Iris cutting control method and system for OLED flexible screen production and storage medium

Abstract

The application relates to the technical field of flexible screen processing, and discloses a special-shaped cutting control method and system for OLED flexible screen production and a storage medium. The method comprises the following steps: obtaining target corner curvature data of a flexible screen; performing stress simulation analysis based on the curvature data to obtain predicted stress distribution of the flexible screen; planning and generating a dynamic cutting track that is adapted to curvature change according to the predicted stress distribution; calculating and obtaining an adapted laser power distribution parameter based on the dynamic cutting track; integrating the laser power parameter and a cutting speed constraint to obtain a cutting execution scheme; executing the cutting scheme, extracting and real-time adjusting edge feature data in the cutting process to obtain a corrected cutting execution scheme; performing cutting quality evaluation according to the corrected scheme to obtain cutting quality feedback data; and finally, performing parameter learning according to the feedback data to determine initial parameters of a subsequent cutting task. The method can realize high-precision adaptive cutting of the corner area of the flexible screen.

Description

Technical Field

[0001] This invention relates to the field of flexible screen processing technology, and in particular to a method, system and storage medium for irregular cutting control in the production of OLED flexible screens. Background Technology

[0002] Flexible screen irregular shape cutting technology originates from advancements in precision laser processing and materials science. It boasts high precision, high efficiency, and high adaptability, and is widely used in OLED display panels, wearable devices, foldable screen phones, and other fields. Its core foundations include laser control technology, dynamic trajectory planning, stress simulation analysis, and real-time quality monitoring. By precisely controlling cutting parameters and paths, high-quality processing of the complex contours of flexible screens is achieved. Currently, in the field of OLED flexible screen manufacturing, irregular shape cutting technology is crucial for improving product yield and reliability, directly affecting the screen's bending resistance, edge display effect, and overall lifespan.

[0003] In one existing technology, laser cutting is based on empirical parameters, completing the process through a preset cutting path and fixed power parameters. This existing method lacks the ability to dynamically adapt to material stress distribution when dealing with different curvature variations, making it difficult to optimize the matching relationship between the cutting trajectory and laser parameters in real time. If high-speed cutting is uniformly adopted to pursue efficiency, the cutting accuracy of the rounded corners will drastically decrease due to heat accumulation effects and other problems, leading to defects such as microcracks. Conversely, if low-speed cutting is used globally to ensure rounded corner accuracy, overall processing efficiency will be sacrificed.

[0004] In summary, existing technologies lack the ability to dynamically and collaboratively optimize cutting parameters, resulting in a tradeoff between cutting accuracy and efficiency. Summary of the Invention

[0005] This invention provides a method, system, and storage medium for irregular cutting control in the production of OLED flexible screens, in order to solve the problem of the inability to balance cutting accuracy and efficiency in irregular cutting.

[0006] In a first aspect, to solve the above-mentioned technical problems, the present invention provides a method for controlling irregular cutting in the production of flexible OLED screens, comprising:

[0007] Obtain the target rounded corner curvature data of the flexible screen to be cut;

[0008] Based on the target fillet curvature data, stress simulation analysis is performed to obtain the predicted stress distribution of the flexible screen;

[0009] Based on the predicted stress distribution, trajectory planning is performed to generate a dynamic cutting trajectory that adapts to the curvature change.

[0010] Based on the dynamic cutting trajectory, power parameters are calculated to obtain laser power distribution parameters that are compatible with the trajectory;

[0011] Based on the laser power distribution parameters and cutting speed constraints, the parameters are integrated to obtain a cutting execution scheme;

[0012] The cutting execution scheme is executed, edge feature data during the cutting process is extracted, and the edge feature data is adjusted in real time to obtain a corrected cutting execution scheme;

[0013] Based on the revised cutting execution plan, a cutting quality assessment is performed to obtain cutting quality feedback data;

[0014] Based on the cutting quality feedback data, parameter learning is performed to determine the initial cutting parameters for subsequent cutting tasks.

[0015] Preferably, the step of performing stress simulation analysis based on the target fillet curvature data to obtain the predicted stress distribution of the flexible screen includes:

[0016] Construct a finite element analysis model based on the target fillet curvature data and a preset material property database;

[0017] Based on the finite element analysis model, data mapping processing is performed to obtain a preliminary stress concentration distribution;

[0018] Based on the preliminary stress concentration distribution, it is compared with the preset stress value and marked to obtain the stress unevenness points;

[0019] Based on the stress unevenness points, the region corresponding to the target fillet curvature data is divided into grids, the stress value of each grid is obtained, and the predicted stress distribution of the flexible screen is obtained by integrating them.

[0020] Preferably, the step of performing trajectory planning based on the predicted stress distribution to generate a dynamic cutting trajectory adapted to the curvature change includes:

[0021] Based on the predicted stress distribution, an initial path is drawn to obtain the initial trajectory distribution points;

[0022] Based on the initial trajectory distribution points, the areas with large curvature changes are smoothed to obtain smoothed trajectory data;

[0023] Based on the smoothed trajectory data and the predicted stress distribution, a final adaptation process is performed to obtain the dynamic cutting trajectory.

[0024] Preferably, the step of calculating power parameters based on the dynamic cutting trajectory to obtain laser power distribution parameters adapted to the trajectory includes:

[0025] Based on the dynamic cutting trajectory, the thermal influence range of each trajectory segment is calculated and obtained;

[0026] Based on the heat-affected zone, a comparison is made with a preset heat-affected threshold. If there is a trajectory segment in the heat-affected zone that exceeds the preset heat-affected threshold, the laser power of that trajectory segment is reduced to obtain the laser power distribution parameters.

[0027] Preferably, the step of integrating parameters based on the laser power distribution parameters and cutting speed constraints to obtain a cutting execution scheme includes:

[0028] Based on the laser power distribution parameters, the velocity limits of each trajectory segment are calibrated to obtain preliminary trajectory planning results;

[0029] Based on the preliminary trajectory planning results, a comparison is made with a preset curvature matching threshold. If the result exceeds the preset curvature matching threshold, the velocity constraints in the preliminary trajectory planning results are optimized to obtain optimized velocity constraint results.

[0030] Based on the optimized velocity constraint results and the laser power distribution parameters, the cutting execution scheme is obtained through integration processing.

