A method for adaptive precision machining of small curvature structures

Abstract

The present invention provides a method for adaptive precision machining of small-curvature structures. The specific steps are as follows: the small-curvature structure to be machined is placed on a platform, and a servo driver controls the platform to move along a machining trajectory; combining platform acceleration waveform monitoring and visual topography detection, the standard deviation of the platform acceleration fluctuation, the maximum position deviation of the machining trajectory, and the percentage of geometric similarity between the preset path and the actual machining path are determined; a determination is made as to whether a target value is met. If not, at least one of the smoothing instruction filter coefficient, the maximum adaptive gain, and the adaptive gain scaling factor in the servo driver's built-in parameters is dynamically adjusted based on the comparison result; the structure is re-machined using the adjusted servo driver's built-in parameters until the target value is met; when the target value is met, a small-curvature structure with a smooth contour is obtained. The present invention has the advantages of good process stability, adaptive servo parameter correction, and high efficiency.

Description

Technical Field

[0001] The present invention relates to the field of laser processing, and in particular to a method for adaptively and precisely processing a small-curvature structure. Background Art

[0002] Quartz devices are currently widely used in aerospace, inertial navigation systems, precision instrumentation, and high-end semiconductor equipment. Their core advantages stem from the unique properties of quartz glass. For example, high purity and low impurities ensure stable device operation in extreme environments, reducing performance degradation caused by material contamination. Excellent optical and thermal stability meet the stringent requirements of high-precision photolithography, laser processing, and other processes for material transmittance and thermal expansion coefficient. High insulation ensures effective reduction of external electromagnetic interference during use. However, quartz glass is a typical hard and brittle material, and stress concentration during processing can easily lead to microcracks or chipping. Therefore, high-quality, high-precision processing of small-curvature structures on hard and brittle materials is difficult.

[0003] Currently, the processing of small-curvature structures of hard and brittle materials mainly relies on the following three methods: mechanical grinding, chemical etching, and laser modification cutting-chemical etching. The following are the existing defects:

[0004] First of all, for the mechanical grinding method, this method uses a grinding wheel to grind small curvature structures, and requires the use of a grinding wheel former to adjust the tool path, but there are often problems such as difficulty in ensuring dimensional consistency, low processing efficiency, and easy formation of micro-damage on the edge of the material.

[0005] Secondly, as for the chemical etching method, this method uses hydrofluoric acid solution to selectively etch the pre-masked small curvature structure area, but it often has problems such as long process time and difficulty in controlling the transition smoothness of complex geometric shapes.

[0006] Finally, as for the laser modification cutting-chemical etching method, this method uses an ultrafast laser (such as a picosecond laser or a femtosecond laser) to penetrate the entire piece of quartz glass to form a modified area, and combines it with wet etching technology for selective etching and forming. It has the advantages of non-contact processing, high precision, and high repeatability. It is suitable for the processing of small curvature structures. However, this method currently has certain technical bottlenecks. For example, the trajectory deviation of the laser processing platform during X / Y axis linkage (such as servo lag and mechanical return clearance) will be amplified as the radius of the small curvature structure decreases, causing the modified line to deviate from the preset path, the connection is not smooth, and "steps" or breakpoints are prone to appear in the transition zone. This problem will further make it difficult for the smoothness and consistency of the quartz hollow small curvature structure obtained after selective etching to meet the preset indicators.

[0007] Therefore, there is an urgent need for a method to correct the motion trajectory of the laser processing platform and improve the multi-axis linkage accuracy to improve the processing accuracy of small curvature structures. Summary of the Invention

[0008] In view of the shortcomings of the existing technology, the present invention provides a method for adaptively processing small curvature structures. During the laser precision processing, the servo waveform monitoring and visual morphology detection are combined to calculate the standard deviation of the platform acceleration fluctuation. , Maximum position deviation of machining trajectory , the percentage C of geometric similarity between the preset path and the actual processing path; dynamically optimize the servo drive parameters through the feedback control algorithm based on the data obtained from detection and calculation; perform multiple rounds of closed-loop iterative processing and detection based on the optimized servo parameters until the parameter indicators of the processing effect meet the target values, and finally obtain a small curvature structure with high smoothness.

[0009] The present invention achieves the above technical objectives through the following technical means.

