Mechanical arm megasonic conformal cleaning control method and device for large aperture optical elements

Abstract

The present application belongs to the technical field of optical processing, and particularly relates to a mechanical arm megasonic wave conformal cleaning control method and device for large-aperture optical elements. The cleaning control method comprises the following steps: S1, three-dimensional reconstruction of a curved surface and local normal vector solution; S2, planning of an acoustic coupling trajectory based on a constant liquid bridge; S3, establishment of a normal vector constraint and a six-axis robot posture control matrix; S4, variable speed scanning control of an equivalent acoustic energy dose; according to the effective output power, effective acoustic field width and target position coordinate sequence of a megasonic cleaning head, the instantaneous feed speed of the megasonic cleaning head when moving along the cleaning trajectory is planned. The cleaning device comprises a six-axis robot, a megasonic cleaning head, a liquid supply system, a recovery system and a control system. The present scheme can realize adaptive conformal scanning of the megasonic cleaning head on a large-aperture complex optical curved surface, and ensure full-aperture cleaning uniformity and non-destructiveness.

Description

Technical Field

[0001] This invention belongs to the field of optical processing technology, and particularly relates to a method and device for controlling the megaacoustic conformal cleaning of large-aperture optical elements by robotic arms. Background Technology

[0002] After the precision polishing process of large-aperture optical components (such as aspherical mirrors, freeform surfaces, etc.), a large amount of polishing liquid (such as cerium oxide, zirconium oxide polishing particles and various chemical additives) is usually left on the surface of the component. These nanoscale particles have extremely strong van der Waals forces and electrostatic adsorption forces with the optical surface, making them extremely difficult to remove. If the cleaning is not thorough, the residual particles will not only seriously affect the accuracy of subsequent interference surface shape detection and cause the detection data to be distorted, but will also cause serious defects such as film defects, pinholes and even film peeling during the coating process, ultimately leading to the scrapping of the component.

[0003] Megasonic cleaning technology (operating at a frequency of around 1 MHz) provides an effective solution to the above problems. Its principle is to use the alternating sound pressure and micro-acoustic flow effect generated by high-frequency sound waves in liquids to remove submicron particles. Compared with traditional low-frequency ultrasonic (20-40 kHz) cleaning, megasonic cleaning produces almost no violent cavitation bubble collapse, thus effectively avoiding physical damage to ultra-smooth optical surfaces. It is an ideal technology for cleaning precision optical components.

[0004] Based on the analysis of existing technical solutions, the implementation methods most similar to this application can be summarized into the following categories: 1. Immersion ultrasonic / megason cleaning tank: The component to be cleaned is immersed in a tank filled with cleaning fluid. The sound field is excited by transducers installed at the bottom or side wall of the tank for cleaning. This solution has significant limitations in the cleaning of large-diameter optical components. The cost of manufacturing a megason cleaning tank that matches the size of the component is extremely high. In addition, the sound field distribution in the large-volume tank is extremely uneven, which can easily generate standing wave effects, resulting in cleaning blind spots and making it difficult to guarantee the cleaning quality.

[0005] 2. Semiconductor single-wafer megasonic cleaning machine: The workpiece is placed on a high-speed rotating tray, and a megasonic nozzle or swing arm is set above it to perform fan-shaped sweep cleaning on the surface of the workpiece in a two-dimensional plane. The limitation of this solution is that its geometric adaptability is extremely poor. It is only suitable for flat or near-flat workpieces. For highly steep and complex curved surfaces such as deep concave spherical surfaces, off-axis aspherical surfaces, and free-form surfaces, the fixed-angle nozzle cannot ensure that the distance between the transducer and each point on the surface of the component is constant. This results in the acoustic coupling medium (liquid flow) not being able to effectively cover or the coupling breaking. The megasonic energy cannot be transferred to the surface, and the cleaning effect deteriorates sharply.

