Aircraft titanium alloy skeleton in-situ damage cleaning method and system based on mobile robot

Abstract

The invention discloses an aircraft titanium alloy skeleton in-situ damage cleaning scheme based on a mobile robot, and the scheme comprises the steps: enabling the mobile robot to shoot through a machine vision module to obtain image information, and transmitting the image information to an upper computer; the upper computer locates the damaged part of the titanium alloy part based on the image information and feeds back the information of the damaged part to the movable robot; the movable robot calls a machine vision module, first point cloud information of the damaged part is obtained through laser scanning of the damaged part, and the first point cloud information is sent to the upper computer; the upper computer constructs a first three-dimensional model corresponding to the damaged part based on the first point cloud information; based on the first three-dimensional model and the posture information of the mechanical arm of the mobile robot, generating cutting planning information; the movable robot controls the laser cutting module to cut and clean the damaged part according to the cutting planning information, the cleaning efficiency is high, and continuous drilling is easy to achieve.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of aircraft damage cleaning, in particular to a method and system for cleaning in-situ damage of aircraft titanium alloy framework based on a mobile robot. BACKGROUND

[0002] Titanium alloy is widely used in aircraft due to its high strength, low density, good corrosion resistance and heat resistance. For example, the tail cover of the third-generation aircraft, the rear fuselage skin of the fourth-generation aircraft, and the framework are mainly made of TC4 titanium alloy or TA15 titanium alloy.

[0003] When the titanium alloy skin and framework are hit by foreign objects, combat units, or human construction errors, cracks, broken holes, and other damages may occur. The traditional method requires replacement and repair, which has a long cycle and high pressure on spare parts support. To quickly restore the combat capability of the damaged aircraft skin and framework, an in-situ repair method must be used, which means repairing the damaged part quickly at the damaged location. To carry out in-situ repair, the residual damage of the skin and framework must be cleaned first to remove residual cracks and sharp edges to prevent further damage and facilitate subsequent repair.

[0004] Currently, the main method for cleaning the damage of the titanium alloy skin and framework is to rely on technical personnel to drill holes manually or to use an industrial robot to "show and cut" first. This method is complex, has low damage cleaning efficiency, and cannot quickly complete the repair. In addition, due to the high strength and poor heat conduction of titanium alloy, drilling and cleaning are difficult and time-consuming. For a 2mm thick 50mm broken hole, the damage cleaning time exceeds 1 hour. For titanium alloy above 4mm, continuous drilling is almost impossible. SUMMARY

[0005] The purpose of the embodiments of the present application is to provide a method and system for cleaning in-situ damage of aircraft titanium alloy framework based on a mobile robot, which can solve the problem of low damage cleaning efficiency and difficulty in continuous drilling in the prior art.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] The embodiments of the present application provide a method for cleaning in-situ damage of aircraft titanium alloy framework based on a mobile robot. The method is applied to a system for cleaning in-situ damage of aircraft titanium alloy framework, which includes a mobile robot integrated with a laser cutting module and a machine vision module, and a host computer. The method includes:

[0008] The mobile robot captures image information through the machine vision module and sends the image information to the host computer;

[0009] The host computer locates a damaged part of a titanium alloy component of an aircraft based on the image information, and feeds back the damaged part information of the titanium alloy component of the aircraft to the movable robot; wherein the image information comprises surface information of the titanium alloy component of the aircraft; wherein the titanium alloy component of the aircraft comprises at least one of a skin and a skeleton;

[0010] The movable robot calls a machine vision module, scans the damaged part by laser to obtain first point cloud information of the damaged part, and sends the point cloud information to the host computer;

[0011] The host computer constructs a first three-dimensional model corresponding to the damaged part based on the first point cloud information;

[0012] The host computer generates cutting planning information based on the first three-dimensional model and the mechanical arm posture information of the movable robot, and sends the cutting planning information to the movable robot, wherein the cutting planning information comprises a cutting planning path and cutting process parameters;

[0013] The movable robot controls the laser cutting module to cut and clean the damaged part according to the cutting planning information.

[0014] Optionally, the step of constructing a first three-dimensional model corresponding to the damaged part based on the first point cloud information by the host computer comprises:

[0015] The host computer pre-processes the first point cloud information to filter out invalid points and outliers to obtain second point cloud information;

[0016] The second point cloud information is filtered by a voxel grid filtering algorithm to obtain third point cloud information;

[0017] The third point cloud information is subjected to point cloud three-dimensional reconstruction by a preset point cloud reconstruction algorithm to obtain the first three-dimensional model.

[0018] Optionally, the step of pre-processing the first point cloud information by the host computer to filter out invalid points and outliers to obtain second point cloud information comprises:

[0019] For each point in the first point cloud information, it is judged whether the coordinate of the point in the first dimension is within a preset first dimension value range;

[0020] If not, the point is regarded as an invalid point, and the invalid point is filtered out;

[0021] Each point in the first point cloud information is traversed, and the average value of the distance between the point and each adjacent point is calculated; it is judged whether the average value is within a preset distance range threshold;

[0022] If not, the point and each point adjacent to the point are determined as outliers, and each outlier is filtered out;

[0023] The first point cloud information after filtering out the invalid points and the discrete points is determined as second point cloud information.

[0024] Optionally, the step of performing point cloud three-dimensional reconstruction on the third point cloud information by using a preset point cloud reconstruction algorithm to obtain the first three-dimensional model comprises:

[0025] performing point cloud three-dimensional reconstruction on the third point cloud information by using a plurality of preset point cloud reconstruction algorithms respectively to obtain a plurality of three-dimensional models;

[0026] selecting a model with the highest score from the plurality of three-dimensional models to determine as the first three-dimensional model.

[0027] Optionally, the step of generating cutting planning information by the upper computer based on the first three-dimensional model and the posture information of the movable robot arm comprises:

[0028] The upper computer performs damage site topography recognition based on the first three-dimensional model to obtain damage site topography.

[0029] determining the boundary size of the damage site topography;

[0030] generating a second three-dimensional model based on the boundary size and a preset boundary correction rule;

[0031] determining a cutting process parameter based on the attribute of the damage site;

[0032] generating a cutting planning path according to the second three-dimensional model and the posture information of the movable robot arm.