[0010] A method for adaptively processing a structure with small curvature, the specific steps are as follows:

[0011] According to the outline drawing of the small curvature structure to be processed, the processing trajectory is obtained; the small curvature structure to be processed is placed on the platform, and the servo drive controls the platform to move along the processing trajectory;

[0012] Determine the target value of the standard deviation of the platform acceleration fluctuation, the target value of the maximum position deviation of the machining trajectory, and the target value of the percentage of geometric similarity between the preset path and the actual machining path;

[0013] In laser processing, after completing a processing trajectory, the standard deviation of platform acceleration fluctuation is determined by combining servo waveform monitoring, visual shape detection and acceleration sensor. , Maximum position deviation of machining trajectory , the geometric similarity percentage C between the preset path and the actual processing path;

[0014] Determine the standard deviation of platform acceleration fluctuation , Maximum position deviation of machining trajectory and whether the geometric shape similarity percentage C meets the target value. If not, dynamically adjust at least one of the smoothing instruction filter coefficient, maximum adaptive gain, and adaptive gain proportional factor in the servo driver built-in parameters according to the comparison result; reprocess using the adjusted servo driver built-in parameters until the target value is met;

[0015] When the target value is met, a small curvature structure with a smooth contour is obtained.

[0016] Furthermore, the built-in parameters of the servo driver are dynamically adjusted according to the judgment conditions of the maximum position deviation, the judgment conditions of the geometric shape similarity percentage, and the judgment conditions of the acceleration fluctuation standard deviation.

[0017] Furthermore, according to the judgment condition of the maximum position deviation, at least one of the smoothing instruction filter coefficient, the maximum adaptive gain and the adaptive gain proportional factor in the built-in parameters of the servo driver is dynamically adjusted, specifically:

[0018] Assume the target value of the maximum position deviation is A2;

[0019] when When >2 A2, the smoothing instruction filter coefficient is adjusted. The adjusted smoothing instruction filter coefficient is recorded as S new , S new =min(S+20%×K, ), S is the current smoothing instruction filter coefficient, is the current maximum adaptive gain; K is the current adaptive gain proportional factor; the adjusted smoothing instruction filter coefficient S new Satisfy the condition S new [20%, 95%];

[0020] when When , the adjusted adaptive gain proportional factor is recorded as K new , K new =1.1K, the adjusted smoothing instruction filter coefficient S new =S+5%; adjusted smoothing instruction filter coefficient S new Satisfy the condition S new [20%, 95%], the adjusted adaptive gain scaling factor K new The condition is K new [0.5, 1.8].

[0021] Furthermore, according to the judgment condition of the geometric shape similarity percentage, at least one of the smoothing instruction filter coefficient, the maximum adaptive gain, and the adaptive gain proportional factor in the built-in parameters of the servo driver is dynamically adjusted, specifically:

[0022] Set the target value of geometric shape similarity percentage to A3;

[0023] When C< A3-5%, the adjusted smoothing instruction filter coefficient S new =S+15%; the maximum adaptive gain after adjustment is recorded as , = max( ×0.9, 100%); adjusted smoothing instruction filter coefficient S new Satisfy the condition S new [20%, 95%], maximum adaptive gain after adjustment The conditions are met [80%,200%];

[0024] when When the adjusted adaptive gain proportional factor K new =1.1K, the adjusted smoothing instruction filter coefficient S new =S-5%; adjusted smoothing instruction filter coefficient S new Satisfy the condition S new [30%, 95%], the adjusted adaptive gain scaling factor K new The condition is K new [0.5, 1.8].

[0025] Furthermore, according to the judgment condition of the acceleration fluctuation standard deviation, at least one of the smoothing instruction filter coefficient, the maximum adaptive gain and the adaptive gain proportional factor in the built-in parameters of the servo driver is dynamically adjusted, specifically:

[0026] Assume that the target value of the acceleration fluctuation standard deviation is A1,

[0027] when When >1.6A1, the maximum adaptive gain after adjustment is recorded as = ×0.7, the adjusted adaptive gain scaling factor K new =0.8K, the adjusted smoothing instruction filter coefficient S new =S+10%; adjusted smoothing instruction filter coefficient S new Satisfy the condition S new [30%, 95%], the adjusted adaptive gain scaling factor K new The condition is K new [0.6, 1.8]; maximum adaptive gain after adjustment The conditions are met [80%, 200%];

[0028] when When the adjusted smoothing instruction filter coefficient S new =S+4%, the maximum adaptive gain after adjustment is recorded as = min( ×1.05, 150%), the smoothing instruction filter coefficient S after adjustment new Satisfy the condition S new [30%, 95%], maximum adaptive gain after adjustment The condition to be satisfied is [80%, 200%].

