Adaptive precision machining method for small-curvature structure

Abstract

An adaptive, precision, laser machining servo driver control method, suitable for small-curvature structures (e.g. in quartz semiconductors), comprises: i) generating a planned machining trajectory based on a drawing; ii) placing a workpiece on a servo-driven platform; and iii) determining target values of standard acceleration fluctuation deviation, maximum position deviation and geometric shape similarity percentage (between a preset path and actual machining path). After completing one laser machining trajectory, if the target values are not met; adjustment is made to a smoothing instruction filtering coefficient, a maximum adaptive gain, or an adaptive gain scale factor (for the servo driver’s control parameters). Laser machining is performed again using the new platform servo driver parameters, until target values are satisfied. Quartz 7 can sit on a high-precision displacement platform 8 having an acceleration sensor, linked to a computer control system 10. An ultrafast picosecond laser system 1 and camera 9, may be used.

Description

TECHNICAL FIELD

[0001] The present invention relates to the field of laser machining, and in particular, to an adaptive precision machining method for a small-curvature structure. BACKGROUND

[0002] Quartz devices are now widely used in fields such as aerospace, inertial navigation systems, precision instruments and meters, and high-end semiconductor equipment, and their core advantages derive from unique properties of quartz glass. For instance, high purity and low impurity content ensure stable operation of a device in an extreme environment and reduce performance degradation caused by material contamination. Additionally, excellent optical and thermal stability meets stringent requirements of processes such as high-precision lithography and laser machining for light transmittance and a thermal expansion coefficient of the material. Furthermore, high insulation effectively mitigates external electromagnetic interference during use. However, the quartz glass is a typical hard and brittle material, which is prone to microcracks or edge chipping due to stress concentration during machining. Therefore, it is difficult to machine a small-curvature structure with high quality and high precision on the hard and brittle material.

[0003] Currently, machining of the small-curvature structure on the hard and brittle material primarily relies on the following three methods: mechanical grinding, chemical etching, and laser modification cutting-chemical etching. Drawbacks of the three methods are separately described below:

[0004] Firstly, the mechanical grinding method uses a grinding wheel to grind the small-curvature structure, requiring a grinding wheel dresser to adjust a tool path. Nevertheless, this method commonly suffers from difficulties in ensuring dimensional consistency, low machining efficiency, susceptibility to micro damages on material edges, and other problems.

[0005] Secondly, the chemical etching method employs hydrofluoric acid solution to selectively etch a pre-masked region of the small-curvature structure, but this method generally involves long processing time and is difficult to control transitional smoothness of a complex geometric shape.

[0006] Thirdly, the laser modification cutting-chemical etching method uses an ultrafast laser (e.g., a picosecond laser or a femtosecond laser) to penetrate an entire quartz glass workpiece to form a modified region, and then combines a wet etching technology for selective etching and shaping. This method offers advantages such as non-contact machining, high precision, and high repeatability, making it suitable for machining the small-curvature structure. However, at present, this method also faces some technical bottlenecks. For example, trajectory deviations (such as servo lag and mechanical backlash) of a laser machining platform during X / Y-axis coordinated motion are amplified as a radius of the small-curvature structure decreases, causing a modified line to deviate from a preset path, resulting in poor stitching and prone to "steps" or "breakpoints" in a transition zone. Such defects further make it difficult for a quartz hollowed-out small-curvature structure obtained after the selective etching to meet preset specifications in terms of smoothness and consistency.

[0007] Therefore, there is an urgent need for a method to correct a motion trajectory of the laser machining platform and enhance precision of multi-axis coordinated motion, thereby improving machining precision of the small-curvature structure. SUMMARY

[0008] In view of the deficiencies in the prior art, the present invention provides an adaptive machining method for a small-curvature structure. During precision laser machining, servo waveform monitoring and visual morphology detection are combined to calculate a standard acceleration fluctuation deviation <ja of a platform, a maximum position deviation 6max of a machining trajectory, and a geometric shape similarity percentage C between a preset path and an actual machining path. Based on the detected and calculated data, parameters of a servo driver are dynamically optimized by a feedback control algorithm. Multi-round closed-loop iterative machining and detection are performed based on optimized servo parameters until a parameter indicator of a machining effect meets a target value, so as to finally obtain a small-curvature structure with high smoothness.

[0009] The present invention achieves the above technical objective through the following technical solutions.

[0010] An adaptive precision machining method for a small-curvature structure includes following steps:

[0011] obtaining a machining trajectory based on an outline drawing of a to-be-machined small-curvature structure, and placing the to-be-machined small-curvature structure on a platform, such that a servo driver controls the platform to move along the machining trajectory;

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

[0013] after completing one machining trajectory during laser machining, determining, in combination with servo waveform monitoring, visual morphology detection, and an acceleration sensor, the standard acceleration fluctuation deviation oa of the platform, the maximum position deviation 8max of the machining trajectory, and the geometric shape similarity percentage C between the preset path and the actual machining path;

[0014] determining whether the standard acceleration fluctuation deviation of the platform aa, the maximum position deviation 8max of the machining trajectory, and the geometric shape similarity percentage C respectively satisfies the target values, and when the target values are not satisfied, dynamically adjusting at least one of a smoothing instruction filtering coefficient, a maximum adaptive gain, and an adaptive gain scale factor in built-in parameters of the servo driver based on a comparison result; and reperforming machining by using adjusted built-in parameters of the servo driver until the target values are satisfied; and

[0015] when the target values are satisfied, obtaining a small-curvature structure with a smooth outline.

