A progressive multi-segment dicing method, system, and apparatus for a dicing saw

Abstract

The application discloses a kind of gradual multi-section cutting method, system and equipment of scriber, it is related to cutting management relevant field, this method includes: identifying cutting task sheet, determine cutting structure diagram;Obtain target material characteristics, trigger first decision maker, with controllable stress distribution executes one section cutting decision based on cutting structure diagram, determine first cutting scheme;With minimum idle stroke ratio as constraint, introduce cutting auxiliary mode, trigger second decision maker to execute gradual segmentation decision and cutting drive decision based on cutting structure diagram, determine second cutting scheme;Timing integration first cutting scheme and second cutting scheme, respond to the cutting management under the multi-axis driving of the central control system of scriber.Solve the technical problem that material damage, cutting efficiency is low due to uneven stress distribution in the cutting process of existing scriber, reaches through gradual multi-section cutting and realizes controllable stress distribution optimization, improves the technical effect of cutting precision and efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of cutting management, in particular to a progressive multi-segment cutting method, system and device of a dicing machine. BACKGROUND

[0002] The dicing machine is a key equipment widely used in the fields of semiconductor packaging, photoelectric element manufacturing and precision material processing, etc., mainly used for cutting wafer, ceramic substrate, sapphire glass and other brittle materials into specified size chips or devices. The traditional dicing process usually adopts single-segment cutting or simple segmented cutting method, and relies on preset path and fixed parameters to perform cutting operation. In most cases, standard cutting parameters are set based on uniform thickness, hardness and structure of the material. However, with the development of the processed material towards ultra-thin, multi-layer composite and high stress sensitivity, the traditional method gradually exposes obvious limitations in practical application.

[0003] When dealing with complex structure materials (such as embedded heterogeneous wafers, micro-via glass substrates, etc.), the existing dicing equipment often causes problems such as edge collapse, cracks, delamination or abnormal wear of the cutter due to inconsistent cutting depth and inaccurate stress control. In order to reduce processing defects, some solutions introduce path planning based on cutting map or tension release preprocessing technology, but still mainly rely on static planning and manual debugging, and cannot realize dynamic adaptive control under the change of material properties or process parameter disturbance. At the same time, the existing cutting system lacks effective mechanism in multi-segment process coordination control, causing too much idle time, incoherent cutting rhythm, low system utilization rate and other problems, which is difficult to meet the current industrial demand of high precision, high efficiency and high reliability. SUMMARY

[0004] The present application provides a progressive multi-segment cutting method, system and device of a dicing machine, which solves the technical problems of material damage and low cutting efficiency caused by uneven stress distribution in the cutting process of the existing dicing machine, and achieves the technical effects of realizing controllable stress distribution optimization through progressive multi-segment cutting, improving cutting precision and efficiency.

[0005] The application provides a progressive multi-segment cutting method of a dicing machine, the method comprising: identifying a cutting task list, determining a cutting structure diagram; obtaining target material characteristics, triggering a first decision maker to perform a one-segment cutting decision based on the cutting structure diagram with controllable stress distribution, and determining a first cutting scheme, wherein the first cutting scheme comprises a geometric limiting scheme based on spatial phase, and a cutter driving scheme based on mechanical driving, the mechanical driving comprising a first trigger decision based on reverse micro-vibration, and a second decision based on frequency resonance; introducing a cutting auxiliary mode with minimum idle stroke ratio as a constraint, triggering a second decision maker to perform progressive segmented decision and cutting driving decision based on the cutting structure diagram, and determining a second cutting scheme; and timing integrating the first cutting scheme and the second cutting scheme, and responding to a central control system of the dicing machine to perform cutting management under multi-axis driving.

[0006] The application also provides a progressive multi-segment cutting system of a dicing machine, the multi-segment cutting system comprising: a task identification module for identifying a cutting task list and determining a cutting structure diagram; a first cutting decision module for obtaining target material characteristics, triggering a first decision maker to perform a one-segment cutting decision based on the cutting structure diagram with controllable stress distribution, and determining a first cutting scheme, wherein the first cutting scheme comprises a geometric limiting scheme based on spatial phase, and a cutter driving scheme based on mechanical driving, the mechanical driving comprising a first trigger decision based on reverse micro-vibration, and a second decision based on frequency resonance; a second cutting decision module for introducing a cutting auxiliary mode with minimum idle stroke ratio as a constraint, triggering a second decision maker to perform progressive segmented decision and cutting driving decision based on the cutting structure diagram, and determining a second cutting scheme; and a cutting management module for timing integrating the first cutting scheme and the second cutting scheme, and responding to a central control system of the dicing machine to perform cutting management under multi-axis driving.

[0007] The application also provides an electronic device comprising: a memory for storing executable instructions; and a processor for executing the executable instructions stored in the memory to implement a progressive multi-segment cutting method of a dicing machine.

