A method and system for controlling output power of UV variable frequency power supply

Abstract

The present invention discloses an output power control method and system for a UV variable frequency power supply, which relates to the field of UV printing technology, and includes the following steps: using a laser scanner to perform high-precision scanning on the ink layer of the printed product, and after obtaining the ink layer scanning data, using image processing technology to extract features from the scanning data; analyzing and processing the extracted features, based on the analyzed and processed ink layer feature data. The present invention uses a pre-trained machine learning model to identify potential over-curing phenomena, detect and handle problems in a timely manner, avoid production line stagnation, and ensure product quality consistency. At the same time, through the analysis of historical data and current process parameters, a linear regression model is used to calculate the expected deviation of the output power, and the exposure time is adjusted in real time in combination with fuzzy control logic to ensure the stability of the UV printing process, improve production efficiency through non-stop printing, and effectively improve product quality.

Description

Technical Field

[0001] The invention relates to the technical field of UV printing, and in particular to an output power control method and system of a UV variable frequency power supply. Background Art

[0002] The output power control of UV (ultraviolet) variable frequency power supply refers to the precise control of the power energy provided by the power supply to the UV light source by adjusting the output frequency, output voltage or other parameters of the variable frequency power supply to achieve the purpose of stabilizing and optimizing the output of the light source. In applications such as UV curing and UV printing, different materials and processes require different UV light intensities and exposure times, so it is necessary to control the output power of the variable frequency power supply to ensure that the energy of the UV light source matches these requirements.

[0003] During the UV printing process, the output power of the UV variable frequency power supply needs to be regulated, which is a key step to ensure printing quality and efficiency. UV printing uses ultraviolet light sources to cure special inks, so the intensity and exposure time of the UV light directly affect the curing effect. By regulating the output power of the UV variable frequency power supply, the intensity of the UV light source can be changed, thereby accurately matching the energy output of the UV light source with the printing requirements, ensuring that the curing effect during the printing process is optimal. This precise power control helps achieve high-quality printing results.

[0004] The prior art has the following deficiencies:

[0005] During the UV printing process, it is common for the output power regulation of the UV variable frequency power supply to deviate. When this deviation leads to potential over-curing of the ink layer, the existing technology usually adopts the method of shutting down for maintenance to deal with the problem of output power deviation. Shutdown for maintenance means that the entire printing production line must be suspended. Especially in a large-scale production environment, this shutdown will lead to a complete stagnation of production. The longer the shutdown time, the more serious the production delays the company will face, which may result in the failure to deliver orders on time.

[0006] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0007] The purpose of the present invention is to provide an output power control method and system for a UV variable frequency power supply, which uses a pre-trained machine learning model to identify potential over-curing phenomena, detect and handle problems in a timely manner, avoid production line stagnation, and ensure product quality consistency and corporate reputation. At the same time, through the analysis of historical data and current process parameters, a linear regression model is used to calculate the expected deviation of output power, and the exposure time is controlled in real time in combination with fuzzy control logic to ensure the stability of the UV printing process, improve production efficiency through non-stop printing, and effectively improve product quality, thereby enhancing corporate competitiveness, so as to solve the problems in the above-mentioned background technology.

[0008] In order to achieve the above object, the present invention provides the following technical solution: an output power control method of a UV variable frequency power supply, comprising the following steps:

[0009] The ink layer of the printed product is scanned with high precision by a laser scanner. After the ink layer scanning data is obtained, the image processing technology is used to extract features from the scanning data.

[0010] Analyze and process the extracted features, and identify potential over-curing phenomena through a pre-trained machine learning model based on the analyzed and processed ink layer feature data;

[0011] When potential over-curing is identified, the expected output power of the variable frequency power supply during UV printing is calculated using a linear regression model by analyzing historical data and combining it with the current process parameter input;

[0012] During the UV printing process, the actual output power of the variable frequency power supply is monitored and recorded in real time;

[0013] The actual output power obtained in real time is compared with the expected output power predicted by the linear regression model, and the expected deviation of the output power is calculated;

[0014] Based on the calculated expected deviation of output power, fuzzy control logic is used to control the exposure time in real time during UV printing;

[0015] According to the control results output by the fuzzy control logic, the exposure time during UV printing is adjusted in real time to ensure the continuity and quality stability of the entire printing process.

[0016] Preferably, the extracted ink layer scanning data includes microcracks and reflected light of the ink layer. Microcracks refer to tiny cracks appearing on the surface of the ink layer, and reflected light refers to the part of light reflected back by the ink layer when a laser beam or other light source irradiates the surface of the ink layer.

[0017] Preferably, the specific steps of extracting features of micro-cracks in the ink layer using image processing technology are as follows:

[0018] Preprocessing the ink layer scanning data obtained by the laser scanner;

[0019] After the pre-processing is completed, edge detection is performed to preliminarily identify micro-cracks on the surface of the ink layer;

[0020] Apply morphological processing techniques to further process the detected edges;

[0021] After morphological processing, connected domain analysis is performed on the microcracks to identify and mark the continuous crack areas in the image.

[0022] Preferably, the specific steps of extracting features of reflected light from the ink layer using image processing technology are as follows:

[0023] Before starting image segmentation, the ink layer scanning data obtained from the laser scanner is first preprocessed to remove noise and interference in the image and improve the accuracy of subsequent image segmentation;

[0024] After image preprocessing, the threshold segmentation method is used to perform preliminary reflection light area recognition. Threshold segmentation divides the pixel values ​​of the image into reflection light area and non-reflection light area by setting a threshold.

[0025] After the initial segmentation, edge detection technology is further used to accurately identify the boundaries of the reflected light area;

[0026] After edge detection, the initially identified reflected light area is refined to eliminate errors and further optimize the contour of the area.

