An ink drying control method for glass production

Abstract

The application discloses an ink drying control method for glass production and belongs to the technical field of ink printing, and specifically comprises the following steps: acquiring ink thickness distribution data and ink formula component data; reading the volatilization activation energy and infrared absorption peak wavelength of each solvent to determine the solvent residue proportion of each position point; dividing the surface of a glass plate into a plurality of drying areas, each area corresponding to the use of wavelength-adjustable infrared radiation heating; determining the corresponding infrared absorption peak wavelength according to the solvent type of each area, and then heating by using the corrected wavelength after correcting the wavelength according to the solvent residue proportion; calculating the evaporation heat according to the ink thickness and the solvent residue proportion and determining the heating intensity; acquiring the back surface temperature value and the volatilization gas concentration change rate of each area, and obtaining an adjustment coefficient through fuzzy rule processing and adjusting the heating intensity according to the coefficient. The application realizes the dynamic matching of the infrared radiation wavelength and the solvent residue state in the ink drying process.

Description

Technical Field

[0001] This invention relates to the field of ink printing technology, and more specifically to an ink drying control method for glass production. Background Technology

[0002] The ink drying process in glass production is a crucial step in the thermal curing of decorative or functional inks printed on the glass surface. This process directly affects the adhesion, abrasion resistance, and chemical corrosion resistance of the ink layer. As glass products develop towards multi-color and high-precision manufacturing, the ink drying process faces increasingly stringent requirements for uniformity. Traditional drying methods employ overall heating, either by using infrared heating tubes within the drying chamber to uniformly heat the entire glass sheet, or by dividing the glass sheet into several strip areas and controlling the heating power of each area separately.

[0003] Currently, some improved drying control methods have incorporated temperature feedback mechanisms. These methods collect temperature distribution data by arranging thermocouple arrays on the surface or back of the glass plate and comparing it with a preset target temperature. The output power of the infrared heating tubes in the corresponding areas is then adjusted based on the temperature deviation. Furthermore, some methods attempt to adjust the heating intensity based on the ink thickness distribution, allocating higher heating power to areas with greater ink thickness. These methods improve the uniformity of drying to some extent.

[0004] However, existing methods do not consider the differences in infrared radiation absorption characteristics among different colored inks and different solvent systems. In actual production, different pigments and solvents have their own infrared absorption peak wavelengths. Using infrared heating with a fixed wavelength will lead to low absorption efficiency for some components and overheating for others. At the same time, the solvent continuously evaporates during the ink drying process, and the absorption characteristics of the remaining solvent will also change dynamically. Existing methods cannot adjust the radiation wavelength in real time according to the residual solvent ratio. Furthermore, the temperature feedback control of existing methods only relies on the deviation between the current temperature value and the set value, ignoring the dynamic coupling relationship between the solvent evaporation heat absorption rate and the heating power. When the evaporation rate changes abruptly, temperature overshoot or response lag can easily occur, affecting the consistency of the final drying quality. Summary of the Invention

[0005] The purpose of this invention is to provide a method for controlling ink drying in glass production, thereby solving the following technical problems:

[0006] Existing methods do not take into account the differences in infrared absorption of different solvent systems and the dynamic changes in absorption characteristics during solvent evaporation. They cannot adjust the radiation wavelength in real time, and the temperature feedback ignores the dynamic coupling between evaporation endothermics and heating, resulting in poor consistency in drying quality.

[0007] The objective of this invention can be achieved through the following technical solutions: A method for controlling ink drying in glass production includes the following steps: S1. Obtain ink thickness distribution data and ink formulation component data. The thickness distribution data includes the thickness value at each location point, and the formulation component data includes the solvent type and initial mass fraction. S2. Read the volatilization activation energy and infrared absorption peak wavelength of each solvent from the solvent characteristic data, and determine the solvent residue ratio at each current location. S3. Divide the glass plate surface into multiple drying zones, and use infrared heating for each drying zone. The infrared heating method uses infrared radiation with adjustable wavelength. S4. For each drying zone, determine the corresponding infrared absorption peak wavelength according to the type of solvent in that drying zone, and correct the wavelength according to the solvent residue ratio, and use the corrected wavelength for infrared radiation heating. S5. Calculate the required evaporation heat for each drying zone based on the ink thickness and solvent residue ratio, and determine the heating intensity of infrared heating based on the evaporation heat. S6. Obtain the temperature value and volatile gas concentration change rate on the back of each drying zone, process the temperature value and change rate through fuzzy rules to obtain the heating intensity adjustment coefficient, and adjust the heating intensity of infrared heating according to the adjustment coefficient.

[0008] As a further aspect of the present invention: the specific process for obtaining ink thickness distribution data in step S1 is as follows: The glass plate surface is traversed along a preset path using laser thickness measurement. Multiple echo times are continuously acquired at each sampling point, and the thickness value is calculated by taking the median value of the echo times. The sampling point distribution density in the edge area is higher than that in the center area. The ink formula component data is read from the batching record of the printing process. The batching record contains the actual solvent addition amount and resin addition amount of each batch of ink. The actual addition amount is divided by the theoretical addition amount to obtain the correction coefficient. The correction coefficient is multiplied by the initial mass fraction in the standard formula to obtain the initial mass fraction of the current batch.

