Glass punching hole site chamfering machine control system and method

Abstract

The invention relates to a glass punching hole site chamfering machine control system and method, and belongs to the technical field of glass processing equipment. Comprising a multi-mode cooperative control module which comprises a glass laser drilling and chamfering composite mode, a pure laser drilling mode and a pure transmission mode, the modes are dynamically switched, shaft control and laser energy resources are distributed through a laser processing resource scheduling algorithm, and idle shaft resources are released; the laser vision-axis control closed-loop compensation module integrates an industrial camera and a three-dimensional space deviation algorithm, calculates the space deviation of a vision coordinate, a laser focus coordinate and a workpiece coordinate by collecting a laser drilling area image, and drives an axis control system to correct a path in real time; through a laser vision-axis control closed-loop compensation module and a parameter self-adaptive compensation algorithm, machining precision control is achieved, deviation caused by clamping errors and laser attenuation is effectively eliminated, and the product yield is remarkably improved.

Description

TECHNICAL FIELD

[0001] The application relates to a glass punching hole site chamfering machine control system and method, and belongs to the technical field of glass processing equipment. BACKGROUND

[0002] Glass punching hole site chamfering is a key process in the processing of precision electronic glass (such as smart phone cover plate, vehicle display glass and photovoltaic glass), and the processing precision directly affects the product assembly tolerance and service life. With the development of glass materials towards ultra-thin (thickness ≤0.5mm) and high strength (hardness ≥7H), the traditional processing technology faces the following core challenges: 1. Insufficient visual positioning accuracy, unable to adapt to the transparent nature of glass Most existing laser punching equipment uses a traditional 2D vision system to achieve positioning through surface contour matching, but the high light transmittance of glass leads to low image contrast, and laser reflection spots are prone to interference.

[0003] 2. Conflict between multi-station resource scheduling, low laser energy utilization Fixed priority scheduling algorithm (only considering station number) is used without dynamic adjustment of laser energy utilization (Ei) and usage frequency (Fi). In actual application, when more than 3 stations compete for the laser source, the resource conflict resolution time exceeds 100ms, the laser idle rate is as high as 25%, and there is no energy threshold triggered resource release mechanism, resulting in serious energy waste during continuous processing.

[0004] 3. Laser parameter compensation mechanism is rigid, without considering aging attenuation For example, Chinese Patent No. CN223222687U discloses a borosilicate glass laser punching machine, which optimizes positioning through a fixed assembly, but does not integrate a life warning algorithm, and the laser head replacement relies on manual experience, leading to an increase in sudden failure downtime.

[0005] 4. Fault station replacement relies on manual intervention, with poor production continuity The replacement process uses "stop-manual switching-reset", and the replacement response time exceeds 5 minutes, and the correlation between historical deviation data of the replacement station and the current laser power is not considered. In a glass continuous production line, a single fault causes a large loss of production capacity.

[0006] Therefore, it is an urgent need in the field of precision glass processing to develop a control system that integrates high-precision visual compensation, dynamic resource scheduling, adaptive parameter adjustment and intelligent replacement decision-making. SUMMARY

[0007] The purpose of the present application is to provide a glass punching hole chamfering machine control system and method, which integrates dynamic resource scheduling, visual closed-loop compensation, parameter adaptive adjustment and intelligent compensation function, aiming at the above problems.

[0008] The glass punching hole chamfering machine control system provided by the present application comprises: The multi-mode cooperative control module includes a glass laser punching + chamfering composite mode, a pure laser punching mode and a pure transmission mode, the above modes are dynamically switched, the laser processing resource scheduling algorithm is used to allocate shaft control and laser energy resources, and idle shaft resource release is realized. The laser vision-shaft control closed-loop compensation module integrates an industrial camera and a three-dimensional space deviation algorithm, calculates the space deviation of the visual coordinates, the laser focus coordinates and the workpiece coordinates by collecting the image of the laser punching area, and drives the shaft control system to correct the path in real time. The laser parameter adaptive compensation module dynamically adjusts the processing parameters based on the laser energy attenuation model through the pulse width correction formula T =T0×(1+k×N), T0 is the reference pulse width, k is the laser head aging coefficient, and N is the cumulative number of laser trigger times. The laser processing intelligent compensation decision module generates a compensation path according to the fault work station deviation, the historical deviation average of the compensation work station and the dynamic correction coefficient when the laser punching work station fails, and selects the optimal compensation work station through the distance sorting algorithm.