[0031] Preferably, the step of executing the cutting execution scheme, extracting edge feature data during the cutting process, and adjusting the edge feature data in real time to obtain a corrected cutting execution scheme includes:

[0032] Images of the cutting edges are acquired during the cutting process to obtain cutting edge images;

[0033] Based on the cut edge image, feature extraction is performed to obtain edge smoothness evaluation data;

[0034] Based on the edge smoothness evaluation data, the trajectory parameters to be adjusted are obtained by comparing them with the preset standard.

[0035] Based on the trajectory parameters to be adjusted, the cutting execution scheme is locally adjusted to obtain the corrected cutting execution scheme.

[0036] Preferably, the step of evaluating the cutting quality according to the modified cutting execution plan to obtain cutting quality feedback data includes:

[0037] The modified cutting execution scheme is executed to extract the feature information of the cutting surface and obtain preliminary roughness prediction data;

[0038] From the preliminary roughness prediction data, data that meets the preset roughness threshold are selected to obtain the roughness prediction benchmark value;

[0039] The roughness prediction benchmark value is compared with the preset surface quality standard. If the prediction benchmark value is lower than the preset surface quality standard, the evaluation basis parameter that needs to be adjusted is determined.

[0040] A comprehensive evaluation is conducted on the evaluation parameters that need to be adjusted to obtain the cutting quality feedback data.

[0041] Preferably, the step of performing parameter learning based on the cutting quality feedback data to determine the initial cutting parameters for subsequent cutting tasks includes:

[0042] Based on the cutting quality feedback data, an optimization parameter library containing the mapping relationship between cutting parameters and quality scores is constructed;

[0043] Obtain the target fillet curvature value and material property parameters for the new cutting task;

[0044] Based on the target fillet curvature value of the new cutting task, candidate parameter combinations are extracted from the optimization parameter library;

[0045] The initial cutting parameters are obtained by performing fusion processing and accuracy verification based on the candidate parameter combination and the material property parameters.

[0046] Secondly, the present invention provides a non-circular cutting control system for the production of OLED flexible screens, comprising:

[0047] The data acquisition module is used to acquire the target rounded corner curvature data of the flexible screen to be cut;

[0048] The stress analysis module is used to perform stress simulation analysis based on the target fillet curvature data to obtain the predicted stress distribution of the flexible screen;

[0049] The trajectory planning module is used to perform trajectory planning based on the predicted stress distribution and generate a dynamic cutting trajectory that adapts to the curvature change.

[0050] The power calculation module is used to calculate the power parameters based on the dynamic cutting trajectory to obtain the laser power distribution parameters that are adapted to the trajectory.

[0051] The scheme integration module is used to integrate parameters based on the laser power distribution parameters and cutting speed constraints to obtain a cutting execution scheme;

[0052] The real-time adjustment module is used to execute the cutting execution plan, extract edge feature data during the cutting process, and make real-time adjustments based on the edge feature data to obtain a corrected cutting execution plan.

[0053] The quality assessment module is used to assess the cutting quality based on the revised cutting execution plan and obtain cutting quality feedback data.

[0054] The parameter learning module is used to learn parameters based on the cutting quality feedback data in order to determine the initial cutting parameters for subsequent cutting tasks.

[0055] Thirdly, the present invention also provides an electronic device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the irregular cutting control method for the production of OLED flexible screens as described in any one of the above.

[0056] Fourthly, the present invention also provides a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform any of the above-described irregular cutting control methods for the production of OLED flexible screens.

[0057] Compared with the prior art, the present invention has the following beneficial effects:

[0058] (1) This invention significantly improves the initial accuracy and adaptability of irregular cutting by introducing dynamic trajectory planning that combines real-time curvature sensing based on sensor arrays with stress models. Based on the curvature data of the target rounded corners of the flexible screen and the database of flexible screen material properties, and combined with the finite element analysis model, the stress concentration points are accurately predicted. Based on these accurate geometric and physical data, the system uses a trajectory generation algorithm to construct and smooth the cutting path, thereby generating a dynamic cutting trajectory that perfectly adapts to the current curvature change. This shift from "experience-based setting" to "real-time sensing and modeling" solves the problem that traditional methods cannot adapt to different curvatures due to fixed parameters, resulting in microcracks or poor edge smoothness in the rounded corner area, ensuring that the cutting path has high accuracy and scientific rigor in the planning stage.

[0059] (2) This invention achieves real-time optimization and quality assurance of the cutting process by establishing a closed-loop collaborative control mechanism for laser power, cutting speed, and online quality monitoring. During cutting, the system not only predicts the heat-affected zone based on the dynamic trajectory and adaptively adjusts the laser power, but also updates the speed constraints simultaneously to ensure precise matching between power and speed. More importantly, the system uses image processing technology to monitor the flatness of the cutting edge online. Once the flatness is found to be lower than the preset standard, the trajectory fine-tuning mechanism is immediately triggered to correct the path parameters in real time. This "planning-execution-monitoring-fine-tuning" process ensures that any deviation during the cutting process can be quickly corrected, effectively solving the technical problem of balancing cutting accuracy and efficiency, and greatly improving the product yield and bending resistance.

[0060] (3) This invention achieves continuous iteration and intelligent upgrading of the cutting process by constructing a self-learning system for cutting quality feedback and parameter co-optimization. After each cutting task is completed, the system comprehensively evaluates the cutting quality and updates the parameter co-optimization database with a complete set of parameters (such as power, speed, trajectory correction data, etc.) containing successful experiences and adjustment records. When faced with a new cutting task, the system can dynamically call and integrate historical optimization data according to the new fillet curvature and material properties, thereby determining more accurate initial cutting parameters. This mechanism enables the equipment to have self-learning and evolution capabilities, reduces the reliance on repeated trial and error by humans, and ensures that it can always start the task with the optimal parameters when facing diverse production needs, significantly improving production efficiency and the long-term stability of cutting quality. Attached Figure Description

[0061] Figure 1 This is a schematic flowchart of a non-circular cutting control method for OLED flexible screen production provided in the first embodiment of the present invention;

[0062] Figure 2 This is a schematic diagram of a non-circular cutting control system for OLED flexible screen production provided in the second embodiment of the present invention. Detailed Implementation

[0063] 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.

[0064] In this invention, the geometric data of the desired final rounded corner profile is first obtained. Then, based on this data and material properties, the stress changes and distribution that may be caused by the dynamic physical process of laser cutting are simulated in advance. Next, based on the simulation prediction results, the cutting trajectory and parameters are dynamically planned, thereby achieving the goal of avoiding quality defects from the source. The implementation method will be explained below with examples.

