A method for controlling the movement of a robot arm in an active search for an occluded target

Abstract

The application discloses a mechanical arm motion control method for actively searching for a blocked target, relates to the technical field of mechanical arm motion planning and control, and comprises the following steps: planning an angle range of circular arc rotary motion of a mechanical arm end tool in a space horizontal plane, an accumulated rotation angle, and a default rotation direction, solving a radius of a plane circular arc track based on position coordinates of a vision camera and a guide point, acquiring and verifying pose information of a target point based on the vision camera, verifying and updating the accumulated rotation angle and the rotation direction, solving a joint configuration of the mechanical arm after single-step rotary motion, and controlling the mechanical arm to move to a new joint configuration and then continue searching for the target. The application actively searches for the blocked target by planning and controlling the motion of the mechanical arm, can exert flexibility and initiative of the motion of the mechanical arm, and can realize accurate identification, positioning and grabbing of the blocking object without increasing the hardware and software costs.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of manipulator motion planning and control, and particularly relates to a manipulator motion control method for actively searching for an occluded target. BACKGROUND

[0002] At present, with the rapid development of automation and intelligent manufacturing, manipulators are increasingly widely used in industrial production, social services, and agriculture-forestry cooperation. However, in actual application scenarios, due to the ubiquitous phenomenon of object occlusion, manipulators based on visual guidance are difficult to perform grasping tasks because they cannot accurately identify and locate target objects. For example, when carrying out fruit picking operations in natural environments, dense foliage occlusion and fruit overlap often interfere with the judgment of detection and recognition algorithms, causing picking manipulators to be unable to accurately grasp target fruits, thereby seriously affecting the success rate and work efficiency of automatic picking. In classification and sorting tasks, because the target object may be occluded by the higher part of the front goods, the identification system fails, and thus the execution unit cannot grasp the specific item. This problem limits the application potential of manipulators in complex environments.

[0003] In related technologies, to solve the above-mentioned occlusion problem: 1) one method is to use more sensors to build a complex perception system, and to segment and locate the target object through multi-source information fusion technology; however, this method significantly increases the cost of equipment. 2) another method is to improve the visual recognition algorithm and increase the diversified data set, and to obtain a better detection and recognition strategy by increasing the model parameters of the learning algorithm; however, this method usually relies on greater computing power, more storage space, and longer training time, greatly increasing the cost of software and hardware.

[0004] Therefore, in fact, by imitating the process of human grasping objects, more environmental information can be obtained by switching different perspectives to search for occluded targets. Therefore, in view of the flexibility of manipulator motion, there is an urgent need for a manipulator motion control method for actively searching for an occluded target to control its active motion to change the perspective to search for the target object, so as to achieve accurate positioning and grasping of the occluded target without increasing the cost. SUMMARY

[0005] The purpose of the present application is to provide a manipulator motion control method for actively searching for an occluded target to solve the problem in related technologies that the above-mentioned method significantly increases the cost of equipment and usually relies on greater computing power, more storage space, and longer training time, greatly increasing the cost of software and hardware.

[0006] The manipulator motion control method for actively searching for an occluded target provided by the present application adopts the following technical solution:

[0007] A method for controlling the movement of a robot arm to search for a hidden target, comprising the following steps:

[0008] S1, planning the angle range of the circular arc rotation movement of the robot arm end tool in the horizontal plane, accumulating the rotation angle, and defaulting the rotation direction;

[0009] S2, solving the radius of the plane circular arc trajectory based on the position coordinates of the vision camera and the guide point;

[0010] S3, acquiring and verifying the pose information of the target point based on the vision camera;

[0011] S4, verifying and updating the accumulated rotation angle and rotation direction;

[0012] S5, solving the joint configuration of the robot arm after single-step rotation movement;

[0013] S6, controlling the robot arm to move to the new joint configuration and continue searching for the target;

[0014] The robot arm comprises one joint degree of freedom, an end tool gripper, and a vision camera, and the vision camera is installed on the robot arm.