[0009] Preferably, the laser processing resource scheduling algorithm specifically comprises the following steps: S11: Real-time scanning of each work station state, collecting inherent priority, laser energy utilization rate and historical use frequency parameters of the work station to be scheduled, and generating a to-be-scheduled queue; S12: Calculate the comprehensive priority of each work station based on the inherent priority, laser energy utilization rate, historical use frequency and corresponding weight coefficient, and sort the comprehensive priority in descending order to form a priority queue; S13: Distribute laser source and shaft control resources according to priority order: if the resources are sufficient, directly distribute and mark the work station as "processing"; if the resources conflict, determine the conflict priority by weighted calculation of laser energy utilization rate and historical use frequency, distribute high-priority work stations, and low-priority work stations enter the waiting queue; S14: Real-time monitoring of the utilization rate of the allocated resources: when the laser energy utilization rate is lower than the set value and there is no processing request for a certain period of time, trigger resource release, including laser source hibernation and shaft control resource marking as "to be allocated"; S15: After completing a certain processing task, recalibrate the historical use frequency of the work station and each weight coefficient to ensure that the algorithm adapts to the change of the processing scene.

[0010] Preferably, the formula is: (3) The spatial deviation of visual coordinates (X 视 ,Y 视 ), laser focus coordinates Z 焦 and workpiece coordinates (X 激 ,Y 激 ,Z 工 ) is calculated, and the drive shaft control system corrects the path in real time.

[0011] Preferably, the three-dimensional spatial deviation algorithm is optimized for the glass transparent characteristics, including a hierarchical contour extraction step: Coarse matching positioning: scanning the laser drilling area image with a 3px step, and extracting the approximate region of the hole contour through the laser spot gradient operator: (4) wherein, represents the gray change rate in the x direction (horizontal direction), represents the gray change rate in the y direction (vertical direction), and the approximate region of the hole contour is extracted; Fine matching optimization: sub-pixel level positioning is performed in the coarse matching region with a 0.1px step, and the formula X 亚 =X 粗 +Δx×(G 左 -G 右 ) / (G 左 +G 右 ) (5) is used, wherein Δx is the step, G 左 and G 右 are the left and right neighborhood gradient values of the spot, respectively, and the sub-pixel coordinates are calculated.

[0012] Preferably, the laser parameter adaptive compensation module further comprises a laser head remaining life warning sub-module: the remaining trigger times are calculated through the formula T 剩余 =(T 临界 -T ) / (k×T0) (6) , wherein T 临界 is the laser effective pulse width threshold, T0×0.7 is taken, and when T 剩余 ≤1000 times, the laser head maintenance reminder is triggered.

[0013] Preferably, the intelligent compensation decision module calculates the historical deviation mean value using a sliding window algorithm: the window size N is configurable, and the latest N times of processing deviation data are processed by arithmetic mean.

[0014] The method of the glass drilling hole chamfering machine control system disclosed by the application comprises the following steps: S1: Load the laser processing axis control configuration file, allocate resources through the shaft-laser source mapping algorithm, call the laser processing resource scheduling algorithm to calculate the priority and solve the conflict; S2: Visual-axis control closed loop compensation: collect laser drilling area image to obtain visual coordinates (X 视 , Y 视 ), record laser focal point coordinates Zfocus and workpiece coordinates (X 激 , Y 激 , Z 工 ), substitute into the three-dimensional space deviation formula to calculate delta, if delta>0.005mm, then execute the modified path; S3: Laser parameter compensation execution: read the cumulative trigger number N of the laser head, adjust the processing parameters through the pulse width correction formula T =T0×(1+k×N), and synchronously monitor the laser power fluctuation ΔP, when |ΔP|>3%×P0, P0 is the reference power, trigger the secondary compensation; S4: Intelligent replacement processing: when detecting work station failure, call the replacement path generation algorithm, and select the nearest replacement work station according to the Euclidean distance sorting.