[0009] As a further aspect of the present invention: in step S2, the specific process for determining the solvent residue ratio at each current location is as follows: The activation temperature is obtained by dividing the evaporation activation energy by the gas constant. The activation temperature is then divided by the thermodynamic temperature of the current drying environment to obtain the dimensionless activation parameter. Multiple solvent concentration values ​​are obtained from different height positions in the drying environment. These multiple concentration values ​​are fitted along the height direction to form an exponential decay curve. The decay coefficient of the exponential decay curve is used as the solvent evaporation rate constant. The evaporation rate constant is multiplied by the dimensionless activation parameter to obtain the corrected evaporation rate. The initial mass fraction is subtracted from the product of the corrected evaporation rate and the drying time, and then divided by the initial mass fraction. The result of this division is used as the solvent residue ratio.

[0010] As a further aspect of the present invention: in step S3, the specific process of dividing the glass plate surface into multiple drying zones is as follows: Coordinate data of each location point is extracted from the ink thickness distribution data. The discrete location points are connected into a grid using a region partitioning algorithm. For regions where the thickness gradient change exceeds a preset gradient threshold, the cells in the grid are subdivided into smaller cells. For regions where the thickness gradient change does not exceed the preset gradient threshold, adjacent cells are merged into a polygonal region. The coordinates of the center point of each drying region are determined by the arithmetic mean of the coordinates of all location points in that region. The number of drying regions is equal to the number of location points.

[0011] As a further aspect of the present invention: in step S4, the specific process of correcting the wavelength based on the solvent residue ratio is as follows: The redshift coefficient and blueshift threshold of the solvent absorption peak are read from the solvent characteristic data. The redshift coefficient is the rate at which the absorption peak wavelength shifts to longer wavelengths when the solvent residual ratio decreases, and the blueshift threshold is the critical residual ratio at which the absorption peak shifts to shorter wavelengths when the solvent residual ratio is lower than the blueshift threshold. The current solvent residual ratio is compared with the blueshift threshold. If the current solvent residual ratio is higher than the blueshift threshold, the reference wavelength is added to the redshift coefficient multiplied by the amount of the decrease in residual ratio to obtain the corrected wavelength. If the current solvent residual ratio is lower than the blueshift threshold, the reference wavelength is subtracted from the amount of the blueshift coefficient multiplied by the blueshift threshold to obtain the corrected wavelength.

[0012] As a further aspect of the present invention: In step S5, the specific process of calculating the required evaporation heat for each drying zone based on the ink thickness and solvent residue ratio is as follows: The resin softening point temperature and solvent-resin interaction parameters are read from the ink formulation component data. The solvent residual ratio is multiplied by the interaction parameters to obtain the activity coefficient. The activity coefficient is multiplied by the saturated vapor pressure of the pure solvent to obtain the actual vapor pressure of the solvent in the ink film. The actual vapor pressure is divided by the ambient partial pressure to obtain the evaporation driving force ratio. The evaporation driving force ratio is multiplied by the latent heat of solvent evaporation and then by the solvent mass per unit area to obtain the dynamic evaporation heat. The dynamic evaporation heat is added to the sensible heat of the resin as it rises from the current temperature to the softening point temperature. The result of the addition is taken as the required evaporation heat.

[0013] As a further aspect of the present invention: in step S5, the specific process of determining the heating intensity of infrared heating from the heat of evaporation is as follows: The radiation efficiency of infrared radiation under the corrected wavelength condition is obtained. The target radiation energy is obtained by dividing the heat of evaporation by the radiation efficiency. The target heating intensity is obtained by dividing the target radiation energy by the drying time.

[0014] As a further aspect of the present invention: In step S6, the specific process of obtaining the heating intensity adjustment coefficient by processing the temperature value and the rate of change through fuzzy rules is as follows: The fuzzy rules are established based on the volatile endothermic characteristics of the solvent. The temperature deviation is fuzzy graded using a Gaussian membership function, and the concentration change rate is fuzzy graded using a trapezoidal membership function. The temperature deviation and concentration change rate are respectively input into the membership function to obtain the membership degree of each fuzzy grade. The centroid method is used to defuzzify and obtain the heating intensity adjustment coefficient. The current heating intensity is multiplied by the heating intensity adjustment coefficient to obtain the adjusted heating intensity.

[0015] The beneficial effects of this invention are: This invention achieves the following beneficial effects by introducing solvent type and solvent residue ratio as core control parameters. Addressing the problem that existing methods ignore the differences in infrared absorption of different solvents and the dynamic changes in absorption characteristics during evaporation, this invention queries the corresponding infrared absorption peak wavelength based on the solvent type and corrects this wavelength based on the real-time calculated solvent residue ratio, ensuring that the infrared radiation wavelength matches the absorption peak of the remaining solvent in the current ink film. Addressing the problem that existing temperature feedback ignores the dynamic coupling between evaporation heat absorption and heating power, this invention calculates the required evaporation heat based on the ink thickness and solvent residue ratio, and uses this evaporation heat as the basis for determining the infrared heating intensity, ensuring that the initial heating intensity adapts to the current heat demand of the ink film; simultaneously, it acquires the temperature value and the rate of change of volatile gas concentration, and obtains the heating intensity adjustment coefficient through fuzzy rule processing, performing closed-loop correction of the heating intensity. The above mechanism realizes a complete process control chain from solvent residue ratio detection and wavelength adaptive adjustment to dynamic correction of heating intensity, improving the consistency of drying quality. Attached Figure Description

[0016] The invention will now be further described with reference to the accompanying drawings.

