Laser cutting optimization control method for intelligent manufacturing of display

Abstract

The present application relates to the technical field of laser cutting, in particular to a laser cutting optimization control method for intelligent manufacturing of a display, comprising: predicting a next cutting position according to an interference factor of each cutting position; determining a target power of a supplementary position according to a power adjustment model; correcting the power adjustment model according to the interference factor of the supplementary position before a target position and determining a target power of the target position; obtaining a predicted power of the supplementary position before the target position according to the corrected power adjustment model, obtaining a predicted power of the target position according to the predicted power of the supplementary position before the target position and an initial power of the cutting position, and obtaining a power of the target position when cutting a defect line according to the target power and the predicted power of the target position. The present application can not only densely distribute a large number of defect lines on a cutting path, but also avoid obvious overheating, so that each glass component of the glass plate is easily separated in order.

Description

Technical Field

[0001] This invention relates to the field of laser cutting technology, and more specifically to a laser cutting optimization control method for intelligent manufacturing of displays. Background Technology

[0002] Glass sheets (such as glass substrates, light guide plates, etc.) are important raw materials for manufacturing displays. Glass sheets often need to be laser-cut to obtain glass components suitable for manufacturing displays. When laser-cutting glass sheets, in order to avoid the high energy concentration of the laser forming an overheated area on the glass surface and causing overheating effects on the brittle glass material, it is often necessary to cut a large number of defect lines at a certain interval on the glass sheet. The defect lines are perpendicular to the upper and lower surfaces of the glass substrate. Then, by applying an additional separation force to the glass sheet, the glass components are neatly separated along the defect lines.

[0003] However, in actual cutting, it is difficult to accurately control the spacing of defect lines. If the spacing is too small, an overheated area will still form on the glass surface, causing overheating to the brittle glass material. When additional separation force is applied to separate the glass components, on the one hand, the glass components may not separate neatly along the defect lines, and on the other hand, the light transmission effect of the glass material will be affected. When the spacing is too large, when additional separation force is applied to separate the glass components, due to the lack of densely distributed defect lines, the glass components may not separate neatly along the defect lines, or the glass components may not be able to separate directly or may be slightly damaged, affecting the light transmission effect of the glass material and the subsequent manufacturing process of the display. Summary of the Invention

[0004] To address the aforementioned problems, this invention provides a laser cutting optimization control method for intelligent manufacturing of displays.

[0005] The laser cutting optimization control method for intelligent manufacturing of displays of the present invention adopts the following technical solution:

[0006] One embodiment of the present invention provides a laser cutting optimization control method for intelligent manufacturing of displays, the method comprising the following steps:

[0007] A laser cutter cuts glass substrates for manufacturing displays along a cutting path at a preset initial power. A defect line is cut at each cutting position along the cutting path, and the defect line is perpendicular to the upper and lower surfaces of the glass substrate.

[0008] After a defect line is cut at each cutting position, the interference factor of each cutting position is obtained based on the interference of two adjacent defect lines on the optical signal. The next cutting position is predicted based on the interference factor, and a defect line is cut at the predicted next cutting position. This process is repeated until the first cutting process is completed along the cutting path.

[0009] A supplementary position is determined between every two adjacent cutting positions, and the target power of the first supplementary position is determined according to the power adjustment model; the power adjustment model of the supplementary position is obtained from the interference factors of the cutting positions adjacent to the supplementary position and the initial power; the target power determined by the power adjustment model is negatively correlated with the difference between the interference factors of the adjacent cutting positions.

[0010] The defect line at the first supplementary position is cut according to the target power, and the supplementary positions after the first supplementary position are recorded as the target positions. When cutting the defect line at the target positions, the power adjustment model is first corrected according to the interference factor of any supplementary position before the target position, and then the target power at the target position is determined using the corrected power adjustment model. The target power of the supplementary positions before the target position is obtained again according to the corrected power adjustment model and recorded as the predicted power. The predicted power of the target position is obtained according to the predicted power of the supplementary positions before the target position and the initial power of the cutting position. Finally, the power of the target position when cutting the defect line is obtained according to the target power of the target position and the predicted power of the target position.

[0011] Preferably, the specific steps for obtaining the interference factor at each cutting position based on the interference of two adjacent defect lines on the optical signal are as follows:

[0012] Before cutting a defect line at each cutting position, the light intensity a1 of the light signal passing through the glass plate perpendicularly to the side of the glass plate is obtained. After cutting a defect line at each cutting position, the light intensity a2 of the light signal passing through the glass plate perpendicularly to the side of the glass plate is obtained. |a1-a2| / (a1+a2) is recorded as the interference factor for each cutting position; || indicates the absolute value sign.