[0027] Preferably, the microcrack density and reflected light uniformity of the ink layer are analyzed and processed to generate a microcrack density index and a reflection uniformity rise index, respectively. The microcrack density index and the reflection uniformity rise index are input into a pre-trained machine learning model to generate a potential curing coefficient, and the potential over-curing phenomenon of the ink layer of the printed product is identified through the potential curing coefficient.

[0028] Preferably, the potential curing coefficient generated after comprehensive analysis of the ink layer characteristic data is compared and analyzed with a preset potential curing coefficient reference threshold to identify the potential over-curing phenomenon in the ink layer. The specific process is as follows:

[0029] If the potential curing coefficient is greater than or equal to the potential curing coefficient reference threshold, the ink layer is identified as a potential over-cured ink layer;

[0030] If the latent curing coefficient is less than the latent curing coefficient reference threshold, the ink layer is identified as a non-potential over-cured ink layer.

[0031] Preferably, the expected deviation of the output power is calculated, and the specific steps are as follows:

[0032] Conduct in-depth analysis of historical data, extract key parameters related to UV printing, and combine them with current process parameters to input P evn (t), P mat (t) and P proc (t) Construct a linear regression model, where P evn (t), P mat (t), P proc (t) represent environmental parameters, material parameters, and process parameters, respectively, and are used to predict the expected output power P of the variable frequency power supply. pred (t), the expression of the linear regression model is: In the formula, as well as They represent the rate of change of environmental parameters, material parameters and process parameters over time, α1, α2 and α3 are regression coefficients, representing the weight of each parameter on the output power, φ k (t) is a nonlinear function of other potential influencing factors, β k is the weight coefficient corresponding to the potential impact factor, and v represents the total number of other potential impact factors;

[0033] During the UV printing process, the actual output power P of the variable frequency power supply is monitored in real time. actual (t), the actual output power is measured by the sensor installed in the equipment, which indicates the actual power supplied by the power supply to the UV lamp under the current printing environment and process conditions;

[0034] The actual output power P obtained in real time actual (t) and the expected output power P predicted by the linear regression model pred (t) and calculate the expected output power deviation ΔP(t). The calculation formula of the expected deviation is as follows:

[0035]

[0036] , where τ represents the time variable, t0 represents the starting time for comparing the actual output power with the expected output power, t1 is the current time, f(τ) is the time weight function, and the expected output power deviation ΔP(t) represents the weighted cumulative amount of the power deviation.

[0037] Preferably, based on the calculated expected deviation of the output power, fuzzy control logic is used to control the exposure time in the UV printing process in real time. The specific steps are as follows:

[0038] Define the input variables and output variables of the fuzzy control system. The input variable is the expected deviation of the calculated output power, and the output variable is the adjustment amount of the exposure time during UV printing.

[0039] Before entering the fuzzy control system, the output power expected deviation is fuzzified, that is, the output power expected deviation is converted into a fuzzy value in the fuzzy set;

[0040] Apply the fuzzified input variables to predefined fuzzy control rules for reasoning and decision making;

[0041] Defuzzify the output of the fuzzy controller. Defuzzification is to convert the fuzzy set into a specific value.

[0042] According to the exposure time adjustment value obtained by defuzzification processing, the exposure time of the UV printing device is adjusted in real time;

[0043] After adjusting the exposure time, monitor the print quality and equipment operating status in real time to evaluate the effect of the adjustment.

[0044] Preferably, the specific steps of analyzing and processing the microcrack density of the ink layer to generate a microcrack density index are as follows:

[0045] The detected microcracks are used to establish a microcrack set, and the microcrack set is represented by L, then: L = {L i}={L1, L2, L3, L4,…,L n}, where i represents the number of microcracks in the microcrack set, and n represents the total number of microcracks in the microcrack set;

[0046] Each crack L i The length is l i Indicates that each crack L i The width is w i It means that based on the detected crack information, a crack density field D(x, y) is generated. D(x, y) is a function of the spatial coordinates (x, y) and represents the crack density of each point on the surface of the ink layer. The crack density field is calculated by the following formula:

[0047]

[0048] , where α is an adjustment coefficient used to control the decay rate of crack density, d i (x, y) represents the distance from point (x, y) to the center of the i-th crack;

[0049] The microcrack density index of the entire ink layer area is calculated based on the crack density field D(x, y). The calculation expression is: Where microcrack DI represents the microcrack density index, A is the entire surface area of ​​the ink layer, and They represent the gradients of the crack density field D(x, y) in the x and y directions respectively, and γ is a nonlinear index used to amplify the influence of crack density.

[0050] Preferably, the specific steps of analyzing and processing the uniformity of reflected light of the ink layer to generate a reflection uniformity rise index are as follows:

[0051] Obtain the reflected light intensity of each pixel in the image, and record the reflected light intensity at (x, y) as I(x, y);

[0052] The reflected light intensity of each pixel in the image is gradient-calculated to obtain the reflected light gradient field vector G(x, y) and direction field vector θ(x, y), where G(x, y) represents the reflected light intensity gradient at (x, y), and θ(x, y) represents the reflected light direction gradient at (x, y). The calculation expression is: and Represent the partial derivatives of the reflected light intensity in the x and y directions respectively;

[0053] The synergy function C(x, y) of the gradient is calculated in the local area of ​​the image. The synergy function is used to quantify the consistency of the gradient direction of all pixels in the area. The calculation expression is: In the formula, θ(x+a, y+b) represents the gradient direction of other pixels in the local area, a is the horizontal offset, that is, the moving distance on the x-axis, and b is the vertical offset, that is, the moving distance on the y-axis;

[0054] Based on the generated synergy function C(x, y), the reflection uniformity rise index is calculated, and the calculation expression is: In the formula, Reflection Unif represents the reflection uniformity rising index, M is the total number of local regions, and β is the weight coefficient for adjusting the influence of the gradient field intensity G(x, y).