[0017] Figure 1 This is a flowchart illustrating the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Please see Figure 1 As shown, the present invention is a method for controlling ink drying in glass production, comprising the following steps: S1. Obtain ink thickness distribution data and ink formulation component data. The thickness distribution data includes the thickness value at each location point, and the formulation component data includes the solvent type and initial mass fraction. S2. Read the volatilization activation energy and infrared absorption peak wavelength of each solvent from the solvent characteristic data, and determine the solvent residue ratio at each current location. S3. Divide the glass plate surface into multiple drying zones, and use infrared heating for each drying zone. The infrared heating method uses infrared radiation with adjustable wavelength. S4. For each drying zone, determine the corresponding infrared absorption peak wavelength according to the type of solvent in that drying zone, and correct the wavelength according to the solvent residue ratio, and use the corrected wavelength for infrared radiation heating. S5. Calculate the required evaporation heat for each drying zone based on the ink thickness and solvent residue ratio, and determine the heating intensity of infrared heating based on the evaporation heat. S6. Obtain the temperature value and volatile gas concentration change rate on the back of each drying zone, process the temperature value and change rate through fuzzy rules to obtain the heating intensity adjustment coefficient, and adjust the heating intensity of infrared heating according to the adjustment coefficient.

[0020] In a preferred embodiment of the present invention, the specific process of S1 is as follows: The laser thickness gauge is fixedly mounted below a horizontally moving support, which is mounted on a linear guide rail at the entrance of the drying chamber. The horizontally moving support is driven by a servo motor, which receives a pulse sequence from the programmable controller (PLC). The pulse frequency determines the moving speed. The speed variation curve is stored in the PLC's read-only memory and consists of three segments: the first segment corresponds to the edge region of the glass plate, with a moving speed of 50 mm per second; the second segment corresponds to the center region of the glass plate, with a moving speed of 100 mm per second; and the third segment again corresponds to the opposite edge region, with the moving speed returning to 50 mm per second. The segmentation of the speed variation curve is based on the division of the glass plate's edge and center regions. The edge region refers to the area within 50 mm of the glass plate boundary, and the center region refers to the area beyond 50 mm of the glass plate boundary.

[0021] The lateral moving support begins its movement from one edge of the glass plate, sequentially passing through the first, second, and third speed segments before stopping at the other edge. At each sampling point, the laser thickness gauge emits a laser pulse perpendicularly towards the glass plate surface. The laser pulse is reflected upon encountering the interface between the ink layer and the glass plate, and the reflected light is received by a photodetector inside the laser thickness gauge. The time-to-digital converter inside the laser thickness gauge records the time difference between the emission and reception times. To eliminate random errors, the laser thickness gauge emits three laser pulses consecutively at the same sampling point, obtaining three time differences. The median of these three time differences is taken as the time reference for that sampling point. The unit of the time reference value is nanoseconds; for example, the time reference for a sampling point on a glass plate of uniform thickness is 15 nanoseconds. Multiplying the time reference by the speed of light constant 299,792,458 meters per second and then dividing by 2 yields the thickness value for that sampling point, in micrometers. The thickness values ​​of all sampling points are arranged in the sampling order to form ink thickness distribution data.

[0022] Ink formulation component data is read from the real-time batching system of the printing process. The real-time batching system consists of multiple batching tanks and mass flow meters, each tank corresponding to one solvent or resin. During the production of each batch of ink, the mass flow meters record the actual mass values ​​of solvent and resin flowing into the mixing tank. The theoretical addition amounts are given in the formulation design document; for example, in a standard ink formulation, the theoretical addition amount of solvent is 1000 grams, and the theoretical addition amount of resin is 500 grams. The batching system records an actual addition amount of 980 grams of solvent and 510 grams of resin. Dividing the actual addition amount by the theoretical addition amount yields a correction factor: the solvent correction factor is 980 divided by 1000, which equals 0.98, and the resin correction factor is 510 divided by 500, which equals 1.02. The initial mass fractions in the standard formulation are stored in the formulation database; the standard formulation specifies an initial solvent mass fraction of 0.667 and an initial resin mass fraction of 0.333. Multiply the correction factor by the initial mass fraction in the standard formulation to obtain the initial mass fraction of the current batch. The initial mass fraction of the solvent is 0.98 multiplied by 0.667, which equals 0.653, and the initial mass fraction of the resin is 1.02 multiplied by 0.333, which equals 0.340. The sum of the two initial mass fractions is 0.993, and the remainder is the initial mass fraction of other additives. The initial mass fraction of the solvent is stored as the initial mass fraction of the solvent for the current batch in a temporary variable for subsequent calculations.

[0023] In another preferred embodiment of the present invention, the specific process of S2 is as follows: The solvent property database is stored on the local hard drive. Each record in the database table includes a solvent name field, a volatile activation energy field, an infrared absorption peak wavelength field, and a first and second coefficient field for the vapor pressure-temperature relationship. The volatile activation energy is measured in joules per mole; for example, the volatile activation energy of ethyl acetate is 35,000 joules per mole. The gas constant is 8.314 joules per mole per Kelvin. Dividing the volatile activation energy by the gas constant yields the activation temperature; 35,000 divided by 8.314 is approximately equal to 4210 Kelvin. The current thermodynamic temperature is read from the temperature sensor inside the drying chamber; for example, if the temperature inside the drying chamber is 353 Kelvin, it is 80 degrees Celsius. Dividing the activation temperature by the current thermodynamic temperature yields the dimensionless activation parameter; 4210 divided by 353 is approximately equal to 11.92.