[0008] A progressive multi-segment cutting method, system and device of a wafer sawing machine are proposed in the present application. A cutting task sheet is identified, and a cutting structure diagram is determined. Then, target material characteristics are obtained, a first decision maker is triggered to perform a one-segment cutting decision based on the cutting structure diagram with controllable stress distribution to determine a first cutting scheme, wherein the first cutting scheme includes a spatial phase-based geometric limiting scheme and a mechanical drive-based cutter drive scheme, the mechanical drive includes a first trigger decision based on reverse micro-vibration and a second decision based on frequency resonance. Subsequently, a cutting auxiliary mode is introduced with a minimum idle stroke ratio as a constraint, a second decision maker is triggered to perform a progressive segment decision and a cutting drive decision based on the cutting structure diagram to determine a second cutting scheme. Finally, the first cutting scheme and the second cutting scheme are integrated in time sequence, and the cutting management under multi-axis drive is performed in response to the central control system of the wafer sawing machine, so as to improve the cutting precision and efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0009] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings of the embodiments of the present application will be briefly introduced below. In the present application, a flowchart is used to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the foregoing or the following operations are not necessarily performed in sequence. On the contrary, various steps can be processed in reverse order or simultaneously as needed. Meanwhile, other operations can be added to these processes, or one or more steps of operations can be removed from these processes.

[0010] Figure 1 A flowchart of a progressive multi-segment cutting method of a wafer sawing machine provided by the embodiments of the present application.

[0011] Figure 2 A structure diagram of a progressive multi-segment cutting system of a wafer sawing machine provided by the embodiments of the present application.

[0012] Figure 3 A structure diagram of an electronic device provided by the embodiments of the present application.

[0013] Explanation of reference numerals: task identification module 11, first cutting decision module 12, second cutting decision module 13, cutting management module 14, processor 21, memory 22, input device 23, output device 24. DETAILED DESCRIPTION

[0014] The above description is only a summary of the technical solutions of the present application. In order to more clearly understand the technical means of the present application, the embodiments of the present application can be implemented according to the content of the description, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following specific embodiments of the present application are described.

[0015] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of this application will be provided in conjunction with the accompanying drawings. The described embodiments should not be considered as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0016] In the following description, references to "some embodiments" describe a subset of all possible embodiments. However, it is understood that "some embodiments" can be the same or different subsets of all possible embodiments and can be combined with each other without conflict. The terms "first" and "second" are used merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules not explicitly listed or inherent to these processes, methods, products, or devices. 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 application belongs. The terminology used herein is for the purpose of describing embodiments of this application only.

[0017] This application provides a progressive multi-segment cutting method for a dicing machine, such as... Figure 1 As shown, the method includes:

[0018] Identify the cutting task order and determine the cutting structure diagram.

[0019] Specifically, before the dicing machine begins its processing operation, it first receives a cutting task order from the host production scheduling system or operating terminal. This cutting task order typically includes the material number, material type, size specifications, batch information, target division unit size (such as chip size), arrangement method (such as grid arrangement or customized path), and other process constraints (such as margin retention, anti-chipping area, etc.). By parsing this information, the cutting structure diagram corresponding to the task can be retrieved from the cutting database. This cutting structure diagram is a geometric path diagram that describes the complete cutting path from the raw material to the target unit division, including the main cutting path, secondary cutting path, tool entry and exit positions, tool change interval, connection relationship between path nodes, division unit size, cutting path depth parameters, etc., providing a clear and executable spatial planning basis for subsequent cutting decisions, enabling cutting to be performed while ensuring processing accuracy.

[0020] In one possible implementation, after determining the cutting structure diagram, the following steps are included:

[0021] Identify the cutting structure diagram and divide it into a first cutting segment and a second cutting segment, wherein the segmentation method is subjective segmentation; wherein the first decision-maker executes the driving decision of the first cutting segment, and the second decision-maker executes the segmentation and driving decision of the second cutting segment based on the first cutting scheme, wherein the segmentation method is adaptive objective segmentation.

[0022] Specifically, in the cutting preparation stage, the cutting structure diagram is first identified. This diagram includes information such as the overall outline of the raw material, the main cutting path, the secondary cutting path, the tool entry and exit positions, the tool change interval, the connection relationships between path nodes, the unit size, and the cutting path depth parameters. Then, the cutting structure diagram is subjectively segmented. Based on pre-set engineering experience and process planning strategies, the complete cutting path is manually divided into a first cutting segment and a second cutting segment. The first decision-maker is responsible for executing the driving decisions for the first cutting segment, including limit correction, tool micro-vibration control, and stress pre-release, ensuring stable and controllable cutting during the initial stage. When the first cutting segment is completed or enters its final stage, the second decision-maker is simultaneously triggered. Using the executed cutting data corresponding to the first cutting segment (such as stress feedback, path stability, and edge integrity) as a reference benchmark, adaptive objective segmentation is performed based on the actual execution results of the first cutting segment (such as cutting resistance fluctuations and residual stress distribution). The driving decisions for subsequent cutting sub-segments are then based on this segmented structure.

[0023] The target material properties are obtained, triggering a first decision-maker to execute a segment of cutting decisions based on the cutting structure diagram with controllable stress distribution, and determining a first cutting scheme. The first cutting scheme includes a geometric limiting scheme based on spatial phase and a tool driving scheme based on mechanical drive. The mechanical drive includes a first triggering decision based on reverse micro-vibration and a second decision based on frequency resonance.