[0015] Preferably, the laser power monitoring in step S3 adopts a photodiode to collect laser energy in real time, and the average power is calculated through the formula: When the power fluctuation exceeds the threshold, the parameter compensation is triggered.

[0016] Preferably, the distance sorting algorithm in step S4 is based on the Euclidean distance of the work station X / Y axis coordinates to select the replacement work station with the smallest D to execute processing.

[0017] Preferably, the image acquisition unit of the laser visual-axis control closed loop compensation module adopts a laser coaxial vision system with a resolution of ≥50 million pixels, is equipped with a narrowband filter to filter laser stray light, and has a frame rate of ≥60fps to ensure image capture at the moment of laser pulse triggering.

[0018] Compared with the existing technology, the glass punching hole chamfering machine control system and method has the following beneficial effects: 1. The processing precision and yield are significantly improved Through the laser visual-axis control closed loop compensation module and the parameter adaptive compensation algorithm, the processing precision control is realized, the deviation caused by clamping error and laser attenuation is effectively eliminated, and the product yield is significantly improved.

[0019] 2. Resource utilization rate and energy consumption optimization The multi-mode collaborative control module dynamically allocates laser and axis control resources, improving equipment uptime and laser energy utilization. The idle resource release mechanism reduces energy consumption in non-processing states, extends the service life of core components, reduces maintenance costs, and achieves efficient and energy-saving production.

[0020] 3. Enhanced production continuity and flexibility The intelligent fault-finding module responds quickly to workstation malfunctions, shortens downtime, and ensures continuous production line operation. It supports dynamic switching between multiple modes such as composite processing, pure drilling, and pure transmission to adapt to different product process requirements, improve the flexibility of production lines, reduce manual intervention, and promote the transformation of the glass processing industry from traditional experience-driven to data-driven, thus contributing to green manufacturing and industrial upgrading. Attached Figure Description

[0021] Figure 1 This is a diagram of the system architecture in this invention; Figure 2 This is a schematic diagram of the optical path structure of a single laser drilling machine in this invention; Figure 3 This is an interface display diagram of the fill function in this invention; Figure 4 This is a diagram showing the interface of the image acquisition unit in this invention. Figure 5 This is a punched image from the present invention. Detailed Implementation

[0022] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0023] Example 1: like Figure 1 As shown, this embodiment discloses a control system for a glass drilling and beveling machine, including: Multi-mode collaborative control module: including glass laser drilling + chamfering composite mode, pure laser drilling mode, and pure transmission mode. The above modes are dynamically switched, and the axis control and laser energy resources are allocated through the laser processing resource scheduling algorithm to release idle axis resources; Laser vision-axis control closed-loop compensation module: integrates industrial camera and 3D spatial deviation algorithm. By acquiring images of the laser drilling area, it calculates the spatial deviation between visual coordinates, laser focus coordinates and workpiece coordinates, and drives the axis control system to correct the path in real time. Laser parameter adaptive compensation module: Based on the laser energy attenuation model, the processing parameters are dynamically adjusted by the pulse width correction formula T = T0 × (1 + k × N), where T0 is the reference pulse width, k is the laser head aging coefficient, and N is the cumulative number of laser triggers. Laser processing intelligent compensation decision module: when the laser drilling station fails, generate a compensation path according to the deviation of the failed station, the historical deviation mean of the compensation station and the dynamic correction coefficient, and select the optimal compensation station through the distance sorting algorithm.