[0016] Further, the built-in parameters of the servo driver are dynamically adjusted separately based on a determination condition of the maximum position deviation, a determination condition of the geometric shape similarity percentage, and a determination condition of the standard acceleration fluctuation deviation

[0017] Further, the at least one of the smoothing instruction filtering coefficient, the maximum adaptive gain, and the adaptive gain scale factor in the built-in parameters of the servo driver is dynamically adjusted based on the determination condition of the maximum position deviation, as follows:

[0018] with the target value of the maximum position deviation set to Az,

[0019] when 8max>2Kz, the smoothing instruction filtering coefficient is adjusted, and an adjusted smoothing instruction filtering coefficient is denoted as Snew, and Snew=min(S+20%xK, Gmax), where S represents a current smoothing instruction filtering coefficient, Gmax represents a current maximum adaptive gain, and K represents a current adaptive gain scale factor; and the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that SnewG[20%, 95%]; or

[0020] when A2 an adjusted adaptive gam scale factor is denoted as Knew, where Knew=l.lK; an adjusted smoothing instruction filtering coefficient Snew is equal to S+5%; and the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that SnewG[20%, 95%], and the adjusted adaptive gain scale factor Knew satisfies a condition that Knewe[0.5, 1.8],

[0021] Further, the at least one of the smoothing instruction filtering coefficient, the maximum adaptive gain, and the adaptive gain scale factor in the built-in parameters of the servo driver is dynamically adjusted based on the determination condition of the geometric shape similarity percentage as follows:

[0022] with the target value of the geometric shape similarity percentage set to A3,

[0023] when C<As^5%, an adjusted smoothing instruction filtering coefficient Snew is equal to S+15%, and an adjusted maximum adaptive gain is denoted as Gmaxnew , where GmaXinew=max(Gmaxx0.9, 100%); and the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that SnewG[20%, 95%], and the adjusted maximum adaptive gain Gmax.new satisfies a condition that Gmax new 6[80%, 200%]; or

[0024] when A3 — 5% <C <A3. an adjusted adaptive gain scale factor Knew is equal to 1.1K, and an adjusted smoothing instruction filtering coefficient Snew is equal to S^5%; and the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that SneWG[30%, 95%], and the adjusted adaptive gain scale factor Knew satisfies a condition that KnewG[0.5, 1.8],

[0025] Further, the at least one of the smoothing instruction filtering coefficient, the maximum adaptive gain, and the adaptive gain scale factor in the built-in parameters of the servo driver is dynamically adjusted based on the determination condition of the standard acceleration fluctuation deviation as follows:

[0026] with the taiget value of the standard acceleration fluctuation deviation set to Ai,

[0027] when aa>1.6Ai, an adjusted maximum adaptive gain is denoted as Gmaxnew=Gmax *0.7, an adjusted adaptive gain scale factor Knew is equal to 0.8K, and an adjusted smoothing instruction filtering coefficient Snew is equal to S+10%; and the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that SnewG[30%, 95%], and the adjusted adaptive gain scale factor Knew satisfies a condition that KnewG[0.6, 1.8], and the adjusted maximum adaptive gain Gmax new satisfies a condition that Gmax new e[80%, 200%]; or

[0028] when At <(Ja <1.6A15 an adjusted smoothing instruction filtering coefficient Snew is equal to S+4%, an adjusted maximum adaptive gain is denoted as Gmaxnew=min(Gmax*A.Q5, 150%), the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that SnewG[30%, 95%], and the adjusted maximum adaptive gain Gmax new satisfies a condition that Gmax,neW e[80%, 200%].

[0029] Further, when two of the determination condition of the maximum position deviation, the determination condition of the geometric shape similarity percentage, and the determination condition of the standard acceleration fluctuation deviation are satisfied simultaneously, the built-in parameters of the servo driver are adjusted in a following priority order:

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

[0031] Further, after the built-in parameters of the servo driver are adjusted based on the determination condition of the maximum position deviation, when C<As^5%, an adaptive gain scale factor adjusted based on the determination condition of the maximum position deviation is first increased by 5%, and the built-in parameters of the servo driver are adjusted based on the determination condition of the geometric shape similarity percentage.

[0032] Further, after the built-in parameters of the servo driver are adjusted based on the determination condition of the geometric shape similarity percentage, when <ja <1.2Ai, an adaptive gain scale factor adjusted based on the determination condition of the geometric shape similarity percentage is first increased by 8%, and the built-in parameters of the servo driver are adjusted based on the determination condition of the standard acceleration fluctuation deviation.