[0013] Preferably, the specific steps for predicting the next cutting position based on the interference factor are as follows:

[0014] The cutting position of the most recently cut defect line is recorded as the current cutting position. The distance between the current cutting position and the previous cutting position is recorded as D. Assuming that the distance between the next cutting position and the current cutting position is D1, let D1 = (1 + K1) × D.

[0015] The method for obtaining K1 is as follows: the interference factor at the current cutting position is denoted as w1. When w1 is less than the first preset threshold, K1 is linearly correlated with (w1-1); when w1 is greater than or equal to the first preset threshold th1, K1 = w1.

[0016] The next cutting position is determined by D1.

[0017] Preferably, the power adjustment model at the supplementary position is obtained from the interference factor of the cutting position adjacent to the supplementary position and the initial power, and includes the following specific steps:

[0018] For any supplementary position x, obtain the power factor w2(x) at the supplementary position x. The power factor w2(x) is positively correlated with the difference in interference factors between the two adjacent cutting positions of the supplementary position.

[0019] The power adjustment model P1(x) represents the target power at the supplementary position x, P1(x) = (K2(x) + y) × P0, where P0 represents the initial power; y is the offset coefficient, and the initial value of y is a preset value; where K2(x) represents the adjustment range of P0 at the supplementary position x, and K2(x) is negatively correlated with w2(x).

[0020] Preferably, the specific steps of correcting the power adjustment model based on the interference factor of any supplementary position before the target position are as follows:

[0021] The set of all supplementary positions before the target position is denoted as S, and any supplementary position in S is denoted as position x1, x1∈s; the power factor at position x1 is denoted as w2(x1); the laser power used to cut the defect line at position x1 is denoted as P(x1). The offset coefficient y in the power adjustment model is regarded as a variable, and w2(x1) is substituted into the power adjustment model to obtain the target power at position x1, denoted as P1(x1, y), which represents the target power obtained after correcting the power adjustment model at position x1 using the offset coefficient y;

[0022] The correction deviation magnitude of the power adjustment model is obtained based on the difference between P1(x1, y) and P(x1) and the interference factor at position x1. The correction deviation magnitude is positively correlated with the difference between P1(x1, y) and P(x1) and negatively correlated with the interference factor at position x1. The offset coefficient y when the correction deviation magnitude is minimized is obtained and denoted as the final offset coefficient y0. The offset coefficient y in the power adjustment model is replaced with y0 to obtain the corrected power adjustment model.

[0023] Preferably, the specific formula for obtaining the correction deviation magnitude of the power adjustment model based on the difference between P1(x1, y) and P(x1) and the interference factor at position x1 is as follows:

[0024]

[0025] Where f(x1) represents the interference factor at position x1, F represents the correction deviation magnitude of the power adjustment model, and || represents the sign of taking the absolute value.

[0026] Preferably, the specific steps for obtaining the predicted power of the target position based on the predicted power of the supplementary position before the target position and the initial power of the cutting position are as follows:

[0027] The position before the target position is recorded as the reference position, and the predicted power of the position before the target position is recorded as the predicted power of the reference position.

[0028] For the initial power of all cutting positions and the predicted power of all reference positions during the first cutting process, the cutting position and the reference position are called sampling points. Each initial power or predicted power is denoted as the first power. The coordinates of each sampling point and the first power constitute the cutting feature.

[0029] Along the cutting path, the sampling points are traversed sequentially until the cutting position before the target position is reached. The cutting features of all the sampling points are used to form a cutting feature sequence. The cutting power of the target position is predicted based on the cutting feature sequence and is denoted as the predicted power of the target position.

[0030] Preferably, the power at the target location when cutting the defect line is equal to the average of the target power and the predicted power at the target location.

[0031] Preferably, the specific steps for obtaining the power factor w2(x) are as follows:

[0032] Obtain the absolute value of the difference between the interference factors of two adjacent cutting positions of the supplement position x. The ratio of the absolute value to the distance between the two cutting positions is denoted as the initial power factor of the supplement position x. Linearly normalize the initial power factors of all supplement positions. The normalized initial power factor is denoted as power factor w2(x).

[0033] Preferably, the specific steps for obtaining the adjustment range of P0 at the supplementary position x are as follows:

[0034] When w2(x) is greater than or equal to the second preset threshold, K2(x) = 1 - w2(x); when w2 is less than the second preset threshold th2, K2(x) = 1 - y.