[0055] An output power control system of a UV variable frequency power supply, comprising an ink layer data acquisition and feature extraction module, an over-curing identification module, an output power estimation module, a real-time power monitoring module, a power deviation calculation module, a fuzzy control adjustment module and an exposure time adjustment module;

[0056] The ink layer data acquisition and feature extraction module uses a laser scanner to perform high-precision scanning of the ink layer of the printed product. After obtaining the ink layer scanning data, it uses image processing technology to extract features from the scanning data.

[0057] The over-curing identification module analyzes and processes the extracted features, and identifies potential over-curing phenomena through a pre-trained machine learning model based on the analyzed and processed ink layer feature data;

[0058] Output power estimation module, when potential over-curing phenomenon is identified, by analyzing historical data and combining it with current process parameter input, a linear regression model is used to calculate the expected output power of the variable frequency power supply during UV printing;

[0059] Real-time power monitoring module, during UV printing process, real-time monitoring and recording of the actual output power of the variable frequency power supply;

[0060] The power deviation calculation module compares the actual output power obtained in real time with the expected output power predicted by the linear regression model, and calculates the expected output power deviation;

[0061] The fuzzy control adjustment module uses fuzzy control logic to adjust the exposure time in the UV printing process in real time based on the calculated expected deviation of the output power;

[0062] The exposure time adjustment module adjusts the exposure time during UV printing in real time according to the control results output by the fuzzy control logic to ensure the continuity and quality stability of the entire printing process.

[0063] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0064] The present invention uses a laser scanner to perform high-precision scanning of the ink layer of the printed product, uses image processing technology to extract features from the scanned data, and uses a pre-trained machine learning model to predict the extracted features. It can effectively identify potential over-curing in the printed product, so as to detect and diagnose the problem when it first occurs, thereby avoiding stagnation of the entire production line. By identifying and handling potential problems at an early stage, it can greatly reduce large-scale product quality problems caused by over-curing of the ink layer, thereby ensuring product consistency and reliability, reducing economic losses, and maintaining the company's customer relationships and market reputation.

[0065] The present invention uses a linear regression model to predict the expected output power of the variable frequency power supply during UV printing through a comprehensive analysis of historical data and current process parameters, monitors the deviation between the actual output power and the predicted value in real time, and uses fuzzy control logic to dynamically adjust the exposure time in UV printing based on this deviation. The introduction of fuzzy control logic enables the system to quickly adjust the exposure time after identifying potential over-curing phenomena, thereby ensuring the stability of the UV printing process and the reliability of product quality. This real-time control method not only improves production efficiency and reduces the need for downtime and maintenance, but also ensures high-quality output of products, further enhancing the competitiveness of enterprises in mass production. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0067] Figure 1 The present invention is a method flow chart of an output power control method of a UV variable frequency power supply.

[0068] Figure 2 The present invention is a schematic diagram of a module of an output power control system of a UV variable frequency power supply. DETAILED DESCRIPTION

[0069] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0070] The present invention provides Figure 1 The output power control method of a UV variable frequency power supply shown includes the following steps:

[0071] The ink layer of the printed product is scanned with high precision by a laser scanner. After the ink layer scanning data is obtained, the image processing technology is used to extract features from the scanning data.

[0072] The extracted ink layer scanning data includes micro cracks and reflected light of the ink layer;

[0073] Microcracks refer to tiny cracks that appear on the surface of the ink layer. These cracks are usually not directly visible to the naked eye, but can be detected by high-precision scanning equipment. The formation of microcracks is usually related to stress concentration during the curing process of the ink layer. Especially in the case of over-curing, these cracks may cause the physical strength of the ink layer to decrease, the adhesion to weaken, and even affect the durability and appearance quality of the printed product.

[0074] Reflected light refers to the part of the light reflected by the ink layer when the laser beam or other light source irradiates the surface of the ink layer. The intensity and uniformity of the reflected light can reflect the surface flatness, smoothness and curing degree of the ink layer. Different curing conditions will cause the reflection characteristics of the ink layer surface to change. For example, an over-cured ink layer may show abnormally high reflection or scattering characteristics.

[0075] The specific steps of using image processing technology to extract features of micro cracks in the ink layer are as follows:

[0076] Preprocessing the ink layer scanning data obtained by the laser scanner;

[0077] This step includes removing noise, correcting uneven illumination, and adjusting image contrast. Noise removal can be achieved through filters, and illumination correction is performed through techniques such as mean smoothing or histogram equalization. The purpose of preprocessing is to improve the accuracy of microcrack detection in subsequent analysis and ensure the quality of data and image clarity.

[0078] After the pre-processing is completed, edge detection is performed to preliminarily identify micro-cracks on the surface of the ink layer;

[0079] Edge detection is a technique for identifying areas with strong brightness changes in an image. Commonly used algorithms include the Sobel operator, Canny edge detector, etc. Through edge detection, the location of possible microcracks on the surface of the ink layer can be preliminarily determined, and a binary image containing these edge information can be generated. This image will be used for subsequent feature extraction.

[0080] Apply morphological processing techniques to further process the detected edges;

[0081] This includes operations such as erosion, dilation, opening and closing to eliminate isolated noise points and connect broken crack edges. Morphological processing helps improve the results of edge detection, making the outlines of microcracks clearer and more coherent, thereby improving the accuracy of feature extraction.

[0082] After morphological processing, the connected domain analysis of microcracks is performed to identify and mark the continuous crack areas in the image;

[0083] Connected domain analysis is a technique for segmenting images based on the connectivity between pixels, which can effectively distinguish crack areas from other irrelevant features. By analyzing these connected domains, the size, shape (such as length and width) and location of each microcrack can be further determined.

[0084] The specific steps of extracting features of reflected light from the ink layer using image processing technology are as follows:

[0085] Before starting image segmentation, the ink layer scanning data obtained from the laser scanner is first preprocessed to remove noise and interference in the image and improve the accuracy of subsequent image segmentation;

[0086] Commonly used preprocessing methods include mean filtering or Gaussian filtering to smooth the image and reduce errors caused by noise. In addition, the brightness and contrast of the image can also be adjusted through histogram equalization to enhance the visibility of the reflected light area and make it more prominent in the image.