[0033] Optionally, the machine vision module is installed at the front end of the mechanical arm of the movable robot;

[0034] The laser cutting module is installed at the bottom end of the mechanical arm.

[0035] The application also provides an aircraft titanium alloy framework in-situ damage cleaning system based on a movable robot, wherein the system comprises a movable robot integrated with a laser cutting module and a machine vision module, and an upper computer;

[0036] The movable robot comprises a control cabinet, a machine vision module, and a laser cutting module.

[0037] The control cabinet is configured to call the machine vision module to capture image information and send the image information to the upper computer.

[0038] The host computer is configured to locate a damage position of a titanium alloy component of an aircraft based on the image information, and feed back the damage position information of the titanium alloy component of the aircraft to the movable robot; wherein the image information comprises surface information of the titanium alloy component; and the titanium alloy component of the aircraft comprises at least one of a skin and a skeleton.

[0039] The control cabinet is further configured to call the machine vision module, scan the damage position by laser to obtain first point cloud information of the damage position, and send the point cloud information to the host computer.

[0040] The host computer is further configured to construct a first three-dimensional model corresponding to the damage position based on the first point cloud information, generate cutting planning information based on the first three-dimensional model and mechanical arm posture information of the movable robot, and send the cutting planning information to the control cabinet of the movable robot, wherein the cutting planning information comprises a cutting planning path and cutting process parameters.

[0041] The control cabinet is further configured to control the laser cutting module to cut and clean the damage position according to the cutting planning information.

[0042] Optionally, the host computer comprises:

[0043] The point cloud processing module is configured to pre-process the first point cloud information to filter out invalid points and outliers to obtain second point cloud information, filter out the second point cloud information by using a voxel grid filtering algorithm to obtain third point cloud information, and perform point cloud three-dimensional reconstruction on the third point cloud information by using a preset point cloud reconstruction algorithm to obtain the first three-dimensional model.

[0044] Optionally, when the point cloud processing module pre-processes the first point cloud information to filter out invalid points and outliers to obtain second point cloud information, the point cloud processing module is specifically configured to:

[0045] For each point in the first point cloud information, it is determined whether the coordinate of the point in the first dimension is within a preset first dimension value range.

[0046] If not, the point is regarded as an invalid point, and the invalid point is filtered out.

[0047] The point cloud processing module is configured to traverse each point in the first point cloud information, calculate an average value of distances between the point and each adjacent point, and determine whether the average value is within a preset distance range threshold.

[0048] If not, the point and each adjacent point of the point are determined as outliers, and each outlier is filtered out.

[0049] The first point cloud information after filtering out the invalid points and outliers is determined as the second point cloud information.

[0050] Optionally, when the point cloud processing module adopts a preset point cloud reconstruction algorithm to perform point cloud three-dimensional reconstruction on the third point cloud information to obtain the first three-dimensional model, the point cloud processing module is specifically configured to:

[0051] The third point cloud information is respectively subjected to point cloud three-dimensional reconstruction by using a plurality of preset point cloud reconstruction algorithms to obtain a plurality of three-dimensional models.

[0052] A model with the highest score is selected from the plurality of three-dimensional models to determine the first three-dimensional model.

[0053] Optionally, the upper computer comprises:

[0054] A cutting planning module is configured to perform damage site topography identification based on the first three-dimensional model to obtain damage site topography, determine the boundary size of the damage site topography, generate a second three-dimensional model based on the boundary size and a preset boundary correction rule, determine a cutting process parameter based on the attribute of the damage site, and generate a cutting planning path according to the second three-dimensional model and the posture information of the movable robot mechanical arm.

[0055] Optionally, the machine vision module is installed at the front end of the mechanical arm of the movable robot.

[0056] The laser cutting module is installed at the bottom end of the mechanical arm.

[0057] The in-situ damage cleaning scheme for the aircraft titanium alloy framework based on the movable robot disclosed in the present application is applied to an in-situ damage cleaning system for the aircraft titanium alloy framework. The system comprises a movable robot integrated with a laser cutting module and a machine vision module and an upper computer. In actual work, the movable robot captures image information by the machine vision module and sends the image information to the upper computer. The upper computer locates the damage site of the aircraft titanium alloy component based on the image information and feeds back the damage site information of the aircraft titanium alloy component to the movable robot. The movable robot calls the machine vision module, scans the damage site by laser, obtains first point cloud information of the damage site, and sends the point cloud information to the upper computer. The upper computer constructs a first three-dimensional model corresponding to the damage site based on the first point cloud information. The upper computer generates cutting planning information based on the first three-dimensional model and the posture information of the mechanical arm of the movable robot, sends the cutting planning information to the movable robot, and controls the laser cutting module to cut and clean the damage site according to the cutting planning information. The in-situ damage cleaning scheme for the aircraft titanium alloy framework based on the movable robot disclosed in the present application directly detects and cuts and cleans the damage site by the movable robot. On the one hand, it does not need to use the “first teaching and then cutting” mode, so that the cutting and cleaning efficiency of the damage site can be improved. On the other hand, the movable robot can work for a long time, and continuous drilling and cleaning can be realized. Attached Figure Description

[0058] Figure 1 This is a flowchart illustrating the steps of an in-situ damage cleaning method for an aircraft titanium alloy skeleton based on a mobile robot, according to an embodiment of this application.

[0059] Figure 2 This is a schematic diagram illustrating the structure of an in-situ damage cleaning system for an aircraft titanium alloy frame according to an embodiment of this application;

[0060] Figure 3 This is a schematic diagram illustrating a mobile repair platform according to an embodiment of this application;

[0061] Figure 4 This is a schematic diagram illustrating the principle of surface point cloud slicing processing in an embodiment of this application;

[0062] Figure 5 This is a schematic diagram illustrating the voxel grid filtering principle of an embodiment of this application;

[0063] Figure 6 This is a schematic diagram illustrating the three-dimensional reconstruction effect of the surface damage point cloud of a 316L steel plate according to an embodiment of this application.