[0035] The beneficial effects of the technical solution of the present invention are:

[0036] This invention, after cutting a defect line at each cutting position, obtains the interference factor for each cutting position based on the interference of adjacent two defect lines on the optical signal. The next cutting position is predicted based on this interference factor, and a defect line is cut at the predicted next cutting position. This process continues until the first cutting process is completed along the cutting path. This process ensures that the laser cuts with a constant initial power and continuously predicts the next cutting position based on the overheating effect of the previous defect line cutting, thus avoiding significant overheating problems to a certain extent.

[0037] Furthermore, the present invention corrects the power adjustment model based on the interference factor of any supplementary position before the target position, and then uses the corrected power adjustment model to determine the target power at the target position, so that when cutting the defect line at the supplementary position with the target power, there is appropriate power, reducing the problem of significant overheating.

[0038] Furthermore, the target power at the target position is determined using the modified power adjustment model. The predicted power of the supplementary position before the target position is then re-acquired based on the modified power adjustment model. The predicted power at the target position is then obtained based on the predicted power of the supplementary position before the target position and the initial power at the cutting position. Finally, the power at the target position when cutting the defect line is obtained based on the target power and the predicted power at the target position. This process continuously modifies the power adjustment model to obtain the target power at the target position, re-acquires the predicted power at the reference position using the modified power adjustment model, and further combines the position-power distribution relationship between the cutting position and the reference position during the first cutting process to finally obtain the cutting power at the target position. This avoids the problem that the modified power adjustment cannot always approach or stabilize at a suitable cutting power, ultimately ensuring that a large number of defect lines are closely distributed along the cutting path while avoiding significant overheating. This makes it easy to neatly separate the various glass components of the glass sheet, preventing the cut glass components from affecting the manufacturing of the display. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 This is a flowchart illustrating the steps of a laser cutting optimization control method for intelligent manufacturing of displays, provided in one embodiment of the present invention. Detailed Implementation

[0041] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the laser cutting optimization control method for intelligent manufacturing of displays proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0042] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0043] The specific solution of the laser cutting optimization control method for intelligent manufacturing of displays provided by the present invention will be described in detail below with reference to the accompanying drawings.

[0044] Please see Figure 1 This illustrates a laser cutting optimization control method for intelligent manufacturing of displays provided by an embodiment of the present invention, the method comprising the following steps:

[0045] Step S001: The laser cutter cuts the glass sheet for manufacturing the display along the cutting path at a preset initial power, and a defect line is cut at each cutting position on the cutting path.

[0046] When manufacturing a monitor, it is often necessary to laser cut the glass sheets used to make the monitor (such as glass substrates, light guide plates, etc.) so that the cut glass sheets have the required size or shape.

[0047] The glass sheet used in this embodiment is rectangular in shape, which is easy to process and fix.

[0048] When a laser cutter's laser head cuts a glass sheet along a pre-set cutting path, the highly concentrated laser energy may create overheated areas on the glass surface, causing overheating effects on the brittle glass material. To solve this problem, during laser cutting, a defect line needs to be cut at regular intervals along the cutting path, with the defect line perpendicular to the upper and lower surfaces of the glass substrate. After cutting, a large number of defect lines are distributed along the cutting path. Then, additional separation forces are applied to separate the glass components. For example, a CO2 laser can be used to irradiate along the cutting path and generate thermal stress, causing the glass components to separate along the defect lines on the cutting path.

[0049] However, if the spacing of the defect lines on the cutting path is not set appropriately, for example, if the spacing is too small, the heat generated by the defect lines in local areas will accumulate, still causing overheating of the brittle glass material. When additional separation force is applied to separate the glass components later, the glass components may not separate neatly along the defect lines, affecting the glass material's light transmission. When the spacing is too large, the lack of densely distributed defect lines may prevent the glass components from separating neatly along the defect lines when additional separation force is applied later, or the glass components may not separate directly or may suffer minor damage (such as edge breaks or cracks), affecting the glass material's light transmission and the subsequent manufacturing process of the display.

[0050] In this embodiment, the laser head of the laser cutter cuts a defect line after moving a certain distance L0 along the cutting path. The position of the laser head after each distance L0 is recorded as the cutting position. The laser head emits an ultrashort pulse laser at a certain power. This ultrashort pulse laser enters perpendicularly from the upper surface of the glass plate at the cutting position and forms a defect line within the glass plate. The power used in this embodiment is just enough to ensure that the laser can penetrate the upper and lower surfaces of the glass plate. This power is recorded as the initial power. In other embodiments, the initial power can be set to other values.