[0029] Furthermore, when two of the judgment conditions of the maximum position deviation, the geometric shape similarity percentage, and the standard deviation of acceleration fluctuation are simultaneously satisfied, the built-in parameters of the servo drive are adjusted according to the following priority order, and the priority order is:

[0030] The judgment condition of the maximum position deviation > the judgment condition of the geometric shape similarity > the judgment condition of the standard deviation of acceleration fluctuation.

[0031] Furthermore, after adjusting the built-in parameters of the servo drive according to the judgment condition of the maximum position deviation, if C < A3 - 5% simultaneously, then first increase the adaptive gain ratio factor after adjusting the judgment condition of the maximum position deviation by 5%, and then adjust the built-in parameters of the servo drive according to the judgment condition of the geometric shape similarity percentage.

[0032] Furthermore, after adjusting the built-in parameters of the servo drive according to the judgment condition of the geometric shape similarity, if simultaneously < 1.2A1, then first increase the adaptive gain ratio factor after adjusting the judgment condition of the geometric shape similarity by 8%, and then adjust the built-in parameters of the servo drive according to the judgment condition of the standard deviation of acceleration fluctuation.

[0033] Furthermore, after adjusting the built-in parameters of the servo drive according to the judgment condition of the standard deviation of acceleration fluctuation, if simultaneously 2 A2, then limit the platform acceleration to 90% of the set value.

[0034] The beneficial effects of the present invention are as follows:

[0035] 1. The method for adaptively machining a small-curvature structure according to the present invention combines servo drive platform acceleration waveform monitoring and visual topography detection during the laser precision machining process, calculates the standard deviation of platform acceleration fluctuation 、the maximum position deviation of the machining trajectory 、the geometric shape similarity percentage C between the preset path and the actual machining path; according to the detected and calculated data, dynamically optimize the parameters of the servo drive through a feedback control algorithm; based on the optimized servo parameters, perform multiple rounds of closed-loop iterative machining and detection until the parameter indicators of the machining effect meet the target value, and finally obtain a small-curvature structure with high smoothness. The small-curvature structure machined by the machining method of the present invention has high geometric accuracy and good process stability.

[0036] 2. The method for adaptively processing small-curvature structures described in this invention can be applied to processing small-curvature structures in a variety of hard and brittle materials, such as borosilicate glass, aluminosilicate glass, quartz glass, and sapphire. It is compatible with a variety of lasers that can be used to process hard and brittle materials and has a wide range of applications.

[0037] 3. The adaptive machining method for small curvature structures described in this invention uses adaptive parameter correction to significantly reduce manual rework, resulting in high machining efficiency. Furthermore, after parameter correction, iterative machining and closed-loop quality verification can be automatically performed, making the operation simple.

[0038] 4. The method for adaptively machining small curvature structures described in this invention breaks through the traditional single feedback mode, fuses servo waveform signals with visual topography data, and implements multi-dimensional parameter adaptive compensation based on a feedback control algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. The drawings described below are some embodiments of the present invention. For ordinary technicians in this field, it is obvious that other drawings can be obtained based on these drawings without paying any creative work.

[0040] Figure 1 This is a flow chart of the method for adaptively processing small curvature structures according to the present invention.

[0041] Figure 2 Schematic diagram of the ultrafast shaping laser processing platform of the present invention.

[0042] Figure 3 A trajectory diagram is preset for the small curvature structure to be processed in the present invention.

[0043] Figure 4 This is a diagram showing the improvement effect of the small curvature structure trajectory line obtained by microscope characterization before the adaptive optimization of the servo parameters of the present invention.

[0044] Figure 5 This is a diagram showing the improvement effect of the small curvature structure trajectory line obtained by characterizing under a microscope after the servo parameters are adaptively optimized in the present invention.

[0045] Figure 6 This is a diagram showing the etching effect of a small-curvature structure obtained under a microscope after the quartz glass obtained by modification and processing before parameter optimization of the present invention is etched with a chemical etchant.

[0046] Figure 7 This is a diagram showing the etching effect of a small-curvature structure obtained under a microscope after etching the quartz glass obtained by the modified processing after parameter optimization of the present invention with a chemical etchant.