[0087] After image preprocessing, the threshold segmentation method is used to perform preliminary reflection light area recognition. Threshold segmentation divides the pixel values ​​of the image into reflection light area and non-reflection light area by setting a threshold.

[0088] The grayscale value of each pixel in the image is compared with the threshold. If the grayscale value of the pixel is greater than or equal to the set threshold, the pixel is classified as a reflected light area (usually displayed as white); if the grayscale value is less than the threshold, the pixel is classified as a non-reflected light area (usually displayed as black). In this way, the image is binarized, and the reflected light area is clearly separated from the background, forming a simplified binary image for subsequent refinement processing.

[0089] After the initial segmentation, edge detection technology is further used to accurately identify the boundaries of the reflected light area;

[0090] Commonly used edge detection algorithms include Canny, Sobel, and Laplacian, which can detect the sharp edges of the reflected light area in the image and extract its contour. The results of edge detection are usually represented in the form of a binary image, and the boundaries of the reflected light area are clearly depicted. The accuracy of this step is crucial because it directly determines the shape and range of the reflected light area.

[0091] After edge detection, the initially identified reflected light area is refined to eliminate errors and further optimize the contour of the area;

[0092] The refinement process may include morphological operations such as erosion and dilation to remove noise points and fill holes in the area. In addition, the real reflected light area can be further filtered and extracted through connected domain analysis to ensure that only meaningful high-reflection areas are retained. Finally, the extracted reflected light area data is stored and used for subsequent feature analysis.

[0093] Analyze and process the extracted features, and identify potential over-curing phenomena through a pre-trained machine learning model based on the analyzed and processed ink layer feature data;

[0094] The micro-crack density and reflected light uniformity of the ink layer are analyzed and processed to generate a micro-crack density index and a reflection uniformity rise index respectively. The micro-crack density index and the reflection uniformity rise index are input into a pre-trained machine learning model to generate a potential curing coefficient. The potential over-curing phenomenon of the ink layer of the printed product is identified through the potential curing coefficient.

[0095] The increase of surface microcracks in the finished ink layer of the printed product usually indicates that the ink layer is potentially over-cured. During the over-curing process, the ink layer may be exposed to high-intensity UV light for a long time, which will increase the internal stress of the material and form microcracks. These microcracks not only affect the mechanical properties of the ink layer, such as flexibility and adhesion, but may also visually cause surface unevenness or changes in optical properties. The appearance of microcracks is usually caused by excessive shrinkage or embrittlement of the material during the curing process, all of which indicate that the curing energy exceeds the range that the material can withstand. Therefore, the increase of microcracks is an important indicator of over-curing of the ink layer.

[0096] The specific steps for analyzing and processing the microcrack density of the ink layer to generate the microcrack density index are as follows:

[0097] The detected microcracks are used to establish a microcrack set, and the microcrack set is represented by L, then: L = {L i}={L1, L2, L3, L4,…,L n}, where i represents the number of microcracks in the microcrack set, and n represents the total number of microcracks in the microcrack set;

[0098] Each crack L i The length is l i Indicates that each crack L i The width is w i It means that based on the detected crack information, a crack density field D(x, y) is generated. D(x, y) is a function of the spatial coordinates (x, y) and represents the crack density of each point on the surface of the ink layer. The crack density field is calculated by the following formula:

[0099]

[0100] , where α is an adjustment coefficient used to control the decay rate of crack density, d i (x, y) represents the distance from point (x, y) to the center of the i-th crack;

[0101] The microcrack density index of the entire ink layer area is calculated based on the crack density field D(x, y). The calculation expression is: Where microcrack DI represents the microcrack density index, A is the entire surface area of ​​the ink layer, and They represent the gradients of the crack density field D(x, y) in the x and y directions respectively. γ is a nonlinear index used to amplify the influence of crack density to more sensitively reflect the aggregation effect of microcracks.

[0102] It can be seen from the microcrack density index that the larger the performance value of the microcrack density index generated by analyzing the microcrack density of the ink layer, the higher the density of microcracks in the ink layer of the printed product. This means that the ink layer may have been subjected to excessive stress or light during the curing process, resulting in more serious crack formation. Therefore, the larger the microcrack density index, the greater the potential risk of over-curing of the ink layer; conversely, the smaller the microcrack density index, the lower the density of microcracks, the more uniform the ink layer, and the smaller the risk of over-curing. Therefore, the microcrack density index can be used as an important indicator to evaluate the curing quality of printed products.

[0103] An abnormally high uniformity of reflected light on the surface of the finished ink layer may indeed indicate potential over-curing of the ink layer. Over-curing usually causes the surface of the ink layer to become abnormally smooth and flat, thereby increasing the uniformity of light reflection. Therefore, an abnormally high uniformity of reflected light is one of the important indicators for detecting over-curing of the ink layer.

[0104] The specific steps for analyzing and processing the uniformity of reflected light of the ink layer to generate the reflection uniformity rise index are as follows:

[0105] Obtain the reflected light intensity of each pixel in the image, and record the reflected light intensity at (x, y) as I(x, y);

[0106] The laser scanner can obtain the reflected light intensity of each pixel in the image. When scanning the ink layer of the printed product, the laser scanner emits a precise laser beam. After the laser beam hits the surface of the ink layer, it will produce reflections according to the surface characteristics. The laser scanner's sensor receives the reflected light signal and converts it into a digital signal, recording the reflected light intensity value of each pixel.