[0064] Figure 7 This is a macroscopic morphology diagram showing the optimal cutting quality of the TC4-M titanium alloy skin in the embodiments of this application;

[0065] Figure 8 This is a schematic diagram showing the cut microstructure characteristics and abnormal microstructure areas of the TC4-M titanium alloy skin with the best cut quality according to the embodiments of this application.

[0066] Figure 9 This is a macroscopic morphology diagram showing the medium cutting quality of the TC4-M titanium alloy skin in the embodiments of this application;

[0067] Figure 10 This is a schematic diagram showing the cut microstructure characteristics and abnormal microstructure areas of a TC4-M titanium alloy skin with moderate cut quality according to an embodiment of this application.

[0068] Figure 11 This is a macroscopic morphological image showing the poor cutting quality of the TC4-M titanium alloy skin in an embodiment of this application;

[0069] Figure 12 This is a schematic diagram showing the cut microstructure characteristics and abnormal microstructure areas of TC4-M titanium alloy skin with poor cut quality according to an embodiment of this application.

[0070] Figure 13 This is a macroscopic morphology diagram showing the TA15-M titanium alloy skin after cutting according to an embodiment of this application;

[0071] Figure 14is a schematic diagram of cutting tissue characteristics and abnormal tissue areas of a TA15-M titanium alloy skin of an embodiment of the application;

[0072] Figure 15 is a structural block diagram of an aircraft titanium alloy framework in-situ damage cleaning system based on a mobile robot according to an embodiment of the application. DETAILED DESCRIPTION

[0073] To make the technical problems, technical solutions and advantages of the application clearer, specific embodiments will be described in detail below with reference to the drawings and specific embodiments.

[0074] An aircraft titanium alloy framework in-situ damage cleaning scheme based on a mobile robot is provided in the embodiment of the application, and the damaged part is cleaned through interaction of a mobile robot integrated with a laser cutting module and a machine vision module and an upper computer. In work, the mobile robot freely moves to the vicinity of the damaged aircraft skin, and based on the mechanical arm driving the laser cutting head, the damaged skin, framework and other titanium alloy parts are cleaned, solving the problems of high difficulty and long cycle of aircraft titanium alloy skin, framework and other titanium alloy parts replacement, mechanical cleaning. Combined with machine vision technology, the problems of complex operation and low repair efficiency of industrial robots "first demonstration, then cutting" are solved.

[0075] The aircraft titanium alloy framework in-situ damage cleaning scheme based on a mobile robot provided by the embodiment of the application will be described in detail below with reference to the drawings, specific embodiments and application scenarios.

[0076] As shown in the accompanying Figure 1 The aircraft titanium alloy framework in-situ damage cleaning method based on a mobile robot of the embodiment of the application includes the following steps:

[0077] Step 101: The mobile robot acquires image information by shooting through the machine vision module, and sends the image information to the upper computer.

[0078] The aircraft titanium alloy framework in-situ damage cleaning method based on a mobile robot provided by the embodiment is executed by an aircraft titanium alloy framework in-situ damage cleaning system, and the damage cleaning system includes a mobile robot integrated with a laser cutting module and a machine vision module and an upper computer.

[0079] Figure 2 is a structural schematic diagram of an aircraft titanium alloy framework in-situ damage cleaning system according to an embodiment of the application. As Figure 2As shown, the system is based on the "robot + machine vision" technology, by integrating the robot, laser cutting module, machine vision module and the like on the mobile mechanism (such as a mobile repair platform), configured into the mobile robot shown in the embodiments of the application, the mobile robot can complete the in-situ damage cleaning of the aircraft titanium alloy parts by interacting with the upper computer. The laser cutting module and the machine vision module are installed on the mechanical arm of the mobile robot to act as the eyes and hands of humans. The vision software system is installed on the upper computer, which can process images and point clouds, and can build three-dimensional models and perform cutting path planning. The control cabinet is the "brain" of the mobile robot, used to control the movement of the mobile repair platform, control the movement of the mechanical arm, and control the work of the laser cutting module and the machine vision module.

[0080] The mobile robot provided by the embodiments of the application can be considered as a mobile repair platform with a mobile platform as a carrier, a plurality of modules in cooperation, strong mobility, and multi-degree-of-freedom operation of the laser in-situ damage cleaning platform. As shown in the figure, Figure 3 As shown, the mobile repair platform can select a wheeled platform with a maximum moving speed of 10Km / h, has a certain obstacle crossing ability, and can be controlled by remote control to move the platform, to ensure that it can quickly respond to the actual damage situation and repair requirements.

[0081] The robot system of the mobile robot is composed of an ABB IRB4600 six-degree-of-freedom industrial robot, an IRC5 control system and a FlexPendant teaching pendant, and has a working range of 2.05m, can realize large-range movement, and meets the complex repair environment on site; the effective arm load is 20Kg, which can ensure the strength requirement of the modules installed at the end of the mechanical arm.

[0082] The machine vision module is based on the characteristics of the vision sensor, and the structured light sensor and the line laser sensor are combined and installed at the front end of the mechanical arm to build an eye-in-hand mode hand-eye system to realize data scanning and acquisition. This module mainly includes a structured light sensor, a line laser sensor and a flange.

[0083] The structured light sensor projects the surface structured light to the workpiece surface, acquires image information through the camera, realizes the detection of the workpiece surface information, has the characteristics of large scanning range and fast scanning speed. The structured light sensor acquires the picture of the damaged part to realize rapid positioning of the damage. An example of a structured light sensor model can be SA-T1000, which has a near-field FOV of 700x600mm and a far-field FOV of 4000x3000mm, can realize rapid positioning of damage under a large viewing angle, and realizes communication connection with the upper computer through a gigabit Ethernet to ensure real-time transmission of data.

[0084] In actual implementation, the machine vision module can be installed at the front end of the mechanical arm of the mobile robot, and the laser cutting module is installed at the bottom end of the mechanical arm.

[0085] Step 102: The host computer locates the damage site of the aircraft titanium alloy component based on the image information, and feeds back the information of the damage site of the aircraft titanium alloy component to the mobile robot.

