Casting cleaning robot self-adaptive cleaning method based on machine vision and trajectory planning

Abstract

The invention discloses a self-adaptive cleaning method of a casting cleaning robot based on machine vision and trajectory planning, which comprises the following steps: S1, acquiring point cloud data of a to-be-cleaned part of a casting by using a camera, and preprocessing the point cloud data; s2, performing hand-eye calibration on the casting cleaning robot, and converting coordinates in a camera coordinate system into a robot coordinate system; and S3, under the robot coordinate system, the cleaning process of the casting cleaning robot is divided into three stages of free space movement, transition space movement and limited space movement according to the corresponding working time, and cleaning tracks of the casting cleaning robot in all the movement stages are planned and set according to a track planning strategy. Therefore, the casting cleaning robot adjusts the cleaning track in a self-adaptive mode. According to the method, the cleaning track of the casting cleaning robot is optimized by using different track planning strategies; it is guaranteed that the casting cleaning robot is stable and free of impact in the moving process, and meanwhile the machining precision and stability of the casting cleaning robot for different curved surfaces are further improved.

Description

Technical Field

[0001] The invention relates to the technical field of machine vision-assisted robot casting cleaning, and in particular to a casting cleaning robot adaptive cleaning method based on machine vision and trajectory planning. Background Art

[0002] With the development of automation technology and robot control algorithms, industrial robots can replace humans in more and more scenarios to complete labor-intensive and harsh working environment work. Castings have low production costs and wide adaptability, covering almost all fields in industry and civil use. The characteristics of the casting process determine that many castings have the characteristics of large weight and volume, complex surface structure, and at the same time, high requirements are placed on the quality of castings in use. At present, in many cases, manual processing of castings is still required, which is not only inefficient, but also has a very harsh working environment. The final quality of the castings after cleaning cannot be guaranteed. Therefore, it is very necessary for industrial robots to replace manual cleaning of impurities on castings after casting, which will also become an important scenario for the application of robot automation technology.

[0003] Robot motion trajectory planning is an efficient automated process. This technology can reduce the downtime and maintenance costs of robots in actual production, improve the robot's operating accuracy and repeatability, and thus improve production efficiency and product quality. In the prior art, since the robot only moves along one trajectory during the automatic cleaning process, firstly, the robot will run for a long time before it contacts the casting, and it is difficult to improve the effective working time and work efficiency; secondly, the possibility of collision between the robot and the working environment or other objects will increase, causing damage or safety risks; then, the lack of reasonable trajectory planning will also increase the burden on the robot system, resulting in abnormalities or failures during operation, affecting the stability and reliability of the equipment, and resulting in the quality of the robot cleaning castings being unstable. Summary of the invention

[0004] The present invention provides an adaptive cleaning method for a casting cleaning robot based on machine vision and trajectory planning, so as to overcome the technical problems in the prior art of low work efficiency, safety risks and poor cleaning effect caused by the robot only moving along one trajectory during the automatic cleaning process.

[0005] In order to achieve the above object, the technical solution of the present invention is:

[0006] An adaptive cleaning method for a casting cleaning robot based on machine vision and trajectory planning, the specific steps comprising:

[0007] S1: using a camera to obtain point cloud data of the part to be cleaned of the casting, and preprocessing the point cloud data to obtain processed point cloud data, i.e., surface position information of the part to be cleaned of the casting;

[0008] S2: Perform hand-eye calibration on the casting cleaning robot to obtain a homogeneous change matrix of the camera and the casting cleaning robot in a base coordinate system, and transform the coordinates in the camera coordinate system into the robot coordinate system based on the homogeneous change matrix;

[0009] S3: In the robot coordinate system, the cleaning process of the casting cleaning robot is divided into three stages: free space movement, transition space movement and confined space movement according to the corresponding working time, and the cleaning trajectory of the casting cleaning robot in each movement stage is set according to the trajectory planning strategy, so that the casting cleaning robot can adaptively adjust the cleaning trajectory. The trajectory planning strategy includes:

[0010] The jtraj function is used to plan the path of the foundry cleaning robot when it is in the free space motion stage;

[0011] The Cartesian space planning method is used to plan the path and posture of the casting cleaning robot in the transition space motion stage.