[0107] The reflected light intensity of each pixel in the image is gradient-calculated to obtain the reflected light gradient field vector G(x, y) and the direction field vector θ(x, y), where G(x, y) represents the reflected light intensity gradient at (x, y), θ(x, y) represents the reflected light directional gradient at (x, y), the gradient field vector represents the rate of change of light intensity, and the direction field vector represents the rate of change of light direction angle. The calculation expression is: and Represent the partial derivatives of the reflected light intensity in the x and y directions respectively. These vectors can be used to evaluate the change in the reflected light intensity around each pixel;

[0108] The synergy function C(x, y) of the gradient (reflected light intensity gradient and reflected light direction gradient) is calculated in a local area of ​​the image (for example, an m×m window). The synergy function is used to quantify the consistency of the gradient direction of all pixels in the area. If the gradient direction is highly consistent, it means that the uniformity of the reflected light in the area is high. The calculation expression of the synergy function C(x, y) is: In the formula, θ(x+a, y+b) represents the gradient direction of other pixels in the local area, a is the horizontal offset, that is, the moving distance on the x-axis, and b is the vertical offset, that is, the moving distance on the y-axis;

[0109] The synergy between directions is calculated by the cosine function. If all directions are consistent, the C(x, y) value is close to 1, indicating high uniformity; if the directions are inconsistent, the C(x, y) value is close to 0, indicating low uniformity.

[0110] Based on the generated synergy function C(x, y), the reflection uniformity rise index is calculated (by combining the synergy and gradient field intensity in the region to evaluate whether the reflected light uniformity has increased abnormally), and the calculation expression is: In the formula, Reflection Unif represents the reflection uniformity increase index, M is the total number of local areas, β is the weight coefficient for adjusting the influence of the gradient field intensity G(x, y). The exponential function exp(β·G(x, y)) amplifies the influence of the area with a larger reflected light intensity gradient on the index. Especially in the case of high synergy, the reflection uniformity increase index is higher, indicating that the uniformity of the reflected light is significantly improved.

[0111] It can be seen from the reflection uniformity rise index that the larger the value of the reflection uniformity rise index generated by analyzing the uniformity of the reflected light of the ink layer, the more the uniformity of the reflected light on the surface of the ink layer has increased significantly, which usually indicates that the ink layer may have experienced over-curing during the UV curing process. Over-curing will cause the surface of the ink layer to become abnormally smooth and flat, making the reflected light more uniform, and the reflection uniformity rise index will therefore increase. Therefore, the larger the reflection uniformity rise index, the greater the possibility that the ink layer of the printed product has a potential over-curing risk; on the contrary, the smaller the reflection uniformity rise index, the more normal the curing process of the ink layer is, and the potential over-curing risk is smaller.

[0112] The machine learning model is not specifically limited here, and can realize the microcrack density index microcrack DI and reflection uniformity rise index Reflection Unif Comprehensive analysis to generate potential curing coefficient Curing Co In order to realize the technical solution of the present invention, the present invention provides a specific implementation method;

[0113] Potential curing coefficient Co The resulting calculation formula is:

[0114]

[0115] , where ξ1 and ξ2 are microcrack density indexes, respectively. DI and reflection uniformity rise index Reflection Unif The preset proportional coefficient, and ξ1 and ξ2 are both greater than 0.

[0116] It can be seen from the potential curing coefficient that the greater the expression value of the microcrack density index generated by analyzing and processing the microcrack density of the ink layer, the greater the expression value of the reflection uniformity rise index generated by analyzing and processing the uniformity of the reflected light of the ink layer. That is, the greater the expression value of the potential curing coefficient generated after comprehensive analysis of the ink layer characteristic data, the greater the potential risk of over-curing of the ink layer, and vice versa.

[0117] The potential curing coefficient generated after comprehensive analysis of the ink layer characteristic data is compared and analyzed with the preset potential curing coefficient reference threshold to identify the potential over-curing phenomenon of the ink layer. The specific process is as follows:

[0118] If the potential curing coefficient is greater than or equal to the potential curing coefficient reference threshold, the ink layer is identified as a potential over-cured ink layer;

[0119] If the latent curing coefficient is less than the latent curing coefficient reference threshold, the ink layer is identified as a non-potential over-cured ink layer.

[0120] When potential over-curing is identified, the expected output power of the variable frequency power supply during UV printing is calculated using a linear regression model by analyzing historical data and combining it with the current process parameter input. During the UV printing process, the actual output power of the variable frequency power supply is monitored and recorded in real time. The actual output power obtained in real time is compared with the expected output power predicted by the linear regression model to calculate the expected output power deviation.

[0121] Calculate the expected output power deviation. The specific steps are as follows:

[0122] Conduct in-depth analysis of historical data to extract key parameters related to UV printing, such as ambient temperature, humidity, ink type, printing speed, etc., and combine them with current process parameters to input P evn (t)(environmental parameters), P mat (t) (material parameters) and P proc (t) (process parameters) to build a linear regression model to predict the expected output power P of the variable frequency power supply pred (t), the expression of the linear regression model is: In the formula, as well as They represent the change rates of environmental parameters, material parameters and process parameters over time, respectively. These change rates can capture the sensitivity of power demand under dynamic environmental and material conditions. α1, α2 and α3 are regression coefficients, indicating the influence weight of each parameter on the output power. φ k (t) is a nonlinear function of other potential influencing factors (such as equipment wear and light source attenuation), β k is the weight coefficient corresponding to the potential impact factor, and v represents the total number of other potential impact factors;

[0123] Environmental parameters refer to various physical and chemical factors related to external environmental conditions during UV printing. These parameters usually include ambient temperature, humidity, air flow, light intensity, etc. In UV printing, temperature and humidity will directly affect the curing speed and effect of ink. For example, high temperature may accelerate the evaporation of ink, resulting in uneven curing, while high humidity may cause condensation on the surface of the ink layer, affecting the curing process. Air flow will affect the transmission of UV light and the volatilization speed of ink during the curing process. Changes in environmental parameters need to be monitored and controlled in real time to ensure the consistency and stability of printing quality.