[0047] In the picture:

[0048] 1-Ultrafast picosecond laser system; 2-Beam expander; 3-Reflecting mirror; 4-Axis lens; 5-Plano-convex lens; 6-Microscope objective; 7-Sample to be processed; 8-High-precision displacement platform; 9-CMOS camera and LED light source; 10-Computer control system. DETAILED DESCRIPTION

[0049] The following describes embodiments of the present invention in detail, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and are not to be construed as limiting the present invention.

[0050] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "axial", "radial", "vertical", "horizontal", "inside", "outside" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first" and "second" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0051] In the present invention, unless otherwise expressly specified or limited, the terms "mounted," "connected," "connect," "fixed," etc. should be understood broadly. For example, they may refer to fixed connection, detachable connection, or integral connection; mechanical connection or electrical connection; direct connection or indirect connection through an intermediate medium; or internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0052] like Figure 1 As shown, the method for adaptively processing a small curvature structure according to the present invention comprises the following specific steps:

[0053] S01: A 4f Bessel beam reduction system is constructed based on the sample thickness of the transparent, hard, and brittle material, ensuring that the focal depth of the focused beam is greater than the glass thickness. The reduction system, consisting of a plano-convex lens and a 10x microscope objective, is used to control the focal depth and increase the energy density. A computer control system controls the optical modulator to adjust the number of output bursts (i.e., the number of sub-pulses within the pulse envelope), the output frequency, the pulse energy, and the spatial spacing between pulses. The z-axis position is also set to ensure that the Bessel beam focal depth extends through the upper and lower surfaces of the sample. Based on the outline of the small-curvature structure to be machined, the platform movement speed and acceleration parameters are set according to a preset trajectory. The computer control system then controls the multi-axis linkage based on the preset path to obtain the machining trajectory for the small-curvature structure. Small curvature refers to a radius of curvature of less than 1 mm.

[0054] S02: Accelerometers are used to collect acceleration fluctuations on the platform where the workpiece to be processed is placed. By collecting acceleration sensor signals in real time, an integrated controller monitors the platform's acceleration fluctuations during the processing phase and calculates the standard deviation of the acceleration fluctuations to ensure the stability of the multi-axis linkage platform's movement. Furthermore, an ultra-high-definition camera and coaxial illumination light source are integrated to capture microscopic images (resolution 1µm / pixel) of the modified lines of the small curvature structure after processing. The processing contour is extracted using an edge detection algorithm, and the maximum position deviation is calculated by comparing it with a preset theoretical path. The geometric shape similarity (i.e., geometric shape percentage) is then calculated using a shape context matching algorithm. The processing platform is driven by a servo drive to move along the small curvature structure processing trajectory. Both the shape context matching algorithm and the edge detection algorithm are currently available.

[0055] S03: Using the acceleration fluctuation standard deviation, maximum position deviation, and geometric shape similarity percentage, dynamically adjust the servo driver's built-in parameters (smoothing instruction filter coefficient, maximum adaptive gain, adaptive gain proportional factor) to correct the trajectory deviation during the actual machining process. After overwriting the original parameters with the corrected servo control parameters, a new round of "machining-detection-analysis-compensation" full closed-loop control process is carried out until the acceleration fluctuation standard deviation, the maximum position deviation of the machining trajectory, and the geometric shape similarity percentage meet the predetermined requirements, as follows:

[0056] Assume that the target value of the acceleration fluctuation standard deviation is A1, the target value of the maximum position deviation is A2, and the target value of the geometric shape similarity percentage is A3; the calculated acceleration fluctuation standard deviation is recorded as , the calculated maximum position deviation , the calculated geometric shape similarity percentage is recorded as C;

[0057] when When >2 A2, the smoothing instruction filter coefficient is adjusted. The adjusted smoothing instruction filter coefficient is recorded as S new , S new=min(S+20%×K, ), S is the current smoothing instruction filter coefficient, is the current maximum adaptive gain; K is the current adaptive gain proportional factor; the adjusted smoothing instruction filter coefficient S new The condition is S new [20%, 95%];

[0058] when When , the adjusted adaptive gain proportional factor is recorded as K new , K new =1.1K, the adjusted smoothing instruction filter coefficient S new =S+5%; adjusted smoothing instruction filter coefficient S new The condition is S new [20%, 95%], the adjusted adaptive gain scaling factor K new The condition is K new [0.5, 1.8];