[0024] Three solvent concentration sensors were installed at different vertical heights above a glass plate. The first sensor was installed 20 mm above the glass plate, the second 50 mm, and the third 100 mm. All three sensors simultaneously collected the current solvent gas concentration value, expressed in milligrams per cubic meter (mg / m³). For example, the three concentration values ​​were 120, 80, and 40 mg / m³. An exponential decay curve was fitted to the three concentration values ​​along the vertical direction. The fitting method was to take the natural logarithm of the concentration value and perform a linear regression with the height value. The logarithm for a height of 20 mm was 4.787, for 50 mm it was 4.382, and for 100 mm it was 3.689. The linear regression yielded a slope of -0.0123. The absolute value of this slope, 0.0123, was used as the decay coefficient of the exponential decay curve, which is the solvent evaporation rate constant under the current conditions, expressed in millimeters per second (m).

[0025] Multiplying the evaporation rate constant by the dimensionless activation parameter yields the corrected evaporation rate: 0.0123 multiplied by 11.92 approximately equals 0.1466, in millimeters per second. In practical applications, the evaporation rate constant needs to be converted from per millimeter to per second and multiplied by the conveyor belt speed. Assuming a conveyor belt speed of 10 millimeters per second, the corrected evaporation rate is 1.466 per second. The drying time is calculated from the moment the glass plate enters the drying chamber; for example, if the current time is 30 seconds. Multiplying the corrected evaporation rate by the drying time yields the evaporated percentage: 1.466 multiplied by 30 equals 43.98. Subtracting the evaporated percentage from the initial mass fraction of 0.653 and dividing by 100 gives the current remaining mass fraction: 0.653 minus 0.4398 equals 0.2132. Dividing the current remaining mass fraction by the initial mass fraction of 0.653 gives the solvent residue percentage: 0.2132 divided by 0.653 approximately equals 0.326. The solvent residue percentage ranges from 0 to 1. 0.326 indicates that 32.6% of the solvent has not yet evaporated. This value is taken as the solvent residue percentage at the current detection point.

[0026] In another preferred embodiment of the present invention, the specific process of S3 is as follows: The lateral and conveying direction coordinates of each sampling point are extracted from the variable-speed sampling records of the laser thickness gauge. The lateral coordinate represents the distance of the sampling point from the left edge of the glass plate, and the conveying direction coordinate represents the distance of the sampling point from the front edge of the glass plate. For example, the lateral coordinate of the first sampling point is 10 mm, and the conveying direction coordinate is 15 mm. The total number of sampling points is the number of lateral samplings multiplied by the number of longitudinal samplings. Assuming that the lateral moving support samples 20 points per stroke and there are 15 strokes in the longitudinal direction, the total number of sampling points is 300. These 300 discrete sampling points are used as vertices and connected into a triangular mesh using the Delaunay triangulation algorithm. The specific execution method of Delaunay triangulation is as follows: first, find the point with the smallest x-coordinate among all sampling points as the starting point, connect this point with the two nearest points to form an initial triangle, and then insert points one by one. Each time a new point is inserted, delete all triangles whose circumcircles contain the new point, and connect the new point with the points on the boundary of the new triangle to form a new triangle. After the triangulation is completed, approximately 550 triangles are obtained, each triangle consisting of three vertices and three sides.

[0027] For each triangle, read the thickness values ​​of the sampling points corresponding to the three vertices and calculate the variance of the thickness values. Adjacent triangles share a common edge. Calculate the thickness variance of the two triangles on either side of the shared edge. If the two variances are in opposite directions (i.e., one triangle's thickness is greater than the adjacent triangle's thickness and the other triangle's thickness is less than the adjacent triangle's thickness), then the temperature gradient directions on either side of the common edge of the two triangles are determined to be opposite. The temperature gradient direction is obtained by calculating the slope of the thickness values ​​within the triangle. When the temperature gradient directions on either side of the common edge of two triangles are opposite, merging these two triangles is prohibited. When the temperature gradient directions are the same, merge the two triangles into a single convex polygon control unit. During merging, delete the common edge and connect the outer edges of the two triangles to form a new polygon. Repeat the merging operation until there are no more adjacent triangle pairs that can be merged. Finally, several convex polygon control units are obtained, the number of control units equal to the number of sampling points, i.e., 300.

[0028] The center point coordinates of each control unit are determined by the arithmetic mean of the coordinates of all sampling points within that control unit. For example, if a control unit contains three sampling points with coordinates 10, 15, 12, 18, and 14, 16, then the x-coordinate of the center point is 10 + 12 + 14 divided by 3, which equals 12, and the y-coordinate is 15 + 18 + 16 divided by 3, which equals 16.33. The center point coordinates are used for positioning the infrared heating lamp in subsequent steps. Each control unit corresponds to one infrared heating lamp, and the installation position of the infrared heating lamp is aligned with the center point coordinates of that control unit. The boundary shape of the control unit determines the heating area coverage of the infrared heating lamp; the heating area is a convex polygon, and the heating lamp only radiates heat to the glass surface area falling within this polygon.

[0029] In another preferred embodiment of the present invention, the specific process of S4 is as follows: The redshift coefficient and blueshift threshold for the current solvent are read from the solvent property database. The redshift coefficient is a floating-point number representing the increase in the infrared absorption peak wavelength for every percentage point decrease in the solvent residual ratio, expressed in nanometers per percentage. For example, the redshift coefficient for ethyl acetate is 0.8 nanometers per percentage, meaning that the absorption peak wavelength increases by 0.8 nanometers when the residual ratio decreases from 100 to 99. The blueshift threshold is a percentage value indicating that when the solvent residual ratio is below this threshold, the interaction between solvent molecules weakens, and the absorption peak begins to shift towards shorter wavelengths. For example, the blueshift threshold is set to 30. The solvent residual ratio of the current control unit is read from the calculation results of step S2, assuming the current residual ratio is 45. The residual ratio of 45 is compared with the blueshift threshold of 30. Since 45 is greater than 30, it is determined that the residual ratio is higher than the blueshift threshold.