[0124] Material parameters involve the characteristics of all materials used in the UV printing process, including the composition, viscosity, surface tension, color density of the ink, as well as the material, absorbency and surface roughness of the printing medium. Different materials will exhibit different curing behaviors under UV light. For example, inks with higher viscosity may require stronger UV light to fully cure, and the reflection and absorption characteristics of UV light of printing media of different materials will also affect the curing effect. Material parameters determine the process settings that should be used in specific printing tasks. Reasonable adjustment of these parameters can improve the quality and efficiency of printing.

[0125] Process parameters refer to the various settings of the control equipment and operating procedures during UV printing, including printing speed, movement trajectory of the print head, ink jet volume, etc. These parameters directly determine the deposition and curing process of the ink on the medium.

[0126] During the UV printing process, the actual output power P of the variable frequency power supply is monitored in real time. actual (t), the actual output power is measured by the sensor installed in the equipment, which indicates the actual power supplied by the power supply to the UV lamp under the current printing environment and process conditions;

[0127] In this step, the data sampling frequency needs to be high enough to ensure that the instantaneous changes in the power supply output can be captured, thereby reflecting the real-time process status.

[0128] The actual output power P obtained in real time actual (t) and the expected output power P predicted by the linear regression model pred (t) and calculate the expected output power deviation ΔP(t). The calculation formula of the expected deviation is as follows:

[0129]

[0130] , where τ represents the time variable, t0 represents the starting time for comparing the actual output power with the expected output power, t1 is the current time, and f(τ) is the time weight function, which can be a certain attenuation function (such as f(τ) = e -τ ), which is used to emphasize the impact of recent output power changes on the deviation. The expected output power deviation ΔP(t) represents the weighted cumulative amount of power deviation, capturing the output power deviation accumulated over time.

[0131] Based on the calculated expected deviation of output power, fuzzy control logic is used to control the exposure time in the UV printing process in real time. According to the control result output by the fuzzy control logic, the exposure time in the UV printing process is adjusted in real time to ensure the continuity and quality stability of the entire printing process.

[0132] Based on the calculated expected deviation of output power, fuzzy control logic is used to control the exposure time in real time during UV printing. The specific steps are as follows:

[0133] Define the input variables and output variables of the fuzzy control system. The input variable is the expected deviation of the calculated output power, and the output variable is the adjustment amount of the exposure time during UV printing.

[0134] At this stage, developers need to design a set of fuzzy control rules, which are usually expressed in the form of "if...then...", for example, "if the power deviation is large and the temperature is high, then the exposure time should be reduced." These rules are based on expert knowledge and experience of the UV printing process and are designed to cope with various possible changes in working conditions.

[0135] In actual operation, the expected output power deviation is usually an accurate value. Therefore, before entering the fuzzy control system, the expected output power deviation is fuzzified, that is, the expected output power deviation is converted into a fuzzy value in the fuzzy set, such as "small deviation", "medium deviation" or "large deviation". This process is achieved through a pre-set fuzzy membership function, which maps the value to one or more fuzzy sets to reflect the degree of actual deviation. Through fuzzification, the control system can respond more flexibly to complex and nonlinear system changes.

[0136] Apply the fuzzified input variables to predefined fuzzy control rules for reasoning and decision making;

[0137] Specifically, the fuzzy controller will judge the applicability of each rule according to the fuzzy value of the input variable and calculate the corresponding control output. Different rules may produce different outputs, which will be combined to form a fuzzy set to represent the adjustment suggestions for the exposure time. The process of fuzzy reasoning is essentially to comprehensively consider multiple complex conditions and generate a suitable adjustment strategy.

[0138] The output of a fuzzy control system is usually a fuzzy set, while the actual control system requires an accurate value to adjust the exposure time. Therefore, the output of the fuzzy controller needs to be defuzzified. Defuzzification is the process of converting a fuzzy set into a specific value. Commonly used methods include the centroid method and the maximum membership method. Through defuzzification, a specific exposure time adjustment value is finally obtained, which will be used to adjust the UV printing equipment in actual operation.

[0139] According to the exposure time adjustment value obtained by defuzzification, the exposure time of the UV printing device is adjusted in real time. This adjustment is done automatically, and the control system directly transmits the command to the printing device. The goal of real-time adjustment is to ensure that the exposure time of each section in the printing process can match the current process and environmental conditions, and compensate for any potential curing problems caused by output power deviation or other influencing factors. This process is continuously cycled through a closed-loop control system to ensure the consistency and reliability of printing quality.

[0140] After adjusting the exposure time, monitor the print quality and equipment operating status in real time to evaluate the effect of the adjustment;

[0141] If quality problems or abnormal equipment status are found during the printing process, the control system will further adjust the fuzzy control rules or parameters for dynamic optimization. The focus of this stage is to gradually improve the control strategy through continuous monitoring and feedback to cope with possible changes and ensure the continuity of the entire printing process and the long-term stability of quality.

[0142] The present invention uses a laser scanner to perform high-precision scanning of the ink layer of the printed product, uses image processing technology to extract features from the scanned data, and uses a pre-trained machine learning model to predict the extracted features, thereby identifying potential over-curing in the printed product, so as to detect and diagnose the problem when it first occurs, thereby avoiding stagnation of the entire production line. By identifying and handling potential problems at an early stage, large-scale product quality problems caused by over-curing of the ink layer can be greatly reduced, thereby ensuring product consistency and reliability, reducing economic losses, and maintaining the company's customer relationships and market reputation.

[0143] The present invention uses a linear regression model to predict the expected output power of the variable frequency power supply during UV printing through a comprehensive analysis of historical data and current process parameters, monitors the deviation between the actual output power and the predicted value in real time, and uses fuzzy control logic to dynamically adjust the exposure time in UV printing based on this deviation. The introduction of fuzzy control logic enables the system to quickly adjust the exposure time after identifying potential over-curing phenomena, thereby ensuring the stability of the UV printing process and the reliability of product quality. This real-time control method not only improves production efficiency and reduces the need for shutdown and maintenance, but also ensures high-quality output of products, further enhancing the competitiveness of enterprises in mass production.