[0015] As a preferred, the circular arc rotation movement refers to that in three-dimensional space, regardless of the way the robot arm is installed, taking the current position of the robot arm end tool as the starting point of the circular arc, the end tool is made to rotate around a certain center in the horizontal plane through reasonable planning and control, and during the movement process, the included angle between the tool pointing axis and the horizontal plane is always kept unchanged.

[0016] As a preferred, the solving of the radius of the plane circular arc trajectory based on the position coordinates of the vision camera and the guide point is implemented as follows:

[0017] S201, calculating the coordinates of a guide point P on the target object in the robot tool coordinate system based on the hand-eye calibration method;

[0018] S202, calculating the included angle between the guide point P and the opening and closing movement plane of the robot arm end tool based on the projection method ;

[0019] S203, calculating the included angle between the opening and closing movement plane of the end tool and the horizontal plane based on the dihedral angle principle ;

[0020] S204, calculating the radius of the plane circular arc trajectory as

[0021] ,

[0022] wherein, is the distance from the guide point P to the origin of the tool system. .

[0023] As preferred, the acquiring and verifying the pose information of the target point by the vision camera means that when the manipulator performs the circular arc rotation motion in the horizontal plane, the vision camera synchronously executes the target detection and recognition algorithm; if the vision camera acquires the valid target point, the manipulator stops the circular arc search motion, and then performs the target grasping and other tasks; otherwise, the active search for the occluded target motion is continuously performed.

[0024] As preferred, the verifying and updating the cumulative rotation angle and the rotation direction include the following steps:

[0025] Firstly, verifying whether the cumulative rotation angle of the manipulator performing the circular arc rotation motion exceeds the planned angle range; if exceeding the angle range limit, updating the rotation direction of the end tool to make it move in the opposite direction along the circular arc trajectory; otherwise, keeping the default rotation direction unchanged.

[0026] Then, updating the cumulative rotation angle of the manipulator performing the circular arc rotation motion as

[0027] ,

[0028] wherein, represents the updated cumulative rotation angle, represents the cumulative rotation angle before updating; corresponds to the clockwise motion of the end tool in the horizontal plane, corresponds to the counterclockwise motion of the end tool in the horizontal plane; is the single-step step length of the manipulator performing the rotation motion.

[0029] As preferred, the joint configuration of the manipulator after performing the single-step rotation motion can be represented by the mapping .

[0030] wherein, represents the dimensional joint variable of the manipulator with joint degrees of freedom; represents the plane circular arc radius; represents the cumulative rotation angle; represents the mapping of the plane circular arc radius and the cumulative rotation angle to the joint variable of the manipulator.

[0031] As preferred, the specific implementation process of the mapping is as follows:

[0032] S501, calculating the homogeneous transformation matrix of the end tool coordinate system of the manipulator relative to the world coordinate system at the starting configuration, denoted as​

[0033] ,

[0034] wherein, are elements of the rotation transformation matrix of the tool coordinate system of the manipulator in the initial configuration with respect to the world coordinate system, , , are the corresponding position coordinates;

[0035] S502, the homogeneous transformation matrix of the tool coordinate system of the manipulator in the world coordinate system for planar circular arc motion is calculated, and the calculation steps are as follows:

[0036] 1) the coordinates of the center of the planar circular arc trajectory C are calculated as

[0037]

[0038] wherein, , ;

[0039] 2) the new position coordinates of the tool coordinate system of the manipulator with respect to the world coordinate system after motion are calculated as

[0040]

[0041] 3) the homogeneous transformation matrix of the tool coordinate system of the manipulator with respect to the world coordinate system after motion is calculated as

[0042] ,

[0043] wherein, and are four-order translation and rotation operators respectively, is a homogeneous rotation matrix composed of the first three-order sub-matrix of the matrix ;

[0044] S503, the homogeneous transformation matrix of the wrist coordinate system with respect to the base coordinate system after the motion of the manipulator is calculated as

[0045] ,

[0046] wherein, represents the inverse of the matrix ; is the homogeneous transformation matrix of the base coordinate system of the manipulator with respect to the world coordinate system, is the homogeneous transformation matrix of the tool coordinate system of the manipulator with respect to the wrist coordinate system, which is usually a constant value;

[0047] S504, the joint configuration after the motion of the manipulator is calculated as ; wherein, Inverse kinematics function of the robot arm.