[0059] When C< A3-5%, the adjusted smoothing instruction filter coefficient S new =S+15%; the maximum adaptive gain after adjustment is recorded as , = max( ×0.9, 100%); adjusted smoothing instruction filter coefficient S new The condition is S new [20%, 95%], maximum adaptive gain after adjustment The conditions are met [80%,200%];

[0060] when When the adjusted adaptive gain proportional factor K new =1.1K, the adjusted smoothing instruction filter coefficient S new =S-5%; adjusted smoothing instruction filter coefficient S new The condition is S new [30%, 95%], the adjusted adaptive gain scaling factor K new The condition is K new [0.5, 1.8];

[0061] when When >1.6A1, the maximum adaptive gain after adjustment is recorded as = ×0.7, the adjusted adaptive gain scaling factor K new =0.8K, the adjusted smoothing instruction filter coefficient S new =S+10%; adjusted smoothing instruction filter coefficient S new The condition is S new [30%, 95%], the adjusted adaptive gain scaling factor K new The condition is K new [0.6, 1.8]; maximum adaptive gain after adjustment The conditions are met [80%, 200%];

[0062] when When the adjusted smoothing instruction filter coefficient S new =S+4%, the maximum adaptive gain after adjustment is recorded as = min( ×1.05, 150%), the smoothing instruction filter coefficient S after adjustment new The condition is S new [30%, 95%], maximum adaptive gain after adjustment The conditions are met [80%,200%];

[0063] If the smoothing instruction filter coefficient, maximum adaptive gain, and adaptive gain proportional factor exceed the boundary limit when calculated separately, the parameter value exceeding the boundary limit will be corrected to the corresponding exceeded boundary value; assuming that the maximum adaptive gain after adjustment is 75%, which exceeds the minimum boundary of 80%, then the maximum adaptive gain after adjustment Change to 80%.

[0064] When multiple error conditions are triggered simultaneously, parameter correction is performed in the order of maximum position deviation > geometric shape similarity > acceleration fluctuation standard deviation. That is, the servo driver's built-in parameters are first adjusted based on the maximum position deviation judgment condition, then the geometric shape similarity judgment condition is added and adjusted, and finally the acceleration fluctuation standard deviation judgment condition is added and adjusted.

[0065] The rules for linkage adjustment are also designed, specifically:

[0066] After adjusting the built-in parameters of the servo driver according to the judgment conditions of the maximum position deviation, if C< A3-5% at the same time, the adjusted adaptive gain proportional factor K new=1.05K, where K is the adaptive gain proportional factor adjusted according to the judgment condition of the maximum position deviation;

[0067] After adjusting the built-in parameters of the servo driver according to the judgment conditions of geometric shape similarity, if <1.2A1, the adjusted adaptive gain proportional factor K new =1.08K, where K is the adaptive gain proportional factor adjusted according to the judgment condition of geometric shape similarity;

[0068] After adjusting the built-in parameters of the servo driver according to the judgment conditions of the acceleration fluctuation standard deviation, if 2 A2, the platform acceleration is limited to 90% of the set value.

[0069] S04: After the closed-loop iterative processing is completed, the small curvature structure of the target sample is processed based on the final optimized servo drive parameters; for the laser precision processing method of direct ablation and vaporization, it is necessary to fill the path according to actual needs based on the preset processing trajectory, or repeat the processing times multiple times; for the laser modification supplemented by wet etching processing method, the modified sample needs to be placed in a chemical etching solution for selective etching, and finally achieve high-smoothness small curvature structure processing.

[0070] Example 1

[0071] The quartz small curvature structure is processed by picosecond laser modification combined with chemical etching, which specifically includes the following steps:

[0072] S01: The laser wavelength of the picosecond laser system is 1064 nm, the pulse width is 10 ps, ​​the repetition rate is set to 100 kHz, the burst number is adjusted to 3 by an optical modulator, the time interval between each burst is 25 ns, the single pulse energy is set to 250 µJ, the laser processing mode is PSO mode, and the pulse step interval is 2.5 µm. The Gaussian beam is expanded by a beam expander with a beam waist radius of 3 mm. The propagation direction is then changed by a reflector and shaped into a Bessel beam by a conical lens with a cone bottom angle of 5°. Finally, the beam radius is reduced by a 4f beam reduction system to increase the energy density, which can achieve the modification effect on quartz glass. The 4f system consists of a plano-convex lens with a focal length of 50 mm and a microscope objective with a focal length of 9 mm. The calculated Bessel beam spot diameter is 1.75 µm and the diffraction-free length is 2.46 mm, which meets the modification requirements of the sample to be processed. The CMOS camera (Baslerace 2) The LED light source (wavelength 630nm) is coaxial with the Bessel optical system to capture the modified morphology. Figure 2 shown.