[0024] The laser processing resource scheduling algorithm specifically includes the following steps: S11: Real-time scanning of each station state, collecting inherent priority, laser energy utilization rate, and historical usage frequency parameters of the to-be-scheduled station, and generating a to-be-scheduled queue; S12: Calculate the comprehensive priority of each station based on the inherent priority, laser energy utilization rate, historical usage frequency and corresponding weight coefficient, and sort the comprehensive priority in descending order to form a priority queue; S13: Distribute laser sources and axis control resources according to priority order: if resources are sufficient, directly distribute and mark the station as "processing"; if resources conflict, determine the conflict priority through weighted calculation of laser energy utilization rate and historical usage frequency, distribute high-priority stations, and low-priority stations enter the waiting queue; S14: Real-time monitoring of the utilization rate of the allocated resources: When the laser energy utilization rate is <10% and there is no processing request for 5 consecutive periods, trigger resource release: Release instruction = laser source hibernation + axis control resource marking "to be allocated"; Step S15: After completing 10 processing tasks, calibrate the historical usage frequency of the station and each weight coefficient to ensure that the algorithm adapts to changes in the processing scene.

[0025] The formula is: (3) Calculate the spatial deviation of the visual coordinates (X 视 ,Y 视 ), laser focus coordinates Z 焦 and workpiece coordinates (X 激 ,Y 激 ,Z 工 ), drive the axis control system to correct the path in real time, and compensate the accuracy ≤±0.005mm.

[0026] The three-dimensional space deviation algorithm is optimized for the transparent characteristics of glass, including the following steps of layered contour extraction: Coarse matching positioning: scan the laser drilling area image with a 3px step, and extract the approximate area of the hole contour through the laser spot gradient operator: (4) Fine matching optimization: sub-pixel level positioning is performed in the coarse matching area with a step of 0.1px, and the formula X​亚 =X 粗 +Δx×(G 左 -G 右 ) / (G 左 +G 右 (5) Where Δx is the step size, G 左 G 右 The gradient values ​​of the left and right neighbors of the light spot are used to calculate the sub-pixel coordinates.

[0027] The laser parameter adaptive compensation module also includes a laser head remaining lifespan early warning submodule: via formula T 剩余 =(T 临界 -T ) / (k×T0)(6) Calculate the remaining number of triggers, where T 临界 Let T0 be the effective pulse width threshold for the laser, and let it be 0.7. When T 剩余 A laser head maintenance reminder will be triggered when the number of uses is ≤1000.

[0028] This control system is suitable for precision glass processing, especially for the automated production of laser drilling and chamfering composite processes. In actual production lines, the equipment needs to meet the following requirements: Multi-mode collaborative control module: It adopts an industrial-grade PLC (model: Siemens S7-1511T) as the main controller, integrates 3 laser source interfaces (adapted to IPG YLR-100 laser generator), 8-axis servo drive (Panasonic A6 series, control accuracy ±0.001mm), and supports "laser drilling + chamfering" composite mode (switching response time ≤200ms), pure drilling mode (single hole processing cycle ≤0.5s) and pure transmission mode (no-load moving speed 500mm / s).

[0029] Laser vision-axis control closed-loop compensation module: Equipped with a laser coaxial vision system (Basler acA2500-14gm camera, 5 megapixels, 60fps), 16mm lens focal length, and a 650nm narrowband filter (transmittance ≥90%), and communicates with the PLC in real time via a gigabit Ethernet port.

[0030] Laser parameter adaptive compensation module: integrates a laser power meter (Ophir Nova II, measurement range 1-100W, accuracy ±1%) to acquire laser output power in real time; the storage module uses a 256GB industrial SD card to record the cumulative number of laser head triggers (resolution 1) and aging coefficient (initial value 1.0, increasing by 0.02 every 100,000 triggers).

[0031] Intelligent position compensation decision module: equipped with an edge computing unit (NVIDIA Jetson Nano), pre-stored historical deviation data of 5 position compensation stations (sampling period 10 ms, storage depth 100,000), and data interaction with the main controller through Ethernet.