[0006] 3. Manual wiping and high-pressure water rinsing: Relying on operators to manually wipe with lint-free cloths and cleaning solvents, or to rinse with high-pressure water guns, manual wiping is not only time-consuming, labor-intensive, and inefficient, but it is also difficult to maintain uniform wiping force. It is very easy to drag residual hard polishing abrasive particles on the optical surface, causing fatal scratches or subsurface damage, which directly affects the optical performance and structural strength of the components.

[0007] In summary, existing cleaning technologies face prominent problems when dealing with the precision cleaning requirements of large-diameter, steep, and complex optical surfaces, such as poor geometric adaptability, high equipment costs, inability to guarantee consistent cleaning quality, and susceptibility to secondary damage. Therefore, there is an urgent need to develop an automated cleaning solution that can balance geometric versatility, cleaning uniformity, and non-destructive properties. Summary of the Invention

[0008] In view of this, the present invention aims to provide a robotic arm megasonic conformal cleaning control method and device for large-aperture optical components, so as to solve the technical problems that existing megasonic cleaning equipment cannot achieve adaptive conformal cleaning of large-aperture complex optical surfaces, resulting in acoustic coupling failure, poor cleaning uniformity and easy generation of cleaning blind zones.

[0009] To achieve the above objectives, the technical solution created by this invention is implemented as follows: This invention provides a method for controlling the megaacoustic conformal cleaning of large-aperture optical elements using a robotic arm, comprising the following steps: S1. Perform 3D reconstruction of the surface and calculation of local normal vectors; S11. Obtain surface morphology data of large-aperture optical elements; S12. Discretize the full-caliber cleaning area into a grid to generate several cleaning retention points; S13. Calculate the local normal vector of the surface being cleaned at each dwell point; S2. Plan the acoustic coupling trajectory based on the constant liquid bridge; using each dwelling point in S13 as a reference, offset the preset acoustic coupling distance along the corresponding local normal vector direction, and calculate the target position coordinates of the tool center point of the mega-sonic cleaning head in space. S3. Establish normal vector constraints and the attitude control matrix of the six-axis robot; S31. Define the acoustic wave emission axis vector of the megaacoustic cleaning head and establish attitude constraints. S32. Based on the target position coordinates and the sound wave emission axis vector, construct the homogeneous transformation matrix of the six-axis robot end effector relative to the base coordinate system; S33. Obtain the target rotation sequence of each joint of the six-axis robot by solving inverse kinematics; S4. Perform variable-speed scanning control of equivalent acoustic energy dose; based on the effective output power, effective sound field width and target position coordinate sequence of the megasonic cleaning head, plan the instantaneous feed speed of the megasonic cleaning head when it moves along the cleaning trajectory.

[0010] In S11, the surface topography data of the optical element is obtained through the theoretical surface equation of the optical element. It can be generated directly; or point cloud data can be obtained by scanning the surface of optical components with a 3D measurement device and then reconstructing the point cloud data by surface fitting.

[0011] The local normal vector in S13 is obtained based on surface gradient theory, and the formula is as follows: ; in This represents the local unit normal vector at the i-th cleaning point; The implicit surface equation representing the surface of the optical element being cleaned; Indicates the first Cleaning and rest points Spatial coordinates; This represents the gradient vector of the implicit surface equation at the i-th cleaning and dwelling point; The Euclidean norm of the gradient vector; , , They represent about , , The first-order partial derivatives of the three coordinate variables.

[0012] In S2, the coordinates of the target position in space of the tool center point of the mega-sonic cleaning head are: The calculation formula is as follows: ; in, This represents the local unit normal vector at the i-th cleaning point; Indicates the cleaning of the designated area; This indicates the optimal acoustic coupling distance.