[0024] Specifically, after generating the cutting structure diagram, the process enters the material identification and initial cutting strategy formulation stage. First, key characteristics of the target material are obtained through methods such as material database matching, surface composition scanning, or manual input. These characteristics include the material's hardness / brittleness level, thickness range, interlayer structure, internal stress distribution characteristics, and brittle fracture response frequency. These characteristics serve as the parameter basis for subsequent cutting decisions. After obtaining the target material characteristics, the first decision-maker is automatically triggered. Using the constructed cutting structure diagram as a path template, it formulates the initial cutting segment, i.e., a cutting strategy. During this process, the first decision-maker not only focuses on the execution of the geometric cutting path itself but also simultaneously considers how to regulate the stress field distribution within the material through physical intervention, thereby minimizing cutting-induced defects, especially suitable for ultra-thin or easily fractured materials. Specifically, the first decision-maker identifies the geometric phase information of the clamped material (such as angular offset, planar warping, or splicing boundaries) through optical positioning and establishes a dynamic geometric constraint path based on this phase information, forming a geometric constraint scheme to ensure that the cutting path is precisely aligned with the material structure in spatial distribution, thus avoiding off-cut or empty cuts caused by deviations. Furthermore, the first decision-maker automatically selects either reverse micro-vibration or frequency resonance as the analysis direction for mechanical drive decision-making based on the material thickness. During the decision-making process, when the material thickness is thin, cutting stress easily induces hidden cracks. In this case, a reverse micro-vibration decision is executed, determining the first trigger decision. This first trigger decision allows the tool to generate a slight reverse micro-vibration at the moment of entry, similar to a negative stiffness response, enabling the tool to mitigate the entry impact in a way that approximates stress dissipation, avoiding the formation of micro-cracks or edge chipping. When the material thickness is thick, a resonance fracture decision is executed, determining the second decision. This second decision adjusts the tool vibration frequency according to the material's natural frequency, causing the two to resonate at a physical level, thereby achieving natural fracture of the material under stress resonance conditions. This method not only reduces cutting resistance but also achieves assisted fracture through stress field reconstruction, significantly improving processing efficiency and edge integrity. By merging the first trigger decision and the second decision into a set, a tool driving scheme is formed. This tool driving scheme, together with the initially determined geometric limit scheme, is merged into a new set as the first cutting scheme. This scheme is used to achieve the first segment cutting based on the coordinated control of spatial matching and mechanical intervention, providing stable and low-stress starting conditions for subsequent progressive cutting segments.

[0025] In one possible implementation, the construction of the first decision-maker before triggering it includes:

[0026] Based on the cutting thickness, a first-order node is deployed, wherein the first-order node performs triggering and driving decisions for reverse micro-vibration; based on frequency resonance, a second-order node is deployed, wherein the second-order node performs controllable stress gradient decisions under resonant fracture; the first-order node and the second-order node are cascaded, and the mechanical driving branch is determined by supervised learning until convergence; the mechanical driving branch and the limit driving branch are parallelized as the first decision-maker, wherein the first decision-maker is embedded in the integrated central control system of the dicing machine.

[0027] Specifically, firstly, using the preset cutting thickness of the target material as a reference parameter, the deployment of a first-order node is initiated. This first-order node handles the micro-stress control problem at the moment of tool entry, executing triggering and driving decisions for reverse micro-vibrations. Because this stage focuses on short-time dynamic response and lightweight judgment logic, a lightweight temporal convolutional-recurrent hybrid model (1D-CNN+GRU) can be selected as the first-order node model. This first-order node can simulate the negative stiffness response mechanism based on subsequently learned knowledge to determine the first-order driving scheme. This first-order driving scheme can introduce low-amplitude, short-period reverse micro-vibrations at the moment the tool begins to enter the material to alleviate stress concentration caused by the initial cutting, preventing microcracks or edge chipping on the material surface or interface due to stress surges. Furthermore, based on the material's natural frequency and stress release behavior characteristics, a second-order node is deployed. This second-order node, guided by frequency resonance, focuses on the continuous cutting behavior after the tool penetrates deep into the material. Its task is to execute controllable stress gradient decisions under resonant fracture. Therefore, this second-order node can be constructed based on a Physically Guided Neural Network (PGNN) model. PGNN embeds prior physical constraints on material stress evolution into a basic deep neural network structure. By jointly learning the material stress field model and the tool resonance spectrum response, it achieves precise control over the material stress release path and low-energy-consumption fracture process. Subsequently, the first-order and second-order nodes are cascaded to construct a complete cutting stress control link. Then, by introducing a supervised learning mechanism, the first-order and second-order nodes in this link are trained based on historical cutting structure diagrams, historical cutting thickness information, historical first-order driving schemes, historical material natural frequencies, historical tool vibration frequencies, and historical second-order driving schemes from historical cutting data. The training process includes forward propagation, loss calculation, backpropagation, and parameter optimization. This training process is iterated until the overall driving strategy converges and stabilizes, ultimately forming a mechanical driving branch with generalization capabilities. Subsequently, this mechanical drive branch is integrated with the previously constructed spatial phase-based limiting drive branch in a parallel structure to form a complete first decision-maker. The construction method of this limiting drive branch is similar to that of the mechanical drive branch, except that it contains only one limiting drive node. This limiting drive node can be constructed based on spatially aware convolutional neural networks (SP-CNN), 3D convolutional neural networks (3D-CNN), etc., and then supervised learning is performed using historical cutting structure diagrams, historical material spatial phase information, and historical geometric limiting schemes from historical cutting data to improve generalization ability. The first decision-maker constructed through integration can not only select the corresponding mechanical drive scheme for different material thicknesses and structural stages, but also achieve synergistic optimization of path and stress by combining positioning constraint strategies. Finally, the first decision-maker is deployed as an embedded component in the integrated central control system of the dicing machine to achieve automated and real-time cutting control response.