[0051] In this embodiment, T0 = 10 micrometers. In other embodiments, T0 can be set to other values. This embodiment does not impose any specific limitations.

[0052] Step S002: After a defect line is cut at each cutting position, the interference factor of each cutting position is obtained based on the interference of two adjacent defect lines on the optical signal.

[0053] In this embodiment, a strip light source (e.g., a strip LED light source) is attached to two adjacent sides of the glass plate. The light beam generated by the strip light source is incident into the glass plate perpendicularly to the side of the glass plate (the direction of the light beam propagation is parallel to the upper and lower surfaces of the glass plate). At the same time, a strip light intensity sensor is attached to the other two adjacent sides of the glass plate. The light intensity sensor is used to detect the light intensity of the light beam emitted from each position on the strip light source after passing through the glass plate.

[0054] Before a defect line is cut at each cutting position, the sum of light intensities a1 obtained from all positions of the strip light intensity sensor is obtained. After a defect line is cut at each cutting position, the sum of light intensities a2 obtained from all positions of the strip light intensity sensor is obtained. |a1-a2| / (a1+a2) is recorded as the interference factor for each cutting position. || indicates the absolute value sign.

[0055] In other embodiments, a laser signal of a certain frequency (e.g., 10 kHz) is emitted at each position on the strip light source. The light signal received at each position on the strip light intensity sensor is used to obtain the received frequency through a waveform counter. Before a defect line is cut at each cutting position, the sum of the received frequencies at all positions of the strip light intensity sensor, b1, is obtained. After a defect line is cut at each cutting position, the sum of the received frequencies at all positions of the strip light intensity sensor, b2, is obtained. |b1-b2| / (b1+b2) is recorded as the interference factor for each cutting position.

[0056] The interference factor at each cutting position described above describes the obstruction of light signal (e.g., light intensity or frequency) transmission in the glass sheet before and after cutting the defect line. The larger the interference factor, the more severe the obstruction, indicating that the overheating effect caused by too small a spacing when cutting the defect line at each cutting position is more severe. When the interference factor is smaller (or equal to 0), it indicates that the overheating effect is less severe (or the overheating effect can be ignored), and the spacing is more appropriate or may be too large.

[0057] Step S003: Predict the next cutting position based on the interference factor, and cut a defect line at the predicted next cutting position; and so on, until the first cutting process is completed along the cutting path.

[0058] After cutting at least two defect lines, the cutting position of the most recently cut defect line is recorded as the current cutting position, and the distance between the current cutting position and the previous cutting position is recorded as D. Assuming that the distance between the next cutting position and the current cutting position is D1, let D1 = (1 + K1) × D.

[0059] The method for obtaining K1 is as follows: the interference factor at the current cutting position is denoted as w1. When w1 is less than the first preset threshold th1, K1 = 0.3 × (w1 - 1); when w1 is greater than or equal to the first preset threshold th1, K1 = w1.

[0060] This embodiment uses th1 = 0.1 as an example for description. In other embodiments, th1 can be set to other values. This embodiment does not impose specific limitations.

[0061] The above formula uses the interference factor of the current cutting position to predict the next cutting position. When w1 is larger, it indicates that the overheating effect after cutting the defect line at the current cutting position is more severe and the spacing is too small. Therefore, the spacing D1 of the next cutting position is larger. When w1 is smaller, it indicates that the overheating effect after cutting the defect line at the current cutting position is less severe. If w1 is less than th1, it means that the overheating effect can be ignored. In this case, D1 is smaller than D to avoid the spacing of the cutting positions being too large.

[0062] In other embodiments, the method for obtaining the distance D1 between the next cutting position and the current cutting position further includes:

[0063]

[0064] Where n is a preset integer. In this embodiment, n=5 is used as an example. In other embodiments, n can be set to other values.

[0065] In some of the above embodiments, nonlinearity is introduced to avoid underfitting of spacing D1. Compared with other embodiments, this embodiment uses a linear relationship to predict the next cutting position using an interference factor, which results in faster calculation speed and is suitable for fast cutting scenarios.

[0066] After obtaining D1, the laser head moves a distance D1 along the cutting path from the current cutting position to the next cutting position, cuts out the defect line at the cutting position, and obtains the interference factor at that position.

[0067] Following this method, continue in the same manner to complete the cutting process, which is recorded as the first cutting process.

[0068] Step S004: Determine a supplementary position between every two adjacent cutting positions, and determine the target power of the first supplementary position according to the power adjustment model; the power adjustment model of the supplementary position is obtained from the interference factor of the cutting position adjacent to the supplementary position and the initial power.