[0032] High dynamic switching: the same production line needs to be compatible with "punching + chamfering" compound processing (such as smart watch cover plate), pure punching (such as camera module), and pure transmission (such as glass substrate handling) three modes.

[0033] As shown in Figure 2 , the laser is the core component of the puncher, and the imported laser used in the equipment ensures the processing quality of the laser, thereby accurately punching out the pattern.

[0034] Camera positioning platform: Drive the camera to move to the corresponding positioning position according to the specifications of different products to realize the CCD positioning function of the product.

[0035] Example 2 The method of the glass punching and chamfering machine control system comprises the following steps: S1: load shaft control configuration file (including shaft card and laser source mapping table: shaft 1→laser source 1, shaft 2→laser source 2), call scheduling algorithm to calculate station priority, laser trigger delay≤10ms; S2: vision-shaft control closed loop compensation: collect laser punching area image to obtain vision coordinates (X 视 , Y 视 ), record laser focal point coordinates Z 激 , Y 激 , Z 工 , substitute into the three-dimensional space deviation formula to calculate Δ, if Δ>0.005mm, then execute path correction; S3: laser parameter compensation execution: read the cumulative trigger number N of the laser head, adjust the processing parameters through the pulse width correction formula T =T0×(1+k×N), and simultaneously monitor the laser power fluctuation ΔP, when |ΔP|>3%×P0, P0 is the reference power, trigger secondary compensation; S4: intelligent position compensation processing: when a station fault is detected, call the position compensation path generation algorithm, and select the nearest position compensation station according to the Euclidean distance.

[0036] The laser power monitoring in step S3 uses a photodiode to collect laser energy in real time, and calculates the average power through the formula: I(t) is the instantaneous light intensity, and when the power fluctuation exceeds the threshold, the parameter compensation is triggered.

[0037] The image acquisition unit of the laser vision-axis control closed-loop compensation module adopts a laser coaxial vision system with a resolution of ≥ 50 million pixels, is equipped with a narrowband filter to filter stray light, and has a frame rate of ≥ 60 fps to ensure image capture at the moment of laser pulse triggering.

[0038] As shown in Figure 4 , the camera parameters can adjust the imaging effect of the current vision. By adjusting 'exposure' and 'gain', the brightness of the interface can be adjusted. The higher the 'exposure' setting, the brighter the real-time picture, and vice versa. The higher the 'gain' setting, the stronger the brightness enhancement of the image. Excessive exposure setting will result in long shooting time, so exposure and gain need to be used in coordination. After modification, click Save Parameters.

[0039] As shown in Figure 5 , the recommended imaging has high contrast between the inside and outside of the hole, and no trace interference around the hole.

[0040] Example calculation: Laser processing resource scheduling algorithm: Station state scanning (S11): Real-time monitoring of the state of 8 stations (0 = idle, 1 = processing, 2 = fault), updating the standby queue every 100 ms. For example, when station 3 (fault), station 5 (processing), and station 7 (idle), the queue only contains station 7.

[0041] Comprehensive priority calculation (S12): Inherent priority (weight 0.4): punching station (priority 3) > chamfering station (priority 2) > standby station (priority 1); Laser energy utilization rate (weight 0.3): calculation formula = (actual processing time / total occupancy time) x 100%, for example, station 7 utilization rate 85%; Historical usage frequency (weight 0.3): number of triggers in the last 1 hour / total number of stations, for example, station 7 triggers 120 times / total 800 times = 15%; Comprehensive priority = 3 x 0.4 + 85% x 0.3 + 15% x 0.3 = 1.2 + 0.255 + 0.045 = 1.5 (the higher the value, the higher the priority).

[0042] Resource allocation and conflict resolution (S13): If laser source 1 is idle, it is directly allocated to station 7 and marked as "processing"; If laser source 1 is occupied by station 2 with a priority of 1.8, station 7 enters the waiting queue, and the priority is recalculated every 500 ms.