[0048] As preferred, the control method continues searching the target after the robot arm moves to the new joint configuration, comprising the following steps:

[0049] Firstly, verify the calculated new joint configuration is a valid value, including verifying whether the dimensionality is consistent with the joint degrees of freedom of the robot arm, and verifying whether each element value is within the rotation angle range of the corresponding joint;

[0050] Then, based on the validity determination result of the new joint configuration, determine the running direction of the control method; that is, if is not a valid value, return to step S4 to verify and update the accumulated rotation angle and rotation direction, and then continue to solve the new joint configuration of the robot arm after the next rotation movement; if is a valid value, control the robot arm to move to the new joint configuration, and then return to step S3 to continue searching the target point and verifying its validity; repeat the above steps until the target searching algorithm is closed.

[0051] Compared with the prior art, the beneficial effects of the present application are:

[0052] The present application proposes a method of actively changing the viewing angle by controlling the movement of the robot arm to search for the occluded target, which fully utilizes the flexibility and initiative of the multi-degree-of-freedom robot arm movement, so that the robot arm can actively search for the occluded target in the grasping task, thereby achieving the effect of accurate identification and positioning of the robot arm.

[0053] Meanwhile, compared with the process of solving the target occlusion problem by improving the visual recognition algorithm, which needs to consume a large amount of computing power and storage resources to train a large data set, the control method of the robot arm which can actively search for the occluded target of the present application only contains common matrix operations and visual recognition techniques, and only requires conventional computing and storage units, effectively reducing the hardware and software costs. BRIEF DESCRIPTION OF DRAWINGS

[0054] Figure 1 is a flowchart of the robot arm movement control method of the present application embodiment which actively searches for the occluded target;

[0055] Figure 2 is a schematic diagram of the robot arm moving in a planar circular arc to search for the target of the present application embodiment;

[0056] Figure 3 is a schematic diagram of the robot arm grasping the occluded target fruit based on the vision camera of the present application embodiment;

[0057] Figure 4 is a simplified schematic diagram of solving the radius of the planar circular arc trajectory of the present application embodiment. DETAILED DESCRIPTION

[0058] The application will be further described in detail below. Figures 1-4 The application will be further described in detail below.

[0059] The application discloses a mechanical arm movement control method for actively searching for a shielding target. Figure 1 In the embodiment, the mechanical arm movement control method comprises the following steps:

[0060] S1, planning an angle range of circular arc rotation movement of a mechanical arm end tool in a horizontal plane, accumulating a rotation angle, and defaulting a rotation direction.

[0061] Specifically, the mechanical arm movement needs to establish a world coordinate system and a tool coordinate system. Generally, a horizontal plane is taken as an X w O w Y w axis of a spatial coordinate system, a vertical direction is taken as a Z w axis of the spatial coordinate system, a reference world coordinate system O w -X w Y w Z w is established based on a right-hand coordinate system rule. Then, a pointing direction of the mechanical arm end tool is taken as a Z t axis, a direction perpendicular to a tool plane and upward is taken as a Y t axis, and a tool opening and closing movement direction is taken as an X t axis, a tool coordinate system O t -X t Y t Z t is established based on the right-hand coordinate system rule.

[0062] More specifically, referring to Figure 2 In the embodiment, the circular arc rotation movement refers to that, in a three-dimensional space, regardless of a mounting manner of a mechanical arm, the mounting manner of the mechanical arm has a normal mounting, an inclined mounting, an upside-down mounting and the like; then, a current position of the mechanical arm end tool is taken as a circular arc starting point, the end tool is caused to make a circular arc rotation movement around a certain center C in the horizontal plane X w O w Y w , and in the movement process, an included angle between the tool pointing axis Z t and the horizontal plane X w O w Y w is kept unchanged. That is, the stability of a triangle O t PC is kept unchanged, wherein point P is a guide point on a target object.