[0065] Reference Figure 1 The first embodiment of the present invention provides a method for controlling irregular cutting in the production of flexible OLED screens, comprising the following steps:

[0066] S11, Obtain the target rounded corner curvature data of the flexible screen to be cut;

[0067] S12, based on the target rounded corner curvature data, perform stress simulation analysis to obtain the predicted stress distribution of the flexible screen;

[0068] S13, Based on the predicted stress distribution, perform trajectory planning to generate a dynamic cutting trajectory that adapts to the curvature change;

[0069] S14, Based on the dynamic cutting trajectory, calculate the power parameters to obtain the laser power distribution parameters that are compatible with the trajectory;

[0070] S15, Based on the laser power distribution parameters and cutting speed constraints, the parameters are integrated to obtain the cutting execution scheme;

[0071] S16, Execute the cutting execution scheme, extract edge feature data during the cutting process, adjust the edge feature data in real time, and obtain the corrected cutting execution scheme;

[0072] S17. Based on the revised cutting execution plan, perform a cutting quality assessment to obtain cutting quality feedback data.

[0073] S18. Based on the cutting quality feedback data, perform parameter learning to determine the initial cutting parameters for subsequent cutting tasks.

[0074] In step S11, the target rounded corner curvature data of the flexible screen to be cut is obtained.

[0075] In this application, the target fillet curvature data refers to the design curvature value of the fillet to be formed during processing. This value originates from a digital design model, extracted from the digital design model (such as a CAD file) of the flexible screen product, or preset by the user for the curvature geometry parameters of the fillet portion to be formed by laser cutting. This data characterizes the shape of the target processing contour, rather than a physical feature already present on the blank before cutting. Before cutting, the entire flexible screen blank itself does not possess this fillet shape, and its internal stress distribution is typically uniform.

[0076] The specific acquisition process is as follows: The system first loads a digital design file containing the final contour information of the flexible screen, such as DXF or GDSII format. It then automatically parses the file content to identify the geometric primitives defining the target fillet, such as arcs or B-spline curves. The system extracts key geometric parameters from these primitives and performs calculations accordingly. For example, when processing a DXF format design drawing, if the system identifies an arc entity used to define the product's fillet, and its radius parameter is recorded as 2.5 mm, the system will then use k= The formula calculates that the constant curvature of the target fillet profile is 0.4. For more complex transition fillets composed of B-spline curves, the system can use differential geometry algorithms to calculate and generate a list containing a series of discrete data points that details how the curvature smoothly changes along the path. Finally, this data, precisely calculated from the design model, is integrated to form complete target fillet curvature data, which serves as the initial geometric input for stress simulation analysis in subsequent step S12.

[0077] In step S12, stress simulation analysis is performed based on the target fillet curvature data to obtain the predicted stress distribution of the flexible screen, including:

[0078] Construct a finite element analysis model based on the target fillet curvature data and a preset material property database;

[0079] Based on the finite element analysis model, data mapping processing is performed to obtain a preliminary stress concentration distribution;

[0080] Based on the preliminary stress concentration distribution, it is compared with the preset stress value and marked to obtain the stress unevenness points;

[0081] Based on the stress unevenness points, the region corresponding to the target fillet curvature data is divided into grids, the stress value of each grid is obtained, and the predicted stress distribution of the flexible screen is obtained by integrating them.

[0082] The stress analysis based on the fillet curvature data to obtain the stress distribution of the flexible screen specifically includes: simulating the laser cutting process in advance using a thermo-mechanical coupling simulation model such as finite element analysis (FEA) based on the target fillet contour curvature data and the mechanical properties of the flexible screen material obtained from a material property database. This simulation calculates a predictive stress distribution cloud map of the material at various points on the target fillet contour due to instantaneous heating, melting, and cooling when the laser scans and cuts along the expected path, thereby identifying potential stress concentration areas. In this application, the predicted stress distribution refers to the stress state that may occur during laser cutting, calculated in advance using a computer simulation model based on the target contour and material properties, and is not a measurement of the current physical stress of the blank.

[0083] In one implementation, firstly, a finite element analysis (FEA) model is constructed based on the target fillet curvature data and a pre-defined material property database. This thermo-mechanical coupling simulation model discretizes the geometry of the target contour into tens of thousands of tiny mesh elements and assigns these elements precise material mechanical and thermal properties (such as elastic modulus, Poisson's ratio, and coefficient of thermal expansion) extracted from the database, accurately simulating the interaction between the laser and the flexible screen material. It should be noted that the material property database stores the mechanical properties (such as elastic modulus, Poisson's ratio, and coefficient of thermal expansion), fatigue limit, and stress response characteristics of the flexible screen material under different curvatures. This database can be pre-established through experimental testing (such as tensile tests and bending tests).

[0084] Based on this model, the system performs data mapping processing by pre-simulating the laser cutting process. Specifically, the FEA solver simulates a virtual laser heat source (with pre-set power, beam quality, and movement speed, adaptable to the flexible screen to be cut, thus qualitatively reflecting the stress distribution during the cutting process; this setting can be based on the operator's experience or retrieved from a historical processing database) and makes it scan and cut along the expected path determined by S11 (this expected path is only used for finite element analysis simulation and is unrelated to the actual laser cutting trajectory; when setting it, it must be ensured that the expected path covers all points on the target rounded contour; for example, the expected path can be set as a continuous target rounded contour line; the setting of this expected path is only to facilitate finite element analysis of the stress distribution generated when laser cutting is performed on each point on the target rounded contour under certain conditions). The solver calculates in real time the instantaneous heating, melting, and cooling processes of each grid node in the model caused by laser energy, as well as the resulting complex stress field. After the simulation runs, the dataset containing the dynamic stress values ​​of all grid nodes output by the solver constitutes the preliminary stress concentration distribution. Subsequently, the system compares the preliminary stress concentration distribution with a preset stress threshold (a standard used to determine whether stress is abnormal, usually based on the material's yield strength, etc.). If the instantaneous stress value in a certain region of the model exceeds the threshold, the system automatically marks that region as a stress inhomogeneity point. For these identified high-stress regions, the system performs a meshing operation, which in finite element analysis is usually manifested as adaptive mesh refinement, i.e., automatically increasing the mesh density in these key areas to improve computational accuracy. The system performs secondary or iterative solutions on the refined regions to obtain a more accurate stress value distribution for each refined mesh. Finally, the system integrates these refined high-stress region data with data from other parts of the model to obtain a high-precision, complete predictive stress distribution.