[0043] Resource release mechanism (S14): When there are no processing requests for station 7 for 5 consecutive cycles (500ms) and the laser energy utilization rate is <10%, the PLC sends a "resource release" command, the laser source 1 goes into sleep mode (power drops to standby value 5W), and the axis control resource is marked as "to be allocated".

[0044] Pulse width correction formula: Parameter values ​​and examples: Reference parameters: T0 = 20 μs (reference pulse width), k = 1.2 × 10 -6 (Laser head aging coefficient, measured through multiple triggering experiments), N=30000 (cumulative number of triggers).

[0045] Calculation results: T = 20 × (1 + 1.2 × 10) -6 (×30000) = 20 × 1.036 = 20.72 μs T 临界 =20×0.7=14μs T 剩余 =(14-20.72) / (1.2×10 -6 ×20)=(-6.72) / (2.4×10 -5 )≈-280000 (At this time, T>T) 临界 (Requires immediate maintenance).

[0046] Vision-axis control closed-loop compensation: In laser drilling, a vision-axis control closed-loop compensation system corrects processing deviations caused by workpiece clamping errors and laser focus drift. Equipment configuration: industrial camera (resolution 2592×1944), laser displacement sensor (Z-axis accuracy ±2μm); processing parameters: laser wavelength 1064nm, drilling diameter 0.1mm, cutting speed 65mm / s.

[0047] Coordinate acquisition Visual coordinates (X) 视 Y 视 The image of the punched area was captured by an industrial camera, and the sub-pixel positioning algorithm was used to calculate (12.3456mm, 8.7654mm). Laser focal point coordinates Zfocal: 5.6789 mm as measured by the laser displacement sensor; Workpiece coordinates (X) 激 Y 激 Z 工 : Preset machining coordinates for the axis control system (12.3500mm, 8.7700mm, 5.6800mm).

[0048] Substituting into the three-dimensional spatial deviation formula, Δ = 0.0065 mm is calculated; Correcting Decisions and Outcomes Deviation judgment: Δ=0.0065mm>0.005mm (threshold), trigger path correction; Axis control correction: Drive X-axis compensation +0.0044mm, Y-axis compensation +0.0046mm, Z-axis compensation +0.0011mm; Corrected deviation: The secondary acquisition calculation yielded Δ=0.0028mm, which meets the accuracy requirement of ±0.005mm.

[0049] Fill-in path: The replacement function is used when a workstation is damaged and cannot be replaced in time. This function can be activated to use other workstations to temporarily replace the damaged workstation.

[0050] like Figure 3 As shown, for example, if the replacement function is enabled at workstation 1, the replacement function is set to 2. This indicates that workstation 1 is damaged and workstation 2 needs to be used to replace workstation 1. Workstation 1 can be set with a separate photo-taking position and fixed compensation.

[0051] For example, in an automotive glass production line (8 stations), station 3 (coordinates X=500mm, Y=300mm) is triggered by an emergency stop due to a servo motor failure. The intelligent compensation decision module needs to be called to perform compensation. The processing task is to drill holes in a 3mm thick windshield (hole diameter φ5mm, chamfer angle 45°).

[0052] Step 1: Collect Deviation Data at Faulty Workstations Fault detection: The PLC monitors the status word of station 3 (0x0002 → fault code) in real time through Profinet and triggers the compensation process (response delay ≤100ms).

[0053] Deviation data extraction: The vision system was used to collect the deviations (unit: mm) of the last 10 processing operations before the failure. [0.003, 0.002, -0.001, 0.004, 0.002, -0.002, 0.003, 0.001, -0.003, 0.002] Calculate the mean deviation of the faulty workstation: δ 故障 = 100.003+0.002+....+0.002 / 10=0.0015mm; Step 2: Calculation of the historical average deviation of the replacement workstation Candidate replacement workstations selection: According to the production process plan, workstation 2 (X=502mm, Y=301mm) and workstation 4 (X=498mm, Y=299mm) are backup replacement workstations.

[0054] Applications of the sliding window algorithm: Glass thickness 3mm→window size N=300 times, call the station 2 nearest 300 times processing deviation data, calculate the historical deviation mean:

[0055] (Bias slightly larger than station 2).

