Grabbing planning method of multi-arm space robot

Abstract

The embodiment of the invention provides a grabbing planning method for a multi-arm space robot, and the method comprises the steps: determining a stable contact quality index according to the stable contact condition of the tail end of the multi-arm space robot and the surface of an object; according to the stable contact quality index, a stable grabbing point selection strategy of the multi-arm space robot for the target object is generated; a three-dimensional model of the multi-arm space robot is utilized to generate kinematics and dynamics equations; determining performance evaluation indexes of the grabbing configuration of the multi-arm space robot according to the stable grabbing point of the multi-arm space robot for the target object and the kinematics and dynamics equations; according to the performance evaluation indexes, a multi-target optimization model of the grabbing configuration of the multi-arm space robot is generated; and according to the stable grabbing point of the multi-arm space robot for the target object and the multi-target optimization model, grabbing planning is achieved. According to the technical scheme provided by the embodiment of the invention, a reference can be provided for the design of the multi-arm space robot grabbing planning method.

Description

[Technical field]

[0001] The invention relates to the technical field of space robots, and in particular to a grasping planning method of a multi-arm space robot. [Background technology]

[0002] With the increasing demand for space exploration, space robots have shown great development potential and have been widely used to perform various tasks of large spacecraft, becoming one of the most important ways of on-orbit service missions. As a key link in on-orbit operation tasks, space extravehicular target grasping is affected by environmental factors and has a series of problems such as contact process disturbance. Compared with astronauts' extravehicular operations, space robot operations have higher safety and flexibility, and can perform tasks more effectively in complex space environments, reduce human risks and improve operational efficiency. Most of the existing space robots are single-arm structures and can perform on-orbit operation tasks such as cabin assembly and equipment maintenance. However, single-arm robots have limitations such as insufficient operational capabilities and limited task types when dealing with diverse on-orbit tasks. For this reason, a multi-arm space robot was proposed. Through the coordinated cooperation between multiple arms, it uses contact-type rigid connections to achieve envelope grasping of target objects, which can significantly improve the flexibility and adaptability of operations.

[0003] Compared with single-arm robots, multi-arm space robots have higher operating accuracy in complex space environments, greatly improving the reliability and efficiency of space target grasping. In space multi-arm grasping tasks, in order to successfully grasp the target object, it is usually necessary to pre-select grasping points and generate grasping configurations, and reasonable grasping points and grasping configurations can effectively improve grasping stability and task success rate. Therefore, grasping planning is of great significance to the grasping task execution of multi-arm space robots, and it is urgent to carry out research on grasping point selection and grasping configuration optimization of multi-arm space robots. [Summary of the invention]

[0004] In view of this, an embodiment of the present invention provides a grasping planning method for a multi-arm space robot, comprising:

[0005] Using the three-dimensional model of the target object, the stable contact quality index is determined according to the stable contact condition between the end of the multi-arm space robot and the surface of the object;

[0006] According to the stable contact quality index, a stable grasping point selection strategy of the multi-arm space robot is generated, wherein the stable grasping point selection strategy includes a stable grasping point of the multi-arm space robot for the target object;

[0007] Using the three-dimensional model of the multi-arm space robot, the kinematic and dynamic equations of the multi-arm space robot are generated;

[0008] Determine the performance evaluation index of the grasping configuration of the multi-arm space robot according to the stable grasping point of the multi-arm space robot for the target object and the kinematic and dynamic equations of the multi-arm space robot; and generate a multi-objective optimization model of the grasping configuration of the multi-arm space robot according to the performance evaluation index;

[0009] According to the stable grasping points of the multi-arm space robot for the target object and the multi-objective optimization model, and with stable contact as the goal, grasping planning is performed between the multi-arm space robot and the target object.