[0013] When establishing attitude constraints in S31, the acoustic wave emission axis vector must be kept parallel to the local normal vector at each dwell point in the opposite direction. When the mega-sonic cleaning head is cleaning in a vertical attitude, the acoustic wave emission axis vector and the local normal vector satisfy the following constraint relationship: ; in, This represents the local unit normal vector at the i-th cleaning point; This represents the vector of the sound wave emission axis.

[0014] The expression for the homogeneous transformation matrix in S32 is: ; in, Represents the homogeneous transformation matrix; Indicates the target's location coordinates; This represents the vector of the sound wave emission axis.

[0015] In S33, the definition is... Let the rotation matrix be derived from the attitude vector. Input the inverse kinematics solver of the six-axis robot controller to obtain the target rotation sequence required for each joint in the six-axis robot. .

[0016] In S4, the equivalent acoustic energy dose per unit area is calculated using the following formula: ; in, Indicates the equivalent acoustic energy dose; Indicates effective output power; Indicates the effective sound field width; This indicates the instantaneous feed rate.

[0017] In S4, the speed control law of the instantaneous feed rate is expressed as: ; in, Indicates instantaneous feed rate; Indicates effective output power; Indicates the effective sound field width; Indicates the target cleaning dose.

[0018] A robotic arm megaacoustic conformal cleaning device for large-aperture optical elements, used to implement the above control method, includes: A six-axis robot, as a motion actuator, has a programmable motion control system; The megasonic cleaning head is mounted on the end effector of a six-axis robot. The megasonic cleaning head integrates a megasonic piezoelectric transducer and a flow guiding cavity, and a nozzle is provided at the bottom of the megasonic cleaning head. The liquid supply system is connected to the flow guide cavity of the megasonic cleaning head through the supply pipeline. It is used to supply the megasonic cleaning head with a constant flow rate and constant temperature of cleaning liquid, so that the cleaning liquid is guided through the flow guide cavity and flows out from the nozzle, forming a liquid bridge between the nozzle and the surface of the optical element to be cleaned. The recycling system is connected to the megasonic cleaning head or cleaning station via a recycling pipeline and is used to collect the waste liquid after cleaning into the storage tank. The control system is electrically connected to the six-axis robot and the megasonic piezoelectric transducer. It is used to plan the cleaning trajectory based on the three-dimensional surface data of the optical element and control the six-axis robot to drive the megasonic cleaning head to move along the cleaning trajectory. When the control system controls the six-axis robot to drive the megasonic cleaning head, it maintains a constant normal orientation and a constant working distance between the nozzle and the surface to be cleaned of the optical element, so that the megasonic energy generated by the megasonic piezoelectric transducer is transmitted to the surface to be cleaned through the fluid coupling formed by the cleaning fluid.

[0019] Compared with the prior art, the present invention can achieve the following beneficial effects: 1. Excellent geometric adaptability, improved cleaning efficiency and automation: This invention, with the multi-degree-of-freedom and highly flexible spatial motion capabilities of a six-axis robot, enables the application of megasonic cleaning technology on highly steep and complex optical components such as deeply concave spherical surfaces, off-axis aspherical surfaces, and freeform surfaces. No matter how complex the surface shape of the component, the cleaning head can effectively cover it, completely breaking through the geometric limitations of traditional megasonic cleaning equipment, which is only suitable for planar or slightly curved workpieces. Moreover, this device can achieve fully automated cleaning operations without manual intervention, greatly reducing the labor intensity of operators, improving cleaning efficiency, and avoiding direct contact between operators and chemical cleaning solvents, thus improving operational safety.

[0020] 2. Uniform cleaning across the entire diameter, eliminating cleaning blind spots and reducing cleaning costs for large-diameter components: This invention uses a rigorous conformal trajectory planning algorithm to ensure that the cleaning head maintains a constant normal orientation and working distance with the component surface across the entire diameter range. This ensures that the sound wave transmission path and incident angle remain consistent across the entire surface, and that the megaphonic energy density received at every point on the component surface is uniform and equal. This completely eliminates local cleaning blind spots caused by uneven sound field distribution or geometric obstruction. Compared to manufacturing and maintaining a dedicated megaphonic cleaning tank that matches the size of large-diameter components, this invention adopts a configuration of "small cleaning head + long-stroke robotic arm," which significantly reduces equipment costs and eliminates the need to customize expensive tank facilities for large-sized components.