[0028] In one possible implementation, a geometrically constrained scheme based on spatial phase includes:

[0029] Connect the optical positioning system to scan the clamped material and determine the material's spatial phase; trigger the limiting drive branch to perform trajectory positioning based on the first cutting segment using the material's spatial phase, and determine the geometric limiting scheme.

[0030] Specifically, before the dicing machine is ready to perform the first cut, an optical positioning system is first connected. This optical positioning system typically includes a high-resolution industrial camera, a laser profilometer, or a structured light scanning device, which is used to scan the surface structure of the material already clamped on the platform. During the scanning process, two-dimensional or three-dimensional image information of key feature points on the material surface (such as chip boundaries, calibration points, dicing grooves, splicing seams, etc.) is acquired. These feature points are then spatially registered in conjunction with the clamping position and reference coordinate system to determine the spatial phase information of the clamped material. This material spatial phase information is a geometric distribution map that represents the material arrangement, boundary morphology, and degree of warping, and can truly reflect the spatial attitude of the material in the current clamping state. After spatial phase extraction is completed, the limit drive branch in the first decision maker is automatically triggered. This limit drive branch takes the extracted material spatial phase as input and performs trajectory matching and offset correction calculations based on spatial geometric relationships in conjunction with the first cutting segment in the cutting structure diagram. It compares the offset, angle error and local warping between the theoretical cutting path and the actual material boundary, and dynamically generates limit adjustment parameters, such as the correction of the tool entry starting position, the fine-tuning trajectory of the cutting path (compensation path), spatial boundary constraints (to prevent miscutting and edge chipping), and multi-axis motion coordination parameters (such as synchronous boundary adjustment of the XY platform), to form a geometric limit scheme. This geometric limit scheme serves as the spatial input condition before the execution of the first cutting segment, ensuring that the tool movement path is accurately aligned with the actual material boundary, avoiding problems such as empty cutting, off-center cutting and band breakage, and providing a stable and controllable geometric reference frame for subsequent mechanical drive strategies (such as reverse micro-vibration or resonant fracture), ensuring that the cutting path is completely matched with the actual state of the material.

[0031] In one possible implementation, a mechanically driven tool driving scheme includes:

[0032] The cutting structure diagram is identified, and the cutting thickness information is determined. The mechanical drive branch is triggered, and the first-order node performs a judgment based on the cutting thickness information to determine the judgment result. If the judgment result is greater than or equal to the preset thickness, the second-order node performs a resonant fracture decision to determine the tool drive scheme.

[0033] Specifically, after initializing the cutting task, the initially acquired cutting structure diagram is identified, and the target cutting thickness information corresponding to the current cutting segment is extracted. This is the vertical cutting path depth from the material surface to the target interface. This parameter is used to determine the stress level the material experiences during cutting and the required drive mode for the tool. Subsequently, the mechanical drive branch in the first decision-maker is triggered, and the first-order node within this branch is called to judge the extracted cutting thickness information. That is, the cutting thickness information is compared with the preset cutting thickness in the first-order node to generate a judgment result, indicating whether the current cutting has entered a high-stress stage. When the judgment result shows that the cutting thickness information is greater than or equal to the set preset cutting thickness threshold (for example, the resonant critical thickness of a certain type of material is 150μm), the second-order node will execute the resonance fracture decision. In this process, the material's natural frequency is matched based on the previously extracted material characteristic data, and this material's natural frequency is combined with the tool vibration frequency and the corresponding cutting segment in the cutting structure diagram. The second-order node analyzes this and generates a second-order drive scheme as the second decision, which is added to the corresponding position in the tool drive scheme. At this time, the first trigger decision in the tool drive scheme is temporarily empty. This second-order drive scheme includes target resonance frequency matching parameters, dynamic amplitude control parameters, resonance maintenance time window parameters, etc., which can jointly guide the tool to complete low-stress cutting in a state of frequency resonance with the material, causing the material to fracture along its natural stress release path, thereby improving cutting efficiency, reducing micro-cracks, edge chipping and tool load, and improving yield.

[0034] In one possible implementation, if the determination result is less than the preset thickness, a reverse micro-vibration decision is performed to determine a first-order driving scheme.