[0056] Step 3: Dynamic correction coefficient α calculation According to the current laser power and historical deviation variance dynamic adjustment compensation precision: Input parameters: The reference power P0=30W, the current monitoring power P=29.1W (fluctuation 3%, trigger secondary compensation); Station 2 historical deviation variance σ 2 =8.5×10 -8 (Data from SD card storage for nearly 24 hours record).

[0057] Correction coefficient formula:

[0058] Through the regression analysis of historical compensation data, 0.6 represents the reference response sensitivity of the compensation system. When the compensation station deviation variance σ² tends to infinity (extreme instability) and the power P=P0, α=0.6, ensure that the basic compensation amount is not less than the process safety threshold.

[0059] Substitute data: α=0.6+(29.130)×(8.5×10 -8 +11)≈0.6+1.03×1≈1.63 Step 4: Compensation path generation and distance sorting Euclidean distance calculation: Station 2 and fault station distance:

[0060] Station 4 and fault station distance:

[0061] When the distance is equal, the default is to select the smaller station 2 Compensation path compensation amount: Compensation amount=

[0062] Generate compensation path coordinates: (X 补位 ,Y 补位 )=(501.99674mm,300.99674mm) Step 5: Compensation execution and result verification Path issuing: the PLC issues the compensation coordinate to the station 2 axis control system through EtherCAT, and the servo motor moves according to the S-shaped acceleration and deceleration curve (the maximum speed is 500 mm / s, and the jerk acceleration is 1000 mm / s²).

[0063] Processing result: the compensation position consumes 1.8 s (≤2 s), the visual system detects the hole position deviation after compensation = 0.003 mm (≤±0.005 mm), and the chamfer angle error is ≤0.5°, which meets the quality requirements.

[0064] The above merely describes the preferred embodiments of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can make equivalent replacements or changes according to the technical scheme and the inventive concept of the present application within the technical range disclosed by the present application, which should be covered in the protection scope of the present application.

Claims

1. A control system for a glass drilling and beveling machine, characterized in that, Comprise: Multi-mode cooperative control module: including glass laser drilling + chamfering composite mode, pure laser drilling mode, pure transmission mode, the above mode is dynamically switched, the shaft control and laser energy resources are allocated through the laser processing resource scheduling algorithm, the idle shaft resource is released; Laser vision-shaft control closed loop compensation module: integrated with industrial camera and three-dimensional space deviation algorithm, through collecting laser drilling area image, calculating the space deviation of visual coordinate, laser focal point coordinate and workpiece coordinate, driving shaft control system to correct path in real time; Laser parameter adaptive compensation module: based on laser energy attenuation model, dynamically adjust processing parameters through pulse width correction formula T =T0×(1+k×N), T0 is the baseline pulse width, k is the laser head aging coefficient, N is the cumulative laser trigger number; Laser processing intelligent compensation decision module: when the laser drilling station fails, generate compensation path according to the deviation of the fault station, the historical deviation mean of the compensation station and the dynamic correction coefficient, and select the optimal compensation station through distance sorting algorithm.

2. The glass hole site chamfering machine control system of claim 1, wherein, The laser processing resource scheduling algorithm specifically comprises the following steps: S11: real-time scanning of each station state, collecting inherent priority, laser energy utilization rate and historical use frequency parameters of the station to be scheduled, and generating a to-be-scheduled queue; S12: based on the inherent priority, laser energy utilization rate, historical use frequency and corresponding weight coefficient of each station, the comprehensive priority of each station is calculated, and the priority queue is formed in descending order of comprehensive priority; S13: allocate laser source and shaft control resources according to priority order: if the resources are sufficient, directly allocate and mark the station as "processing"; if the resources conflict, determine the conflict priority by weighted calculation of laser energy utilization rate and historical use frequency, allocate high priority station, and low priority station enters the waiting queue; S14: real-time monitoring of the utilization rate of the allocated resources: when the laser energy utilization rate is lower than the set value and there is no processing request for a certain period of time, trigger resource release, including laser source hibernation and shaft control resource marking as "to be allocated"; S15: after completing a certain processing task, recalibrate the historical use frequency of the station and each weight coefficient to ensure that the algorithm adapts to the change of the processing scene.

