceability code uniquely binds the aforementioned target process parameters, target pinhole data, equipment status data, and environmental monitoring data through a mapping relationship between timestamps and spatial coordinates, forming a traceable full-chain data archive.

[0066] The source tracing results obtained through source tracing analysis using a spatiotemporal correlation model combined with machine learning algorithms include:

[0067] The spatiotemporal correlation model establishes time axis correlation information based on the timestamps of each process, mapping the three-dimensional coordinates of the target pinhole in the base film preparation process to the corresponding spatial positions in the vacuum coating process and the slitting process, forming the spatiotemporal trajectory of the pinhole's entire life cycle.

[0068] The spatiotemporal correlation model establishes a time correlation formula based on the key timestamps of the base film preparation process and the thin film transport parameters of each process:

[0069] For the vacuum coating process, the unwinding speed of the base film roll is crucial. The transport distance of the base film from its preparation to the coating machine Calculate the timestamp of the target pinhole reaching the coating area. ,in This refers to the start time of the base film extrusion molding process. This is a correction value for the storage delay of the base film roll, which is dynamically adjusted according to changes in roll diameter.

[0070] For the slitting process, based on the traction speed of the film after coating. The transmission distance from the coating machine to the slitting machine Calculate the timestamp of the pinhole reaching the cutting position: ,in The compensation value for transmission time fluctuations caused by tension adjustment is corrected in real time through feedback from the tension sensor.

[0071] The spatiotemporal trajectory and the real-time process parameter sequence of each process are input into a machine learning algorithm. The model learns the correlation between historical pinholes and process parameter anomalies and outputs the source tracing results.

[0072] The spatiotemporal trajectory of the target pinhole is input into a spatiotemporal attention machine learning algorithm along with real-time process parameter sequences such as extrusion temperature, die pressure, and traction speed in the base film preparation process; vacuum degree, evaporation rate, and coating thickness in the vacuum coating process; and slitting tension, tool speed, and ambient humidity in the slitting process. The model learns and identifies the spatiotemporal correlation between pinhole generation and process parameter anomalies by comparing the abnormal process parameter patterns corresponding to similar pinholes in the historical database. Finally, the algorithm outputs a source tracing result that includes the key abnormal process that caused the target pinhole, the specific abnormal process parameter, the time window in which the parameter anomaly occurred, and the contribution of the impact.

[0073] Based on the traceability results, the process closed-loop control mechanism is activated, and parameter control instructions are sent to the production equipment of the corresponding process.

[0074] The process closed-loop control mechanism includes a three-level response:

[0075] When the number of target pinholes in a single batch is less than or equal to a first preset threshold, a fine-tuning command is sent.

[0076] When the number of target pinholes in a single batch is greater than the first preset threshold and less than or equal to the second preset threshold, a mid-level adjustment command is sent.

[0077] When the number of pinholes in a single batch exceeds a second preset threshold, a shutdown inspection command is sent and an alarm is triggered.