[0021] 3. Zero physical damage, ensuring the integrity of ultra-smooth surfaces, and highly consistent and repeatable cleaning quality: This invention replaces traditional mechanical brushing, manual wiping, or high-pressure jet cleaning with the non-contact physical action of megasonic waves, fundamentally eliminating contact scratches and subsurface damage to soft optical materials or ultra-smooth surfaces during the cleaning process. Furthermore, relying on the precise programmable control capabilities of a six-axis robot, the cleaning path and process parameters can be standardized and digitized. Multiple batches of cleaning for components of the same specification exhibit highly consistent results, effectively reducing random errors introduced by manual operation. Attached Figure Description

[0022] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments and descriptions of the invention are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is an overall flowchart of the cleaning method in an embodiment of the present invention; Figure 2 This is a schematic diagram of the cleaning device in an embodiment of the present invention.

[0023] Attached reference numerals: 1. Six-axis robot; 2. Megasonic cleaning head; 3. Optical components; 4. Cleaning platform; 5. Supply pipeline; 6. Liquid supply motor; 7. Recovery pipeline; 8. Liquid storage tank. Detailed Implementation

[0024] To make the purpose, technical solution, and advantages of this invention clearer, the following description is provided in conjunction with the appendix. Figure 1-2 The present invention will be further described in detail below with reference to specific embodiments. It should be understood that the specific embodiments described herein are only for explaining the present invention and do not constitute a limitation thereof.

[0025] This invention provides a robotic arm megasonic conformal cleaning control method for large-aperture optical elements and a robotic arm megasonic conformal cleaning device for implementing the control method. It is mainly used for large-aperture optical elements with complex curved surfaces. The cleaning device includes a six-axis robot 1, a megasonic cleaning head 2, a liquid supply system, a recovery system, and a control system.

[0026] The six-axis robot 1 serves as a motion actuator and has a programmable motion control system. A cleaning platform 4 is located on the side of the six-axis robot 1, and an optical element 3 is placed on the cleaning platform 4. A megasonic cleaning head 2 is located on the end effector of the six-axis robot 1. The megasonic cleaning head 2 integrates a megasonic piezoelectric transducer and a flow guiding cavity, and a nozzle is located at the bottom of the megasonic cleaning head 2.

[0027] The liquid supply system is connected to the guide cavity of the megasonic cleaning head 2 through the supply pipeline 5. It is used to supply the megasonic cleaning head 2 with a constant flow rate and constant temperature of cleaning liquid. The cleaning liquid is guided through the guide cavity and flows out from the nozzle, forming a liquid bridge between the nozzle and the surface of the optical element 3 to be cleaned. The liquid supply system is controlled by the liquid supply motor 6.

[0028] The recycling system is connected to the megasonic cleaning head 2 or the cleaning station through the recycling pipeline 7. It is used to recycle the waste liquid after cleaning to the storage tank 8. The control system is electrically connected to the six-axis robot 1 and the megasonic piezoelectric transducer. It is used to plan the cleaning trajectory based on the three-dimensional surface data of the optical element 3 and control the six-axis robot 1 to drive the megasonic cleaning head 2 to move along the cleaning trajectory.

[0029] When the control system controls the six-axis robot 1 to drive the megasonic cleaning head 2, it maintains a constant normal orientation and a constant working distance between the nozzle and the surface to be cleaned of the optical element 3, so that the megasonic energy generated by the megasonic piezoelectric transducer is transmitted to the surface to be cleaned through the fluid coupling formed by the cleaning fluid.