3. The glass drilling and chamfering machine control system according to claim 1, characterized in that the formula is used. The three-dimensional space deviation algorithm is optimized for the transparency of glass, including the following steps of hierarchical contour extraction: （3） Calculate the spatial deviation of visual coordinates (X 视 ,Y 视 ), laser focus coordinates Z 焦 and workpiece coordinates (X 激 ,Y 激 ,Z 工 ), and drive shaft control system to correct the path in real time.

4. The glass hole deburring machine control system of claim 1, wherein, Coarse matching positioning: scanning the laser drilling area image with a 3px step, and using a laser spot gradient operator: Fine matching optimization: sub-pixel level positioning is performed in the coarse matching area with a 0.1px step, and the formula is used. （4） wherein represents a rate of change of the gray scale in the x direction (horizontal direction), represents a rate of change of the gray scale in the y direction (vertical direction), and a large region of the hole contour is extracted; The laser parameter adaptive compensation module further comprises a laser head remaining life warning sub-module: the formula is used. X 亚 =X 粗 +Δx×(G 左 -G 右 ) / (G 左 +G 右 ) (5) where Δx is the step size, G 左 , G 右 are the left and right neighborhood gradient values, respectively, of the light spot.

5. The glass hole site chamfering machine control system of claim 1, wherein, The intelligent compensation decision module uses a sliding window algorithm to calculate the historical deviation mean: the window size N is configurable, and the recent N times of processing deviation data are processed by arithmetic mean. T 剩余 = (T 临界 -T ) / (k×T0) (6) The remaining trigger times are calculated, wherein, T 临界 is the laser effective pulse width threshold, take T0×0.7, when T 剩余 The laser head maintenance reminder is triggered when T ≤1000 times.

6. The glass hole site chamfering machine control system of claim 1, wherein, The method comprises the following steps:

7. A method of controlling a glass hole chamfering machine according to any one of claims 1-6, wherein, S1: load the laser processing shaft control configuration file, allocate resources through the shaft-laser source mapping algorithm, call the laser processing resource scheduling algorithm to calculate the priority and solve the conflict; ​ S2: vision-axis control closed loop compensation: collect laser drilling area image to obtain vision coordinates (X 视 , Y 视 ), record laser focus coordinates Zfocusand workpiece coordinates (X 激 , Y 激 , Z 工 ), substitute into three-dimensional space deviation formula to calculate Δ, if Δ>0.005mm, then execute correction path; S3: laser parameter compensation execution: read the cumulative number of laser head triggers N, adjust the processing parameters through the pulse width correction formula T = T0×(1+k×N), and monitor the laser power fluctuation ΔP synchronously. When |ΔP|>3%×P0, P0 is the reference power, trigger secondary compensation; S4: intelligent compensation processing: when detecting a work station failure, call the compensation path generation algorithm, and select the nearest compensation work station according to the Euclidean distance sorting.

8. The method of claim 7, wherein, The laser power monitoring in step S3 uses a photodiode to collect laser energy in real time, and calculates the average power through the formula: I(t) is the instantaneous light intensity, and parameter compensation is triggered when the power fluctuation exceeds the threshold.

9. The method of claim 7, wherein, The distance sorting algorithm in step S4 calculates the Euclidean distance based on the X / Y axis coordinates of the stations , and selects the station with the smallest D to perform the processing.

10. The method of claim 7, wherein, The image acquisition unit of the laser vision-axis control closed-loop compensation module adopts a laser coaxial vision system, with a resolution of ≥5 million pixels, equipped with a narrowband filter to filter stray light, and a frame rate of ≥60fps to ensure image capture at the moment of laser pulse triggering.

Images