[0012] The improved Bezier curve and inverse kinematics method are used to plan the path and posture of the casting cleaning robot when it is in the confined space motion stage.

[0013] Furthermore, the free space movement stage is a process in which the casting cleaning robot moves from any posture to directly above the cleaning starting point, and the casting cleaning robot itself is not restricted in its movement during this process;

[0014] The process of using the jtraj function to plan the path of the casting cleaning robot in the free space motion stage is:

[0015] The initial point position, end point position, interpolation steps, initial speed and end speed of the casting cleaning robot are set, and the jtraj function is input to obtain the cleaning trajectory of the casting cleaning robot. The expression of the jtraj function is:

[0016]

[0017] in, represents the trajectory after calculation, represents the speed after calculation, represents the acceleration after calculation, represents the initial point position, Indicates the end point position, step indicates the number of interpolation steps, Indicates the set initial speed, Indicates the set end speed.

[0018] Further, the transitional space motion stage is a transitional stage between the free space motion stage and the confined space motion stage;

[0019] The process of planning the path and posture of the casting cleaning robot in the transition space motion stage using the Cartesian space planning method is as follows:

[0020] The end point position of the casting cleaning robot in the free space motion phase is used as the starting point, and the end point position of the casting cleaning robot is set;

[0021] Setting the path trajectory of the casting cleaning robot from the starting point to the end point, and performing fitting processing on the path trajectory;

[0022] The fitted path trajectory is discretized into a number of path points, and the position information of the casting cleaning robot at the path point is obtained by using the trajectory difference algorithm and solving the three-dimensional spatial position information of each path point;

[0023] The position information of the casting cleaning robot at each path point is solved by the inverse kinematics method to obtain each joint angle of the casting cleaning robot;

[0024] The casting cleaning robot is controlled to move according to the path points and the corresponding joint angles.

[0025] Furthermore, the confined space movement stage is a stage in which the casting cleaning robot cleans various parts to be cleaned on the surface of the casting;

[0026] The process of planning the path and posture of the casting cleaning robot in the confined space motion stage using the improved Bezier curve and inverse kinematics method is as follows:

[0027] The improved Bezier curve is used to plan the path of the casting cleaning robot. The specific form of the improved Bezier curve is expressed as follows:

[0028]

[0029] Where p0, p1, …, p i ,…,p n Represents n control points on the Bezier curve, p i is the i-th control point; t is the parameter controlling the shape of the curve, and t=i / n+1;

[0030] Constructing auxiliary point p based on improved Bezier curve Cn , expressed as:

[0031]

[0032] According to the surface position information of the part to be cleaned of the casting, the inverse kinematics method is used to solve the posture information of the casting cleaning robot during the cleaning process, that is, the angles of each joint.

[0033] Furthermore, in S1, the point cloud data of the part to be cleaned of the casting is obtained, and the point cloud data is preprocessed to obtain the processed point cloud data.

[0034] S11: Use a 3D binocular camera to take photos of the casting to be cleaned to obtain corresponding point cloud data;

[0035] S12: performing filtering processing, noise reduction processing, edge feature extraction processing of the casting to-be-cleaned portion, and identification processing of the casting to-be-cleaned portion on the point cloud data in sequence, thereby obtaining surface position information of the casting to-be-cleaned portion.

[0036] Furthermore, in S2, the formula for converting the coordinates in the camera coordinate system to the robot coordinate system based on the homogeneous transformation matrix is:

[0037] p c =T CE T EB p B

[0038] In the formula, p B is the coordinate of the marker selected on the calibration plate in the calibration plate coordinate system; p c is the coordinate in the camera coordinate system; T CE T EB is the homogeneous transformation matrix; T CE is the transformation matrix between the robot end effector and the camera; T EB Represents the transformation matrix between the calibration plate coordinate system and the robot end effector coordinate system.