[0085] For example, when performing stress simulation analysis on the target rounded corner profile, the system first constructs a 50,000-element FEA model based on its geometric data and material properties. By simulating the laser cutting process, the solver generates a preliminary stress concentration distribution map. Assuming a preset stress threshold of 120 MPa, the post-processing module detects that the instantaneous stress in a certain region of the model reaches 130 MPa after laser cutting and marks it as a stress non-uniformity point. Subsequently, the system automatically refines the mesh of this high-stress region, increasing the number of elements in the region to four times the original, and performs a secondary local solution. After more accurate calculation, the peak stress in this region is corrected to 135 MPa. Finally, the system integrates this locally optimized data with the model's global stress data to obtain the predicted stress distribution of the flexible screen, which is used to guide subsequent trajectory planning.

[0086] In step S13, based on the predicted stress distribution, trajectory planning is performed to generate a dynamic cutting trajectory adapted to the curvature change, including:

[0087] Based on the predicted stress distribution, an initial path is drawn to obtain the initial trajectory distribution points;

[0088] Based on the initial trajectory distribution points, the areas with large curvature changes are smoothed to obtain smoothed trajectory data;

[0089] Based on the smoothed trajectory data and the predicted stress distribution, a final adaptation process is performed to obtain the dynamic cutting trajectory.

[0090] In this application, the dynamic cutting trajectory refers to a laser scanning path that has been optimized and adjusted based on the predicted stress distribution and real-time feedback information. Its cutting speed and laser power can dynamically change with the path position. In one implementation, the system constructs a cutting trajectory planning path for locations with uneven stress distribution. The system initially draws high-stress areas in the predicted stress distribution map by executing a path planning algorithm (such as the A-star algorithm) and performs position calibration based on the region division results to determine the initial trajectory distribution points. It should be noted that the initial trajectory distribution points are obtained by extracting reference trajectories for key positions from a pre-established trajectory strategy library based on relevant data of high-stress areas in the predicted stress distribution map, and then performing region division processing on the locations to obtain preliminary curvature variation range data. Specifically, the trajectory strategy library contains historical optimized trajectory data for different stress distributions and geometric contours. For example, when the system identifies a high-stress area with a stress value of 135 MPa from the predicted stress distribution map, it divides the high-stress area into several sub-regions according to preset region division rules (such as based on stress gradient).

[0091] The system uses the A-Star path planning algorithm to draw the initial path. The A-Star algorithm comprehensively considers the actual cost (e.g., cutting distance) and the estimated cost (e.g., straight-line distance to the target point). Starting from the starting point, within the divided sub-regions, it selects the node with the lowest cost for expansion based on the predicted stress and target contour of each sub-region, until an initial cutting path from the starting point to the ending point is found. For example, in a sub-region with significant changes in predicted stress, the system sets eight initial trajectory distribution points. The A-Star algorithm will draw a preliminary trajectory line based on these points and the stress change trend. Subsequently, position calibration is performed based on the region division results to ensure that the trajectory points match the actual stress distribution.

[0092] Subsequently, the system will dynamically adjust the curvature of regions with significant changes by executing a smooth interpolation algorithm (such as cubic B-spline interpolation) based on the initial trajectory distribution points. If the curvature value of the region reaches 2.7... The value exceeds the preset threshold range of 1.5 to 2.5. Then, a second interpolation process is performed on the points to obtain suitable smooth trajectory data. It is worth noting that the smooth interpolation process is to avoid the instability of the cutting caused by abrupt changes in curvature, thereby improving the adaptability of the trajectory.

[0093] Finally, the system performs final adaptation processing on the cutting trajectory based on the adapted smooth trajectory data by executing a trajectory generation algorithm (such as particle swarm optimization). It is worth noting that the particle swarm optimization algorithm is used here to fine-tune the smoothed trajectory to find the optimal balance point among multiple performance indicators. In this algorithm, each "particle" represents a potential trajectory correction scheme; its position is a vector containing the fine-tuning displacements of all trajectory control points, and its velocity corresponds to the rate of change of these displacements. All particles search within a bounded displacement range, which constitutes the algorithm's multi-dimensional search space. The algorithm's goal is to find the particle position that maximizes the overall cutting quality score, defined by a weighted multi-objective function. The main evaluation terms include: a stress term that penalizes trajectories close to predicted high-stress areas, a thermal effect term that rewards trajectories with small cumulative thermal effects, and a smoothness term that rewards smooth trajectories with curvature changes. During the optimization process, all particles must also satisfy a key geometric constraint: the maximum deviation between the corrected trajectory and the original target contour must not exceed a preset product tolerance. Any particle that violates this constraint will be assigned a very low penalty score. The algorithm optimizes through iteration. In each iteration, each particle updates its velocity and position based on its own historical best position and the global best position of the entire particle swarm. The iteration terminates when the preset maximum number of iterations is reached or the fitness value of the global best solution converges. Finally, the control point displacement vector corresponding to the global best position at convergence is used as the final trajectory correction scheme to generate the final dynamic cutting trajectory.

[0094] In step S14, power parameters are calculated based on the dynamic cutting trajectory to obtain laser power distribution parameters adapted to the trajectory, including:

[0095] Based on the dynamic cutting trajectory, the thermal influence range of each trajectory segment is calculated and obtained;

[0096] Based on the heat-affected zone, a comparison is made with a preset heat-affected threshold. If there is a trajectory segment in the heat-affected zone that exceeds the preset heat-affected threshold, the laser power of that trajectory segment is reduced to obtain the laser power distribution parameters.

[0097] In this application, the laser power distribution parameter refers to a sequence of laser power values ​​that matches each segment of the dynamic cutting trajectory, used to ensure that the optimal energy density is applied at different locations on the trajectory (especially in regions where the velocity changes).

[0098] In one implementation, the system calculates the thermal influence range of each trajectory segment based on the dynamic cutting trajectory obtained in S13 and the thermal accumulation difference analysis.

[0099] It should be noted that the system extracts the thermal distribution information of each trajectory segment from a pre-established thermal accumulation information database. This database typically stores historical thermal distribution data of different trajectory segments during the flexible screen cutting process, covering various cutting scenarios and parameter configurations. The system compares the thermal impact range of each trajectory segment one by one, and uses difference analysis methods (such as standard deviation analysis) to evaluate the fluctuations, thereby determining the thermal impact range of the trajectory segment.