[0086] The image information includes the surface information of the aircraft titanium alloy component, and the aircraft titanium alloy component includes a skin, a skeleton and the like.

[0087] Step 103: The mobile robot calls the machine vision module, scans the damage site by laser to obtain the first point cloud information of the damage site, and sends the point cloud information to the host computer.

[0088] The line laser sensor is composed of a laser emitter and a camera, and has the characteristics of simple structure, high precision and easy image processing. Based on these characteristics, after the structured light sensor has positioned the damage site, the project scans the damage site by the line laser sensor and takes pictures of the line laser state in real time, and obtains the three-dimensional point cloud information of the damage site through the change of the laser stripe information.

[0089] Step 104: The host computer constructs a first three-dimensional model corresponding to the damage site based on the first point cloud information.

[0090] The vision software system is composed of a camera acquisition module, a system calibration module, a point cloud processing module and auxiliary software. Through the vision algorithm, automatic recognition of broken hole damage, automatic planning of cutting path and automatic programming can be realized. The camera acquisition module is used to acquire the image information obtained by the camera; the system calibration module is used to unify the image coordinates of the camera to the robot base coordinates; the point cloud processing module is used to analyze the three-dimensional point cloud information collected by the line laser sensor, i.e. the first point cloud information.

[0091] The visual software system runs on the host computer. The system first realizes image capture, color setting, image saving and other operations through the camera during runtime. Then, the system calibration is completed through camera calibration, light plane calibration and hand-eye calibration, and the image coordinates of the camera are unified to the robot base coordinates, which is the premise of realizing accurate measurement of damaged objects by the robot. The point cloud processing module includes point cloud filtering function, point cloud visualization function and point cloud size and color parameter function. Common point cloud data processing methods include point cloud data preprocessing and three-dimensional point cloud reconstruction. Based on the data processed by the point cloud processing module, the three-dimensional triangular mesh of the point cloud file is edited, rendered and screened by using a mesh processing system such as Meshlab. At the same time, three-dimensional modeling of the damaged part is realized based on reverse engineering software, and layered slices are generated, which lays a foundation for the generation of cladding repair path. The principle diagram of surface point cloud slice processing is shown in FIG. 8. Figure 4

[0092] An optional way for the host computer to construct a first three-dimensional model corresponding to the damaged part based on the first point cloud information can include the following sub-steps:

[0093] Sub-step 1: The host computer pre-processes the first point cloud information to filter out invalid points and outliers to obtain second point cloud information.

[0094] An example way for the host computer to pre-process the first point cloud information to filter out invalid points and outliers to obtain second point cloud information can be as follows:

[0095] For each point in the first point cloud information, it is judged whether the coordinate of the point in the first dimension is within a preset first dimension value range. If not, the point is regarded as an invalid point, and if yes, it is regarded as a valid point. The invalid points are filtered out. The average value of the distance between the point and each adjacent point is calculated by traversing each point in the first point cloud information. It is judged whether the average value is within a preset distance range threshold. If not, the point and each adjacent point of the point are determined as outliers, and each outlier is filtered out. The first point cloud information after filtering out the invalid points and outliers is determined as the second point cloud information.

[0096] It should be noted that the specific values of the preset first dimension value range and the preset distance range threshold can be flexibly set by those skilled in the art, and the present application does not make specific limitations thereto.

[0097] ​Sub-step 1 is point cloud data preprocessing: irrelevant points and outliers can be filtered out based on two methods of straight-through filtering and statistical filtering in the point cloud library. The principle of the irrelevant point filtering algorithm of the straight-through filtering is to specify one dimension of X, Y and Z and the value range in this dimension, traverse each point in the point cloud, judge whether the point is in the value range, and remove the points outside the value range. The principle of the statistical filtering algorithm is to traverse each point in the point cloud, calculate the average distance of all neighboring points, and remove outliers by comparing the average distance with the standard range threshold to retain the filtered point cloud data. The removal of outliers is realized by the statistical filter in the point cloud library.

[0098] Sub-step 2: the second point cloud information is filtered by a voxel grid filtering algorithm to obtain third point cloud information.

[0099] Sub-step 2 is further simplification of the preprocessed point cloud data.

[0100] A large number of redundant data points will increase the algorithm running time and thus affect the three-dimensional reconstruction efficiency, so the preprocessed point cloud data needs to be simplified. For example, a voxel grid filtering algorithm can be used. The principle of the algorithm is to divide the point cloud data by creating a fixed-size voxel grid, calculate the center of gravity of all points in the grid, delete the grid without point cloud data, and approximate all points in each grid as the center of gravity. The set of all voxel centers represents the point cloud data filtered by the voxel grid, which can preserve the original geometric structure of the point cloud without changing the shape characteristics of the point cloud. The principle of the voxel grid filtering is shown in the accompanying Figure 5

[0101] Sub-step 3: a preset point cloud reconstruction algorithm is used to perform point cloud three-dimensional reconstruction on the third point cloud information to obtain a first three-dimensional model.

[0102] An optional way of using a preset point cloud reconstruction algorithm to perform point cloud three-dimensional reconstruction on the third point cloud information to obtain a first three-dimensional model can be as follows:

[0103] A plurality of point cloud reconstruction algorithms are used to perform point cloud three-dimensional reconstruction on the third point cloud information to obtain a plurality of three-dimensional models; the model with the highest score is selected from the plurality of three-dimensional models to determine the first three-dimensional model.

[0104] The plurality of preset point cloud reconstruction algorithms can include but are not limited to Delaunay, Possion and rolling ball method.