[0078] First, the definition of a single batch is based on the production batch of the base film roll. For example, each roll of base film corresponds to one production batch, encompassing the entire product process from base film preparation to slitting. The first and second preset thresholds are set based on the pinhole number distribution of historical qualified batches and product quality standards. For instance, the first preset threshold corresponds to the upper limit of minor defects allowed in a single batch, and the second preset threshold corresponds to the critical value of moderate defects requiring intervention. When the number of target pinholes in a single batch is less than or equal to the first preset threshold, it is determined to be an occasional defect caused by slight fluctuations in process parameters. The system sends a fine-tuning instruction to the key processes identified by the traceability analysis, such as fine-tuning the traction speed in the base film preparation process to stabilize the base film thickness, or fine-tuning the evaporation source power in the vacuum coating process to optimize coating uniformity. After fine-tuning, the system continuously monitors the change in the number of pinholes through the detection unit to ensure that the parameters converge to the normal range. When the number of target pinholes in a single batch is greater than the first preset threshold but less than or equal to the second preset threshold, it is determined to be a systematic deviation in process parameters. A medium-adjustment instruction is sent, such as stepwise adjustment of the die pressure in the base film preparation process. During the adjustment or slitting process, the tool tension is adjusted and the tool positioning accuracy is simultaneously calibrated. In the mid-adjustment process, the correlation curve between process parameters and pinhole data is collected in real time until the number of pinholes falls below the first preset threshold. When the number of target pinholes in a single batch exceeds the second preset threshold, it is determined that there is a risk of equipment abnormality or process failure. A shutdown inspection command is immediately sent to the corresponding production equipment, and a multi-level alarm mechanism is triggered at the same time: the workshop's audible and visual alarm device is activated, and the production management system pushes alarm information containing abnormal processes, pinhole distribution heat maps, and preliminary traceability results to the equipment maintenance personnel. After the technicians have investigated and repaired the equipment faults or environmental abnormalities, normal production can only be restarted after three small-batch trial productions to verify that there are no abnormalities. This achieves full-level closed-loop control from minor adjustments to emergency intervention, ensuring product quality stability.

[0079] The basic concepts have been described above. Obviously, for those skilled in the art, the detailed disclosure above is merely illustrative and does not constitute a limitation of this specification. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and corrections to this specification. Such modifications, improvements, and corrections are suggested in this specification and therefore remain within the spirit and scope of the exemplary embodiments described herein.

[0080] Furthermore, unless expressly stated in the claims, the order of processing elements and sequences, the use of numbers and letters, or other names described in this specification are not intended to limit the flow and sequence of the laminar flow shield. Although various examples have been discussed in the foregoing disclosure of some embodiments that are currently considered useful, it should be understood that such details are for illustrative purposes only, and the appended claims are not limited to the disclosed embodiments; rather, the claims are intended to cover all modifications and equivalent combinations that conform to the substance and scope of the embodiments described in this specification. For example, while the system components described above can be implemented using hardware devices, they can also be implemented solely using software solutions, such as installing the described system on existing servers or mobile devices.

[0081] Similarly, it should be noted that, in order to simplify the description disclosed herein and thus aid in the understanding of one or more embodiments of the invention, the foregoing description of embodiments in this specification may sometimes combine multiple features into a single embodiment, drawing, or description thereof. However, this method of disclosure does not imply that the subject matter of this specification requires more features than those mentioned in the claims. In fact, the embodiments contain fewer features than all the features of a single embodiment disclosed above.

[0082] Finally, it should be understood that the embodiments described in this specification are merely illustrative of the principles of the embodiments described herein. Other variations may also fall within the scope of this specification. Therefore, alternative configurations of the embodiments described herein are intended to be illustrative rather than limiting, and are considered consistent with the teachings of this specification. Accordingly, the embodiments described herein are not limited to those explicitly introduced and described herein.

Claims

1. A method for monitoring and controlling a capacitor manufacturing process, characterized in that, The method includes: A detection unit is deployed, which identifies candidate pinholes by analyzing the distortion characteristics of interference fringes generated by laser irradiation on the thin film surface, verifies and screens target pinholes from the candidate pinholes, and simultaneously records the three-dimensional coordinates of the target pinholes. A traceability analysis library is constructed, which includes traceability codes. Based on the traceability codes, the target process parameters of the base film preparation, vacuum coating and slitting processes are bound together. The traceability results are obtained by performing traceability analysis through a spatiotemporal correlation model and machine learning algorithms. Based on the traceability results, the process closed-loop control mechanism is activated, and parameter control instructions are sent to the production equipment of the corresponding process.