[0039] Beneficial effects: The present invention divides the cleaning process of the casting cleaning robot into three stages: free space movement, transition space movement and confined space movement, and uses different trajectory planning strategies to optimize the cleaning trajectory of the casting cleaning robot, so that the casting cleaning robot can adaptively clean castings; an improved Bezier curve is used to plan the cleaning trajectory of the casting cleaning robot in the confined space movement stage, ensuring that the casting cleaning robot is stable and impact-free during movement, reducing safety risks; and at the same time, further improving the processing accuracy, stability and work efficiency of the casting cleaning robot facing different curved surfaces. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0041] Figure 1 The present invention is a flowchart of an adaptive cleaning method of a casting cleaning robot based on machine vision and trajectory planning. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] This embodiment provides a casting cleaning robot adaptive cleaning method based on machine vision and trajectory planning, such as Figure 1 As shown, the specific steps include:

[0044] S1: using a 3D binocular camera to obtain point cloud data of the part to be cleaned of the casting, and preprocessing the point cloud data to obtain processed point cloud data, i.e., surface position information of the part to be cleaned of the casting;

[0045] In a specific embodiment, in S1, the point cloud data of the part to be cleaned of the casting is obtained, and the point cloud data is preprocessed to obtain the processed point cloud data.

[0046] S11: Use a 3D binocular camera to take photos of the casting to be cleaned to obtain corresponding point cloud data;

[0047] S12: performing filtering processing, noise reduction processing, edge feature extraction processing of the casting to be cleaned part and identification processing of the casting to be cleaned part on the point cloud data in sequence, so as to obtain surface position information of the casting to be cleaned part, that is, three-dimensional position information of the casting to be cleaned part.

[0048] Specifically, this embodiment obtains the three-dimensional position information of different parts to be cleaned on the casting surface by means of machine vision, ensuring that the robot moves smoothly without impact, so that the robot can achieve adaptive processing of different parts to be cleaned.

[0049] S2: Perform hand-eye calibration on the casting cleaning robot to obtain a homogeneous change matrix of the camera and the casting cleaning robot in a base coordinate system, and transform the coordinates in the camera coordinate system into the robot coordinate system based on the homogeneous change matrix;

[0050] In a specific embodiment, in S2, the formula for converting the coordinates in the camera coordinate system to the robot coordinate system based on the homogeneous transformation matrix is:

[0051] p c =T CE T EB p B

[0052] In the formula, p B is the coordinate of the marker selected on the calibration plate in the calibration plate coordinate system; p c is the coordinate in the camera coordinate system; T CE T EB is the homogeneous transformation matrix; T CE is the transformation matrix between the robot end effector and the camera; T EB Represents the transformation matrix between the calibration plate coordinate system and the robot end effector coordinate system.

[0053] S3: In the robot coordinate system, the cleaning process of the casting cleaning robot is divided into three stages: free space movement, transition space movement and confined space movement according to the corresponding working time, and the cleaning trajectory of the casting cleaning robot in each movement stage is set according to the trajectory planning strategy, so that the casting cleaning robot can adaptively adjust the cleaning trajectory. The trajectory planning strategy includes:

[0054] The jtraj function is used to plan the path of the foundry cleaning robot when it is in the free space motion stage;

[0055] In a specific embodiment, the process of using the jtraj function in the robotic toolbox to plan the path of the casting cleaning robot when it is in the free space motion stage is:

[0056] The initial point position, end point position, interpolation steps, initial speed and end speed of the casting cleaning robot are set, and the jtraj function is input to obtain the cleaning trajectory of the casting cleaning robot. The expression of the jtraj function is:

[0057]

[0058] in, represents the trajectory after calculation, represents the speed after calculation, represents the acceleration after calculation, represents the initial point position, Indicates the end point position, Step indicates the number of interpolation steps, Indicates the set initial speed, Indicates the set end speed.

[0059] Specifically, the free space motion stage is the process in which the casting cleaning robot moves from any posture to directly above the cleaning starting point. During this process, the casting cleaning robot's own motion is not restricted, so there is no need for posture planning.

[0060] The Cartesian space planning method is used to plan the path and posture of the casting cleaning robot in the transition space motion stage.