[0105] ​Common point cloud three-dimensional reconstruction algorithms mainly include Delaunay, Possion and rolling ball method. Delaunay is a kind of triangulation algorithm, which inserts a point into a selected region containing the point set, searches and deletes adjacent triangles to form a Delaunay cavity, and then connects the point with each vertex in the cavity to form a new triangular mesh. Possion is a reconstruction method based on Possion mesh, which is to fit the surface indicator function of the point cloud with normal data, and to extract the isosurface by setting the threshold to realize the three-dimensional reconstruction of the point cloud. The rolling ball method is a local region growing three-dimensional reconstruction algorithm, which first defines a ball, and then defines the ball in the seed triangle to rotate and roll until it contacts the next point. The edge and the point form a triangle, and the surface triangle reconstruction can be completed by continuous iteration. For the point cloud data of the structural part damage pretreatment, the point cloud surface three-dimensional reconstruction is carried out based on the three point cloud reconstruction algorithms of Delaunay, Possion and rolling ball method, the reconstruction effect is compared and analyzed, and the optimal three-dimensional point cloud reconstruction algorithm is determined. Figure 6 Fig. 1 is a schematic diagram of the point cloud surface three-dimensional reconstruction effect of the 316L steel plate surface damage according to an embodiment of the present application, Figure 6 (a) is a schematic diagram of the reconstruction effect of the Delaunay reconstruction method, Figure 6 (b) is a schematic diagram of the reconstruction effect of the Possion reconstruction method, Figure 6 (c) is a schematic diagram of the reconstruction effect of the rolling ball method.

[0106] Step 105: The host computer generates cutting planning information based on the first three-dimensional model and the movable robot mechanical arm posture information, and sends the cutting planning information to the movable robot.

[0107] The cutting planning information includes a cutting planning path and a cutting process parameter.

[0108] An optional way for the host computer to generate cutting planning information based on the first three-dimensional model and the movable robot mechanical arm posture information can include the following sub-steps:

[0109] Sub-step 1: The host computer identifies the damage site topography based on the first three-dimensional model to obtain the damage site topography.

[0110] The damage site topography can include but is not limited to: groove type, flat type, curved surface type and hole type.

[0111] Sub-step 2: Determine the boundary size of the damage site topography.

[0112] Sub-step 3: Generate a second three-dimensional model based on the boundary size and a preset boundary correction rule.

[0113] The boundary correction rule can be flexibly set by those skilled in the art, for example, set to be indented along the boundary by a preset size, which can be 2 cm, 1 cm or 5 mm, etc., and the present application embodiment does not make specific limitation thereto.

[0114] Sub-step 4: determining the cutting process parameters based on the attribute of the damage site;

[0115] The attribute of the damage site can be determined based on the type of the workpiece where the damage site is located and the morphology of the damage site. Different attributes correspond to different cutting process parameters, and the system has a preset correspondence between different attributes and cutting process parameters. After determining the attribute of the damage site, the cutting process parameters can be determined based on the preset correspondence.

[0116] For the aircraft TC4-M titanium alloy material with a thickness of 2 mm to 5 mm, the process parameter range of laser cutting is shown in Table 1:

[0117] Table 1: Laser cutting process parameters of TC4-M titanium alloy

[0118]

[0119] For the aircraft TA15-M titanium alloy material with a thickness of 2 mm to 7 mm, the process parameter range of laser cutting is shown in Table 2:

[0120] Table 2: Laser cutting process parameters of TA15-M titanium alloy

[0121]

[0122] In the cutting size identification, a circle (such as 20 mm, 30 mm, 40 mm, 50 mm and 60 mm, etc.), an oblong or a rectangle with a specific size rule is drawn according to the principle that the broken hole exceeds 5 mm on one side, so as to remove the residual micro-cracks and corners on the edge of the broken hole; after cutting, the surface oxide film and attached residues are removed, and the shape is polished according to the requirements of smooth transition and smoothness of the cutting surface.

[0123] Sub-step 5: generating a cutting planning path according to the second three-dimensional model and the posture information of the movable robot mechanical arm.

[0124] Step 106: The movable robot controls the laser cutting module to cut and clean the damage site according to the cutting planning information.

[0125] The method for cleaning in-situ damage of an aircraft titanium alloy framework based on a mobile robot provided by the embodiment of the application is applied to an aircraft titanium alloy framework in-situ damage cleaning system. The damage cleaning system comprises a mobile robot integrated with a laser cutting module and a machine vision module and a host computer. In actual work, the mobile robot captures image information through the machine vision module and sends the image information to the host computer. The host computer locates the damage position of the aircraft titanium alloy component based on the image information and feeds back the damage position information of the aircraft titanium alloy component to the mobile robot. The mobile robot calls the machine vision module, scans the damage position through a laser, obtains first point cloud information of the damage position, and sends the point cloud information to the host computer. The host computer constructs a first three-dimensional model corresponding to the damage position based on the first point cloud information. The host computer generates cutting planning information based on the first three-dimensional model and the mechanical arm posture information of the mobile robot, and sends the cutting planning information to the mobile robot. The mobile robot controls the laser cutting module to cut and clean the damage position according to the cutting planning information. The aircraft titanium alloy framework in-situ damage cleaning scheme based on a mobile robot disclosed in the application can directly detect and cut and clean the damage position through the mobile robot. On the one hand, the cutting and cleaning efficiency of the damage position can be improved without using the "first demonstration and then cutting" mode. On the other hand, the mobile robot can work for a long time and can realize continuous drilling and cleaning.

[0126] The method for cleaning in-situ damage of an aircraft titanium alloy framework based on a mobile robot provided by the embodiment of the application is applied to an aircraft titanium alloy framework in-situ damage cleaning system. The damage cleaning system comprises a mobile robot integrated with a laser cutting module and a machine vision module and a host computer. In actual work, the mobile robot captures image information through the machine vision module and sends the image information to the host computer. The host computer locates the damage position of the aircraft titanium alloy component based on the image information and feeds back the damage position information of the aircraft titanium alloy component to the mobile robot. The mobile robot calls the machine vision module, scans the damage position through a laser, obtains first point cloud information of the damage position, and sends the point cloud information to the host computer. The host computer constructs a first three-dimensional model corresponding to the damage position based on the first point cloud information. The host computer generates cutting planning information based on the first three-dimensional model and the mechanical arm posture information of the mobile robot, and sends the cutting planning information to the mobile robot. The mobile robot controls the laser cutting module to cut and clean the damage position according to the cutting planning information. The aircraft titanium alloy framework in-situ damage cleaning scheme based on a mobile robot disclosed in the application can directly detect and cut and clean the damage position through the mobile robot. On the one hand, the cutting and cleaning efficiency of the damage position can be improved without using the "first demonstration and then cutting" mode. On the other hand, the mobile robot can work for a long time and can realize continuous drilling and cleaning.