[0030] The cleaning method includes the following steps: S1. Perform 3D reconstruction of the surface and calculation of local normal vectors.

[0031] S11. Obtain the surface morphology data of optical element 3. The surface morphology data of optical element 3 is obtained through the theoretical surface equation of optical element 3. Point cloud data can be obtained either by directly generating the data or by scanning the surface of the optical element 3 with a 3D measuring device and then performing surface fitting and reconstruction on the point cloud data.

[0032] S12. Discretize the full-caliber cleaning area into a grid, generating several cleaning retention points. The set of retention points is as follows: , among which, the i The spatial coordinates of each station are: .

[0033] S13. To ensure the perpendicular incidence of the cleaning acoustic waves, the local normal vector of the surface being cleaned at each dwell point is calculated. The local normal vector is obtained based on the surface gradient theory, and the formula is expressed as: ;in, This represents the local unit normal vector at the i-th cleaning point; The implicit surface equation representing the surface of the optical element being cleaned; Indicates the first Cleaning and rest points Spatial coordinates; This represents the gradient vector of the implicit surface equation at the i-th cleaning and dwelling point; The Euclidean norm of the gradient vector; , , They represent about , , The first-order partial derivatives of the three coordinate variables.

[0034] S2. Plan the acoustic coupling trajectory based on the constant liquid bridge; using each dwelling point in S13 as a reference, offset the preset acoustic coupling distance along the corresponding local normal vector direction, and calculate the target position coordinates of the tool center point of the mega-sonic cleaning head 2 in space.

[0035] Megasonic energy attenuates extremely rapidly in air, necessitating a pure water flow between the nozzle and the workpiece as the acoustic coupling medium. To ensure consistent energy intensity upon reaching the workpiece surface, the liquid bridge length must be kept constant. The target position coordinates of the tool center point of the megasonic cleaning head 2 in space are defined as follows: The calculation formula is as follows: ;in, This represents the local unit normal vector at the i-th cleaning point; Indicates the cleaning of the designated area; This indicates the optimal acoustic coupling distance.

[0036] This formula ensures that the Euclidean distance between the nozzle end face connected to the megasonic cleaning head 2 and the optical surface remains locked at a constant value when facing any steep curved surface. This ensures stable transmission of sound pressure.

[0037] S3. Cleaning requires not only accurate positioning but also correct posture (angle) to prevent total reflection of sound waves at the solid-liquid interface. In this step, normal vector constraints and the posture control matrix of the six-axis robot are established.

[0038] S31. Define the acoustic wave emission axis vector of the mega-sonic cleaning head 2 and establish attitude constraints.

[0039] When establishing attitude constraints, the acoustic wave emission axis vector and the local normal vector at each dwell point must be kept parallel to each other in opposite directions. When the mega-sonic cleaning head 2 is cleaning in a vertical attitude, the acoustic wave emission axis vector and the local normal vector satisfy the following constraint relationship: ;in, This represents the local unit normal vector at the i-th cleaning point; This represents the vector of the sound wave emission axis.

[0040] S32. Based on the target position coordinates and the sound wave emission axis vector, construct the homogeneous transformation matrix of the end effector of the six-axis robot 1 relative to the base coordinate system.

[0041] The expression for the homogeneous transformation matrix is: ;in, This represents the 4×4 homogeneous transformation matrix of the robot's end effector relative to the base coordinate system; Indicates the target's location coordinates; This represents the vector of the sound wave emission axis.

[0042] S33. Obtain the target rotation angle sequence of each joint of the six-axis robot 1 by solving inverse kinematics, and define... Let the rotation matrix be derived from the attitude vector. Input the inverse kinematics solver of the six-axis robot 1 controller to obtain the target rotation sequence required for each joint in the six-axis robot 1. .