[0063] Meanwhile, in the step S1, the angle range of the planned circular arc rotation movement refers to an initialized expected circular arc angle, which is denoted as . wherein, is the maximum rotation angle of the single-side rotation starting from the current position of the tool, and the counterclockwise rotation direction is positive and the clockwise rotation direction is negative.

[0064] The cumulative rotation angle of the planned circular arc rotation motion refers to the cumulative rotation angle of the initialization end tool circular arc rotation motion, denoted as , and always satisfies .

[0065] The default rotation direction of the planned circular arc rotation motion refers to the initialization end tool first making a planar circular arc motion in the clockwise direction to search for the target, and the clockwise rotation direction is denoted as , and the corresponding counterclockwise rotation direction is denoted as .

[0066] S2, solving the radius of the planar circular arc trajectory based on the position coordinates of the vision camera and the guide point.

[0067] Specifically, referring to Figure 3 , in this embodiment, the manipulator includes one joint degree of freedom, an end tool gripper, and a vision camera, and the vision camera is installed on the manipulator in the manner of "eye on hand". By using the manipulator equipped with the vision camera and the end tool gripper, the target fruit can be automatically recognized and grasped, for example, the target fruit is a pineapple; in this way, the automatic and intelligent harvesting of fruits can be realized.

[0068] At the same time, according to the biological and mechanical characteristics of the pineapple plant, when the manipulator accurately grasps the target point Q on the fruit, and the target point Q is located at the connection between the pineapple crown bud and the fruit body, the fruit can be easily picked off by the lateral breaking or cutting knife breaking method. However, due to the obstruction of the dense branches and leaves, the vision camera usually cannot directly detect and recognize the target point Q, especially cannot accurately obtain the depth distance information of the target point, which seriously affects the success rate of the manipulator grasping. Then considering that the top of the fruit crown is usually the highest point of the pineapple plant and is not easily obstructed by branches and leaves, therefore, taking the top of the crown as the guide point P, the spatial coordinates of the guide point P can be obtained by the vision camera, and then guiding the end tool of the manipulator to make circular motion in the horizontal plane to search for the target point Q; once the target point Q is obtained after passing through the obstructed branches and leaves, the manipulator stops the circular search motion and then performs the grasping action.

[0069] More specifically, referring to Figure 4 , in this embodiment, in order to realize the above-mentioned rotation motion, it is necessary to determine the radius of the planar circular arc trajectory , so the radius of the planar circular arc trajectory based on the position coordinates of the vision camera and the guide point is solved, and the specific solving process is as follows:

[0070] S201. Calculate the coordinates of a guide point P on the target object in the robot tool coordinate system based on the hand-eye calibration method. These coordinates are:

[0071] ,

[0072] in, It is the homogeneous transformation matrix of the robot tool coordinate system relative to the vision camera coordinate system, which is generally obtained by the hand-eye calibration algorithm; This represents the homogeneous position coordinates of the guide point P in the visual camera coordinate system, which are generally obtained by image feature extraction methods or visual recognition algorithms such as deep learning.

[0073] S202. Calculate the angle between the guide point P and the opening / closing motion plane of the robot's end effector based on the projection method. Specifically, the coordinates of the guide point P and the robot tool system coordinate plane X are calculated based on the projection method. t O t Z t The included angle is

[0074] .

[0075] S203. Calculate the angle between the opening and closing motion plane of the end effector and the horizontal plane based on the dihedral angle principle. Specifically, the calculation tool system coordinate plane X t O t Z t With the world coordinate plane X w O w Y w The included angle of the dihedral angle formed is

[0076] ,

[0077] in, Represents plane X t O t Z t normal vector With plane X w O w Y w normal vector The inner product, and These are the magnitudes of the corresponding vectors, and they have... .