[0127] In the embodiment of the application, cutting of the damaged position on the TC4-M titanium alloy skin is taken as an example for illustration.

[0128] The best cutting effect of the TC4-M titanium alloy skin cutting (4mm thickness) is as follows:

[0129] 1) Macroscopic morphology of cutting

[0130] Figure 7 is a macroscopic morphology diagram of the TC4-M titanium alloy skin cutting with the best cutting quality in the embodiment of the application;

[0131] The cross-section macroscopic morphology diagram is taken by a body microscope. It can be seen that the structure of the cross section is nitrided at high temperature and the color becomes gray-black. According to the different cross-section roughness, it can be divided into three typical regions from top to bottom: smooth region, rough region and slag region. The cross section with the best cutting quality is smoother and has less slag.

[0132] Figure 8 is a cutting structure feature and abnormal structure area diagram of the TC4-M titanium alloy skin with the best cutting quality in the embodiment of the application;

[0133] Microstructure of the laser cutting section includes three typical regions: remelted zone, heat affected zone and base material. The metallographic sample of the cutting section is prepared, the sample is inlaid by using a Q-2B type metallographic sample inlaid machine, is polished by using sandpaper with a mesh size of 60 to 3000, is etched after polishing by using an MP-1S type metallographic polishing machine, and the microstructure diagram is observed by using a metallographic microscope. As can be seen, the base material structure of the TC4 plate is primary equiaxed alpha phase and intergranular beta phase; the remelted zone is fine needle-shaped martensite, and no beta phase is found; the typical structure of the heat affected zone is primary equiaxed alpha phase, black beta phase and a small amount of needle-shaped alpha' phase, and the needle-shaped alpha' phase is finer. Phase transition occurs in the remelted zone and the heat affected zone, which can be collectively referred to as the abnormal structure zone. When the laser cutting quality is the best, the width of the abnormal structure zone produced is also the smallest.

[0134] The cutting static strength impact data is shown in Table 4, and the tensile strength of the TC4 base material (without damage) is 1026 MPa

[0135] Table 4

[0136]

[0137] The cutting effect of the TC4-M titanium alloy skin cutting (4mm thickness) with medium cutting quality is as follows:

[0138] Figure 9 is a macroscopic morphology diagram representing the TC4-M titanium alloy skin cutting with medium cutting quality according to the embodiment of the present application; Figure 10 is a cutting structure feature and abnormal structure zone diagram of the TC4-M titanium alloy skin with medium cutting quality according to the embodiment of the present application; the cutting static strength impact data is shown in Table 5:

[0139] Table 5

[0140]

[0141] The cutting effect of the TC4-M titanium alloy skin cutting (4mm thickness) with the worst cutting quality is as follows:

[0142] Figure 11 is a macroscopic morphology diagram representing the TC4-M titanium alloy skin cutting with poor cutting quality according to the embodiment of the present application; Figure 12 is a cutting structure feature and abnormal structure zone diagram of the TC4-M titanium alloy skin with poor cutting quality according to the embodiment of the present application; the cutting static strength impact data is shown in Table 6:

[0143] Table 6

[0144]

[0145] The method for cleaning in-situ damage of a titanium alloy framework of an aircraft based on a mobile robot provided by the embodiment of the application is described below with a specific example.

[0146] The embodiment of the application is described by taking cutting of a damaged part on a TA15-M titanium alloy skin as an example.

[0147] The cutting effect of the TA15-M titanium alloy skin (4mm thickness) is as follows:

[0148] Figure 13 is a macroscopic morphology diagram of the TA15-M titanium alloy skin after cutting according to the embodiment of the application; the cross-sectional macroscopic morphology diagram is taken by a stereo microscope, and it can be seen that the microstructure of the cross section is nitrided at high temperature, and the color becomes gray-black. According to the different roughness of the cross section, it can be divided into three typical regions from top to bottom: smooth region, rough region and slag hanging region. The cross section with the best cutting quality is smoother, and the amount of slag is less

[0149] Figure 14 is a diagram of the cutting microstructure characteristics and abnormal microstructure region of the TA15-M titanium alloy skin according to the embodiment of the application; the microstructure of the cross section after laser cutting includes three typical regions: remelted region, heat affected zone and matrix. The cross section is metallographic sample preparation, the sample is inlaid by using a Q-2B type metallographic sample inlaying machine, the sandpaper with a mesh size of 60 to 3000 is used for polishing, and the microstructure diagram is observed by using a metallographic microscope after etching by using an MP-1S type metallographic polishing machine. It can be seen that the matrix microstructure of the TA15 plate is equiaxed microstructure with a large amount of α phase, as shown in (b); the remelted region is acicular martensite, the typical microstructure of the heat affected zone is primary α phase and a small amount of acicular martensite, and the acicular martensite phase is smaller, as shown in (c). The phase change occurs in the remelted region and the heat affected zone, which can be collectively referred to as the abnormal microstructure region. When the laser cutting quality is the best, the width of the abnormal microstructure region is also the smallest.

[0150] The cutting static strength influence data is shown in Table 6:

[0151] Table 6

[0152]

[0153] The aircraft TC4-M / TA15-M titanium alloy framework damage cleaning method provided by the specific example can be implemented in situ on the damaged aircraft, thereby overcoming the problems of high difficulty, long cycle and great spare part support pressure of traditional spare part repair; secondly, the method can replace the in-situ mechanical cleaning method of the aircraft TA15 titanium alloy framework, thereby solving the problems of high difficulty and long cycle of titanium alloy mechanical cleaning and the problem of programming and teaching of traditional lasers; thirdly, the cutting efficiency of the method is high, for a 50mm hole, the program can be automatically called, and the cutting can be completed in 2min, thereby greatly improving the cutting efficiency, and the influence on the static strength after cutting is within 10%, which is within the acceptable range in practical application.

[0154] Figure 15 To realize the aircraft titanium alloy framework in-situ damage cleaning system based on a mobile robot according to an embodiment of the present application, a structural block diagram is provided.