[0100] Subsequently, if the thermally affected area (TWA) of a certain trajectory exceeds a preset TWA threshold (this threshold is pre-set based on the material's heat resistance and deformation capacity and the product's edge quality requirements, such as a diameter of 2.0 mm), the system will adaptively reduce the laser power corresponding to that trajectory to determine the optimized power distribution parameters. The system will use a PID (Proportional-Integral-Derivative) control algorithm to recalculate the power value and obtain the adjusted power distribution information. Specifically, the PID controller takes the deviation between the current TWA and the preset TWA threshold as input, and outputs a power adjustment amount through proportional, integral, and derivative operations. For example, when the TWA is 2.5 mm, exceeding the 2.0 mm threshold, the PID algorithm reduces the laser power proportionally according to the deviation, and fine-tunes it by combining historical deviation accumulation and deviation change rate until the TWA converges within the threshold range. For example, the original power is iteratively reduced from 10 watts to 8.5 watts, with each power reduction of 0.2 watts, until the TWA is reduced to below 2.0 mm. Subsequently, the system will perform overall optimization parameter calibration for the trajectory segment, and verify it in conjunction with the distribution of the thermal influence range to determine the optimization parameter data that meets the requirements.

[0101] In step S15, based on the laser power distribution parameters and cutting speed constraints, parameter integration is performed to obtain a cutting execution scheme, including:

[0102] Based on the laser power distribution parameters, the velocity limits of each trajectory segment are calibrated to obtain preliminary trajectory planning results;

[0103] Based on the preliminary trajectory planning results, a comparison is made with a preset curvature matching threshold. If the result exceeds the preset curvature matching threshold, the velocity constraints in the preliminary trajectory planning results are optimized to obtain optimized velocity constraint results.

[0104] Based on the optimized velocity constraint results and the laser power distribution parameters, the cutting execution scheme is obtained through integration processing.

[0105] In this application, the cutting execution scheme refers to a complete instruction set (such as G-code) that integrates dynamic cutting trajectory, laser power distribution and speed constraints and can be directly executed by the cutting equipment.

[0106] In one implementation, the system adjusts the laser power output in real time based on the optimized power distribution parameters obtained in S14, and updates the speed constraints in the cutting trajectory planning in a synchronous manner to recalibrate the speed limits of each trajectory segment and obtain preliminary trajectory planning results.

[0107] It should be noted that the system extracts the corresponding laser output value from a pre-established power information database. This database typically stores laser output value records under different cutting parameters during the flexible screen cutting process. By conducting pre-cutting experiments on flexible screens of various materials and thicknesses, the actual cutting effects and corresponding laser output values ​​under different laser powers, cutting speeds, and auxiliary gas pressures can be recorded, thus establishing a mapping relationship between power and cutting effect. Speed ​​constraints in trajectory planning directly affect cutting accuracy and stability. The system recalibrates the speed limits of each trajectory segment by executing an energy density-based speed calibration algorithm. This algorithm calculates the maximum speed required for effective cutting at a given power based on laser power, material absorptivity, material thickness, and the required energy density. For example, if the laser power of a certain trajectory segment is 7.8 watts, based on the material's heat capacity model and cutting quality constraints, the system calculates a new speed limit of 4.5 mm / s to ensure consistency with curvature matching requirements.

[0108] Subsequently, if the curvature matching value of a segment in the initial trajectory planning results exceeds the preset curvature matching threshold range (this threshold is preset based on the geometric tolerance and bending resistance requirements of the OLED flexible screen; for example, if the preset curvature matching threshold is 0.3, and the curvature matching value of a certain trajectory segment reaches 0.4, it is clearly outside the range), the system will perform secondary optimization of the speed constraint of the segment by executing a constraint adjustment algorithm (such as a fuzzy PID control algorithm) to obtain the optimized speed constraint result. The fuzzy PID control algorithm combines the intelligent decision-making capability of fuzzy logic with the precise adjustment capability of PID control. Based on the fuzzy rules of curvature matching deviation (e.g., "large deviation" means "significantly reduce speed"), it dynamically adjusts the cutting speed to ensure that the cutting trajectory closely matches the curvature of the flexible screen. For example, the speed is further adjusted from 4.5 mm / s to 4.0 mm / s to reduce the curvature deviation.

[0109] Finally, based on the optimized speed constraint results and the real-time power distribution adjustment data, the system integrates the overall cutting trajectory through a G-code generation and simulation verification algorithm to determine the final execution scheme information that meets the accuracy requirements. This algorithm converts the optimized power and speed parameters, along with the corrected trajectory coordinates, into a G-code instruction sequence recognizable by the laser cutting equipment. Simultaneously, an integrated simulation module performs virtual cutting simulation on the generated G-code to verify whether its dimensional tolerances and geometric accuracy meet the requirements. For example, if the final power of a certain trajectory segment is 7.8 watts and the speed is 4.0 mm / s, the system verifies through simulation that this scheme can achieve 0.02... Based on the cutting accuracy, it was determined that the solution met the accuracy requirements.

[0110] In step S16, the cutting execution scheme is executed, edge feature data during the cutting process is extracted, and the edge feature data is adjusted in real time to obtain a corrected cutting execution scheme, including:

[0111] Images of the cutting edges are acquired during the cutting process to obtain cutting edge images;

[0112] Based on the cut edge image, feature extraction is performed to obtain edge smoothness evaluation data;

[0113] Based on the edge smoothness evaluation data, the trajectory parameters to be adjusted are obtained by comparing them with the preset standard.

[0114] Based on the trajectory parameters to be adjusted, the cutting execution scheme is locally adjusted to obtain the corrected cutting execution scheme.

[0115] In this application, the modified cutting execution scheme refers to the local, real-time parameter fine-tuning of the original cutting execution scheme based on online monitoring data during the actual cutting process, in order to cope with the uncertainties in the physical process.

[0116] In one implementation, the system integrates a closed-loop control system to monitor the edge smoothness value online during the cutting process, based on the cutting execution scheme obtained in S15. It also acquires images of the cutting edges using image processing techniques (e.g., based on a CCD camera and image sensor) to obtain cutting edge images. Specifically, the system performs grayscale conversion and noise filtering on the acquired flexible screen edge images using an image processing algorithm based on the OpenCV library, thereby obtaining the original set of edge feature data. Specifically, a Gaussian filtering algorithm (e.g., setting the filter window size to 3x3) is used to smooth the image to eliminate random noise, and then an adaptive grayscale threshold segmentation algorithm is used to convert the image into a binary image, thus obtaining the original set of edge feature data. For example, for an original image with a resolution of 1920×1080 pixels, after preliminary processing, the system extracts the set of pixel coordinates of the edge contours, forming an original feature data set containing approximately 2000 key pixels. The purpose of the image processing algorithm is to preprocess the image, such as reducing noise and enhancing contrast, to facilitate more accurate extraction of edge features later.