[0030] Calculate the decrease in residual ratio. The decrease in residual ratio equals the initial residual ratio minus the current residual ratio. The initial residual ratio is assumed to be 100, or it can be read from the ink formulation component data at the start of drying as the initial value. Assuming the initial residual ratio is 100, the decrease is 100 minus 45, which equals 55. Multiply the red shift coefficient of 0.8 nm per percentage by the decrease of 55 to obtain the wavelength shift of 44 nm. Read the reference wavelength from the solvent properties database. The reference wavelength is the peak infrared absorption wavelength of the pure solvent at a residual ratio of 100. For example, the reference wavelength for ethyl acetate is 5.7 μm, or 5700 nm. Add the wavelength shift of 44 nm to the reference wavelength of 5700 nm to obtain the corrected wavelength of 5744 nm. If the residual ratio is lower than the blue shift threshold, for example, a residual ratio of 25, the decrease is still calculated as 75 based on the initial residual ratio of 100. However, in this case, it is necessary to calculate the excess of the blue shift threshold, which equals the blue shift threshold of 30 minus the current residual ratio of 25, which equals 5. The blue shift coefficient is read from the solvent property database. The blue shift coefficient represents the amount of reduction in absorption peak wavelength for each percentage decrease after the residual ratio is below the blue shift threshold. Assuming the blue shift coefficient is 1.2 nanometers per percentage, the excess amount of 5 multiplied by 1.2 gives an offset of 6 nanometers. The corrected wavelength is equal to the reference wavelength of 5700 nanometers minus 6 nanometers, which gives 5694 nanometers.

[0031] The filter is a liquid crystal tunable filter, which consists of a liquid crystal layer between two transparent electrodes. Applying different voltages changes the alignment of the liquid crystal molecules, altering the birefringence of the liquid crystal layer and thus changing the center wavelength of the transmitted light. A one-to-one correspondence exists between the driving voltage and the transmission wavelength of the liquid crystal tunable filter; this curve is provided by the filter manufacturer and stored in the non-volatile memory of the programmable controller. Using a corrected wavelength of 5744 nm as the target wavelength, the corresponding driving voltage value is found on the voltage-wavelength relationship curve, for example, a driving voltage of 7.2 volts. The programmable controller outputs this voltage value to the input of the voltage-controlled oscillator (VCO) of the liquid crystal tunable filter via a digital-to-analog converter (DAC). The VCO generates an AC signal of the corresponding frequency based on the voltage value and applies it to the liquid crystal layer, locking the peak transmission wavelength of the filter at 5744 nm. This corrected wavelength matches the actual absorption peak at the current solvent residue level. Infrared radiation emitted by the infrared heating lamp, after passing through the filter, only allows narrowband light near the peak wavelength to pass through; this narrowband light is efficiently absorbed by the remaining solvent in the ink film.

[0032] In another preferred embodiment of the present invention, the specific process of calculating the required evaporation heat for each drying zone in step S5 based on the ink thickness and solvent residue ratio is as follows: The resin softening point temperature and solvent-resin interaction parameters are retrieved from the ink formulation component data. The resin softening point temperature refers to the temperature at which the resin begins to soften and flow; for example, the softening point temperature of a certain acrylic resin is 95 degrees Celsius. The interaction parameter is a dimensionless floating-point number representing the strength of the intermolecular forces between the solvent and resin, typically ranging from 0 to 2; for example, the interaction parameter between ethyl acetate and acrylic resin is 1.2. Multiplying the current residual solvent ratio of 0.45 by the interaction parameter 1.2 yields an activity coefficient of 0.54. The saturated vapor pressure of the pure solvent is retrieved from the solvent property database. Since saturated vapor pressure varies with temperature, the saturated vapor pressure of ethyl acetate is found to be 45 kPa based on the current thermodynamic temperature of 353 Kelvin in the drying chamber. Multiplying the activity coefficient 0.54 by the saturated vapor pressure of 45 kPa yields the actual vapor pressure of the solvent in the ink film, 24.3 kPa.

[0033] The ambient partial pressure inside the drying chamber is read from the ambient pressure sensor. The ambient partial pressure equals atmospheric pressure minus the partial pressure of the solvent gas. Assuming atmospheric pressure is 101.3 kPa and the solvent gas partial pressure is 5 kPa, the ambient partial pressure is 96.3 kPa. Dividing the actual vapor pressure (24.3 kPa) by the ambient partial pressure (96.3 kPa) yields a volatile driving force ratio of 0.252. The latent heat of volatilization of the solvent is read from the solvent property database; the latent heat of volatilization for ethyl acetate is 360 joules per gram. The ink thickness value of the current control unit is read from the ink thickness distribution data, assuming a thickness of 20 micrometers (0.002 cm). Multiplying the thickness value by the ink's solid density (assuming a solid density of 1.2 g / cm³) yields an ink mass of 0.0024 g / cm². Multiplying the solvent residue ratio (0.45) by the ink mass per unit area yields a solvent mass of 0.00108 g / cm². Multiplying the evaporation driving force ratio of 0.252 by the latent heat of evaporation of 360 joules per gram and then by the solvent mass per unit area of ​​0.00108 grams per square centimeter, we get the dynamic heat of evaporation of 0.098 joules per square centimeter.