[0155] The aircraft titanium alloy framework in-situ damage cleaning system based on a mobile robot provided by the embodiments of the present application comprises a mobile robot 201 integrated with a laser cutting module and a machine vision module and a host computer 202.

[0156] The mobile robot 201 comprises a control cabinet 2011, a machine vision module 2012 and a laser cutting module 2013.

[0157] The control cabinet 2011 is configured to call the machine vision module 2012 to capture image information and send the image information to the host computer 202.

[0158] The host computer 202 is configured to locate the damage site of the aircraft titanium alloy component based on the image information and feed back the information of the damage site of the titanium alloy component to the mobile robot, wherein the image information comprises surface information of the aircraft titanium alloy component, and the aircraft titanium alloy component comprises at least one of a skin and a framework.

[0159] The control cabinet 2011 is further configured to call the machine vision module 2012 to obtain first point cloud information of the damage site by laser scanning the damage site and send the point cloud information to the host computer 202.

[0160] The host computer 202 is further configured to construct a first three-dimensional model corresponding to the damage site based on the first point cloud information, generate cutting planning information based on the first three-dimensional model and mechanical arm posture information of the mobile robot, and send the cutting planning information to the control cabinet of the mobile robot, wherein the cutting planning information comprises a cutting planning path and cutting process parameters.

[0161] The control cabinet 2011 is further configured to control the laser cutting module 2013 to cut and clean the damaged part according to the cutting planning information.

[0162] Optionally, the upper computer comprises:

[0163] The point cloud processing module is configured to: pre-process the first point cloud information to filter out invalid points and outliers to obtain second point cloud information; filter the second point cloud information using a voxel grid filtering algorithm to obtain third point cloud information; and perform point cloud three-dimensional reconstruction on the third point cloud information using a preset point cloud reconstruction algorithm to obtain the first three-dimensional model.

[0164] Optionally, when the point cloud processing module pre-processes the first point cloud information to filter out invalid points and outliers to obtain second point cloud information, the point cloud processing module is specifically configured to:

[0165] For each point in the first point cloud information, it is determined whether the coordinate of the point in the first dimension is within a preset first dimension value range.

[0166] If not, the point is regarded as an invalid point, and the invalid point is filtered out.

[0167] The point cloud processing module is configured to: pre-process the first point cloud information to filter out invalid points and outliers to obtain second point cloud information; filter the second point cloud information using a voxel grid filtering algorithm to obtain third point cloud information; and perform point cloud three-dimensional reconstruction on the third point cloud information using a preset point cloud reconstruction algorithm to obtain the first three-dimensional model.

[0168] If not, the point and each adjacent point of the point are determined as outliers, and each outlier is filtered out.

[0169] The first point cloud information after filtering out invalid points and outliers is determined as the second point cloud information.

[0170] Optionally, when the point cloud processing module performs point cloud three-dimensional reconstruction on the third point cloud information using a preset point cloud reconstruction algorithm to obtain the first three-dimensional model, the point cloud processing module is specifically configured to:

[0171] The third point cloud information is reconstructed using a plurality of preset point cloud reconstruction algorithms to obtain a plurality of three-dimensional models.

[0172] The model with the highest score is selected from the plurality of three-dimensional models to determine the first three-dimensional model.

[0173] Optionally, the upper computer comprises:

[0174] The cutting planning module is configured to perform damage site topography identification based on the first three-dimensional model to obtain a damage site topography, determine a boundary size of the damage site topography, generate a second three-dimensional model based on the boundary size and a preset boundary correction rule, determine a cutting process parameter based on an attribute of the damage site, and generate a cutting planning path according to the second three-dimensional model and the movable robot mechanical arm posture information.

[0175] Optionally, the machine vision module is installed at a front end of a mechanical arm of the movable robot, and the laser cutting module is installed at a bottom end of the mechanical arm.

[0176] The movable robot obtains image information by shooting through the machine vision module and sends the image information to the upper computer. The upper computer locates a damage site of the aircraft titanium alloy component based on the image information and feeds back the damage site information of the aircraft titanium alloy component to the movable robot. The movable robot calls the machine vision module, scans the damage site by laser, obtains first point cloud information of the damage site, and sends the point cloud information to the upper computer. The upper computer constructs a first three-dimensional model corresponding to the damage site based on the first point cloud information. The upper computer generates cutting planning information based on the first three-dimensional model and the movable robot mechanical arm posture information, sends the cutting planning information to the movable robot, and controls the laser cutting module to cut and clean the damage site according to the cutting planning information. Since the movable robot directly detects and cuts and cleans the damage site, on the one hand, the cutting and cleaning efficiency of the damage site can be improved without using the mode of “first demonstration and then cutting”, and on the other hand, the movable robot can work for a long time and can realize continuous drilling and cleaning.

[0177] The embodiment of the present application further provides an electronic device, including a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete mutual communication through the communication bus.

[0178] The memory is used for storing a computer program.

[0179] The processor is used for executing the program stored on the memory to realize each step of the aircraft titanium alloy framework in-situ damage cleaning performed by the upper computer in the method embodiment.

[0180] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For the convenience of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or only one type of bus.

[0181] The communication interface is used for communication between the terminal and other devices.

[0182] The memory can include a Random Access Memory (RAM) and can also include a non-volatile memory, such as at least one disk memory. Optionally, the memory can also be at least one storage device located away from the aforementioned processor.

[0183] The processor mentioned above can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc. It can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component.

[0184] In another embodiment provided by the application, a computer readable storage medium is also provided, which stores instructions, when running on a computer, causes the computer to implement the method steps performed by the host computer in any of the above embodiments.

[0185] It should be noted that, in this document, the terms "comprise", "contain" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such a process, method, article or device. Without more limitations, the element defined by the statement "comprises a" does not exclude the presence of another identical element in the process, method, article or device that includes the element.

[0186] The above describes the preferred embodiments of the present application. It should be noted that, for those skilled in the art, several improvements and refinements can be made without departing from the principles of the present application, and these improvements and refinements should also be considered as falling within the scope of the present application.