[0061] In a specific embodiment, the transitional space motion stage is a transitional stage between the free space motion stage and the confined space motion stage;

[0062] Specifically, the transition space movement stage is a transition link in which the casting cleaning robot changes from a state of no contact with the casting to a state of about to start cleaning work.

[0063] Specifically, the process of planning the path and posture of the casting cleaning robot in the transition space motion stage using the Cartesian space planning method is as follows:

[0064] The end point position of the casting cleaning robot in the free space motion phase is used as the starting point, and the end point position of the casting cleaning robot is set;

[0065] The path trajectory of the casting cleaning robot from the starting point to the end point is set, and the path trajectory is fitted; specifically, in this embodiment, the process of fitting the path trajectory is a conventional prior art and will not be described in detail here.

[0066] The fitted path trajectory is discretized into a number of path points, and the position information of the casting cleaning robot at the path point is obtained by using the trajectory difference algorithm and solving the three-dimensional spatial position information of each path point;

[0067] The position information of the casting cleaning robot at each path point is solved by the inverse kinematics method to obtain each joint angle of the casting cleaning robot;

[0068] The casting cleaning robot is controlled to move according to the path points and the corresponding joint angles.

[0069] Specifically, in the present embodiment, the end point of the free space movement stage is 150 mm above the initial point to be contacted for cleaning. During the free space movement stage, the casting cleaning robot descends vertically and maintains low-speed movement.

[0070] The improved Bezier curve and inverse kinematics method are used to plan the path and posture of the casting cleaning robot when it is in the confined space motion stage.

[0071] In a specific embodiment, the confined space movement stage is a stage in which the casting cleaning robot cleans various parts to be cleaned on the surface of the casting; an improved Bezier curve is used to plan the path of the casting cleaning robot, and the specific form of the improved Bezier curve is expressed as:

[0072]

[0073] Where p0, p1, …, p i ,…,p n Represents n control points on the Bezier curve, p i is the i-th control point; t is the parameter controlling the shape of the curve, and t=i / n+1;

[0074] In this embodiment, n control points are used to control the shape of the trajectory curve, and the curve passes through the starting point p0 and the end point p n , and construct auxiliary point p Cn Make the trajectory curve smoother and construct auxiliary point p based on the improved Bezier curve Cn , expressed as:

[0075]

[0076] According to the surface position information of the part to be cleaned of the casting, the inverse kinematics method is used to solve the posture information of the casting cleaning robot during the cleaning process, that is, the angles of each joint. Specifically, the casting surface point cloud position information and the robot inverse kinematics method are used to solve the angles of each joint of the robot, and the optimized trajectory is combined to realize the adaptive cleaning of different parts of the casting by the casting cleaning robot.

[0077] Specifically, in this embodiment, in the confined space motion stage, the posture information is obtained by inverse kinematics, and the path information is obtained by using the improved Bezier curve. The use of the improved Bezier curve to plan the path can improve the problem of the sharp change in the curvature of the cleaning trajectory, make the casting cleaning robot move more smoothly, and ensure the smooth progress of the cleaning process.

[0078] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A casting cleaning robot adaptive cleaning method based on machine vision and trajectory planning, characterized in that: The specific steps include: S1: using a camera to obtain point cloud data of the part to be cleaned of the casting, and preprocessing the point cloud data to obtain processed point cloud data, i.e., surface position information of the part to be cleaned of the casting; S2: Perform hand-eye calibration on the casting cleaning robot to obtain a homogeneous change matrix of the camera and the casting cleaning robot in a base coordinate system, and transform the coordinates in the camera coordinate system into the robot coordinate system based on the homogeneous change matrix; S3: In the robot coordinate system, the cleaning process of the casting cleaning robot is divided into three stages: free space movement, transition space movement and confined space movement according to the corresponding working time, and the cleaning trajectory of the casting cleaning robot in each movement stage is set according to the trajectory planning strategy, so that the casting cleaning robot can adaptively adjust the cleaning trajectory. The trajectory planning strategy includes: The jtraj function is used to plan the path of the foundry cleaning robot when it is in the free space motion stage; The Cartesian space planning method is used to plan the path and posture of the casting cleaning robot in the transition space motion stage. The improved Bezier curve and inverse kinematics method are used to plan the path and posture of the casting cleaning robot when it is in the confined space motion stage.