[0033] Further, after the built-in parameters of the servo driver are adjusted based on the determination condition of the standard acceleration fluctuation deviation, when 6max >2Ki, an acceleration of the platform is limited to 90% of a set value.

[0034] The present invention has the following advantages:

[0035] 1. According to the adaptive machining method for a small-curvature structure in the present invention, during precision laser machining, acceleration waveform monitoring of a servo drive platform and visual morphology detection are combined to calculate a standard acceleration fluctuation deviation oa of the platform, a maximum position deviation 8max of a machining trajectory, and a geometric shape similarity percentage C between a preset path and an actual machining path. Based on the detected and calculated data, parameters of a servo driver are dynamically optimized by using a feedback control algorithm. Multi-round closed-loop iterative machining and detection are performed based on optimized servo parameters until a parameter indicator of a machining effect meets a target value, so as to finally obtain a small-curvature structure with high smoothness. The small-curvature structure obtained according to the machining method in the present invention has high geometric precision and good process stability.

[0036] 2. The adaptive machining method for a small-curvature structure in the present invention can be applied to machine a small-curvature structure on a variety of hard and brittle materials, such as borosilicate glass, aluminosilicate glass, quartz glass, and sapphire. The adaptive processing method is compatible with various lasers suitable for machining the hard and brittle materials, offering a wide range of applications.

[0037] 3. The adaptive machining method for a small-curvature structure in the present invention supports adaptive parameter correction, which significantly reduces manual reworks and achieves high machining efficiency. Furthermore, after parameter correction, iterative machining and closed-loop quality verification can be performed automatically, which is simple.

[0038] 4. The adaptive machining method for a small-curvature structure in the present invention breaks through a traditional single feedback mode by integrating a servo waveform signal with visual morphology data, and achieves multi-dimensional adaptive parameter compensation based on a feedback control algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] To describe the technical solutions in the embodiments of the present invention or in the prior art more clearly, the accompanying drawings required for describing the embodiments or the prior art will be briefly described below. Apparently, the accompanying drawings in the following description show some embodiments of the present invention, and a person of ordinary skill in the art may still derive other accompanying drawings from these accompanying drawings without creative efforts.

[0040] FIG. 1 is a flowchart of an adaptive machining method for a small-curvature structure according to the present invention;

[0041] FIG. 2 is a schematic diagram of an ultrafast shaping and laser machining platform according to the present invention;

[0042] FIG. 3 shows a preset trajectory of a to-be-machined small-curvature structure according to the present invention;

[0043] FIG. 4 shows a modification effect of a trajectory line of a small-curvature structure that is characterized under a microscope before adaptive servo parameter optimization according to the present invention;

[0044] FIG. 5 shows a modification effect of a trajectory line of a small-curvature structure that is characterized under a microscope after adaptive servo parameter optimization according to the present invention;

[0045] FIG. 6 shows an etching effect of a hollowed-out small-curvature structure obtained under a microscope after quartz glass obtained through modification and machining before parameter optimization is etched with a chemical etchant according to the present invention; and

[0046] FIG. 7 shows an etching effect of a hollowed-out small-curvature structure obtained under a microscope after quartz glass obtained through modification and machining after parameter optimization is etched with a chemical etchant according to the present invention.

[0047] Reference numerals:

[0048] 1 : ultrafast picosecond laser system; 2: beam expander; 3: reflecting mirror; 4: axicon lens; 5: plano-convex lens; 6: microscope objective; 7: to-be-machined sample; 8: high-precision displacement platform; 9: complementary metal-oxide-semiconductor (CMOS) camera and light emitting diode (LED) light source; 10: computer control system. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] The embodiments of the present invention are described below in detail. Examples of the embodiments are shown in the accompanying drawings. Hie same or similar numerals represent the same or similar elements or elements having the same or similar functions throughout the specification. The embodiments described below with reference to the accompanying drawings are exemplary, and are intended to explain the present invention but should not be construed as a limitation to the present invention.

[0050] It should be understood that, in the description of the present invention, the terms such as "central", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "axial", "radial", "vertical", "horizontal", "inner", and "outer" are intended to indicate orientations or positional relationships based on the accompanying drawings. These terms are merely intended to facilitate description of the present invention and simplify the description, rather than to indicate or imply that the mentioned apparatus or element must have a specific orientation and must be constructed and operated in a specific orientation. Therefore, these terms should not be construed as a limitation to the present invention. In addition, the terms "first" and "second" are merely for a purpose of description, and shall not be understood as an indication or implication of relative importance or implicit indication of a quantity of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, "a plurality of' means two or more, unless otherwise specifically defined.