[0078] S204. Calculate the radius of the planar circular arc trajectory.

[0079] ,

[0080] in, is the distance from the guide point P to the origin of the tool system, , The simplified schematic and geometric meaning of each parameter are shown in Figure 4 .

[0081] S3, acquiring and verifying the pose information of the target point based on the visual camera.

[0082] Specifically, in the present embodiment, the acquiring and verifying the pose information of the target point based on the visual camera refers to that when the robot arm performs a circular arc rotation motion in the spatial horizontal plane, the visual camera synchronously executes a target detection and recognition algorithm; if the visual camera acquires an effective target point, the robot arm stops the circular arc search motion and then performs a target grasping task; otherwise, the active search for the occluded target motion is continued.

[0083] More specifically, when the robot arm performs a circular arc rotation motion in the spatial horizontal plane with the point P as the guide point and as the radius, the visual camera synchronously executes a target detection and recognition algorithm and verifies and judges in real time whether an effective target point is obtained. The effective target point refers to that the data information of the target point Q acquired based on the visual recognition algorithm at least contains three-dimensional position coordinates, and the coordinate values of the target point and those of the nearby feature points are not significantly different, so as to avoid visual noise points and non-target point interference. If the visual camera acquires an effective target point, the robot arm stops the circular arc search motion and then performs a target grasping task. Otherwise, the following steps are continued.

[0084] S4, verifying and updating the cumulative rotation angle and the rotation direction.

[0085] Specifically, in the present embodiment, the verifying and updating the cumulative rotation angle and the rotation direction includes the following steps:

[0086] Firstly, it is verified whether the cumulative rotation angle of the robot arm performing the circular arc rotation motion exceeds the planned angle range; if the angle range limit is exceeded, the rotation direction of the end tool is updated to make it move in the opposite direction along the circular arc trajectory; otherwise, the default rotation direction remains unchanged.

[0087] More specifically, the angle range is , i.e., if , the rotation direction of the robot arm is updated to be counterclockwise, i.e., . If , the rotation direction of the robot arm is updated to be clockwise, i.e., .

[0088] Then, the cumulative rotation angle of the robot arm performing the circular arc rotation motion is updated, and the cumulative rotation angle is

[0089] ​ ,

[0090] wherein, denotes the updated cumulative rotation angle, denotes the pre-updated cumulative rotation angle; corresponds to the clockwise movement of the end-effector in the horizontal plane, corresponds to the counterclockwise movement of the end-effector in the horizontal plane; is the single-step step length of the robot for the rotational movement, which is usually a given prior value.

[0091] S5, solving the joint configuration of the robot after the single-step rotational movement.

[0092] Specifically, in the present embodiment, after the robot makes a single-step rotational movement along the circular arc trajectory in the spatial horizontal plane, the solving of the joint configuration of the robot after the single-step rotational movement can be represented by the mapping . Wherein, denotes the dimensional joint variable of the robot with joints; denotes the radius of the planar circular arc; denotes the cumulative rotation angle; denotes the mapping of the radius of the planar circular arc and the cumulative rotation angle to the joint variable of the robot.

[0093] More specifically, the specific implementation process of the mapping is as follows:

[0094] S501, calculating the homogeneous transformation matrix of the end-effector coordinate system of the robot relative to the world coordinate system in the starting configuration, denoted as

[0095] ,

[0096] wherein, is the element of the rotation transformation matrix of the tool coordinate system of the robot relative to the world coordinate system in the starting configuration; , , are the corresponding position coordinates.

[0097] In addition, ,

[0098] wherein, is the homogeneous transformation matrix of the base coordinate system of the robot relative to the world coordinate system; is the homogeneous transformation matrix of the tool coordinate system of the robot relative to the wrist coordinate system, which are usually constant values; is the homogeneous transformation matrix of the wrist coordinate system of the manipulator in the initial configuration relative to the base coordinate system.