[0034] The resin specific heat capacity is read from the ink formulation composition data, assumed to be 1.5 joules per gram per degree Celsius. The current temperature of the ink film is read from the temperature sensor, assumed to be 30 degrees Celsius. The resin softening point temperature is 95 degrees Celsius, and the temperature difference is 65 degrees Celsius. The resin mass per unit area equals the ink mass per unit area multiplied by the initial resin mass fraction of 0.34, i.e., 0.0024 multiplied by 0.34 equals 0.000816 grams per square centimeter. Multiplying the resin mass per unit area by the resin specific heat capacity and then by the temperature difference, 0.000816 multiplied by 1.5 and then by 65 yields the sensible heat of 0.0796 joules per square centimeter. Adding the dynamic evaporation heat of 0.098 to the sensible heat of 0.0796 yields the currently required evaporation heat of 0.1776 joules per square centimeter. This value is used as the evaporation heat of this control unit.

[0035] In another preferred embodiment of the present invention, the specific process of determining the heating intensity of infrared heating from the heat of evaporation in step S5 is as follows: The radiative efficiency curve is retrieved from the spectral radiation database of the infrared heating lamp. The radiative efficiency curve is a two-dimensional data table; the first column is the wavelength value in nanometers, and the second column is the radiative efficiency at the corresponding wavelength, i.e., the proportion of electrical energy converted into infrared radiation at that wavelength by the infrared heating lamp, with values ​​ranging from 0 to 1. Using the corrected transmission wavelength of 5744 nm obtained in step S4 as the lookup key, the closest wavelength point is found in the radiative efficiency curve, and the corresponding radiative efficiency value is read. Assuming the lookup result is 0.85, it means that at a wavelength of 5744 nm, the electrothermal conversion efficiency is 85%.

[0036] Dividing the evaporation heat of 0.1776 joules per square centimeter calculated in step S5 by the radiation efficiency of 0.85 yields an energy requirement of 0.209 joules per square centimeter. The drying time is read from the process parameter configuration file. The drying time refers to the time the glass plate remains under the infrared heating lamp, which is determined by the conveyor belt speed and the length of the heating zone. Assuming a conveyor belt speed of 10 mm / s and a heating zone length of 500 mm, the drying time is 50 seconds. Dividing the energy requirement of 0.209 joules per square centimeter by the drying time of 50 seconds yields a target heating power density of 0.00418 watts per square centimeter. Multiplying this power density by the area of ​​a control unit, calculated from the area of ​​the polygon in step S3, assuming an area of ​​25 square centimeters, yields a target heating power value of 0.1045 watts.

[0037] The rated power of the infrared heating lamp is read from the equipment parameter table; for example, the rated power of a certain infrared heating lamp is 200 watts. The output power is adjusted using pulse width modulation (PWM). The duty cycle is equal to the target heating power divided by the rated power. 0.1045 divided by 200 equals 0.0005225, or 0.05225. The duty cycle is multiplied by the value of the pulse width modulator's period register to obtain the value of the comparison register. If the period register is set to 10000, then the comparison register is set to 10000 multiplied by 0.0005225, which equals 5.225, rounded down to the nearest integer. This value is written into the pulse width modulator's comparison register. Within each cycle, the pulse width modulator outputs a high level when the counter value is less than the comparison register value; otherwise, it outputs a low level. During the high-level period, the infrared heating lamp operates at full power; during the low-level period, it does not operate, thus achieving an average output power equal to the target heating power value.

[0038] In another preferred embodiment of the present invention, the specific process of S6 is as follows: The fuzzy rule table is stored in the read-only memory of the programmable controller. This table is a two-dimensional matrix. The rows of the matrix correspond to five fuzzy levels of temperature deviation, and the columns correspond to three fuzzy levels of concentration change rate. The five fuzzy levels of temperature deviation are named negative large, negative small, zero, positive small, and positive large, respectively. The three fuzzy levels of concentration change rate are named slow, normal, and fast, respectively. Each matrix cell stores a power adjustment coefficient value, which ranges from 0.5 to 1.5.

[0039] The fuzzy rule table is generated through offline genetic algorithm pre-training. The training process of the genetic algorithm is completed on a regular computer, and the results are burned into the read-only memory of the programmable controller after training. During training, a large number of data samples from the drying process are collected. Each sample contains three input variables and one output variable. The three input variables are temperature deviation, concentration change rate, and actual optimal power adjustment coefficient. The actual optimal power adjustment coefficient is obtained through multiple trials. In each trial, other conditions are kept constant, and the power adjustment coefficient is adjusted to achieve the best drying quality. The chromosome encoding of the genetic algorithm is a binary string of length 17. The first 5 segments each have 3 bits corresponding to the power adjustment coefficients for the five levels of temperature deviation, the middle 3 segments each have 2 bits corresponding to the power adjustment coefficients for the three levels of concentration change rate, and the last 9 segments each have 1 bit corresponding to the center value and width parameter of the membership function. 200 chromosomes are randomly generated during initialization. The fitness function is equal to the reciprocal of the mean square error between the model output power adjustment coefficient and the actual optimal power adjustment coefficient. In each generation, the 40 chromosomes with the highest fitness are selected for crossover and mutation operations, with a crossover probability of 0.8 and a mutation probability of 0.01. After 500 iterations, the chromosome with the highest fitness is selected for decoding to obtain the coefficient values ​​in the fuzzy rule table, as well as the center and width values ​​of the membership function.