[0144] The present invention provides Figure 2 An output power control system of a UV variable frequency power supply shown includes an ink layer data acquisition and feature extraction module, an over-curing identification module, an output power estimation module, a real-time power monitoring module, a power deviation calculation module, a fuzzy control adjustment module, and an exposure time adjustment module;

[0145] The ink layer data acquisition and feature extraction module uses a laser scanner to perform high-precision scanning of the ink layer of the printed product. After obtaining the ink layer scanning data, it uses image processing technology to extract features from the scanning data.

[0146] The over-curing identification module analyzes and processes the extracted features, and identifies potential over-curing phenomena through a pre-trained machine learning model based on the analyzed and processed ink layer feature data;

[0147] Output power estimation module, when potential over-curing phenomenon is identified, by analyzing historical data and combining it with current process parameter input, a linear regression model is used to calculate the expected output power of the variable frequency power supply during UV printing;

[0148] Real-time power monitoring module, during UV printing process, real-time monitoring and recording of the actual output power of the variable frequency power supply;

[0149] The power deviation calculation module compares the actual output power obtained in real time with the expected output power predicted by the linear regression model, and calculates the expected output power deviation;

[0150] The fuzzy control adjustment module uses fuzzy control logic to adjust the exposure time in the UV printing process in real time based on the calculated expected deviation of the output power;

[0151] The exposure time adjustment module adjusts the exposure time in the UV printing process in real time according to the control results output by the fuzzy control logic to ensure the continuity and quality stability of the entire printing process;

[0152] An output power control method of a UV variable frequency power supply provided in an embodiment of the present invention is implemented by the output power control system of the above-mentioned UV variable frequency power supply. The specific method and process of the output power control system of a UV variable frequency power supply are detailed in the embodiment of the output power control method of the above-mentioned UV variable frequency power supply, which will not be repeated here.

[0153] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0154] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0155] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for controlling the output power of a UV variable frequency power supply, characterized in that: The following steps are involved: The ink layer of the printed product is scanned with high precision by a laser scanner. After the ink layer scanning data is obtained, the image processing technology is used to extract features from the scanning data. Analyze and process the extracted features, and identify potential over-curing phenomena through a pre-trained machine learning model based on the analyzed and processed ink layer feature data; When potential over-curing is identified, the expected output power of the variable frequency power supply during UV printing is calculated using a linear regression model by analyzing historical data and combining it with the current process parameter input; During UV printing, the actual output power of the variable frequency power supply is monitored and recorded in real time; The actual output power obtained in real time is compared with the expected output power predicted by the linear regression model, and the expected deviation of the output power is calculated; Based on the calculated expected deviation of output power, fuzzy control logic is used to control the exposure time in real time during UV printing; According to the control results output by the fuzzy control logic, the exposure time during UV printing is adjusted in real time to ensure the continuity and quality stability of the entire printing process.

2. The output power control method of a UV variable frequency power supply according to claim 1, characterized in that: The extracted ink layer scanning data includes micro cracks and reflected light of the ink layer. Micro cracks refer to the tiny cracks on the surface of the ink layer, and reflected light refers to the part of the light reflected back by the ink layer when the laser beam or other light source irradiates the surface of the ink layer. The specific steps of using image processing technology to extract features of micro cracks in the ink layer are as follows: Preprocessing the ink layer scanning data obtained by the laser scanner; After the pre-processing is completed, edge detection is performed to preliminarily identify micro-cracks on the surface of the ink layer; Apply morphological processing techniques to further process the detected edges; After morphological processing, connected domain analysis is performed on the microcracks to identify and mark the continuous crack areas in the image.

3. The output power control method of a UV variable frequency power supply according to claim 2, characterized in that: The specific steps of extracting features of reflected light from the ink layer using image processing technology are as follows: Before starting image segmentation, the ink layer scanning data obtained from the laser scanner is first preprocessed to remove noise and interference in the image and improve the accuracy of subsequent image segmentation; After image preprocessing, the threshold segmentation method is used to perform preliminary reflection light area recognition. Threshold segmentation divides the pixel values ​​of the image into reflection light area and non-reflection light area by setting a threshold. After the initial segmentation, edge detection technology is further used to accurately identify the boundaries of the reflected light area; After edge detection, the initially identified reflected light area is refined to eliminate errors and further optimize the contour of the area.

4. The output power control method of a UV variable frequency power supply according to claim 2, characterized in that: The micro-crack density and reflected light uniformity of the ink layer are analyzed and processed to generate a micro-crack density index and a reflection uniformity rise index respectively. The micro-crack density index and the reflection uniformity rise index are input into a pre-trained machine learning model to generate a potential curing coefficient. The potential over-curing phenomenon of the ink layer of the printed product is identified through the potential curing coefficient.

5. The output power control method of a UV variable frequency power supply according to claim 4, characterized in that: The potential curing coefficient generated after comprehensive analysis of the ink layer characteristic data is compared and analyzed with the preset potential curing coefficient reference threshold to identify the potential over-curing phenomenon of the ink layer. The specific process is as follows: If the potential curing coefficient is greater than or equal to the potential curing coefficient reference threshold, the ink layer is identified as a potential over-cured ink layer; If the latent curing coefficient is less than the latent curing coefficient reference threshold, the ink layer is identified as a non-potential over-cured ink layer.