[0069] In the first cutting process described above, a constant power (i.e., initial power) is used to avoid severe overheating of the glass sheet by adjusting the spacing of the defect lines in real time. While cutting with a constant initial power avoids severe overheating to some extent, it also results in a less dense distribution of defect lines at all cutting locations (or, in other words, it sacrifices the density of defect line distribution to avoid severe overheating), affecting the subsequent neat separation of the glass components.

[0070] Furthermore, in this embodiment, a supplementary position is determined between every two adjacent cutting positions. The second cutting process in this embodiment involves cutting defect lines at all supplementary positions.

[0071] As an example, the supplementary position is the midpoint between two adjacent cutting positions.

[0072] Furthermore, a power adjustment model is obtained based on the interference factor of the cutting position adjacent to the supplementary position and the initial power, and then the target power of the first supplementary position is determined based on the power adjustment model.

[0073] As a preferred example, the method for determining the target power at the first replenishment position based on the power adjustment model is as follows:

[0074] For any supplementary position x, obtain the two adjacent cutting positions to the left and right of supplementary position x, and obtain the absolute value of the difference between the interference factors of these two cutting positions. The ratio of this absolute value to the distance between these two cutting positions is denoted as the initial power factor of supplementary position x. In this embodiment, the initial power factor of all supplementary positions is linearly normalized to eliminate the dimension of the initial power factor. The normalized initial power factor is denoted as power factor w2(x). The larger the power factor, the greater the overheating effect change that occurred after cutting in a small local area during the cutting process, indicating that the cutting power in this local area (i.e., at supplementary position x) needs to be adjusted more significantly.

[0075] The target power at position x is P1(x) = (K2(x) + y) × P0, where P0 represents the initial power described in step S001. y is the offset coefficient, with an initial value of 0.

[0076] Where K2(x) represents the adjustment magnitude of P0 at position x. K2(x) is obtained as follows: when w2(x) is greater than or equal to the second preset threshold th2, K2(x) = 1 - w2(x); when w2 is less than the second preset threshold th2, K2(x) = 1 - y.

[0077] The above P1(x) represents the power adjustment model. A larger power factor w2(x) at the supplementary position x indicates a greater likelihood of significant overheating in a localized area when cutting with the initial power P0. In this case, the target power adjustment at the supplementary position x should be smaller. Conversely, a smaller power factor w2(x) at the supplementary position x indicates that no significant overheating occurred in a localized area when cutting with the initial power P0. When w2(x) is less than th2, the overheating effect can be ignored, and the target power P1 = P0, allowing the laser head sufficient power to penetrate the glass substrate.

[0078] This embodiment uses th2 = 0.1 as an example for description. In other embodiments, th2 can be set to other values. This embodiment does not impose specific limitations.

[0079] The target power P1(x) obtained from the power factor w2(x) at position x represents the power adjustment model.

[0080] The target power at the first supplementary position of the power adjustment model is denoted as P1(1).

[0081] As another example, the method for determining the target power at the first replenishment position based on the power adjustment model is as follows:

[0082] Let the target power at position x be added.

[0083] In the other example above, the underfitting problem of P1(x) is avoided by introducing nonlinearity. Compared with the other example, the preferred example uses a power factor to build a power adjustment model by introducing a linear relationship, which is faster to calculate and suitable for scenarios with rapid cutting.

[0084] Furthermore, after obtaining the target power P1(1) at the first replenishment position, the defect line is cut at the first replenishment position with the target power P1(1).

[0085] Step S005: The next supplementary position after the first supplementary position is recorded as the target position; when cutting the defect line at the target position, first correct the power adjustment model according to the interference factor of any supplementary position before the target position, and then use the corrected power adjustment model to determine the target power at the target position.

[0086] The power adjustment model is based on the interference factor correction of the supplementary positions before the target position. Its purpose is to ensure that the cutting power is appropriate when cutting the defect line at each supplementary position (e.g., at the second supplementary position or other supplementary positions afterward), so as to avoid further overheating of the glass sheet when adding the defect line.

[0087] As an example, the power adjustment model, which corrects the interference factor based on the supplementary position before the target position, includes the following methods:

[0088] The set of all supplementary positions preceding the target position is denoted as S. Any supplementary position in S is denoted as position x1, i.e., x1∈S. The power factor at position x1 is denoted as w2(x1). The laser power P(x1) used to cut the defect line at position x1 is obtained. The offset coefficient y in the above power adjustment model is regarded as a variable (i.e., an unknown). Substituting w2(x1) into the power adjustment model, the target power at position x1 is obtained as P1(x1, y). The obtained P1(x1, y) is a function model of the offset coefficient y, representing the target power obtained after correcting the power adjustment model using the offset coefficient y.