2. The self-adaptive cleaning method of a casting cleaning robot based on machine vision and trajectory planning according to claim 1, characterized in that: The free space movement stage is the process in which the casting cleaning robot moves from any posture to directly above the cleaning starting point, during which the casting cleaning robot's own movement is not restricted; The process of using the jtraj function to plan the path of the casting cleaning robot in the free space motion stage is: The initial point position, end point position, interpolation steps, initial speed and end speed of the casting cleaning robot are set, and the jtraj function is input to obtain the cleaning trajectory of the casting cleaning robot. The expression of the jtraj function is: in, represents the trajectory after calculation, represents the speed after calculation, represents the acceleration after calculation, represents the initial point position, Indicates the end point position, step indicates the number of interpolation steps, Indicates the set initial speed, Indicates the set end speed.

3. The self-adaptive cleaning method of a casting cleaning robot based on machine vision and trajectory planning according to claim 2, characterized in that: The transitional space motion stage is a transitional stage between the free space motion stage and the confined space motion stage; The process of planning the path and posture of the casting cleaning robot in the transition space motion stage using the Cartesian space planning method is as follows: The end point position of the casting cleaning robot in the free space motion phase is used as the starting point, and the end point position of the casting cleaning robot is set; Setting the path trajectory of the casting cleaning robot from the starting point to the end point, and performing fitting processing on the path trajectory; The fitted path trajectory is discretized into a number of path points, and the position information of the casting cleaning robot at the path point is obtained by using the trajectory difference algorithm and solving the three-dimensional spatial position information of each path point; The position information of the casting cleaning robot at each path point is solved by the inverse kinematics method to obtain each joint angle of the casting cleaning robot; The casting cleaning robot is controlled to move according to the path points and the corresponding joint angles.

4. The self-adaptive cleaning method of a casting cleaning robot based on machine vision and trajectory planning according to claim 3, characterized in that: The confined space movement stage is a stage in which the casting cleaning robot cleans various parts to be cleaned on the surface of the casting; The process of planning the path and posture of the casting cleaning robot in the confined space motion stage using the improved Bezier curve and inverse kinematics method is as follows: The improved Bezier curve is used to plan the path of the casting cleaning robot. The specific form of the improved Bezier curve is expressed as follows: Where p0, p1, …, p i ,…,p n Represents n control points on the Bezier curve, p i is the i-th control point; t is a parameter controlling the shape of the curve, and t=i / n+1; Constructing auxiliary point p based on improved Bezier curve Cn , expressed as: According to the surface position information of the part to be cleaned of the casting, the inverse kinematics method is used to solve the posture information of the casting cleaning robot during the cleaning process, that is, the angles of each joint.

5. The method for adaptive cleaning of a casting cleaning robot based on machine vision and trajectory planning according to claim 4, characterized in that: In S1, the point cloud data of the part to be cleaned of the casting is obtained, and the point cloud data is preprocessed to obtain the processed point cloud data: S11: Use a 3D binocular camera to take photos of the casting to be cleaned to obtain corresponding point cloud data; S12: performing filtering processing, noise reduction processing, edge feature extraction processing of the casting to-be-cleaned portion, and identification processing of the casting to-be-cleaned portion on the point cloud data in sequence, thereby obtaining surface position information of the casting to-be-cleaned portion.

6. The self-adaptive cleaning method of a casting cleaning robot based on machine vision and trajectory planning according to claim 5, characterized in that: In S2, the formula for converting the coordinates in the camera coordinate system to the robot coordinate system based on the homogeneous transformation matrix is: p c =T CE T EB p B In the formula, p B is the coordinate of the marker selected on the calibration plate in the calibration plate coordinate system; p c is the coordinate in the camera coordinate system; T CE T EB is the homogeneous transformation matrix; T CE is the transformation matrix between the robot end effector and the camera; T EB Represents the transformation matrix between the calibration plate coordinate system and the robot end effector coordinate system.

Images