[0040] The temperature deviation is calculated as follows: The temperature value on the back of the current control unit is read from the thermocouple, for example, a measured temperature of 82 degrees Celsius. The optimal drying temperature for the solvent is read from the process parameter configuration file; for example, the optimal drying temperature for ethyl acetate is 80 degrees Celsius. The temperature deviation equals the measured temperature minus the optimal temperature, which is 2 degrees Celsius. The fuzzy classification of the temperature deviation uses a Gaussian membership function. The expression for the Gaussian membership function is the input value minus the square of the center value, divided by twice the square of the width value, and then the negative exponent is taken. Each fuzzy classification corresponds to a center value and a width value. For example, the center value for the negative large classification is -10 degrees Celsius, and the width value is 3; the center value for the negative small classification is -3 degrees Celsius, and the width value is 2; the center value for the zero classification is 0 degrees Celsius, and the width value is 1.5; the center value for the positive small classification is 3 degrees Celsius, and the width value is 2; and the center value for the positive large classification is 10 degrees Celsius, and the width value is 3. The temperature deviation of 2 degrees Celsius is substituted into each of the five Gaussian membership functions to calculate the membership degree for each classification. When substituting into the zero level, 2 minus 0 squared equals 4, 2 multiplied by 1.5 squared equals 4.5, the negative exponent is -0.889, and the membership degree is 0.411. When substituting into the positive small level, 2 minus 3 squared equals 1, 2 multiplied by 2 squared equals 8, the negative exponent is -0.125, and the membership degree is 0.882. The membership degrees of the negative large, negative small, and positive large levels are all close to 0. Therefore, temperature deviation mainly activates the zero level and the positive small level, with membership degrees of 0.411 and 0.882, respectively.

[0041] The concentration change rate is calculated as follows: Concentration values ​​are continuously read twice from three solvent concentration sensors, with the first sampling time being t1 and the second sampling time being t2. The difference between the two concentration values ​​is divided by the time interval to obtain the change rate. For example, the concentration change rate is 0.5 mg / m³ / s. The fuzzy classification of the concentration change rate uses a trapezoidal membership function, which is determined by four parameters: left shoulder start point, left shoulder end point, right shoulder start point, and right shoulder end point. The parameters for slow classification are set as follows: left shoulder start point -∞, left shoulder end point -0.1, right shoulder start point 0.1, right shoulder end point 0.3; normal classification parameters are set as follows: left shoulder start point 0.1, left shoulder end point 0.3, right shoulder start point 0.6, right shoulder end point 0.8; and fast classification parameters are set as follows: left shoulder start point 0.6, left shoulder end point 0.8, right shoulder start point 1.2, right shoulder end point positive infinity. The concentration change rate of 0.5 is substituted into the trapezoidal function for slow classification. The slow fractional fraction starts at 0.1 and ends at 0.3. Since 0.5 is greater than 0.3, the rate of concentration change is in the descending segment of the right shoulder, and the membership degree of the slow fractional fraction is 0. Substituting into the trapezoidal function for the normal fractional fraction, the left shoulder ends at 0.3, and the right shoulder starts at 0.6. 0.5 lies within the interval between the left shoulder end and the right shoulder start. The membership degree of the trapezoidal membership function is 1 in this interval, therefore the membership degree of the normal fractional fraction is 1. Substituting into the trapezoidal function for the fast fractional fraction, the left shoulder starts at 0.6, and 0.5 is less than 0.6, therefore the membership degree of the fast fractional fraction is 0. The rate of concentration change only activates the normal fractional fraction, and the corresponding membership degree is 1.

[0042] The centroid method is used to defuzzify and obtain the power adjustment coefficient. The centroid method is calculated as follows: multiply the power adjustment coefficient corresponding to each activated fuzzy level by the membership degree of that level, sum the products, and divide by the sum of the membership degrees. The power adjustment coefficient at the intersection of the zero temperature deviation level and the normal concentration change rate level is read from the fuzzy rule table; let's assume this value is 1.0. The power adjustment coefficient at the intersection of the small positive temperature deviation level and the normal concentration change rate level is read; let's assume this value is 0.9. Multiplying 1.0 by the membership degree of the zero temperature deviation level (0.411) gives 0.411, and multiplying 0.9 by the membership degree of the small positive temperature deviation level (0.882) gives 0.794; adding these two gives 1.205. The sum of the membership degrees is 0.411 plus 0.882, which equals 1.293. Dividing 1.205 by 1.293 gives the power adjustment coefficient 0.932. Multiplying the current output power by the power adjustment coefficient 0.932 yields the new output power. The programmable controller converts the new output power value into an analog voltage signal of 0 to 10 volts via a digital-to-analog converter. For example, an output power of 100 watts corresponds to a voltage of 5 volts, and a new output power of 93.2 watts corresponds to a voltage of 4.66 volts. This analog voltage signal is output to the input terminal of the linear power amplifier of the infrared heating lamp. The linear power amplifier adjusts the output current linearly according to the input voltage, thereby realizing the continuous adjustment of the output power of the infrared heating lamp.