[0035] Specifically, after the cutting structure diagram is identified and the first-order node determines the cutting thickness, if the determination result shows that the cutting thickness is less than the preset cutting thickness, it indicates that the current cutting process belongs to the cutting stage of highly brittle and highly sensitive materials. At this time, directly using conventional or high-frequency vibration cutting can easily lead to stress concentration in the material structure, thereby inducing adverse consequences such as microcracks, edge chipping, and hidden defects, affecting the final yield. Therefore, the first-order node will continue to execute the reverse micro-vibration decision. In this process, the first-order node will analyze the corresponding cutting segment and cutting thickness information in the cutting structure diagram based on the learned knowledge, and determine the specific first-order driving scheme, including the micro-vibration triggering time, reverse micro-vibration amplitude, and reverse micro-vibration frequency. Through these parameters, the dicing machine can control the tool to generate a low-frequency, low-amplitude micro-vibration signal with the opposite direction to the feed direction along the perpendicular incident direction at the moment the tool cuts into the material, to prevent cutting defects caused by excessive cutting stress and to establish a good physical condition foundation for subsequent segmented cutting.

[0036] In one possible implementation, the second-order node performs a resonant fracture decision to determine the tool drive scheme, including:

[0037] Based on the target material properties, the material's natural frequency is determined according to the material stress distribution; taking the resonant fracture based on the cutting structure diagram as the target and the resonance between the material's natural frequency and the tool's vibration frequency as the guide, the tool mechanical drive decision is made to determine a second-order drive scheme; wherein, the tool drive scheme includes the first-order drive scheme and the second-order drive scheme.

[0038] Specifically, after determining the cutting thickness, if the determined cutting thickness is greater than or equal to the preset cutting thickness, it indicates that a cutting depth region capable of accommodating higher vibration energy has been entered. At this point, resonance-assisted fracture can be used to improve cutting efficiency and reduce tool load. To this end, firstly, based on the target material properties, such as the material's elastic modulus, density, internal microstructure arrangement (e.g., single crystal, polycrystalline, or composite layered structure), and residual stress distribution, combined with a material database or real-time measurement data, the natural frequency of the material is determined. The natural frequency refers to the characteristic frequency at which a material naturally generates a resonant response under excited conditions, and is closely related to its geometry, boundary conditions, and stress distribution. Subsequently, the resonant fracture of the cutting structure diagram is used as the cutting target. This resonant fracture is a method of fracturing induced by internal stress under low external force, suitable for structures with strong self-cracking ability, high brittleness, and concentrated energy response. To achieve this goal, the material's natural frequency, the tool's vibration frequency, and the corresponding cutting segment of the cutting structure diagram are input into a second-order node to execute mechanical drive decisions. The second-order node, through learned knowledge, uses the resonance between the material's natural frequency and the tool's vibration frequency as a guide to calculate target resonance frequency matching parameters, dynamic amplitude control parameters, and resonance sustaining time window parameters, ensuring that these parameters match the material's natural frequency or form a harmonic relationship. This ensures low-energy-consumption pyrolysis processing within the resonance range. Finally, the generated second-order drive scheme is stored as the second decision in the corresponding position of the tool drive scheme. Together with the first-order drive scheme, which forms the first trigger decision, a complete tool drive scheme is created. This achieves a high degree of integration between mechanical control logic and physical response laws, ensuring cutting accuracy and efficiency.

[0039] With the minimum idle distance ratio as a constraint, a cutting assistance method is introduced to trigger the second decision-maker to execute progressive segmentation decision and cutting-driven decision based on the cutting structure diagram, and determine the second cutting scheme.

[0040] Specifically, after the first cutting segment is completed and the initial stress intervention is generated by the first decision-maker, the second stage of cutting control begins. The goal of this stage is to further optimize the cutting path efficiency and tool movement state while ensuring material cutting quality. Therefore, the minimum backlash ratio is used as the core constraint. This minimum backlash ratio refers to the ratio of the actual cutting path length of the tool per unit path to its total running length (including the return path and non-cutting movement segments). Higher backlash means lower equipment resource utilization, slower processing cycle time, and increased tool fatigue risk. Therefore, ineffective cutting paths need to be compressed as much as possible in the second cutting segment. To meet this constraint, cutting assistance methods are introduced based on material characteristics. For example, in high-temperature materials or areas of intense stress release, a water-cooling auxiliary device is activated to reduce the temperature rise in the tool-material contact area through a precisely controlled water spray system, while simultaneously stabilizing the material edge structure and mitigating thermal stress concentration. Furthermore, depending on the material characteristics, negative pressure adsorption, auxiliary air jetting, and micro-lubrication can also be selected to enhance the stability of the tool movement environment and ensure low backlash cutting. Subsequently, the second decision-maker is triggered. Based on the identified second cutting segment, feedback data from the execution of the first cutting scheme (such as cutting resistance data and residual material stress), and the parameter control range of the selected cutting assistance method (such as temperature, duration, and spray angle), the second decision-maker initiates a progressive segmented decision-making mechanism. Specifically, the segmented decision branch in the second decision-maker analyzes the feedback data from the second cutting segment and the first cutting scheme, as well as the parameter control range of the selected cutting assistance method, to generate a second cutting scheme with the goal of minimizing the number of segments. This segmented decision branch can be trained using supervised learning on a deep neural network. This second cutting scheme includes toolpath speed, local excitation frequency, and water-cooling start / stop control, which can minimize the return stroke between each cutting path segment or achieve continuous cutting between adjacent paths as much as possible, thereby reducing the overall idle stroke ratio.