[0051] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected with", "connected to", and "fixed" should be understood in a broad sense. For example, the connection may be a fixed connection, a detachable connection or an integrated connection, may be a mechanical connection or an electrical connection, may be a direct connection or an indirect connection through an intermediate medium, or may be intercommunication between two elements. Those of ordinary7 skill in the art may understand specific meanings of the above terms in the present invention based on a specific situation.

[0052] As shown in FIG. 1, an adaptive machining method for a small-curvature structure in the present invention includes the following specific steps:

[0053] SOI: A 4f beam-reducing system for a Bessel beam is constructed based on a thickness of a sample of a transparent hard and brittle material, ensuring that a focal depth of a focused beam is greater than a glass thickness. The beam-reducing system consists of a plano-convex lens and a 10x microscope objective, and is configured to control the focal depth and increase an energy density. A computer control system is employed to control an optical modulator to adjust a quantity of output bursts (i.e., a quantity of sub-pulses within a pulse envelope), an emission frequency, pulse energy, and a spatial interval of pulse emission. Meanwhile, a z-axis position is set, such that a focal depth of the Bessel beam penetrates upper and lower surfaces of the sample. A movement speed and an acceleration parameter of a platform are set based on an outline drawing of a to-be-machined small-curvature structure and a preset trajectory. Subsequently, the computer control system controls multi-axis coordinated motion based on a preset path to obtain a machining trajectory' of the small-curvature structure. Herein, a small curvature refers to a curvature radius of less than 1 mm.

[0054] S02: An acceleration sensor is used to acquire an acceleration fluctuation of the platform on which the to-be-machined workpiece is placed. Through real-time acquisition of an acceleration sensor signal, an integrated controller monitors the acceleration fluctuation of the platform at a machining stage and calculates a standard acceleration fluctuation deviation to ensure movement stability of the platform during the multi-axis coordinated motion. In addition, with an integrated ultra-high-definition camera and a coaxial illumination source, a microscopic image (with a resolution of 1 um / pixel) of a modified line of the small-curvature structure is captured after the machining. A machining contour is extracted based on an edge detection algorithm, a maximum position deviation is calculated through comparison with a preset theoretical path, and a geometric shape similarity (which is also referred to as a geometric shape similarity percentage) is calculated based on a shape context matching algorithm. The machining platform is driven by a servo driver to move along the machining trajectory of the small-curvature structure. Both the shape context matching algorithm and the edge detection algorithm are existing algorithms.

[0055] S03: Built-in parameters (a smoothing instruction filtering coefficient, a maximum adaptive gain, and an adaptive gain scale factor) of the servo driver are dynamically adjusted based on the standard acceleration fluctuation deviation, the maximum position deviation, and the geometric shape similarity percentage, so as to correct a trajectory deviation in an actual machining process. After a corrected servo control parameter overwrites an original parameter, a new full closed-loop control cycle of "machining-detection-analysis-compensation" is executed until the standard acceleration fluctuation deviation, the maximum position deviation of the machining trajectory', and the geometric shape similarity percentage meet predetermined requirements, as specifically described below:

[0056] A target value of the standard acceleration fluctuation deviation is set to Ai, a target value of the maximum position deviation is set to A2, and a target value of the geometric shape similarity percentage is set to A3. The calculated standard acceleration fluctuation deviation is denoted as aa, the calculated maximum position deviation is denoted as 6max, and the calculated geometric shape similarity percentage is denoted as C.

[0057] When 6max>2A2, the smoothing instruction filtering coefficient is adjusted, and an adjusted smoothing instruction filtering coefficient is denoted as Snew, where Snew=min(S+20%xK, Gmax), where S represents a current smoothing instruction filtering coefficient, Gmax represents a current maximum adaptive gain, and K represents a current adaptive gain scale factor; and the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that SnewG[20%, 95%].

[0058] When A2 <6max <2A2, an adjusted adaptive gain scale factor is denoted as Knew, where Knew=l.lK; an adjusted smoothing instruction filtering coefficient Snew is equal to S+5%; and the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that SnewG[20%, 95%], and the adjusted adaptive gain scale factor Knew satisfies a condition that KnewG[0.5, 1.8],

[0059] When C<A3-5%, an adjusted smoothing instruction filtering coefficient Snew is equal to S+15%, and an adjusted maximum adaptive gain is denoted as GmaXinew , where Gmax,new=m&x(GmaxxQ.9, 100%); and the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that SnewG[20%, 95%], and the adjusted maximum adaptive gain Gmax,new satisfies a condition that Gmax>new e[80%, 200%].

[0060] When A3 — 5% <C <A3, an adjusted adaptive gain scale factor Knew is equal to 1.1K, and an adjusted smoothing instruction filtering coefficient Snew is equal to S-5%; and the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that SnewG[30%, 95%], and the adjusted adaptive gain scale factor Knew satisfies a condition that Knew6[0.5. 1.8],

[0061] When aa >1.6Ai, an adjusted maximum adaptive gain is denoted as Gmax,new=Gmax x0.7, an adjusted adaptive gain scale factor Knew is equal to 0.8K, and an adjusted smoothing instruction filtering coefficient Snew is equal to S+10%; and the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that SnewG[30%, 95%], and the adjusted adaptive gain scale factor Knew satisfies a condition that KnewG[0.6, 1.8], and the adjusted maximum adaptive gain Gmaxnew satisfies a condition that Gmax new G[80%, 200%].