[0089] Let F = ∑ x1∈S(1-f(x1))×|P(x1)-P1(x1,y)|, where F represents the correction deviation magnitude of the power adjustment model, and P1(x1,y) represents the target power obtained after the power adjustment model is corrected by using the offset coefficient y at position x1, and |P1(x1)-P1(x1,y)| represents the error between the target power obtained after correction and the target power when cutting the defect line.

[0090] f(x1) represents the interference factor at position x1 (obtained using the same method as step S002). The smaller f(x1), the smaller the overheating effect caused by cutting the defect line at position x1 with power P1(x1). In this case, we pay more attention to |P1(x1)-P1(x1,y)| so that the target power obtained after correction approaches P1(x1). The larger f(x1), the greater the overheating effect caused by cutting the defect line at position x1 with power P1(x1). In this case, we pay less attention to |P1(x1)-P1(x1,y)| so as to avoid the target power obtained after correction from approaching P1(x1).

[0091] Obtain the offset coefficient y when F is minimized, and denote it as the final offset coefficient y0. Replace the offset coefficient in the above power adjustment model with y0 to obtain the corrected power adjustment model.

[0092] Obtain the power factor at the target location, and then input the power factor at the target location into the modified power adjustment model to obtain the target power at the target location.

[0093] The target power obtained above is based on the laser cutting situation of the previous supplementary position, which makes the target position cut the defect line with the target power with high accuracy and avoids the existence of obvious overheating effect area due to inaccurate target power calculation.

[0094] Step S006: Re-obtain the predicted power of the supplementary position before the target position according to the corrected power adjustment model, obtain the predicted power of the target position according to the predicted power of the supplementary position before the target position and the initial power of the cutting position, and finally obtain the power of the target position when cutting the defect line according to the target power of the target position and the predicted power of the target position.

[0095] In some other embodiments of the present invention, a defect line is cut at a target position using a target power, and then the next supplementary position is recorded as the target position. The power adjustment model is then corrected again based on the interference factor of any supplementary position before the target position, and the target power at the target position is determined using the corrected power adjustment model. The defect line is then cut at the target position again using the target power, and so on, until the laser cutting process is completed after the defect line is cut at all supplementary positions.

[0096] In the other embodiments described above, by continuously refining the power adjustment model, the appropriate cutting power is determined for different replenishment positions based on the overheating effect of their local areas (i.e., the interference factor of the cutting positions adjacent to the replenishment position). This helps to avoid the recurrence of significant overheating areas while replenishing defect lines. However, as cutting progresses, the refined power adjustment model may never reach or stabilize at a suitable cutting power, resulting in suboptimal power adjustment of the laser head during the cutting process, making it impossible to further reduce the overheating effect.

[0097] The difference between this embodiment and other embodiments is that:

[0098] After obtaining the target power at the target position in step S005, the supplementary position before the target position is recorded as the reference position. The target power at the reference position is re-obtained based on the power factor of the reference position using the modified power adjustment model, and recorded as the predicted power of the reference position. The predicted power represents the cutting power that can significantly eliminate the overheating effect at the reference position (this cutting power may not be equal to the power when the reference position actually cuts the defect line).

[0099] The initial power at all cutting positions and the predicted power at all reference positions are obtained during the first cutting process. These cutting positions and reference positions are collectively referred to as sampling points. Each initial or predicted power is denoted as the first power. The coordinates of each sampling point and the first power constitute the cutting feature (in this embodiment, the vector formed by the coordinates and the first power is denoted as the cutting feature). The coordinates of a sampling point refer to the path distance from the sampling point to the cutting start point along the cutting path.

[0100] On the cutting path, according to the cutting direction of the reference point (for example, from a top-down view, when the cutting path is a straight line distributed from left to right, the cutting direction is from left to right), the sampling points are traversed sequentially until the cutting position before the target position is reached. The cutting features of all the sampling points are used to form a cutting feature sequence. The cutting power of the target position is predicted based on the cutting feature sequence and is denoted as the predicted power of the target position.