[0041] The first cutting scheme and the second cutting scheme are integrated in sequence, and the cutting management under multi-axis drive is performed in response to the central control system of the dicing machine.

[0042] Specifically, after obtaining the first and second cutting schemes, the process enters the integration and execution phase of the cutting control flow. At this stage, the generated cutting schemes are integrated sequentially. Specifically, the first cutting scheme (aiming at stress control and precise tool entry) and the second cutting scheme (aiming at path optimization and backlash reduction) are ordered along a unified timeline and connected to the process segments, constructing a continuous, conflict-free, full-segment cutting control flow. After integration, the complete integrated scheme is sent to the dicing machine's central control system. The central control system, under a multi-axis drive architecture (such as the XY platform, Z-axis tool entry / exit mechanism, and θ-axis rotation compensation mechanism), performs high-precision synchronous control between each axis, thereby achieving closed-loop execution of the complete, multi-segment, and intelligent cutting process. This improves cutting accuracy and efficiency, ensuring continuous, efficient, and low-damage processing of the entire material.

[0043] In one possible implementation, the central control system of the dicing machine performs multi-axis coupled cutting management, including:

[0044] The first cutting scheme and the second cutting scheme are integrated in sequence to determine the progressive cutting scheme; for the progressive cutting scheme, the drive shaft of the dicing machine is set and the scheme is decomposed and distributed to perform multi-axis cascaded orderly cutting management.

[0045] Specifically, after generating the first and second cutting schemes, the execution sequence, physical path distribution, and tool movement logic of these two schemes are sequentially integrated and mapped into a continuous, conflict-free machining process, generating a complete progressive cutting scheme. This progressive scheme not only retains the advantages of stress control and high tool entry accuracy in the first stage but also integrates the control logic of segmented path optimization and auxiliary mechanism matching in the second stage, forming a hierarchical, flexible, multi-segment machining process. Subsequently, based on this progressive cutting scheme, precise configuration and control parameters are issued to each drive axis of the dicing machine. The dicing machine is typically equipped with a multi-axis linkage system, including XY axes, Z axes, θ axes or U axes, and platform movement axes (such as T axes). Among them, the XY axes are responsible for material positioning and tool path movement in the two-dimensional plane; the Z axis controls the up and down movement of the tool to achieve cutting and disengagement; the θ axis or U axis is used for micro-rotation adjustment of the material or tool to achieve angle compensation; and the platform movement axis (such as the T axis) is used for production line transfer of the workpiece from the first machining unit to the second machining unit. Subsequently, based on the stage to which the cutting segment belongs, each segment's path is mapped to a specific axis control task. For example, the first cutting segment is executed by drive axes of group A (such as X1-Y1-Z1), completing the micro-vibration and initial limit cutting operations of the first stage. After cutting, the processed workpiece is transferred to the second cutting station via the platform's moving axis. The second cutting segment is taken over by drive axes of group B (such as X2-Y2-Z2), automatically adjusting the initial position of the drive axes (such as zeroing, aligning corner points, and repositioning), loading the second segment's drive parameters and auxiliary control logic, and continuing to execute subsequent cutting tasks. This division of labor achieves orderly cutting management of multi-axis cascade, thereby ensuring that each stage's drive axis group has independent control authority, accurate path execution boundaries, and clear task logic, avoiding inter-axis conflicts or processing cycle delays, and improving collaborative processing capabilities.

[0046] In the above text, refer to Figure 1 A progressive multi-segment cutting method for a dicing machine according to an embodiment of the present invention is described in detail. Next, reference will be made to... Figure 2 A progressive multi-segment cutting system for a dicing machine according to an embodiment of the present invention is described.

[0047] A progressive multi-segment cutting system for a dicing machine according to an embodiment of the present invention addresses the technical problems of material damage and low cutting efficiency caused by uneven stress distribution during the cutting process in existing dicing machines. It achieves the technical effect of optimizing controllable stress distribution through progressive multi-segment cutting, thereby improving cutting accuracy and efficiency. The progressive multi-segment cutting system for a dicing machine includes: a task identification module 11, a first cutting decision module 12, a second cutting decision module 13, and a cutting management module 14.

[0048] Task identification module 11: Identifies the cutting task order and determines the cutting structure diagram; First cutting decision module 12: Obtains the target material properties, triggers the first decision-maker, and executes a segment cutting decision based on the cutting structure diagram with controllable stress distribution to determine the first cutting scheme, wherein the first cutting scheme includes a geometric limiting scheme based on spatial phase and a tool driving scheme based on mechanical drive, wherein the mechanical drive includes a first triggering decision based on reverse micro-vibration and a second decision based on frequency resonance; Second cutting decision module 13: With minimum backlash ratio as a constraint, introduces a cutting assistance method, triggers the second decision-maker to execute progressive segmentation decision and cutting drive decision based on the cutting structure diagram to determine the second cutting scheme; Cutting management module 14: Integrates the first cutting scheme and the second cutting scheme in sequence, and performs multi-axis joint drive cutting management in response to the central control system of the dicing machine.