[0062] When At <aa <1.6A1; an adjusted smoothing instruction filtering coefficient Snew is equal to S+4%, and an adjusted maximum adaptive gain is denoted as GmaXineiv=min(Gmaxxl.05, 150%); and the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that Snew6[30%, 95%], and the adjusted maximum adaptive gain Gmax,new satisfies a condition that Gmaxnew e[80%, 200%].

[0063] If the smoothing instruction filtering coefficient, the maximum adaptive gain, and the adaptive gain scale factor are calculated separately and exceed respective boundary limits, the parameter values exceeding the boundary limits will be corrected to corresponding boundary' values that are exceeded. Assuming the adjusted maximum adaptive gain Gmaxnew is 75% and exceeds a minimum boundary of 80%, the adjusted maximum adaptive gain Gmax>new is changed to 80%.

[0064] When a plurality of error conditions are triggered simultaneously, parameter correction is carried out in the following priority order: the maximum position deviation>the geometric shape similarity>the standard acceleration fluctuation deviation. That is, the built-in parameters of the servo driver are first adjusted based on a determination condition of the maximum position deviation, then the built-in parameters of the servo driver are adjusted based on a determination condition of the geometric shape similarity, and finally the built-in parameters of the servo driver are adjusted based on a determination condition of the standard acceleration fluctuation deviation.

[0065] A rule for coordinated adjustment is also designed. Specifically:

[0066] After the built-in parameters of the servo driver are adjusted based on the determination condition of the maximum position deviation, if C<Ai^5%, the adjusted adaptive gain scale factor Knew is equal to 1.05K. In this case, K is an adaptive gain scale factor adjusted based on the determination condition of the maximum position deviation.

[0067] After the built-in parameters of the servo driver are adjusted based on the determination condition of the geometric shape similarity, if cra<1.2Ai, the adjusted adaptive gain scale factor Knew is equal to 1.08K. In this case, K is an adaptive gain scale factor adjusted based on the determination condition of the geometric shape similarity.

[0068] After the built-in parameters of the servo driver are adjusted based on the determination condition of the standard acceleration fluctuation deviation, if 6max >2 A2. an acceleration of the platform is limited to 90% of the set value.

[0069] S04: Upon completion of the closed-loop iterative machining, a target small-curvature structure sample is machined based on final optimized servo drive parameters. For a precision laser machining method involving direct ablation and vaporization, it is necessary' to perform path filling based on the preset machining trajectory and an actual requirement, or repeat the machining for a plurality of times. For a machining method that combines laser modification with wet etching, a modified sample needs to be placed in chemical etching solution for selective etching, ultimately machining a small-curvature structure with high smoothness.

[0070] Embodiment 1

[0071] A quartz small-curvature structure is machined by a method that combines picosecond laser modification and chemical etching, which specifically includes the following steps:

[0072] SOI: A picosecond laser system has a laser wavelength of 1064 nm and a pulse width of 10 ps. A repetition rate is set to 100 kHz, a quantity of bursts is adjusted to 3 by an optical modulator, an emission time interval per burst is set to 25 ns, single-pulse energy is set to 250 pj, a laser machining mode is a position synchronized output (PSO) mode, and a pulse step interval is set to 2.5 pm. After being expanded by a beam expander, a Gaussian beam has a waist radius of 3 mm, a propagation direction of the Gaussian beam is then changed by a reflecting mirror, and the Gaussian beam is reshaped into a Bessel beam after passing through an axicon lens with a conical base angle of 5°. Finally, a beam radius is reduced and an energy density is increased via a 4f beam-reducing system, thereby achieving a modification effect of quartz glass. Hie 4f beam-reducing 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. A calculated spot diameter of the Bessel beam is 1.75 pm and a non-diffracting length is 2.46 mm, which meet modification requirements for a to-be-machined sample. A CMOS camera (Basler ace 2) and an LED light source (with a wavelength of 630 nm) are coaxial with a Bessel optical system for capturing modified morphology. An optical path of the machining is shown in FIG. 2.

[0073] As shown in FIG. 2, a picosecond laser machining apparatus includes an ultrafast picosecond laser system 1, a beam expander 2, a reflecting mirror 3, an axicon lens 4, a plano-convex lens 5, a 10x microscope objective 6, a CMOS camera and an LED light source 9. The to-be-machined sample 7 is placed at a center of a high-precision displacement platform 8, an acceleration sensor is installed inside the platform, and an acquisition module is connected to an accelerometer and a computer control system 10. After the ultrafast laser system generates the Gaussian beam, an output power of a laser is controlled by an internal attenuator. The computer control system controls a Q-switch in the laser system to adjust the repetition rate, and a quantity of pulses in a burst mode is controlled through the optical modulator. After being expanded by the beam expander to enlarge a spot radius, the Gaussian beam is shaped into the Bessel beam by the axicon lens. An adjusted laser beam is then focused onto the sample on a high-precision translation platform through the plano-convex lens and the 10 microscope objective.