[0117] Subsequently, the system extracts features from the original set of edge feature data by executing the Canny edge detection algorithm combined with the curvature extremum method. The Canny edge detection algorithm first applies Gaussian filtering to smooth the image, then calculates the image gradient and performs non-maximum suppression and double thresholding to obtain a refined edge contour. Based on this, the curvature extremum method is used to identify key points in the image; points where the rate of change of edge curvature exceeds a preset threshold (e.g., 8%) are identified as key inflection points. For example, in a rounded corner region, the system identified 15 key inflection points, of which 12 points met the preset standard for curvature continuity, and 3 points showed slight deviations. Through a filtering mechanism, the system ultimately determined the acceptable edge flatness evaluation data, with a value of 0.82.

[0118] Finally, the system compares the edge smoothness evaluation data with a preset standard (e.g., the average value based on historical cutting task data or the industry-specified minimum smoothness, with a preset standard of 0.85). If the edge smoothness value is lower than the preset standard, the system triggers a trajectory fine-tuning mechanism to determine the range of trajectory parameters that need correction and locally adjusts the cutting execution plan to obtain the corrected cutting execution plan. The trajectory fine-tuning mechanism analyzes the speed and power parameters in the cutting execution plan, determining that the range of trajectory parameters that need correction covers a 3.2 mm long rounded corner segment with a current cutting speed of 4.0 mm / s and a power of 7.8 watts. The system then locally adjusts the cutting execution plan by executing a gradient descent-based path optimization algorithm. This algorithm uses edge smoothness as the objective function and iteratively calculates and adjusts trajectory parameters (e.g., radius of curvature, cutting speed) to minimize the objective function value. For example, if the radius of curvature of the original trajectory at a certain rounded corner is 0.8 mm, after local adjustments, the system optimizes the radius of curvature of the trajectory at that point to 0.9 mm, while simultaneously adjusting the corresponding cutting speed from 4.0 mm / s to 3.6 mm / s to ensure improved edge smoothness. The system will comprehensively verify the adjusted trajectory parameters. For example, if the expected edge smoothness value increases to 0.86, exceeding the preset threshold of 0.85, the system will generate a complete execution plan including the new trajectory coordinates, adjusted speed parameters, and corresponding power settings.

[0119] In step S17, a cutting quality assessment is performed based on the revised cutting execution plan to obtain cutting quality feedback data, including:

[0120] The modified cutting execution scheme is executed to extract the feature information of the cutting surface and obtain preliminary roughness prediction data;

[0121] From the preliminary roughness prediction data, data that meets the preset roughness threshold are selected to obtain the roughness prediction benchmark value;

[0122] The roughness prediction benchmark value is compared with the preset surface quality standard. If the prediction benchmark value is lower than the preset surface quality standard, the evaluation basis parameter that needs to be adjusted is determined.

[0123] A comprehensive evaluation is conducted on the evaluation parameters that need to be adjusted to obtain the cutting quality feedback data.

[0124] In this application, the cutting quality feedback data refers to the dataset obtained after quantitatively evaluating the final quality of a complete cutting task, which includes the correlation information between various quality indicators of the final product and all process parameters used.

[0125] In one implementation, the system extracts feature information of the cut surface based on the corrected cutting trajectory data in S16 by executing an image processing and feature extraction algorithm based on wavelet transform. It then obtains the correspondence between historical cut surface features and roughness predictions from a pre-established historical cut surface feature and roughness prediction database, thus obtaining a preliminary roughness prediction dataset. This historical cut surface feature and roughness prediction database is established by precisely cutting a large number of flexible screen cutting samples and measuring their cut surface roughness (Ra, Rz, etc.) and surface texture features using high-precision metrology equipment (such as atomic force microscopes and white light interferometers). These actual measurement data are then associated and stored with the corresponding cutting parameters and microscopic image features of the cut surface. It should be noted that the feature information of the cut surface may include key indicators such as surface texture and edge undulations. The system captures microscopic images of the cut surface using high-resolution imaging equipment (such as laser confocal microscopes) and analyzes the average distribution density and undulation height of the surface texture.

[0126] Subsequently, the system will classify and filter the preliminary roughness prediction dataset through cluster analysis to obtain the range of predicted roughness values ​​that meet a preset roughness threshold (e.g., a roughness value between 0.1 and 0.3 mm is considered acceptable), thus determining the benchmark data for subsequent evaluation. For example, the system will remove data points that significantly deviate from the threshold range (e.g., data with roughness values ​​higher than 0.5 mm), retaining the predicted value range that meets the threshold as benchmark data. Data clustering analysis, by grouping similar data points into one category, helps to identify valid data that meets preset standards from a large amount of data. The preset roughness threshold range is pre-set based on the actual application requirements of flexible screens and industry quality standards (e.g., ISO 4287 standard) to ensure that the cut surface meets functional and aesthetic requirements.

[0127] Finally, the surface quality assessment model (such as a support vector machine-based surface quality assessment model) is executed, combining the benchmark data with the cut surface feature information for evaluation. The support vector machine model learns the complex nonlinear relationship between cut surface features and surface quality standards during the training phase, enabling it to accurately predict roughness based on input features. If the predicted value (e.g., a predicted value of 0.3 mm) is lower than the preset surface quality standard (e.g., an industry standard or customer-required Ra value not exceeding 0.2 mm, set based on the display effect, bending life, and reliability requirements of flexible screens), the model will trigger a secondary analysis of the data to determine the evaluation criteria parameters that need adjustment. The secondary analysis may focus on local area features of the cut surface to determine whether local defects cause the overall predicted value to be too high. Based on the evaluation criteria parameters that need adjustment, the system will compare the analysis results with historical feedback results by executing a data integration algorithm to obtain the final cut quality feedback data and determine the specific content of the feedback results. For example, if historical feedback shows that similar cut surface defects can be resolved by adjusting the cutting speed, and the current analysis also points to a problem with the speed parameter, then the final feedback data will suggest reducing the cutting speed, for example, from 500 mm per minute to 400 mm per minute. The data integration algorithm is responsible for cleaning, transforming, and merging data from different sources and of different types for unified analysis and comparison.