2. The method according to claim 1, characterized in that, The detection unit includes a first detection unit, a second detection unit, and a third detection unit: The first detection unit corresponds to the base film preparation process. The first detection unit uses a continuous laser of a preset wavelength to irradiate the surface of the base film and extracts the first distortion feature by Fourier transform through interference fringes. The first distortion feature includes stripe offset, rate of curvature change, and break length; The second detection unit corresponds to the vacuum coating process. It uses a tunable laser to irradiate the surface of the coated film and extracts the second distortion feature through the phase demodulation algorithm of the interference fringes. The second distortion feature includes the stripe contrast attenuation rate, the phase shift of the discontinuous region of the metal layer, and the diffusion radius of the annular stripe; The third detection unit corresponds to the slitting process. It uses linearly polarized laser to irradiate along the parallel direction of the slitting edge and extracts the third distortion feature through a stripe edge detection algorithm. The third distortion feature includes the abrupt change in stripe density near the slit edge, the length of the serrated distortion segment, and the stripe breakage spacing across the slit line.

3. The method according to claim 1, characterized in that, The target pinholes include a first target pinhole, a second target pinhole, and a third target pinhole. The step of verifying and screening target pinholes from candidate pinholes includes: For the first candidate pinhole identified by the first detection unit, its first distortion feature and the base film thickness data of the corresponding area are obtained; in response to the stripe offset reaching the preset offset threshold, the curvature change rate exceeding the preset curvature threshold, the fracture length meeting the preset fracture threshold, and the base film thickness data deviating from the preset thickness threshold of the normal area, it is determined to be the first target pinhole. For the second candidate pinhole identified by the second detection unit, its second distortion feature and the coating layer thickness uniformity data of the corresponding area are obtained; in response to the contrast attenuation rate reaching the preset attenuation threshold, the phase offset exceeding the preset phase threshold, the annular stripe diffusion radius meeting the preset diffusion threshold, and the coating layer thickness uniformity deviating from the preset uniformity threshold, it is determined to be the second target pinhole. For the third candidate pinhole identified by the third detection unit, its third distortion feature and the flatness data of the cutting edge of the corresponding region are obtained; in response to the stripe density mutation value reaching the preset mutation threshold, the length of the serrated distortion segment exceeding the preset distortion length threshold, the cross-cutting line breakage distance meeting the preset distance threshold, and the flatness of the cutting edge deviating from the preset flatness threshold, it is determined to be the third target pinhole.

4. The method according to claim 1, characterized in that, The source tracing analysis library also includes the following related data: The target pinhole data for each process includes the three-dimensional coordinates, number and distribution density of the first target pinhole, the damaged area and location information of the coating layer of the second target pinhole, and the cutting edge correlation data of the third target pinhole. The equipment status data for the entire process includes the wear of the die head of the base film preparation equipment, the stability of the evaporation source power of the vacuum coating machine, and the tool wear parameters of the slitting machine; Environmental monitoring data, including the cleanliness level of the base film preparation workshop, the residual gas composition in the vacuum coating chamber, and the temperature and humidity fluctuation curve of the slitting workshop; The traceability code uniquely binds the aforementioned target process parameters, target pinhole data, equipment status data, and environmental monitoring data through a mapping relationship between timestamps and spatial coordinates, forming a traceable full-chain data archive.

5. The method according to claim 1, characterized in that, The source tracing results obtained through source tracing analysis using a spatiotemporal correlation model combined with machine learning algorithms include: The spatiotemporal correlation model establishes time axis correlation information based on the timestamps of each process, mapping the three-dimensional coordinates of the target pinhole in the base film preparation process to the corresponding spatial positions in the vacuum coating process and the slitting process, forming the spatiotemporal trajectory of the pinhole's entire life cycle. The spatiotemporal trajectory and the real-time process parameter sequence of each process are input into a machine learning algorithm. The model learns the correlation between historical pinholes and process parameter anomalies and outputs the source tracing results.

6. The method according to claim 1, characterized in that, The process closed-loop control mechanism includes a three-level response: When the number of target pinholes in a single batch is less than or equal to a first preset threshold, a fine-tuning command is sent. When the number of target pinholes in a single batch is greater than the first preset threshold and less than or equal to the second preset threshold, a mid-level adjustment command is sent. When the number of pinholes in a single batch exceeds a second preset threshold, a shutdown inspection command is sent and an alarm is triggered.

Images