[0074] A transparent hard and brittle material selected for an experiment is the quartz glass (with a purity of >99.99%). A thickness of a glass sample is 1 mm and its refractive index is 1.51. The glass sample is fixed on the high-precision translation platform. A machining path is generated based on an outline of a small-curvature structure to be modified and machined. A morphology of the to-be-detected small-curvature structure consists of a main arc and two arc transition lines, with corresponding curvature radii being 270 pm, 680 pm, and 1060 pm. The three arcs are centered at different positions and designed to achieve a smooth transition. A preset trajectory is shown in FIG. 3. A speed and an acceleration of a multi-axis coordinated motion platform are initially set to 10 mm / s to 20 mm / s2 respectively. Trial machining is then performed on a laser machining platform along a preset path, where the Bessel beam penetrates upper and lower surfaces of the glass to form a continuous modified region. In this process, a region directly irradiated by the Bessel beam transforms from an original silicon-oxygen six-membered ring structure into silicon-oxygen three-membered and four-membered ring structures.

[0075] S02: From start to completion of the machining, an integrated data acquisition (DAQ) module is used to acquire a signal in real time from a micro-electro-mechanical system (MEMS) accelerometer (Kistler 8704B) located at a key position of the platform, with a sampling rate of 10 kHz. Acceleration fluctuations at start and stop stages in a sample machining process are detected, and the data is transmitted to the computer control system. A standard acceleration fluctuation deviation aa during a first round of machining is calculated, which is equal to 0.064 g, where g represents a gravitational acceleration, and lg«9.8m / s2. Meanwhile, after the machining is finished, the LED light source is kept continuously on, and the CMOS camera (with a resolution of 1 pm / pixel) automatically performs visual morphology detection of a currently modified trajectory. After a microscopic image of the small-curvature structure is captured, a contour of a modified line is extracted by using a Canny edge detection algorithm, and a maximum position deviation is calculated, namely 6max =12 pm. Based on an obtained actual path and the preset path, a geometric shape similarity percentage C is calculated by using a shape context matching algorithm, which is equal to 87.1%.

[0076] S03: A target value of the standard acceleration fluctuation deviation is <0.05 g, a target value of the maximum position deviation is <3 pm, and a target value of a geometric shape similarity percentage is >95%. To facilitate calculation, it is set that the target value Ai of the standard acceleration fluctuation deviation =0.05g, the target value A2 of the maximum position deviation =3 pm, and the target value A3 of the geometric shape similarity percentage=95%.

[0077] Built-in parameters (a smoothing instruction filtering coefficient, a maximum adaptive gain, an adaptive gain scale factor) of a servo driver are dynamically adjusted based on the standard acceleration fluctuation deviation, the maximum position deviation, and the geometric shape similarity percentage, so as to correct a trajectory-' deviation in an actual machining process.

[0078] In this embodiment, the standard acceleration fluctuation deviation, the maximum position deviation, and the geometric shape similarity percentage all fail to reach respective target values. Since a plurality of error conditions are triggered, parameter correction is performed in the following priority order: the maximum position deviation>the geometric shape similarity>the standard acceleration fluctuation deviation, as shown in Table 1.

[0079] An initial smoothing instruction filtering coefficient S, an initial adaptive gain scale factor K, and an initial maximum adaptive gain Gmax of the servo driver of the platform are respectively 50%, 0.8, and 120%.

[0080] During a first round of optimization:

[0081] Since the maximum position deviation 6max =7.2 pm. which is greater than 6 pm, an adjusted smoothing instruction filtering coefficient Snew is equal to min(50%+20%x0.8, 120%)=66%. The adjusted smoothing instruction filtering coefficient Snew meets a condition that SnewG[20%, 95%]. After the built-in parameters of the servo driver are based on a determination condition of the maximum position deviation, since C=87.1%, which is less than 90%, an adjusted adaptive gain scale factor Knew is equal to 1.05K, namely 0.8^ 105%=0.84.

[0082] Since the geometric shape similarity percentage C is equal to 87.1%, which is less than 90%, the adjusted smoothing instruction filtering coefficient Snew is equal to S+15%, namely 66%+15%=81%; and an adjusted maximum adaptive gain Gmaxnew is equal to max(Gmazx0.9, 100%), namely max(120%x0.9, 100%)=108%.

[0083] Since the standard acceleration fluctuation deviation meets the following condition: 0.05g< oa< 0.08g, the adjusted smoothing instruction filtering coefficient Snew is equal to S+4%. namely 81%+4%=85%, and the adjusted maximum adaptive gain is denoted as Gmaz,neiv=min(Gmaxx1.05, 150%), namely min(108%x 1.05, 150%)=113.4%. Therefore, aresult of this round of servo parameter optimization is as follows: Snew=85%, Knew=0.84, and Gmaznew=113.4%, and servo parameters do not exceed boundary limits.