Claims

1. A method for in-situ damage cleaning of aircraft titanium alloy frames based on mobile robots, characterized in that, The application is applied to an in-situ damage cleaning system of a titanium alloy framework of an airplane, and the system comprises a movable robot integrated with a laser cutting module and a machine vision module and a host computer, and the method comprises the following steps: The movable robot shoots and acquires image information through the machine vision module, and sends the image information to the host computer; The host computer locates a damage position of a titanium alloy component of the airplane based on the image information, and feeds back the damage position information of the titanium alloy component of the airplane to the movable robot; wherein, the image information comprises surface information of the titanium alloy component of the airplane; wherein, the titanium alloy component of the airplane comprises at least one of a skin and a framework; The movable robot calls the machine vision module, scans the damage position through a laser, obtains first point cloud information of the damage position, and sends the point cloud information to the host computer; The host computer constructs a first three-dimensional model corresponding to the damage position based on the first point cloud information; The host computer generates cutting planning information based on the first three-dimensional model and mechanical arm posture information of the movable robot, and sends the cutting planning information to the movable robot, wherein, the cutting planning information comprises a cutting planning path and cutting process parameters; The movable robot controls the laser cutting module to cut and clean the damage position according to the cutting planning information.

2. The method of claim 1, wherein, The step of constructing a first three-dimensional model corresponding to the damage position based on the first point cloud information by the host computer comprises the following steps: The host computer pre-processes the first point cloud information to filter out invalid points and outliers, and obtains second point cloud information; A voxel grid filtering algorithm is adopted to filter the second point cloud information, and third point cloud information is obtained; A preset point cloud reconstruction algorithm is adopted to perform point cloud three-dimensional reconstruction on the third point cloud information, and the first three-dimensional model is obtained.

3. The method of claim 2, wherein, The step of pre-processing the first point cloud information by the host computer to filter out invalid points and outliers, and obtaining second point cloud information comprises the following steps: For each point in the first point cloud information, it is judged whether the coordinate of the point in the first dimension is within a preset first dimension value range; If not, the point is regarded as an invalid point, and the invalid point is filtered out; The average value of the distance between the point and each adjacent point in the first point cloud information is calculated, and it is judged whether the average value is within a preset distance range threshold; If not, the point and each adjacent point are determined as outliers, and each outlier is filtered out; The first point cloud information after filtering out invalid points and outliers is determined as the second point cloud information.

4. The method of claim 2, wherein, The step of adopting a preset point cloud reconstruction algorithm to perform point cloud three-dimensional reconstruction on the third point cloud information, and obtaining the first three-dimensional model comprises the following steps: A plurality of point cloud reconstruction algorithms are adopted to perform point cloud three-dimensional reconstruction on the third point cloud information, and a plurality of three-dimensional models are obtained; The model with the highest score is selected from the plurality of three-dimensional models, and is determined as the first three-dimensional model.

5. The method of claim 1, wherein, The step of generating cutting planning information by the host computer based on the first three-dimensional model and the mechanical arm posture information of the movable robot comprises the following steps: The host computer identifies a damage site morphology based on the first three-dimensional model, and obtains the damage site morphology; determining the boundary size of the damage site morphology; generating a second three-dimensional model based on the boundary size and a preset boundary correction rule; determining the cutting process parameters based on the attributes of the damage site; generating a cutting planning path according to the second three-dimensional model and the movable robot mechanical arm posture information.

6. The method of claim 1, wherein, The machine vision module is installed at the front end of the mechanical arm of the movable robot; The laser cutting module is installed at the bottom end of the mechanical arm.

7. A mobile robot-based in-situ damage cleaning system for aircraft titanium alloy frames, comprising: The system comprises a movable robot integrated with a laser cutting module and a machine vision module, and a host computer; The movable robot comprises a control cabinet, a machine vision module, and a laser cutting module: The control cabinet is used to call the machine vision module to capture image information and send the image information to the host computer; The host computer is used to locate the damage site of the aircraft titanium alloy component based on the image information, and feed back the damage site information of the aircraft titanium alloy component to the movable robot; wherein the image information comprises surface information of the aircraft titanium alloy component; wherein the aircraft titanium alloy component comprises at least one of a skin and a skeleton; The control cabinet is also used to call the machine vision module to obtain first point cloud information of the damage site by laser scanning the damage site, and send the point cloud information to the host computer; The host computer is also used to construct a first three-dimensional model corresponding to the damage site based on the first point cloud information, generate cutting planning information based on the first three-dimensional model and the movable robot mechanical arm posture information, and send the cutting planning information to the control cabinet of the movable robot, wherein the cutting planning information comprises a cutting planning path and cutting process parameters; The control cabinet is also used to control the laser cutting module to cut and clean the damage site according to the cutting planning information.

8. The system of claim 7, wherein, The host computer comprises: A point cloud processing module is used to pre-process the first point cloud information to filter out invalid points and outliers, obtain second point cloud information, filter the second point cloud information using a voxel grid filtering algorithm to obtain third point cloud information, and perform point cloud three-dimensional reconstruction on the third point cloud information using a preset point cloud reconstruction algorithm to obtain the first three-dimensional model.

9. The system of claim 8, wherein, When the point cloud processing module pre-processes the first point cloud information to filter out invalid points and outliers to obtain second point cloud information, it is specifically used for: For each point in the first point cloud information, it is judged whether the coordinates of the point in the first dimension are within a preset first dimension value range; If not, the point is regarded as an invalid point, and the invalid point is filtered out; Iterate through each point in the first point cloud information, calculate the average value of the distance between the point and each adjacent point, and judge whether the average value is within a preset distance range threshold; If not, the point and each adjacent point are determined as outliers, and each outlier is filtered out; The first point cloud information after filtering out invalid points and outliers is determined as the second point cloud information.

10. The system of claim 7, wherein, The host computer comprises: The cutting planning module is configured to perform damage site topography identification based on the first three-dimensional model to obtain a damage site topography, determine a boundary size of the damage site topography, generate a second three-dimensional model based on the boundary size and a preset boundary correction rule, determine a cutting process parameter based on an attribute of the damage site, and generate a cutting planning path according to the second three-dimensional model and the movable robot mechanical arm posture information.

Images