[0073] like Figure 2 As shown, the picosecond laser processing device includes an ultrafast picosecond laser system 1, a beam expander 2, a reflector 3, an aconic lens 4, a plano-convex lens 5, a 10× microscope objective 6, a CMOS camera and an LED light source 9. The sample to be processed 7 is placed at the center of a high-precision displacement platform 8, the acceleration sensor is placed inside the platform, and the acquisition module is connected to the accelerometer and a computer control system 10. After the ultrafast laser system generates Gaussian light, the laser output power is controlled by an internal attenuator, the computer control system controls the Q switch in the laser system to adjust the repetition frequency, and the optical modulator is used to control the number of pulses in the Burst mode. The beam spot radius is enlarged by the beam expander and shaped into a Bessel beam after passing through the aconic lens. The adjusted laser beam is focused on the sample located on the high-precision translation stage through the plano-convex lens and the microscope objective.

[0074] The transparent, hard, and brittle material used in the experiment was quartz glass (purity ≥99.99%). The glass sample was 1 mm thick and had a refractive index of 1.51. The glass sample was fixed on a high-precision translation stage. A machining path was generated based on the shape of the small-curvature structure to be modified. The small-curvature structure to be tested consisted of a main arc and two arc transition lines, with corresponding curvature radii of 270µm, 680µm, and 1060µm, respectively. The three arcs were based on different center positions to achieve a smooth transition in design. The preset trajectory diagram is shown in the attached figure. Figure 3 The multi-axis linkage platform was initially set to a speed of 10 mm / s and an acceleration of 20 mm / s². A trial process was then performed on the laser processing platform along a preset path. The Bessel beam penetrated the upper and lower surfaces of the glass, forming a continuous modified region. During this process, the area directly affected by the Bessel beam transformed from the original six-membered silicon-oxygen ring structure to three-membered and four-membered silicon-oxygen ring structures.

[0075] S02: From the start to the end of processing, the integrated DAQ module collects real-time signals from MEMS accelerometers (Kistler 8704B) placed at key positions on the platform (sampling rate of 10 kHz). The acceleration fluctuations during the start and stop phases of sample processing are detected and transmitted to the computer control system to calculate the standard deviation of the acceleration fluctuations during the first round of processing. =0.064g, where g represents the acceleration due to gravity, 1g≈9.8 At the same time, after the processing is completed, the LED light source is kept on, and the CMOS camera (resolution 1µm / pixel) will automatically perform visual morphology detection of the modified trajectory. After taking a microscopic image of the small curvature structure, the Canny edge detection algorithm will be used to extract the modified line contour and calculate the maximum position deviation. 7.2µm, based on the actual path and the preset path, the geometric shape similarity percentage C=87.1% is calculated using the shape context matching algorithm.

[0076] S03: Target value of acceleration fluctuation standard deviation 0.05g, target value of maximum position deviation 3µm, target value for geometric similarity percentage 95%; for the convenience of calculation, the target of the acceleration fluctuation standard deviation is set to A1 0.05g, target value of maximum position deviation A2 3µm, target value of geometric similarity percentage A3 95%;

[0077] Using the acceleration fluctuation standard deviation, maximum position deviation and geometric shape similarity percentage, the built-in parameters of the servo drive (smoothing instruction filter coefficient, maximum adaptive gain, adaptive gain proportional factor) are dynamically adjusted to correct the trajectory deviation during the actual processing;

[0078] In the embodiment, the acceleration fluctuation standard deviation, maximum position deviation, and geometric shape similarity percentage did not meet the target. Since multiple error conditions were triggered, the parameters were corrected in the order of maximum position deviation > geometric shape similarity > acceleration fluctuation standard deviation, as shown in Table 1.