[0043] S4. Perform variable-speed scanning control of equivalent acoustic energy dose. In the cleaning of complex curved surfaces, the distribution of polishing fluid residue and the liquid bridge drainage resistance are different in different areas. In order to ensure the cleanliness of the entire diameter is consistent, the instantaneous feed speed of the megasonic cleaning head 2 when it moves along the cleaning trajectory is planned according to the effective output power, effective sound field width and target position coordinate sequence of the megasonic cleaning head 2.

[0044] The equivalent acoustic energy dose per unit area is calculated using the following formula: ;in, Indicates the equivalent acoustic energy dose; Indicates effective output power; Indicates the effective sound field width; This indicates the instantaneous feed rate.

[0045] The speed control law for instantaneous feed rate is expressed as: ;in, Indicates instantaneous feed rate; Indicates effective output power; Indicates the effective sound field width; Indicates the target cleaning dose.

[0046] The above embodiments use a six-axis serial industrial robot as an example to illustrate the motion actuator of the present invention. However, those skilled in the art should understand that the motion actuator of the present invention is not limited thereto. In actual industrial applications, the actuator can be completely replaced by a five-axis linkage CNC machining center, a gantry-type Cartesian coordinate robot equipped with a double-swing gimbal, or a parallel robot (such as a Stewart platform) or a hybrid robot with high rigidity. As long as the platform can achieve translation and attitude deflection in three-dimensional space to maintain acoustic normal constraints, it is an equivalent protected object of the present invention.

[0047] Furthermore, the above embodiments mainly operate in the mode of "cleaning head movement - workpiece fixation". As another equivalent implementation, the mode of "workpiece movement - cleaning head fixation" can also be adopted: that is, the megasonic cleaning head 2 is fixedly set, and a lightweight optical workpiece is clamped by a multi-axis turntable or robot to perform spatial pose conformal interpolation motion. This mode is only the inverse transformation of the kinematic relationship of the present invention, and is completely equivalent to the present invention in principle, and should fall within the protection scope of the present invention.

[0048] Regarding the method of maintaining the constant liquid bridge distance in S2, in addition to adopting pure open-loop control based on offline trajectory planning, as a further preferred implementation, a high-precision laser displacement sensor or a non-contact ultrasonic ranging module can be integrated on the side of the end of the megasonic cleaning head 2. During the cleaning process, the actual physical distance between the nozzle end face and the workpiece surface is detected in real time by the sensor, and the deviation signal is fed back to the robot controller to perform dynamic Z-axis (normal) compensation for the robot end position. This closed-loop control strategy of sensing measurement combined with real-time compensation can effectively overcome the influence of large-diameter workpiece installation error and robot absolute positioning error on cleaning consistency, which is a natural evolution of the technical solution of this invention.

[0049] Regarding the megasonic sound field and liquid flow distribution structure, as alternative implementation schemes: First, to prevent waste liquid with polishing slurry residue from flowing under gravity and causing secondary pollution to the lower half of the workpiece, the nozzle periphery can be changed to a coaxial structure with a negative pressure suction ring, that is, a megasonic pure water column is sprayed in the center and the outer ring suctions the waste liquid in real time, forming a dynamically updated liquid bridge microenvironment in the local area; Second, for polishing residues with a wide particle size distribution (from tens of nanometers to several micrometers), the piezoelectric transducer built into the cleaning end can be replaced with a dual-frequency or multi-frequency broadband transducer (such as alternately outputting 0.8MHz and 3.0MHz sound waves) to excite acoustic flow effects of different scales and further improve the cleaning coverage.

[0050] The mega-sonic conformal cleaning method for large-diameter complex curved surface robotic arms proposed in this invention not only breaks through the size limitations of traditional ultrasonic cleaning tanks, but also solves the problem of non-destructive cleaning of three-dimensional curved surfaces. Besides its application in removing residues after polishing high-end optical components, it has broad industrial application prospects, including but not limited to the following fields: Non-destructive cleaning of extreme ultraviolet (EUV) lithography masks and large-size wafers: In advanced semiconductor processes, EUV mask surfaces are extremely expensive and have "zero tolerance" for micron / nanoscale physical scratches. The suspension bridge megasonic cleaning technology of this invention can precisely remove submicron-level dust from the surface of masks or large-size (12 inches and above) silicon wafers in a non-contact, cavitation-free manner, meeting the stringent requirements of the semiconductor industry for ultra-cleanliness.