[0128] In step S18, based on the cutting quality feedback data, parameter learning is performed to determine the initial cutting parameters for subsequent cutting tasks, including:

[0129] Based on the cutting quality feedback data, an optimization parameter library containing the mapping relationship between cutting parameters and quality scores is constructed;

[0130] Obtain the target fillet curvature value and material property parameters for the new cutting task;

[0131] Based on the target fillet curvature value of the new cutting task, candidate parameter combinations are extracted from the optimization parameter library;

[0132] The initial cutting parameters are obtained by performing fusion processing and accuracy verification based on the candidate parameter combination and the material property parameters.

[0133] First, the system constructs and dynamically updates an optimized parameter library containing the mapping relationship between cutting parameters and quality scores based on the cutting quality feedback data generated in S17. For each successful cutting task (i.e., a quality score higher than a preset threshold, such as 95 points), the complete parameter set used, including the initially planned trajectory, power, and speed, as well as the correction data adjusted in real-time in S16, is stored as a data point in this parameter library along with its corresponding target fillet curvature value, material properties, and final quality score. The system can learn and establish a mapping relationship through regression analysis of which parameter combinations can stably obtain high-quality cutting results under specific geometric and material conditions.

[0134] When a new cutting task is received, the system first retrieves the target fillet curvature value and material property parameters from the design file. Then, the system uses this data as a query index to search the optimization parameter library, extracting one or more historical successful cases that best match the new task. The parameter sets from these cases constitute the candidate parameter combinations.

[0135] Finally, the system performs fusion processing and accuracy verification based on the candidate parameter combinations and the material property parameters of the current task. The fusion processing may employ a weighted average algorithm, assigning different weights to different candidate parameters based on the similarity between historical cases and the current task (e.g., the closeness of curvature and material properties), thereby generating a more targeted set of recommended initial parameters. The subsequent accuracy verification is a crucial step. The system uses these recommended initial parameters to quickly perform a simplified S12 stress simulation analysis to predict the stress results that may occur when cutting with these parameters. If the prediction results meet preset quality standards (e.g., the maximum predicted stress is below the material safety threshold), these parameters are ultimately determined as the initial cutting parameters for the new task and are sent to subsequent S13 to S15 steps, thus starting the new cutting plan with a more optimized starting point. If the prediction results are unsatisfactory, the system will fine-tune the parameters and verify them again until a suitable starting point is found. In this application, the initial cutting parameters refer to a set of optimized process parameters pre-set by the system by learning from historical data before executing a new cutting task, serving as the starting point for the new task planning stage.

[0136] It should be noted that the stress prediction based on simulation described above is an initial, forward-looking planning basis. Due to the extremely complex thermo-mechanical effects during actual cutting, this invention constructs a real-time closed-loop feedback system through steps S16 and S17. This system uses image sensing technology to monitor the stress effect characteristics (such as edge morphology) generated during actual cutting online, thereby dynamically correcting the cutting execution plan. This means that this invention does not rely on a one-time static prediction, but rather uses a control system with real-time feedback to address and compensate for the inherent dynamic uncertainties in laser cutting.

[0137] In summary, this invention achieves high-precision adaptive cutting of the rounded corner area of ​​the flexible screen by real-time monitoring of the change in the curvature of the flexible screen's rounded corners and the corresponding material stress distribution, combined with thermal impact analysis and online evaluation of edge flatness, and adaptively adjusting the laser cutting parameters and trajectory, thus significantly improving cutting quality and efficiency.

[0138] Reference Figure 2 The second embodiment of the present invention provides a non-circular cutting control system for the production of OLED flexible screens, comprising:

[0139] The data acquisition module is used to acquire the target rounded corner curvature data of the flexible screen to be cut;

[0140] The stress analysis module is used to perform stress simulation analysis based on the target fillet curvature data to obtain the predicted stress distribution of the flexible screen;

[0141] The trajectory planning module is used to perform trajectory planning based on the predicted stress distribution and generate a dynamic cutting trajectory that adapts to the curvature change.

[0142] The power calculation module is used to calculate the power parameters based on the dynamic cutting trajectory to obtain the laser power distribution parameters that are adapted to the trajectory.

[0143] The scheme integration module is used to integrate parameters based on the laser power distribution parameters and cutting speed constraints to obtain a cutting execution scheme;

[0144] The real-time adjustment module is used to execute the cutting execution plan, extract edge feature data during the cutting process, and make real-time adjustments based on the edge feature data to obtain a corrected cutting execution plan.

[0145] The quality assessment module is used to assess the cutting quality based on the revised cutting execution plan and obtain cutting quality feedback data.

[0146] The parameter learning module is used to learn parameters based on the cutting quality feedback data in order to determine the initial cutting parameters for subsequent cutting tasks.

[0147] It should be noted that the irregular cutting control system for OLED flexible screen production provided in this embodiment of the invention is used to execute all the process steps of the irregular cutting control method for OLED flexible screen production described in the above embodiment. The working principles and beneficial effects of the two are one-to-one, so they will not be described again.

[0148] This invention also provides an electronic device. The electronic device includes a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a non-circular cutting control program for OLED flexible screen production. When the processor executes the computer program, it implements the steps in the various embodiments of the non-circular cutting control method for OLED flexible screen production described above, for example... Figure 1 The step S11 shown. Alternatively, when the processor executes the computer program, it implements the functions of each module / unit in the above-described device embodiments, such as the data acquisition module.

[0149] For example, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the electronic device.

[0150] The electronic device may be a desktop computer, laptop, handheld computer, or smart tablet, etc. The electronic device may include, but is not limited to, a processor and memory. Those skilled in the art will understand that the above components are merely examples of electronic devices and do not constitute a limitation on the electronic device. It may include more or fewer components than described above, or combine certain components, or different components. For example, the electronic device may also include input / output devices, network access devices, buses, etc.

[0151] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the electronic device, connecting all parts of the electronic device via various interfaces and lines.