[0010] In the above method, the three-dimensional model of the target object is a point cloud model containing shape information of the target object, and each point contains corresponding normal vector information;

[0011] The stable contact conditions between the end of the multi-arm space robot and the surface of the object include:

[0012] for There exists f that satisfies:

[0013]

[0014] Among them, w ext is the force spiral generated by the external force, F S is the comprehensive force spiral generated by each contact point at the center of mass of the target object, G = [G 1 G k ] is the comprehensive grasping matrix of the multi-arm space robot, is the grasping force vector generated by the end of each manipulator of the multi-arm space robot, and k is the number of contact points;

[0015] The force at the contact point between the multi-arm space robot and the target object needs to meet the following requirements:

[0016]

[0017] Among them, μ f is the Coulomb friction coefficient, f in is the force at the end of the robot arm along the normal direction of the contact surface, f io 、f it is the force along the tangential direction.

[0018] In the above method, the stable contact quality index includes the force spiral space ellipsoid volume index and the maximum resistance spiral;

[0019] Among them, the force spiral space ellipsoid volume index is:

[0020]

[0021] Among them, G is the comprehensive capture matrix, M fIt is the volume index of the ellipsoid in the force spiral space;

[0022] Maximum resistance spiral quality indicators:

[0023]

[0024] Among them, M r is the maximum resistance spiral quality index, r max (a,b) represents the calculation function of the maximum inscribed sphere radius in the convex hull b with a as the center. Represents the convex hull of the linearized force spiral construction for the combination l.

[0025] In the above method, generating a stable grasping point selection strategy for a multi-arm space robot according to the stable contact quality index includes:

[0026] Construct a random combination of n-1 points and construct the corresponding force spiral set G asc ={ω 1 ,ω 2 ,...,ω n-1} and W asc ={ω 1,1 ,ω 1,2 ,...,ω 1,m ,ω 2,1 ,...,ω n-1,m}∪{O}. The force spiral set G asc Dimensionality reduction processing is performed to obtain the reduced-dimensional force spiral set Y Gmds :

[0027]

[0028] Among them, Θ G3 With Λ G3 is a dual-centered matrix The first three eigenvectors and eigenvalues ​​of D G 2 is the square distance matrix obtained by calculating the square Euclidean distance between pairs of data points; H is the centering matrix;

[0029] Force spiral set W asc Dimensionality reduction processing is performed to obtain the reduced-dimensional force spiral set Y Wmds , for Y Gmds With Y Wmds Recombined to form Y Gasc With Y Wasc , and reconstruct Coh(Y Wasc ), calculate the convex hull supporting hyperplane set U passing through point O;

[0030] like Then point O is located at Coh(YWasc ) inside, for All meet the gripping force closure, constructing G σ ={ω 1 ,ω 2 ,...,ω n-1 ,ω c};

[0031] like Point O is located at Coh(Y Wasc ) surface, Constructing a force spiral collection

[0032] like for Constructed G σ ={ω 1 ,ω 2 ,...,ω n-1 ,ω t}Satisfy the force closure condition;

[0033] like Randomly select ω rmds ∈{N 1 ∪N 2}, replace Y Gmds Zhong and ω rmds The closest ω to the Euclidean distance imds , reformed And rebuild the corresponding and Restart the algorithm calculation;

[0034] Based on the obtained G σ , a multi-criteria decision operation based on entropy weight method and distance comprehensive evaluation method is performed on the force spiral space ellipsoid volume and the maximum resistance spiral to select a stable grasping point set:

[0035]

[0036] in, is the stable grasping point at the end of the i-th robot arm in the l-th combination, i = 1, 2, …, k.

[0037] In the above method, the three-dimensional model of the multi-arm space robot is composed of multiple 7-DOF mechanical arms, and the kinematic and dynamic equations of the multi-arm space robot are generated by using the three-dimensional model of the multi-arm space robot, including:

[0038] Construct the kinematic equations of the multi-arm space robot:

[0039]

[0040] in, Represents the velocity vector of the end of the multi-arm robot; represents the base velocity vector of the multi-arm robot system; Indicates the speed of each joint of the multi-arm robot; J b , J m Represent the Jacobian matrices related to the base and the multi-manipulator respectively;

[0041] Construct the forward dynamics equations of the multi-arm space robot:

[0042]

[0043] Among them, H(Θ) is the comprehensive generalized inertia matrix of the multi-arm space robot, τ is the comprehensive generalized driving torque of the multi-arm space robot, Divide the system into linear terms.