[0084] Table 1

[0085] Error type Adjustment rule for a to-be-optimized parameter Parameter boundary limit Rule for coordinated adjustment Smax>^ SneM / =min(S+20%xK, Gmax) Se[20%, 95%] If C<90%, Kf5%. 6pm>Smaz>3pm Knew=Kxl.l; 5^=8+5% SG[0.5,1.8] C<90% ^max,new~ Hiax(G'max><0.9, 100%) SG[20%, 95%]; 6mazG[80%, 200%] If aa<0.06 g, Kf8%. 90%<C<95% K^Kxl.l; Snew = ^% SG[30%, 95%]; KG[0.5,1.8] <ra>0.08g ^max.new umax Knew=Kx0.8; Snew=S+10% Gmax G [80%, 200%]; KG[0.6,1.8]; SG[30%, 95%] If ^max>6 1™, 311 acceleration of the platform is limited to 90% of the set value. 0.05g<cra<0.08g Snew=S+4%; SG[30%, 95%] ^max.new X 1 150%) Gmaxe [80%, 200%]

[0086] S04: Based on Snew=85%, Knew=0.84, and Gmaxnew=113.4%, the servo driver of the platform proceeds with next-round machining, effect detection, and analysis of the modified line and re-correction of the servo parameters, until optimized machining trajectory parameters reach predetermined specifications. After four iterations, the following machining effect is achieved: standard acceleration fluctuation deviation aa=0.045g, maximum position deviation 8max =2.7 pm, and geometric shape similarity percentage C=95.4%. The corresponding servo drive parameters are as follows: S=90%, K=1.4640, and Gmax=101.16%. FIG. 4 and FIG. 5 compare modification effects of a trajectory line of the small-curvature structure before and after adaptive optimization of the servo parameters.

[0087] Quartz glass that has been modified and machined without adaptive parameter correction and quartz glass that has been modified and machined with servo parameter correction are both ultrasonically cleaned in anhydrous ethanol for 10 minutes to remove surface impurities. After being dried in a 70°C drying oven, they are separately immersed in hydrofluoric acid etching solution. Based on selective etching characteristics of a modified region, a hollowed-out small-curvature structure of a quartz oscillating plate is obtained. An etching solution concentration is 15%, an initial temperature is set to 30°C, and etching time is 15 minutes. During the etching, ultrasonic oscillation is introduced to assist the etching, enhancing etching efficiency and uniformity. After the etching is completed, the samples are first placed in clarified lime water to neutralize residual hydrofluoric acid on a surface, and then ultrasonically cleaned again in anhydrous ethanol and dried. Thus, machining of the small-curvature structure of the quartz oscillating plate is completed. FIG. 6 and FIG. 7 compare etching effects of the small-curvature structure before and after tire adaptive optimization of the servo parameters.

[0088] It should be understood that although this specification is described in accordance with the embodiments, not every embodiment only includes one independent technical solution. Such a description of this specification is for the sake of clarity only. Those skilled in the art should take this specification as a whole, and the technical solutions in the embodiments can also be appropriately combined to form other implementations that can be understood by those skilled in the art.

[0089] The series of detailed description listed above are only specific illustration of feasible embodiments of the present invention, rather than limiting the protection scope of the present invention. All equivalent embodiments or changes made without departing from the technical spirit of the present invention should be included in the protection scope of the present invention.

Claims

1. An adaptive precision machining method for a small-curvature structure, characterized by comprising following steps:obtaining a machining trajectory based on an outline drawing of a to-be-machined small-curvature structure, and placing the to-be-machined small-curvature structure on a platform, such that a servo driver controls the platform to move along the machining trajectory;determining a target value of a standard acceleration fluctuation deviation of the platform, a target value of a maximum position deviation of the machining trajectory, and a target value of a geometric shape similarity percentage between a preset path and an actual machining path;after completing one machining trajectory during laser machining, determining, in combination with acceleration waveform monitoring and visual morphology detection, the standard acceleration fluctuation deviation oa of the platform, the maximum position deviation 8max of the machining trajectory, and the geometric shape similarity percentage C between the preset path and the actual machining path;detennining whether a determination condition of the maximum position deviation, a determination condition of the geometric shape similarity percentage, and a determination condition of the standard acceleration fluctuation deviation are satisfied, and when the conditions are not satisfied, dynamically adjusting at least one of a smoothing instruction filtering coefficient, a maximum adaptive gain, and an adaptive gain scale factor in built-in parameters of the servo driver based on a comparison result; and reperfonning machining by using adjusted built-in parameters of the servo driver until the target values are satisfied, wherein the determination condition of the maximum position deviation is as follows:with the target value of the maximum position deviation set to A2,when 6max>2A2, the smoothing instruction filtering coefficient is adjusted, and an adjusted smoothing instruction filtering coefficient is denoted as Snew, and Snew=min(S+20%*K, Gmax), wherein S represents a current smoothing instruction filtering coefficient, Gmax represents a current maximum adaptive gain, and K represents a current adaptive gain scale factor; and the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that SnewG[20%, 95%]; orwhen A2 <3max <2A2, an adjusted adaptive gain scale factor is denoted as Knew, wherein Knew=l.lK; an adjusted smoothing instruction filtering coefficient Snew is equal to S+5%; and the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that SnewG[20%, 95%], and the adjusted adaptive gain scale factor Knew satisfies a condition that KnewG[0.5, 1.8]; andwhen the target values are satisfied, obtaining a small-curvature structure with a smoothoutline.