[0043] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A method for controlling ink drying in glass production, characterized in that, Includes the following steps: S1. Obtain ink thickness distribution data and ink formulation component data. The thickness distribution data includes the thickness value at each location point, and the formulation component data includes the solvent type and initial mass fraction. S2. Read the volatilization activation energy and infrared absorption peak wavelength of each solvent from the solvent characteristic data, and determine the solvent residue ratio at each current location. S3. Divide the glass plate surface into multiple drying zones, and use infrared heating for each drying zone. The infrared heating method uses infrared radiation with adjustable wavelength. The specific process of dividing the glass plate surface into multiple drying zones is as follows: The coordinate data of each location point is extracted from the ink thickness distribution data. The discrete location points are connected into a grid using a region partitioning algorithm. For regions where the thickness gradient change exceeds a preset gradient threshold, the cells in the grid are subdivided into smaller cells. For regions where the thickness gradient change does not exceed the preset gradient threshold, adjacent cells are merged into a polygonal region. The center point coordinates of each drying region are determined by the arithmetic mean of the coordinates of all location points in that region. The number of drying regions is equal to the number of location points. S4. For each drying zone, determine the corresponding infrared absorption peak wavelength according to the type of solvent in that drying zone, and correct the wavelength according to the solvent residue ratio, and use the corrected wavelength for infrared radiation heating. S5. Calculate the required evaporation heat for each drying zone based on the ink thickness and solvent residue ratio, and determine the heating intensity of infrared heating based on the evaporation heat. The specific process for calculating the required evaporation heat for each drying zone based on ink thickness and solvent residue ratio is as follows: The resin softening point temperature and solvent-resin interaction parameters are read from the ink formulation component data. The solvent residual ratio is multiplied by the interaction parameters to obtain the activity coefficient. The activity coefficient is multiplied by the saturated vapor pressure of the pure solvent to obtain the actual vapor pressure of the solvent in the ink film. The actual vapor pressure is divided by the ambient partial pressure to obtain the volatilization driving force ratio. The volatilization driving force ratio is multiplied by the latent heat of solvent volatilization and then by the solvent mass per unit area to obtain the dynamic evaporation heat. The dynamic evaporation heat is added to the sensible heat of the resin from the current temperature to the softening point temperature. The result of the addition is taken as the required evaporation heat. S6. Obtain the temperature value and volatile gas concentration change rate on the back of each drying zone, process the temperature value and change rate through fuzzy rules to obtain the heating intensity adjustment coefficient, and adjust the heating intensity of infrared heating according to the adjustment coefficient.

2. The ink drying control method for glass production according to claim 1, characterized in that, In step S1, the specific process for obtaining ink thickness distribution data is as follows: The glass plate surface is traversed along a preset path by laser thickness measurement. Multiple echo times are continuously acquired at each sampling point, and the thickness value is calculated by taking the median value of the echo times. The sampling point distribution density in the edge area is higher than that in the center area. The ink formulation component data is read from the batching record of the printing process. The batching record contains the actual amount of solvent and resin added for each batch of ink. The actual amount added is divided by the theoretical amount added to obtain the correction factor. The correction factor is multiplied by the initial mass fraction in the standard formulation to obtain the initial mass fraction of the current batch.

3. The ink drying control method for glass production according to claim 1, characterized in that, In step S2, the specific process for determining the solvent residue ratio at each current location point is as follows: The activation temperature is obtained by dividing the evaporation activation energy by the gas constant. The activation temperature is then divided by the thermodynamic temperature of the current drying environment to obtain the dimensionless activation parameter. Multiple solvent concentration values ​​are obtained from different height positions in the drying environment. These multiple concentration values ​​are fitted along the height direction to form an exponential decay curve. The decay coefficient of the exponential decay curve is used as the solvent evaporation rate constant. The evaporation rate constant is multiplied by the dimensionless activation parameter to obtain the corrected evaporation rate. The initial mass fraction is subtracted from the product of the corrected evaporation rate and the drying time, and then divided by the initial mass fraction. The result of this division is used as the solvent residue ratio.

4. The ink drying control method for glass production according to claim 1, characterized in that, In step S4, the specific process of correcting the wavelength based on the solvent residue ratio is as follows: The redshift coefficient and blueshift threshold of the solvent absorption peak are read from the solvent characteristic data. The redshift coefficient is the rate at which the absorption peak wavelength shifts to longer wavelengths when the solvent residual ratio decreases, and the blueshift threshold is the critical residual ratio at which the absorption peak shifts to shorter wavelengths when the solvent residual ratio is lower than the blueshift threshold. The current solvent residual ratio is compared with the blueshift threshold. If the current solvent residual ratio is higher than the blueshift threshold, the reference wavelength is added to the redshift coefficient multiplied by the amount of the decrease in residual ratio to obtain the corrected wavelength. If the current solvent residual ratio is lower than the blueshift threshold, the reference wavelength is subtracted from the amount of the blueshift coefficient multiplied by the blueshift threshold to obtain the corrected wavelength.

5. The ink drying control method for glass production according to claim 1, characterized in that, In step S5, the specific process of determining the heating intensity of infrared heating from the heat of evaporation is as follows: The radiation efficiency of infrared radiation under the corrected wavelength condition is obtained. The target radiation energy is obtained by dividing the heat of evaporation by the radiation efficiency. The target heating intensity is obtained by dividing the target radiation energy by the drying time.

6. The ink drying control method for glass production according to claim 1, characterized in that, In step S6, the specific process of obtaining the heating intensity adjustment coefficient by processing the temperature value and the rate of change through fuzzy rules is as follows: The fuzzy rules are established based on the volatile endothermic characteristics of the solvent. The temperature deviation is fuzzy graded using a Gaussian membership function, and the concentration change rate is fuzzy graded using a trapezoidal membership function. The temperature deviation and concentration change rate are respectively input into the membership function to obtain the membership degree of each fuzzy grade. The centroid method is used to defuzzify and obtain the heating intensity adjustment coefficient. The current heating intensity is multiplied by the heating intensity adjustment coefficient to obtain the adjusted heating intensity.

Images