[0152] The memory can be used to store the computer programs and / or modules. The processor implements various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory and by calling data stored in the memory. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0153] Wherein, if the modules / units integrated in the electronic device are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0154] It should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0155] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A method for controlling irregular cutting in the production of flexible OLED screens, characterized in that, Executed by a computer, including: Obtain the target rounded corner curvature data of the flexible screen to be cut; Based on the target fillet curvature data, stress simulation analysis is performed to obtain the predicted stress distribution of the flexible screen; Based on the predicted stress distribution, trajectory planning is performed to generate a dynamic cutting trajectory that adapts to the curvature change. Based on the dynamic cutting trajectory, power parameters are calculated to obtain laser power distribution parameters that are compatible with the trajectory; Based on the laser power distribution parameters and cutting speed constraints, the parameters are integrated to obtain a cutting execution scheme; The cutting execution scheme is executed, edge feature data during the cutting process is extracted, and the edge feature data is adjusted in real time to obtain a corrected cutting execution scheme; Based on the revised cutting execution plan, a cutting quality assessment is performed to obtain cutting quality feedback data; Based on the cutting quality feedback data, parameter learning is performed to determine the initial cutting parameters for subsequent cutting tasks; The step of performing trajectory planning based on the predicted stress distribution to generate a dynamic cutting trajectory adapted to the curvature change includes: Based on the predicted stress distribution, an initial path is drawn to obtain the initial trajectory distribution points; Based on the initial trajectory distribution points, the areas with large curvature changes are smoothed to obtain smoothed trajectory data; Based on the smoothed trajectory data and the predicted stress distribution, a final adaptation process is performed to obtain the dynamic cutting trajectory, including: The cutting trajectory is finally adapted by executing a trajectory generation algorithm. Each "particle" represents a potential trajectory correction scheme, and its position is a vector containing the fine-tuning displacement of all trajectory control points. The goal of the algorithm is to find the particle position that maximizes the overall cutting quality score. This score is defined by a weighted multi-objective function, in which the main evaluation terms include: a stress term that penalizes trajectories close to the predicted high-stress area, a thermal effect term that rewards trajectories with small cumulative thermal effects, and a smoothness term that rewards trajectories with smooth curvature changes.

2. The irregular cutting control method for OLED flexible screen production according to claim 1, characterized in that, The step of performing stress simulation analysis based on the target fillet curvature data to obtain the predicted stress distribution of the flexible screen includes: Construct a finite element analysis model based on the target fillet curvature data and a preset material property database; Based on the finite element analysis model, data mapping processing is performed to obtain a preliminary stress concentration distribution; Based on the preliminary stress concentration distribution, it is compared with the preset stress value and marked to obtain the stress unevenness points; Based on the stress unevenness points, the region corresponding to the target fillet curvature data is divided into grids, the stress value of each grid is obtained, and the predicted stress distribution of the flexible screen is obtained by integrating them.

3. The irregular cutting control method for OLED flexible screen production according to claim 1, characterized in that, The step of calculating power parameters based on the dynamic cutting trajectory to obtain laser power distribution parameters adapted to the trajectory includes: Based on the dynamic cutting trajectory, the thermal influence range of each trajectory segment is calculated and obtained; Based on the heat-affected zone, a comparison is made with a preset heat-affected threshold. If there is a trajectory segment in the heat-affected zone that exceeds the preset heat-affected threshold, the laser power of that trajectory segment is reduced to obtain the laser power distribution parameters.

4. The irregular cutting control method for OLED flexible screen production according to claim 1, characterized in that, The step of integrating parameters based on the laser power distribution parameters and cutting speed constraints to obtain a cutting execution scheme includes: Based on the laser power distribution parameters, the velocity limits of each trajectory segment are calibrated to obtain preliminary trajectory planning results; Based on the preliminary trajectory planning results, a comparison is made with a preset curvature matching threshold. If the result exceeds the preset curvature matching threshold, the velocity constraints in the preliminary trajectory planning results are optimized to obtain optimized velocity constraint results. Based on the optimized velocity constraint results and the laser power distribution parameters, the cutting execution scheme is obtained through integration processing.

5. The irregular cutting control method for OLED flexible screen production according to claim 1, characterized in that, The step of executing the cutting execution scheme, extracting edge feature data during the cutting process, and adjusting the edge feature data in real time to obtain a corrected cutting execution scheme includes: Images of the cutting edges are acquired during the cutting process to obtain cutting edge images; Based on the cut edge image, feature extraction is performed to obtain edge smoothness evaluation data; Based on the edge smoothness evaluation data, the trajectory parameters to be adjusted are obtained by comparing them with the preset standard. Based on the trajectory parameters to be adjusted, the cutting execution scheme is locally adjusted to obtain the corrected cutting execution scheme.

6. The irregular cutting control method for OLED flexible screen production according to claim 1, characterized in that, The step of evaluating the cutting quality according to the revised cutting execution plan and obtaining cutting quality feedback data includes: The modified cutting execution scheme is executed to extract the feature information of the cutting surface and obtain preliminary roughness prediction data; From the preliminary roughness prediction data, roughness prediction values ​​within a preset threshold range are selected to obtain roughness prediction benchmark values. The roughness prediction benchmark value is compared with the preset surface quality standard. If the prediction benchmark value is lower than the preset surface quality standard, the evaluation basis parameter that needs to be adjusted is determined. A comprehensive evaluation is conducted on the evaluation parameters that need to be adjusted to obtain cutting quality feedback data.

7. The irregular cutting control method for OLED flexible screen production according to claim 1, characterized in that, The step of learning parameters based on the cutting quality feedback data to determine the initial cutting parameters for subsequent cutting tasks includes: Based on the cutting quality feedback data, an optimization parameter library containing the mapping relationship between cutting parameters and quality scores is constructed; Obtain the target fillet curvature value and material property parameters for the new cutting task; Based on the target fillet curvature value of the new cutting task, candidate parameter combinations are extracted from the optimization parameter library; The initial cutting parameters are obtained by performing fusion processing and accuracy verification based on the candidate parameter combination and the material property parameters.

8. A shaped cutting control system for OLED flexible screen production, used to implement the shaped cutting control method for OLED flexible screen production as described in any one of claims 1 to 7, characterized in that, include: The data acquisition module is used to acquire the target rounded corner curvature data of the flexible screen to be cut; The stress analysis module is used to perform stress simulation analysis based on the target fillet curvature data to obtain the predicted stress distribution of the flexible screen; The trajectory planning module is used to perform trajectory planning based on the predicted stress distribution and generate a dynamic cutting trajectory that adapts to the curvature change. The power calculation module is used to calculate the power parameters based on the dynamic cutting trajectory to obtain the laser power distribution parameters that are adapted to the trajectory. The scheme integration module is used to integrate parameters based on the laser power distribution parameters and cutting speed constraints to obtain a cutting execution scheme; The real-time adjustment module is used to execute the cutting execution plan, extract edge feature data during the cutting process, and make real-time adjustments based on the edge feature data to obtain a corrected cutting execution plan. The quality assessment module is used to assess the cutting quality based on the revised cutting execution plan and obtain cutting quality feedback data. The parameter learning module is used to learn parameters based on the cutting quality feedback data in order to determine the initial cutting parameters for subsequent cutting tasks.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform a non-circular cutting control method for the production of OLED flexible screens as described in any one of claims 1 to 7.

Images