2. The adaptive precision machining method for the small-curvature structure according to claim 1, characterized in that the at least one of the smoothing instruction filtering coefficient, the maximum adaptive gain, and the adaptive gain scale factor in the built-in parameters of the servo driver is dynamically adjusted based on the determination condition of the geometric shape similarity percentage as follows:with the target value of the geometric shape similarity percentage set to A3,when C<Aa-5%, an adjusted smoothing instruction filtering coefficient Snew is equal to S+15%, and an adjusted maximum adaptive gain is denoted as Gmaxnew , wherein Gm(lznew=max(Gmaxx0.9, 100%); and the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that Sne«G[20%, 95%], and the adjusted maximum adaptive gain Gmax.new satisfies a condition that Gmax new e[80%, 200%]; orwhen A3 — 5% <C <A3, an adjusted adaptive gain scale factor Knew is equal to 1.1K, and an adjusted smoothing instruction filtering coefficient Snew is equal to S-5%; and the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that SnewG[30%, 95%], and the adjusted adaptive gain scale factor Knew satisfies a condition that KnewG[0.5, 1.8],3. The adaptive precision machining method for the small-curvature structure according to claim 1, characterized in that the at least one of the smoothing instruction filtering coefficient, the maximum adaptive gain, and the adaptive gain scale factor in the built-in parameters of the servo driver is dynamically adjusted based on the determination condition of the standard acceleration fluctuation deviation as follows:with the target value of the standard acceleration fluctuation deviation set to Ai,when cra>1.6Ai, an adjusted maximum adaptive gain is denoted as Gmax>new=Gmax *0.7, an adjusted adaptive gain scale factor Knew is equal to 0.8K, and an adjusted smoothing instruction filtering coefficient Snew is equal to S+10%; and the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that SnewG[30%, 95%], and the adjusted adaptive gain scale factor Knew satisfies a condition that KnewG[0.6, 1.8], and the adjusted maximum adaptive gain Gmax,new satisfies a condition that Gmax new e[80%, 200%]; orwhen Aj <aa <1.6A15 an adjusted smoothing instruction filtering coefficient Snew is equal to S+4%, an adjusted maximum adaptive gain is denoted as Gmaxnew=min(Gmax*l.Q5, 150%), the adjusted smoothing instruction filtering coefficient Snew satisfies a condition that SnewG[30%, 95%], and the adjusted maximum adaptive gain Gmaxnew satisfies a condition that Gmax,neW e[80%, 200%].

4. The adaptive precision machining method for the small-curvature structure according toclaim 1, characterized in that when two of the determination condition of the maximum position deviation, the determination condition of the geometric shape similarity percentage, and the determination condition of the standard acceleration fluctuation deviation are satisfied simultaneously, the built-in parameters of the servo driver are adjusted in a following priority order:the determination condition of the maximum position deviation>the detennination condition of the geometric shape similarity percentage>the determination condition of the standard acceleration fluctuation deviation.

5. The adaptive precision machining method for the small-curvature structure according to claim 4, characterized in that after the built-in parameters of the servo driver are adjusted based on the determination condition of the maximum position deviation, when C<As^5%, an adaptive gain scale factor adjusted based on the determination condition of the maximum position deviation is first increased by 5%, and the built-in parameters of the servo driver are adjusted based on the determination condition of the geometric shape similarity percentage, wherein A3 represents the target value of the geometric shape similarity percentage.

6. The adaptive precision machining method for the small-curvature structure according to claim 4, characterized in that after the built-in parameters of the servo driver are adjusted based on the determination condition of the geometric shape similarity percentage, when oa<\2K\, an adaptive gain scale factor adjusted based on the determination condition of the geometric shape similarity percentage is first increased by 8%, and the built-in parameters of the servo driver are adjusted based on the determination condition of the standard acceleration fluctuation deviation, wherein Ai represents the target value of the standard acceleration fluctuation deviation.

7. The adaptive precision machining method for the small-curvature structure according to claim 4, characterized in that after the built-in parameters of the servo driver are adjusted based on the determination condition of the standard acceleration fluctuation deviation, when 8max >2A2, an acceleration of the platform is limited to 90% of a set value.18

Images