[0051] Ultra-clean treatment of conformal optical windows and fairings in aerospace: New generation aircraft widely use conformal radomes and infrared windows made of special materials such as sapphire, zinc sulfide or spinel. These components have large curvature and are asymmetrical. Before applying high-quality antireflection coating, this invention can be used to perform full-surface conformal megasonic cleaning without dead angles, which can greatly improve the adhesion of the film and the optical transmittance.

[0052] High-end new display panel and curved glass cleaning scenarios: This invention is applicable to the cleaning of flexible OLED substrates and the base plate before the mass transfer of Mini / Micro-LEDs, as well as the removal of micro-dust from large-size automotive 3D curved cover glass. The wide-area scanning capability of the robotic arm perfectly matches the processing requirements of large-size panels and can effectively replace the traditional tank-type ultrasonic cleaning solution.

[0053] Microscopic cleanliness control of high-precision medical implants: After machining, complex and porous titanium alloy artificial joints or orthopedic implants are prone to the adhesion of cutting fluid and tiny metal debris on their surface. This invention can use mega-sonic micro-acoustic flow to penetrate deep into the porous structure, achieving efficient cleaning of complex curved implants and avoiding material wear caused by traditional brush cleaning.

[0054] It should be understood that the various forms of processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this invention disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this invention can be achieved, and this is not limited herein.

[0055] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for controlling the megaacoustic conformal cleaning of large-aperture optical elements using a robotic arm, characterized in that, Includes the following steps: S1. Perform 3D reconstruction of the surface and calculation of local normal vectors; S11. Obtain surface morphology data of optical components; S12. Discretize the full-caliber cleaning area into a grid to generate several cleaning retention points; S13. Calculate the local normal vector of the surface being cleaned at each dwell point; S2. Plan the acoustic coupling trajectory based on a constant liquid bridge; Using each dwelling point in S13 as a reference, offset by a preset acoustic coupling distance along the corresponding local normal vector direction, calculate the target position coordinates of the tool center point of the mega-sonic cleaning head in space; S3. Establish normal vector constraints and the attitude control matrix of the six-axis robot; S31. Define the acoustic wave emission axis vector of the megaacoustic cleaning head and establish attitude constraints. S32. Based on the target position coordinates and the sound wave emission axis vector, construct the homogeneous transformation matrix of the six-axis robot end effector relative to the base coordinate system; S33. Obtain the target rotation sequence of each joint of the six-axis robot by solving inverse kinematics; S4. Perform variable-speed scanning control of equivalent acoustic energy dose; based on the effective output power, effective sound field width and target position coordinate sequence of the megasonic cleaning head, plan the instantaneous feed speed of the megasonic cleaning head when it moves along the cleaning trajectory.

2. The robotic arm megaacoustic conformal cleaning control method for large-aperture optical elements according to claim 1, characterized in that, In S11, the surface topography data of the optical element is obtained through the theoretical surface equation of the optical element. It can be generated directly; or point cloud data can be obtained by scanning the surface of optical components with a 3D measurement device and then reconstructing the point cloud data by surface fitting.

3. The robotic arm megaacoustic conformal cleaning control method for large-aperture optical elements according to claim 1, characterized in that, The local normal vector in S13 is obtained based on surface gradient theory, and the formula is as follows: ； in, This represents the local unit normal vector at the i-th cleaning point; The implicit surface equation representing the surface of the optical element being cleaned; Indicates the first Cleaning and rest points Spatial coordinates; This represents the gradient vector of the implicit surface equation at the i-th cleaning and dwelling point; The Euclidean norm of the gradient vector; , , They represent about , , The first-order partial derivatives of the three coordinate variables.