[0101] During the first cut, the spacing between the cutting positions is changed, and the initial power is used to cut at the cutting positions, which eliminates the overheating effect to a certain extent. In addition, the predicted power corresponding to the reference position can also eliminate the overheating effect to a certain extent. Therefore, the cutting feature sequence describes the position-power distribution that can eliminate or avoid the overheating effect to a certain extent. The predicted power of the target position obtained based on the cutting feature sequence can describe the cutting power required to avoid introducing a significant overheating effect at the target position.

[0102] The cutting power at the target position is predicted based on the cutting feature sequence. The prediction algorithm used is the extended Kalman filter algorithm. In other embodiments, the ARIMA algorithm or the LSTM neural network can also be used for prediction.

[0103] Furthermore, in this embodiment, the laser power required to cut the defect line at the target position is obtained based on the target power at the target position and the predicted power at the target position.

[0104] As an example, the laser power required to cut the defect line at the target location is the average of the target power at the target location and the predicted power at the target location.

[0105] Then, the next supplementary position is designated as the target position. The power adjustment model is corrected again based on the interference factor of any supplementary position before the target position. The target power and predicted power at the target position are then determined using the corrected power adjustment model. The cutting power required to cut the defect line at the target position is obtained again using the target power and predicted power. This process is repeated until the defect line is cut at all supplementary positions, completing the second cutting process, which uses different powers to cut the defect line at each supplementary position.

[0106] This process obtains the target power at the target location by continuously modifying the power adjustment model, and then uses the modified power adjustment model to re-obtain the predicted power at the reference location. Furthermore, it combines the position-power distribution relationship between the cutting location and the reference location during the first cutting process to ultimately obtain the cutting power at the target location. This avoids the problem of the modified power adjustment never reaching or stabilizing at a suitable cutting power, ensuring that a large number of defect lines are closely distributed along the cutting path while avoiding significant overheating. This also facilitates the neat separation of the various glass components from the glass sheet, preventing the cut glass components from affecting the manufacturing of the display.

[0107] The method to make it easy to neatly separate the various components of the glass sheet is to use a CO2 laser to irradiate along the cutting path. In this embodiment, the CO2 laser generates a 200-watt CO2 laser with a wavelength of 9 micrometers and a spot size of 8 millimeters. The defect line generates thermal stress under the irradiation of the CO2 laser, causing the glass sheet to separate along the defect line on the cutting path to obtain a glass component of the required size and shape. The glass component can then be processed by processes such as chamfering, edge grinding, and cleaning, and then used to manufacture a display.

[0108] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A laser cutting optimization control method for intelligent manufacturing of displays, characterized in that, The method includes the following steps: A laser cutter cuts glass substrates for manufacturing displays along a cutting path at a preset initial power. A defect line is cut at each cutting position along the cutting path, and the defect line is perpendicular to the upper and lower surfaces of the glass substrate. After a defect line is cut at each cutting position, the interference factor of each cutting position is obtained based on the interference of two adjacent defect lines on the optical signal. The next cutting position is predicted based on the interference factor, and a defect line is cut at the predicted next cutting position. This process is repeated until the first cutting process is completed along the cutting path. A supplementary position is determined between every two adjacent cutting positions, and the target power of the first supplementary position is determined according to the power adjustment model; the power adjustment model of the supplementary position is obtained from the interference factors of the cutting positions adjacent to the supplementary position and the initial power; the target power determined by the power adjustment model is negatively correlated with the difference between the interference factors of the adjacent cutting positions. The defect line at the first supplementary position is cut according to the target power, and the supplementary positions after the first supplementary position are recorded as the target positions. When cutting the defect line at the target positions, the power adjustment model is first corrected according to the interference factor of any supplementary position before the target position. Then, the target power of the target position is determined using the corrected power adjustment model. The target power of the supplementary positions before the target position is obtained again according to the corrected power adjustment model and recorded as the predicted power. The predicted power of the target position is obtained according to the predicted power of the supplementary positions before the target position and the initial power of the cutting position. Finally, the power of the target position when cutting the defect line is obtained according to the target power of the target position and the predicted power of the target position. The specific steps for obtaining the interference factor at each cutting position based on the interference of two adjacent defect lines on the optical signal are as follows: Before cutting a defect line at each cutting position, the light intensity a1 of the light signal passing through the glass plate perpendicularly to the side of the glass plate is obtained. After cutting a defect line at each cutting position, the light intensity a2 of the light signal passing through the glass plate perpendicularly to the side of the glass plate is obtained. |a1-a2| / (a1+a2) is recorded as the interference factor for each cutting position; || indicates the absolute value sign. The specific steps involved in predicting the next cutting position based on the interference factor are as follows: The cutting position of the most recently cut defect line is recorded as the current cutting position. The distance between the current cutting position and the previous cutting position is recorded as D. Assuming that the distance between the next cutting position and the current cutting position is D1, let D1 = (1 + K1) × D. The method for obtaining K1 is as follows: the interference factor at the current cutting position is denoted as w1. When w1 is less than the first preset threshold, K1 is linearly correlated with (w1-1); when w1 is greater than or equal to the first preset threshold th1, K1 = w1. The next cutting position is determined by D1; The power adjustment model at the supplementary position is obtained from the interference factor of the cutting position adjacent to the supplementary position and the initial power, and includes the following specific steps: For any supplementary position x, obtain the power factor w2(x) at the supplementary position x. The power factor w2(x) is positively correlated with the difference in interference factors between the two adjacent cutting positions of the supplementary position. The power adjustment model P1(x) represents the target power at the supplementary position x, P1(x) = (K2(x) + y) × P0, where P0 represents the initial power; y is the offset coefficient, and the initial value of y is a preset value; where K2(x) represents the adjustment range of P0 at the supplementary position x, and K2(x) is negatively correlated with w2(x).