[0049] The specific configuration of the task identification module 11 will be described in detail below. As mentioned above, after determining the cutting structure diagram, the task identification module 11 may further include: identifying the cutting structure diagram and dividing it into a first cutting segment and a second cutting segment, wherein the segmentation method is subjective segmentation; wherein the first decision-maker executes the driving decision of the first cutting segment, and the second decision-maker executes the segmentation and driving decision of the second cutting segment based on the first cutting scheme, wherein the segmentation method is adaptive objective segmentation.

[0050] The specific configuration of the first cutting decision module 12 will be described in detail below. As mentioned above, before triggering the first decision-maker, the construction of the first decision-maker, the first cutting decision module 12 may further include: deploying a first-order node based on the cutting thickness, wherein the first-order node performs triggering and driving decisions for reverse micro-vibration; deploying a second-order node based on frequency resonance, wherein the second-order node performs controllable stress gradient decisions under resonant fracture; cascading the first-order node and the second-order node, and determining the mechanical driving branch through supervised learning to convergence; and paralleling the mechanical driving branch and the limit driving branch as the first decision-maker, wherein the first decision-maker is embedded in the integrated central control system of the dicing machine.

[0051] Among them, the geometric limiting scheme based on spatial phase, the first cutting decision module 12 may further include: connecting to the optical positioning system, scanning the clamping material, determining the material spatial phase; triggering the limiting drive branch, using the material spatial phase to perform trajectory positioning based on the first cutting segment, and determining the geometric limiting scheme.

[0052] In the tool driving scheme based on mechanical drive, the first cutting decision module 12 may further include: identifying the cutting structure diagram and determining the cutting thickness information; triggering the mechanical drive branch, the first-order node performing a judgment based on the cutting thickness information and determining the judgment result; if the judgment result is greater than or equal to a preset thickness, the second-order node performing a resonant fracture decision and determining the tool driving scheme.

[0053] The first cutting decision module 12 may further include: if the determination result is less than the preset thickness, performing reverse micro-vibration decision to determine a first-order driving scheme.

[0054] Wherein, the second-order node performs resonant fracture decision-making to determine the tool driving scheme, and the first cutting decision module 12 may further include: determining the material natural frequency based on the material stress distribution for the target material characteristics; taking the resonant fracture based on the cutting structure diagram as the target and the resonance between the material natural frequency and the tool vibration frequency as the guide, performing tool mechanical driving decision-making to determine the second-order driving scheme; wherein, the tool driving scheme includes the first-order driving scheme and the second-order driving scheme.

[0055] The specific configuration of the cutting management module 14 will be described in detail below. As mentioned above, in response to the cutting management under multi-axis drive by the central control system of the dicing machine, the cutting management module 14 may further include: sequentially integrating the first cutting scheme and the second cutting scheme to determine the progressive cutting scheme; setting and decomposing the drive shafts of the dicing machine for the progressive cutting scheme, and executing the orderly cutting management of multi-axis cascade.

[0056] The progressive multi-segment cutting system for a dicing machine provided in this embodiment of the invention can execute the progressive multi-segment cutting method for a dicing machine provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0057] Although this application makes various references to certain modules in the system according to the embodiments of this application, any number of different modules can be used and run on user terminals and / or servers. The various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of each functional unit are only for easy distinction between each other and are not used to limit the scope of protection of this invention.

[0058] Based on the foregoing embodiments, this application also provides an electronic device. Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention, showing a block diagram of an exemplary electronic device suitable for implementing the embodiments of the present invention. Figure 3The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of the present invention. This electronic device is in the form of a general-purpose computing device, and its components may include, but are not limited to, a processor 21, a memory 22, an input device 23, and an output device 24. The processor 21 may be one or more; the memory 22 may include a computer-readable medium and at least one program product having a set (at least one) of program modules configured to perform the functions of the embodiments of this application.

[0059] The memory 22 shown in this embodiment of the invention can be any combination of one or more computer-readable media. The computer-readable storage medium can be, but is not limited to, infrared, semiconductor systems, devices or components, or any combination thereof, used to store software programs, computer-executable programs and modules, such as the program instructions / modules corresponding to the progressive multi-segment cutting method of a dicing machine in this embodiment of the invention. The processor 21 executes various functional applications and data processing of the computer device by running the software programs, instructions and modules stored in the memory 22, thereby realizing the progressive multi-segment cutting method of the dicing machine described above.