4. The robotic arm megaacoustic conformal cleaning control method for large-aperture optical elements according to claim 1, characterized in that, In S2, the coordinates of the target position in space of the tool center point of the megasonic cleaning head are: The calculation formula is as follows: ； in, This represents the local unit normal vector at the i-th cleaning point; Indicates the cleaning of the designated area; This indicates the optimal acoustic coupling distance.

5. The robotic arm megaacoustic conformal cleaning control method for large-aperture optical elements according to claim 1, characterized in that, When establishing attitude constraints in S31, the acoustic wave emission axis vector must be kept parallel to the local normal vector at each dwell point in the opposite direction. When the mega-sonic cleaning head is cleaning in a vertical attitude, the acoustic wave emission axis vector and the local normal vector satisfy the following constraint relationship: ； in, This represents the local unit normal vector at the i-th cleaning point; This represents the vector of the sound wave emission axis.

6. The robotic arm megaacoustic conformal cleaning control method for large-aperture optical elements according to claim 5, characterized in that, The expression for the homogeneous transformation matrix in S32 is: ； in, Represents the homogeneous transformation matrix; Indicates the target's location coordinates; This represents the vector of the sound wave emission axis.

7. The robotic arm megaacoustic conformal cleaning control method for large-aperture optical elements according to claim 6, characterized in that, In S33, the definition is... Let the rotation matrix be derived from the attitude vector. Input the inverse kinematics solver of the six-axis robot controller to obtain the target rotation sequence required for each joint in the six-axis robot. .

8. The robotic arm megaacoustic conformal cleaning control method for large-aperture optical elements according to claim 1, characterized in that, In S4, the equivalent acoustic energy dose per unit area is calculated using the following formula: ； in, Indicates the equivalent acoustic energy dose; Indicates effective output power; Indicates the effective sound field width; This indicates the instantaneous feed rate.

9. The robotic arm megaacoustic conformal cleaning control method for large-aperture optical elements according to claim 1, characterized in that, In S4, the speed control law of the instantaneous feed rate is expressed as: ； in, Indicates instantaneous feed rate; Indicates effective output power; Indicates the effective sound field width; Indicates the target cleaning dose.

10. A robotic arm megaacoustic conformal cleaning device for large-aperture optical elements, used to implement the control method described in any one of claims 1-9, characterized in that, include: A six-axis robot, as a motion actuator, has a programmable motion control system; The megasonic cleaning head is mounted on the end effector of a six-axis robot. The megasonic cleaning head integrates a megasonic piezoelectric transducer and a flow guiding cavity, and a nozzle is provided at the bottom of the megasonic cleaning head. The liquid supply system is connected to the flow guide cavity of the megasonic cleaning head through the supply pipeline. It is used to supply the megasonic cleaning head with a constant flow rate and constant temperature of cleaning liquid, so that the cleaning liquid is guided through the flow guide cavity and flows out from the nozzle, forming a liquid bridge between the nozzle and the surface of the optical element to be cleaned. The recycling system is connected to the megasonic cleaning head or cleaning station via a recycling pipeline and is used to collect the waste liquid after cleaning into the storage tank. The control system is electrically connected to the six-axis robot and the megasonic piezoelectric transducer. It is used to plan the cleaning trajectory based on the three-dimensional surface data of the optical element and control the six-axis robot to drive the megasonic cleaning head to move along the cleaning trajectory. When the control system controls the six-axis robot to drive the megasonic cleaning head, it maintains a constant normal orientation and a constant working distance between the nozzle and the surface to be cleaned of the optical element, so that the megasonic energy generated by the megasonic piezoelectric transducer is transmitted to the surface to be cleaned through the fluid coupling formed by the cleaning fluid.

Images