[0099] S502, the homogeneous transformation matrix of the manipulator tool system in the world system for planar circular motion is calculated, and the calculation steps are as follows:

[0100] 1) the coordinates of the center of the planar circular arc trajectory C are calculated as

[0101]

[0102] wherein, , .

[0103] 2) the new position coordinates of the tool system relative to the world system after the movement of the manipulator are calculated as

[0104]

[0105] 3) the homogeneous transformation matrix of the tool system relative to the world system after the movement of the manipulator is calculated as

[0106] ,

[0107] wherein, and are four-order translation and rotation operators respectively; is a homogeneous rotation matrix composed of the first three-order sub-matrix of the matrix , that is,

[0108] .

[0109] S503, the homogeneous transformation matrix of the wrist coordinate system relative to the base coordinate system after the movement of the manipulator is calculated as

[0110] ,

[0111] wherein, represents the inverse of the matrix ; is the homogeneous transformation matrix of the manipulator base coordinate system relative to the world coordinate system, is the homogeneous transformation matrix of the manipulator tool coordinate system relative to the wrist coordinate system, which is usually a constant value.

[0112] S504, the joint configuration after the movement of the manipulator is calculated as ; wherein, represents the inverse kinematics function of the manipulator.

[0113] S6, the manipulator is controlled to move to the new joint configuration and continue to search for the target.

[0114] Specifically, in the present embodiment, the control continues to search for the target after the robot arm moves to the new joint configuration, including the following steps:

[0115] First, the calculated new joint configuration is verified to be a valid value, including verifying whether the dimensionality is consistent with the joint degrees of freedom of the robot arm, and verifying whether each element value is within the rotation angle range of the corresponding joint;

[0116] Next, based on the validity determination result of the new joint configuration, the running direction of the control method is determined; that is, if the new joint configuration is not a valid value, the accumulated rotation angle and rotation direction are verified and updated, and then the new joint configuration of the robot arm after the next rotation movement is solved; if the new joint configuration is a valid value, the robot arm is controlled to move to the new joint configuration, and then step S3 is returned to continue to search for the target point and verify its validity; the above steps are repeated until the target search algorithm is closed.

[0117] The above is only an embodiment of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation using the content of the present application specification and drawings, or direct or indirect application in other related technical fields, is also included in the patent protection scope of the present application.​​​

Claims

1. A motion control method for a robotic arm that actively searches for occluded targets, characterized in that, Includes the following steps: S1. Plan the angular range of the robotic arm's end effector to make circular rotational motion in the horizontal plane of space, the cumulative rotation angle, and the default rotation direction; S2. Calculate the radius of the planar circular arc trajectory based on the position coordinates of the visual camera and the guide point; S3. Acquire and verify the pose information of the target point based on the visual camera; S4. Verify and update the cumulative rotation angle and rotation direction; S5. Solve for the joint configuration of the robot arm after it performs a single-step rotational motion; S6. After controlling the robotic arm to move to the new joint configuration, continue searching for the target; The robotic arm includes The robotic arm has one joint degree of freedom, an end effector gripper, and a vision camera, wherein the vision camera is mounted on the robotic arm; The specific implementation process for calculating the radius of the planar circular arc trajectory based on the position coordinates of the visual camera and the guide point is as follows: S201. Calculate the coordinates of a guide point P on the target object in the coordinate system of the robotic arm tool based on the hand-eye calibration method; S202. Calculate the angle between the guide point P and the opening / closing motion plane of the robot's end effector based on the projection method. ; S203. Calculate the angle between the opening and closing motion plane of the end effector and the horizontal plane based on the dihedral angle principle. ; S204. Calculate the radius of the planar circular arc trajectory. ， in, Let P be the distance from the guiding point P to the origin of the tool system. ; The acquisition and verification of the target point's pose information based on the vision camera means that when the robot arm performs a circular rotation in the horizontal plane of space, the vision camera simultaneously executes a target detection and recognition algorithm; if the vision camera acquires a valid target point, the robot arm stops its circular search motion and then performs tasks such as target grasping. Otherwise, continue actively searching for occluded targets.