[0060] The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application. In some cases, the actions or steps described in this application can be performed in a different order than that shown in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A progressive multi-segment cutting method for a dicing machine, characterized in that, The method includes: Identify the cutting task sheet and determine the cutting structure diagram; The target material properties are obtained, triggering the first decision-maker to execute a cutting decision based on the cutting structure diagram with controllable stress distribution, and determining the first cutting scheme. The first cutting scheme includes a geometric limiting scheme based on spatial phase and a tool driving scheme based on mechanical drive. The mechanical drive includes a first triggering decision based on reverse micro-vibration and a second decision based on frequency resonance. With the minimum idle ratio as a constraint, a cutting assistance method is introduced to trigger the second decision-maker to execute progressive segmented decision-making and cutting drive decision based on the cutting structure diagram, and determine the second cutting scheme. The minimum idle ratio refers to the ratio of the actual cutting path length of the tool in a unit path to its total running length. The first cutting scheme and the second cutting scheme are integrated in sequence, and the cutting management under multi-axis drive is performed in response to the central control system of the dicing machine; After determining the cutting structure diagram, the process includes: Identify the cutting structure diagram and divide it into a first cutting segment and a second cutting segment, wherein the segmentation method is subjective segmentation; The first decision-maker executes the driving decision for the first segmentation, and the second decision-maker executes the segmentation and driving decision for the second segmentation based on the first segmentation scheme, wherein the segmentation method is adaptive objective segmentation.

2. The progressive multi-segment cutting method for a dicing machine as described in claim 1, characterized in that, Before triggering the first decision-maker, the construction of the first decision-maker includes: A first-order node is deployed based on the cutting thickness, wherein the first-order node performs triggering and driving decisions for reverse micro-vibrations; Guided by frequency resonance, second-order nodes are deployed, wherein the second-order nodes perform controllable stress gradient decisions under resonant fracture. The first-order node and the second-order node are cascaded, and the mechanical drive branch is determined through supervised learning until convergence. The mechanical drive branch and the limit drive branch are used in parallel as the first decision-maker, wherein the first decision-maker is embedded in the integrated central control system of the dicing machine.

3. The progressive multi-segment cutting method for a dicing machine as described in claim 2, characterized in that, Geometric restraint schemes based on spatial phase include: Connect the optical positioning system to scan the clamped material and determine its spatial phase; The limiting drive branch is triggered to perform trajectory positioning based on the first cutting segment using the material space phase, thereby determining the geometric limiting scheme.

4. The progressive multi-segment cutting method for a dicing machine as described in claim 3, characterized in that, Mechanically driven tool driving solutions include: Identify the cutting structure diagram and determine the cutting thickness information; The mechanical drive branch is triggered, and the first-order node performs a determination based on the cutting thickness information to determine the determination result. If the determination result is greater than or equal to the preset thickness, the second-order node performs a resonant fracture decision to determine the tool driving scheme.

5. The progressive multi-segment cutting method for a dicing machine as described in claim 4, characterized in that, If the determination result is less than the preset thickness, reverse micro-vibration decision is executed to determine the first-order driving scheme.

6. The progressive multi-segment cutting method for a dicing machine as described in claim 5, characterized in that, The second-order node performs a resonant fracture decision to determine the tool drive scheme, including: For the target material properties, determine the material's natural frequency based on the material stress distribution; With the resonant fracture based on the cutting structure diagram as the target, and guided by the resonance between the material's natural frequency and the tool's vibration frequency, the tool's mechanical drive decision is made to determine the second-order drive scheme. The tool driving scheme includes the first-order driving scheme and the second-order driving scheme.

7. The progressive multi-segment cutting method for a dicing machine as described in claim 1, characterized in that, The central control system of the dicing machine performs multi-axis joint drive cutting management, including: The first and second cutting schemes are sequentially integrated to determine a progressive cutting scheme; For the aforementioned progressive cutting scheme, the drive shafts of the dicing machine are configured and the scheme is decomposed and distributed to implement orderly cutting management through multi-axis cascading.

8. A progressive multi-segment cutting system for a dicing machine, characterized in that, The multi-segment cutting system is used to implement the progressive multi-segment cutting method of a dicing machine according to any one of claims 1-7, the multi-segment cutting system comprising: Task recognition module: Recognizes cutting task orders and determines the cutting structure diagram; First cutting decision module: acquires target material properties, triggers first decision-maker, executes a cutting decision based on the cutting structure diagram with controllable stress distribution, and determines first cutting scheme, wherein the first cutting scheme includes a geometric limiting scheme based on spatial phase and a tool driving scheme based on mechanical drive, wherein the mechanical drive includes a first triggering decision based on reverse micro-vibration and a second decision based on frequency resonance; The second cutting decision module: with the minimum idle ratio as a constraint, a cutting assistance method is introduced to trigger the second decision-maker to execute progressive segmented decision and cutting drive decision based on the cutting structure diagram to determine the second cutting scheme. The minimum idle ratio refers to the ratio of the actual cutting path length of the tool in a unit path to its total running length. Cutting Management Module: Integrates the first cutting scheme and the second cutting scheme in sequence, and performs cutting management under multi-axis drive in response to the central control system of the dicing machine.

9. An electronic device, characterized in that, The electronic device includes: Memory, used to store executable instructions; The processor, when executing executable instructions stored in the memory, implements the progressive multi-segment cutting method of the dicing machine according to any one of claims 1 to 7.

Images