[0079] The initial smoothing command filter coefficient S of the platform's servo drive is 50%, the initial adaptive gain proportional factor K is 0.8, and the initial maximum adaptive gain =120%;

[0080] During the first round of optimization:

[0081] Due to the maximum position deviation 7.2µm>6µm, then the adjusted smoothing instruction filter coefficient =min(50%+20%×0.8, )=66%, the adjusted smoothing instruction filter coefficient S new The condition is S new [20%, 95%]; After adjusting the built-in parameters of the servo driver according to the judgment conditions of the maximum position deviation, since C = 87.1% <90% at the same time, the adjusted adaptive gain proportional factor K new =1.05K=0.8×105%=0.84;

[0082] Since the geometric shape similarity percentage C = 87.1% < 90%, the adjusted smoothing instruction filter coefficient S new =S+15%= 66% + 15%=81%; the maximum adaptive gain after adjustment =max( ×0.9, 100%)= max( ×0.9, 100%)=108%;

[0083] Since the standard deviation of acceleration fluctuation is 0.05g ≤ 0.08g, then the adjusted smoothing instruction filter coefficient S new =S+4%= 81%+ 4%=85%, the maximum adaptive gain after adjustment is recorded as = min( ×1.05, 150%)=min(108%×1.05, 150%)=113.4%, so the result of this round of servo parameter optimization is: S new =85%, K new =0.84, =113.4%, the servo parameters are within the boundary limits.

[0084] Table 1

[0085]

[0086] S04: The servo drive of the platform is based on S new =85%, K new =0.84, =113.4% to carry out the next round of processing, effect detection, analysis and servo parameter correction of the modified line until the optimized processing trajectory parameters reach the predetermined indicators; after four rounds of iteration, the standard deviation of acceleration fluctuation is obtained. =0.045g, maximum position deviation The processing effect of 2.7µm and geometric similarity percentage C=95.4% is achieved. The corresponding servo drive parameters are S=90% and K=1.4640. =107.16%. The comparison of the trajectory effect of the small curvature structure before and after the servo parameter adaptive optimization is shown in the figure below. Figure 4 and Figure 5 shown.

[0087] The quartz glass that has not been modified by adaptive parameter correction and the quartz glass that has been modified by servo parameter correction were placed in anhydrous ethanol for ultrasonic cleaning for 10 minutes to remove impurities on the glass surface. After drying in a drying oven, they were placed in a hydrofluoric acid etching solution, and based on the selective etching characteristics of the modified area, the hollow small curvature structure of the quartz pendulum was obtained; the etching solution concentration was 15%, the initial temperature was set to 30 ° C, and the etching time was 15 minutes. At the same time, ultrasonic oscillation was introduced during etching to assist etching and accelerate etching efficiency and etching uniformity; after etching, the sample was placed in clear lime water to neutralize the residual hydrofluoric acid on the surface, and then cleaned and dried with anhydrous ethanol assisted by ultrasonic cleaning and drying to complete the processing of the small curvature structure of the quartz pendulum. The comparison of the etching effect of the small curvature structure before and after the adaptive optimization of the servo parameters is shown in the figure below. Figure 6 and Figure 7 shown.

[0088] It should be understood that although this specification is described according to various embodiments, not every embodiment contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

[0089] The series of detailed descriptions listed above are only specific descriptions of feasible embodiments of the present invention. They are not intended to limit the scope of protection of the present invention. Any equivalent embodiments or changes that do not deviate from the technical spirit of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for adaptive precision machining of small curvature structures, characterized in that: The specific steps are as follows: According to the outline drawing of the small curvature structure to be processed, the processing trajectory is obtained; the small curvature structure to be processed is placed on the platform, and the servo drive controls the platform to move along the processing trajectory; Determine the target value of the standard deviation of the platform acceleration fluctuation, the target value of the maximum position deviation of the machining trajectory, and the target value of the percentage of geometric similarity between the preset path and the actual machining path; In laser processing, after completing a processing trajectory, the standard deviation of platform acceleration fluctuation is determined by combining acceleration waveform monitoring and visual appearance detection. , Maximum position deviation of machining trajectory , the geometric similarity percentage C between the preset path and the actual processing path; Whether the conditions are met is determined based on the judgment conditions of the maximum position deviation, the judgment conditions of the geometric shape similarity percentage, and the judgment conditions of the acceleration fluctuation standard deviation. If not, at least one of the smoothing instruction filter coefficient, the maximum adaptive gain, and the adaptive gain proportional factor in the servo driver built-in parameters is dynamically adjusted based on the comparison results; the servo driver built-in parameters after adjustment are used to reprocess until the target value is met; the judgment conditions of the maximum position deviation are as follows: Assume the target value of the maximum position deviation is A2; when When >2 A2, the smoothing instruction filter coefficient is adjusted. The adjusted smoothing instruction filter coefficient is recorded as S new , S new =min(S+20%×K, ), S is the current smoothing instruction filter coefficient, is the current maximum adaptive gain; K is the current adaptive gain proportional factor; the adjusted smoothing instruction filter coefficient S new Satisfy the condition S new [20%, 95%]; when When , the adjusted adaptive gain proportional factor is recorded as K new , K new =1.1K, the adjusted smoothing instruction filter coefficient S new =S+5%; adjusted smoothing instruction filter coefficient S new Satisfy the condition S new [20%, 95%], adjusted adaptive gain scaling factor K new The condition is K new [0.5, 1.8]; When the target value is met, a small curvature structure with a smooth contour is obtained.