2. The method for motion control of a robotic arm actively searching for occluded targets according to claim 1, characterized in that, The aforementioned circular rotation motion refers to a three-dimensional motion in which, regardless of how the robotic arm is installed, the current position of the end effector tool is taken as the starting point of the circular arc, and through reasonable planning and control, the end effector tool rotates around a certain center in the horizontal plane, while maintaining the angle between the tool pointing axis and the horizontal plane unchanged throughout the motion.

3. The method for motion control of a robotic arm actively searching for occluded targets according to claim 1, characterized in that, The verification and updating of the cumulative rotation angle and rotation direction includes the following steps: First, verify the cumulative rotation angle of the robotic arm performing circular rotational motion. Check if the angle exceeds the planned range; if it does, update the rotation direction of the end tool so that it moves in the opposite direction along the arc trajectory; otherwise, keep the default rotation direction unchanged. Next, update the cumulative rotation angle of the robotic arm's circular rotation motion to... ， in, This represents the updated cumulative rotation angle. This represents the cumulative rotation angle before the update; The corresponding end tool moves clockwise in the horizontal plane. The corresponding end tool moves counterclockwise in the horizontal plane; The single-step length for a robotic arm to perform rotational motion.

4. The method for motion control of a robotic arm actively searching for occluded targets according to claim 1, characterized in that, The solution to the joint configuration after a single-step rotational motion of the robotic arm can be obtained using mapping. To indicate; in, Indicates having A robotic arm with one joint degree of freedom Joint variables; Indicates the radius of a planar circular arc; Indicates the cumulative rotation angle; Indicates the radius of the planar arc. and cumulative rotation angle A function that maps to the joint variables of the robotic arm.

5. The method for motion control of a robotic arm actively searching for occluded targets according to claim 4, characterized in that, The mapping The specific implementation process is as follows: S501. Calculate the homogeneous transformation matrix of the robot end-tool coordinate system relative to the world coordinate system under the initial configuration, denoted as: ， in, These are the elements of the rotation transformation matrix of the robot's tool coordinate system relative to the world coordinate system in its initial configuration. , , These are the corresponding position coordinates; S502. Calculate the homogeneous transformation matrix of the tool system of the robot undergoing planar circular arc motion in the world frame. The calculation steps are as follows: 1) Calculate the coordinates of the center C of the planar circular arc trajectory. in, , ; The new position coordinates of the tool system relative to the world system after the robot arm's movement are calculated. 3) Calculate the homogeneous transformation matrix of the tool system relative to the world system after the robot arm has moved. ， in, and These are fourth-order translation and rotation operators, It is a matrix The homogeneous rotation matrix formed by the first three submatrices; S503. Calculate the homogeneous transformation matrix of the wrist coordinate system relative to the base coordinate system after the robot arm has moved. ， in, Representation matrix The reverse; It is the homogeneous transformation matrix of the robot's base coordinate system relative to the world coordinate system. These are the homogeneous transformation matrices of the robotic tool coordinate system relative to the wrist coordinate system; they are usually constant values. S504, Calculate the joint configuration of the robot arm after movement. ;in, This represents the inverse kinematics function of the robotic arm.

6. The method for motion control of a robotic arm actively searching for occluded targets according to claim 5, characterized in that, After the controlled robotic arm moves to the new joint configuration, it continues to search for the target, including the following steps: First, verify the new joint configuration obtained from the solution. Whether it is a valid value includes verifying whether its dimension is consistent with the joint degrees of freedom of the robot hand, and verifying whether the value of each element is within the rotation angle range of the corresponding joint. Next, based on the effectiveness judgment result of the new joint configuration, the execution direction of the control method is determined; that is, if If the value is not valid, return to step S4 to verify and update the accumulated rotation angle and rotation direction, then continue to solve for the new joint configuration after the robot arm performs the next rotational motion; if If the value is valid, the robot arm is controlled to move to the new joint configuration, and then the process returns to step S3 to continue searching for the target point and verifying its validity until a valid target point is obtained.

Images