6. The output power control method of a UV variable frequency power supply according to claim 1, characterized in that: Calculate the expected output power deviation. The specific steps are as follows: Conduct in-depth analysis of historical data and combine it with current process parameters to input P evn (t), P mat (t) and P proc (t) Construct a linear regression model, where P evn (t), P mat (t), P proc (t) represent environmental parameters, material parameters, and process parameters, respectively, and are used to predict the expected output power P of the variable frequency power supply. pred (t), the expression of the linear regression model is: In the formula, as well as They represent the rate of change of environmental parameters, material parameters and process parameters over time, α1, α2 and α3 are regression coefficients, indicating the influence weight of each parameter on the output power, φ k (t) is a nonlinear function of other potential influencing factors, β k is the weight coefficient corresponding to the potential impact factor, and v represents the total number of other potential impact factors; During the UV printing process, the actual output power P of the variable frequency power supply is monitored in real time. actual (t), the actual output power is measured by the sensor installed in the equipment, which indicates the actual power supplied by the power supply to the UV lamp under the current printing environment and process conditions; The actual output power P obtained in real time actual (t) and the expected output power P predicted by the linear regression model pred (t) and calculate the expected output power deviation ΔP(t). The calculation formula of the expected deviation is as follows: Wherein, τ represents the time variable, t0 represents the starting time for comparing the actual output power with the expected output power, t1 is the current time, f(τ) is the time weight function, and the expected output power deviation ΔP(t) represents the weighted cumulative amount of the power deviation.

7. The output power control method of a UV variable frequency power supply according to claim 1, characterized in that: Based on the calculated expected deviation of output power, fuzzy control logic is used to control the exposure time in real time during UV printing. The specific steps are as follows: Define the input variables and output variables of the fuzzy control system. The input variable is the expected deviation of the calculated output power, and the output variable is the adjustment amount of the exposure time during UV printing. Before entering the fuzzy control system, the output power expected deviation is fuzzified, that is, the output power expected deviation is converted into a fuzzy value in the fuzzy set; Apply the fuzzified input variables to predefined fuzzy control rules for reasoning and decision making; Defuzzify the output of the fuzzy controller. Defuzzification is to convert the fuzzy set into a specific value. According to the exposure time adjustment value obtained by defuzzification processing, the exposure time of the UV printing device is adjusted in real time; After adjusting the exposure time, monitor the print quality and equipment operating status in real time to evaluate the effect of the adjustment.

8. The output power control method of a UV variable frequency power supply according to claim 4, characterized in that: The specific steps for analyzing and processing the microcrack density of the ink layer to generate the microcrack density index are as follows: The detected microcracks are used to establish a microcrack set, and the microcrack set is represented by L, then: L = {L i }={L1, L2, L3, L4,…,L n }, where i represents the number of microcracks in the microcrack set, and n represents the total number of microcracks in the microcrack set; Each crack L i The length is l i Indicates that each crack L i The width is w i It means that based on the detected crack information, a crack density field D(x, y) is generated. D(x, y) is a function of the spatial coordinates (x, y) and represents the crack density of each point on the surface of the ink layer. The crack density field is calculated by the following formula: Where α is an adjustment coefficient used to control the decay rate of crack density, d i (x, y) represents the distance from point (x, y) to the center of the i-th crack; The microcrack density index of the entire ink layer area is calculated based on the crack density field D(x, y). The calculation expression is: Where microcrack DI represents the microcrack density index, A is the entire surface area of ​​the ink layer, and They represent the gradients of the crack density field D(x, y) in the x and y directions respectively, and γ is a nonlinear index used to amplify the influence of crack density.

9. The output power control method of a UV variable frequency power supply according to claim 4, characterized in that: The specific steps for analyzing and processing the uniformity of reflected light of the ink layer to generate the reflection uniformity rise index are as follows: Obtain the reflected light intensity of each pixel in the image, and record the reflected light intensity at (x, y) as I(x, y); The reflected light intensity of each pixel in the image is gradient-calculated to obtain the reflected light gradient field vector G(x, y) and direction field vector θ(x, y), where G(x, y) represents the reflected light intensity gradient at (x, y), and θ(x, y) represents the reflected light direction gradient at (x, y). The calculation expression is: and Represent the partial derivatives of the reflected light intensity in the x and y directions respectively; The synergy function C(x, y) of the gradient is calculated in the local area of ​​the image. The synergy function is used to quantify the consistency of the gradient direction of all pixels in the local area. The calculation expression is: Wherein, θ(x+a, y+b) represents the gradient direction of other pixel points in the local area, a is the horizontal offset, that is, the moving distance on the x-axis, b is the vertical offset, that is, the moving distance on the y-axis, and the local area is an m×m window; Based on the generated synergy function C(x, y), the reflection uniformity rise index is calculated, and the calculation expression is: In the formula, Reflection Unif represents the reflection uniformity rising index, M is the total number of local regions, and β is the weight coefficient for adjusting the influence of the gradient field intensity G(x, y).

10. An output power control system of a UV variable frequency power supply, used to implement an output power control method of a UV variable frequency power supply as described in any one of claims 1 to 9, characterized in that: It includes ink layer data acquisition and feature extraction module, over-curing identification module, output power estimation module, real-time power monitoring module, power deviation calculation module, fuzzy control adjustment module and exposure time adjustment module; The ink layer data acquisition and feature extraction module uses a laser scanner to perform high-precision scanning of the ink layer of the printed product. After obtaining the ink layer scanning data, it uses image processing technology to extract features from the scanning data. The over-curing identification module analyzes and processes the extracted features, and identifies potential over-curing phenomena through a pre-trained machine learning model based on the analyzed and processed ink layer feature data; Output power estimation module, when potential over-curing phenomenon is identified, by analyzing historical data and combining it with current process parameter input, a linear regression model is used to calculate the expected output power of the variable frequency power supply during UV printing; Real-time power monitoring module, during UV printing process, real-time monitoring and recording of the actual output power of the variable frequency power supply; The power deviation calculation module compares the actual output power obtained in real time with the expected output power predicted by the linear regression model, and calculates the expected output power deviation; The fuzzy control adjustment module uses fuzzy control logic to adjust the exposure time in the UV printing process in real time based on the calculated expected deviation of the output power; The exposure time adjustment module adjusts the exposure time during UV printing in real time according to the control results output by the fuzzy control logic to ensure the continuity and quality stability of the entire printing process.

Images