2. The method for adaptively and precisely machining a small curvature structure according to claim 1, characterized in that: According to the judgment condition of the geometric shape similarity percentage, at least one of the smoothing instruction filter coefficient, the maximum adaptive gain and the adaptive gain proportional factor in the built-in parameters of the servo driver is dynamically adjusted, specifically: Set the target value of geometric shape similarity percentage to A3; When C< A3-5%, the adjusted smoothing instruction filter coefficient S new =S+15%; the maximum adaptive gain after adjustment is recorded as , = max( ×0.9, 100%); adjusted smoothing instruction filter coefficient S new Satisfy the condition S new [20%, 95%], maximum adaptive gain after adjustment The conditions are met [80%,200%]; when When the adjusted adaptive gain proportional factor K new =1.1K, the adjusted smoothing instruction filter coefficient S new =S-5%; adjusted smoothing instruction filter coefficient S new The condition is S new [30%, 95%], the adjusted adaptive gain scaling factor K new The condition is K new [0.5, 1.8].

3. The method for adaptively and precisely machining a small curvature structure according to claim 1, wherein: According to the judgment condition of the acceleration fluctuation standard deviation, at least one of the smoothing instruction filter coefficient, the maximum adaptive gain and the adaptive gain proportional factor in the built-in parameters of the servo driver is dynamically adjusted, specifically: Assume that the target value of the acceleration fluctuation standard deviation is A1, when When >1.6A1, the maximum adaptive gain after adjustment is recorded as = ×0.7, the adjusted adaptive gain scaling factor K new =0.8K, the adjusted smoothing instruction filter coefficient S new =S+10%; adjusted smoothing instruction filter coefficient S new Satisfy the condition S new [30%, 95%], the adjusted adaptive gain scaling factor K new The condition is K new [0.6, 1.8]; maximum adaptive gain after adjustment The conditions are met [80%, 200%]; when When the adjusted smoothing instruction filter coefficient S new =S+4%, the maximum adaptive gain after adjustment is recorded as = min( ×1.05, 150%), the adjusted smoothing instruction filter coefficient S new Satisfy the condition S new [30%, 95%], maximum adaptive gain after adjustment The conditions are met [80%, 200%].

4. The method for adaptively and precisely machining a small curvature structure according to claim 1, characterized in that: When two of the following conditions are met simultaneously: the maximum position deviation, the geometric shape similarity percentage, and the acceleration fluctuation standard deviation, the built-in parameters of the servo drive are adjusted in the following order of priority: The judgment condition of maximum position deviation > the judgment condition of geometric shape similarity > the judgment condition of acceleration fluctuation standard deviation.

5. The method for adaptively and precisely machining a small curvature structure according to claim 4, characterized in that: After adjusting the built-in parameters of the servo driver based on the judgment condition of the maximum position deviation, if C < A3-5% at the same time, first increase the adaptive gain proportional factor after the adjustment of the judgment condition of the maximum position deviation by 5%, and then adjust the built-in parameters of the servo driver based on the judgment condition of the geometric shape similarity percentage, where A3 is the target value of the geometric shape similarity percentage.

6. The method for adaptively and precisely machining a small curvature structure according to claim 4, characterized in that: After adjusting the built-in parameters of the servo driver according to the judgment conditions of geometric shape similarity, if When < 1.2A1, first increase the adaptive gain proportional factor after adjusting the judgment condition of geometric shape similarity by 8%, and then adjust the built-in parameters of the servo driver according to the judgment condition of acceleration fluctuation standard deviation. A1 is the target value of acceleration fluctuation standard deviation.

7. The method for adaptively and precisely machining a small curvature structure according to claim 4, characterized in that: After adjusting the built-in parameters of the servo driver according to the judgment conditions of the acceleration fluctuation standard deviation, if 2 A2, the platform acceleration is limited to 90% of the set value.

Images