2. The laser cutting optimization control method for intelligent manufacturing of displays according to claim 1, characterized in that, The specific steps of correcting the power adjustment model based on the interference factor of any supplementary position before the target position are as follows: The set of all supplementary positions before the target position is denoted as S, and any supplementary position in S is denoted as position x1, x1∈S; the power factor at position x1 is denoted as w2(x1); the laser power used to cut the defect line at position x1 is denoted as P(x1). The offset coefficient y in the power adjustment model is regarded as a variable, and w2(x1) is substituted into the power adjustment model to obtain the target power at position x1, denoted as P1(x1, y), which represents the target power obtained after correcting the power adjustment model at position x1 using the offset coefficient y; The correction deviation magnitude of the power adjustment model is obtained based on the difference between P1(x1, y) and P(x1) and the interference factor at position x1. The correction deviation magnitude is positively correlated with the difference between P1(x1, y) and P(x1) and negatively correlated with the interference factor at position x1. The offset coefficient y when the correction deviation magnitude is minimized is obtained and denoted as the final offset coefficient y0. The offset coefficient y in the power adjustment model is replaced with y0 to obtain the corrected power adjustment model.

3. The laser cutting optimization control method for intelligent manufacturing of displays according to claim 2, characterized in that, The specific formulas for obtaining the correction deviation of the power adjustment model based on the difference between P1(x1, y) and P(x1) and the interference factor at position x1 are as follows: Where f(x1) represents the interference factor at position x1, F represents the correction deviation magnitude of the power adjustment model, and || represents the sign of taking the absolute value.

4. The laser cutting optimization control method for intelligent manufacturing of displays according to claim 1, characterized in that, The specific steps for obtaining the predicted power of the target position based on the predicted power of the supplementary positions before the target position and the initial power of the cutting position are as follows: The position before the target position is recorded as the reference position, and the predicted power of the position before the target position is recorded as the predicted power of the reference position. For the initial power of all cutting positions and the predicted power of all reference positions during the first cutting process, the cutting position and the reference position are called sampling points. Each initial power or predicted power is denoted as the first power. The coordinates of each sampling point and the first power constitute the cutting feature. Along the cutting path, the sampling points are traversed sequentially until the cutting position before the target position is reached. The cutting features of all the sampling points are used to form a cutting feature sequence. The cutting power of the target position is predicted based on the cutting feature sequence and is denoted as the predicted power of the target position.

5. The laser cutting optimization control method for intelligent manufacturing of displays according to claim 1, characterized in that, The power at the target location when cutting the defect line is equal to the average of the target power and the predicted power at the target location.

6. The laser cutting optimization control method for intelligent manufacturing of displays according to claim 1, characterized in that, The specific steps for obtaining the power factor w2(x) are as follows: Obtain the absolute value of the difference between the interference factors of two adjacent cutting positions of the supplement position x. The ratio of the absolute value to the distance between the two cutting positions is denoted as the initial power factor of the supplement position x. Linearly normalize the initial power factors of all supplement positions. The normalized initial power factor is denoted as power factor w2(x).

7. The laser cutting optimization control method for intelligent manufacturing of displays according to claim 1, characterized in that, The specific steps for obtaining the adjustment range of P0 at the supplementary position x are as follows: When w2(x) is greater than or equal to the second preset threshold, K2(x) = 1 - w2(x); when w2 is less than the second preset threshold th2, K2(x) = 1 - y.

Images