[0044] In the above method, the performance evaluation index of the grasping configuration of the multi-arm space robot is determined according to the stable grasping point of the multi-arm space robot on the target object and the kinematic and dynamic equations of the multi-arm space robot, including:

[0045] According to the kinematic equation of the multi-arm space robot, the operability performance evaluation index M is constructed. o ;

[0046] According to the dynamic equation of the multi-arm space robot, the motion energy consumption performance evaluation index M is constructed. s ;

[0047] According to the change of joint torque in the joint space planning process of the multi-arm space robot, the joint torque peak performance evaluation index M is constructed. τ ;

[0048] According to the performance evaluation index, a multi-objective optimization model of the grasping configuration of the multi-arm space robot is generated, including: according to the operability performance evaluation index M o , Sports energy consumption performance evaluation index M s and joint torque peak performance evaluation index M τ , as well as the motion constraints of the multi-arm space robot, a multi-objective optimization model of the grasping configuration of the multi-arm space robot is generated.

[0049] In the above method, the operability performance evaluation index M is constructed based on the kinematic equation of the multi-arm space robot. o ,include:

[0050] According to the joint speed of the multi-arm space robot during movement Limit value, construct the joint angular velocity constraint weight matrix:

[0051]

[0052] in, is the joint angular velocity constraint weight matrix, is the joint angular velocity of joint j of robot arm i, is the joint angular velocity constraint value of joint j of robot arm i, satisfying

[0053] And based on the joint angular velocity constraint weight matrix, the operability performance evaluation index is constructed:

[0054]

[0055] Among them, M o is the operability performance evaluation index, G is the comprehensive grasping matrix of the multi-arm space robot, is the pseudo-inverse matrix of the Jacobian matrix J; under a fixed basis, J = J m , J m is the Jacobian matrix of the robot arm.

[0056] In the above method, the motion energy consumption performance evaluation index M is constructed based on the dynamic equation of the multi-arm space robot. s ,include:

[0057] According to the initial state configuration of the multi-arm space robot Configuration with target state Use quintic polynomial difference to plan joint space and get the joint angular velocity at each moment

[0058] According to the forward dynamics equation of the multi-arm space robot, the time-varying function τ(t) of the torque of each joint of the multi-arm space robot is obtained, and the motion energy consumption performance evaluation index of the multi-arm space robot is constructed according to the changing parameters:

[0059]

[0060] Among them, M s is the motion energy consumption performance evaluation index, λ is the multi-arm space robot from Θ init to Θ end The number of segments in joint space trajectory planning, t i-1 represents the start time of the i-th segment, t i Represents the end time of the i-th segment.

[0061] In the above method, the joint torque peak performance evaluation index M is constructed according to the change of joint torque in the joint space planning process of the multi-arm space robot. τ ,include:

[0062] According to the forward dynamics equation of the multi-arm space robot and the time-varying function τ(t) of the joint torque of the multi-arm space robot, the joint torque peak performance evaluation index is constructed:

[0063]

[0064] Among them, M τ is the joint torque peak performance evaluation index, and for any time, Represents the joint torque of joint j of robot arm i.

[0065] In the above method, the motion constraint conditions of the multi-arm space robot are the joint angle constraint conditions of the multi-arm space robot; the method further comprises:

[0066] According to the joint angle restrictions of the multi-arm space robot, the joint angle constraint conditions of the multi-arm space robot are established:

[0067] Θ min ≤Θ≤Θ max

[0068] in, is the vector composed of the joint angles of the multi-arm space robot, is the vector of the maximum value of each joint angle of the multi-arm space robot, The vector of the minimum limit values ​​of each joint angle of the multi-arm space robot;

[0069] The operability performance evaluation index M o , Sports energy consumption performance evaluation index M s and joint torque peak performance evaluation index M τ , and the motion constraints of the multi-arm space robot, a multi-objective optimization model of the grasping configuration of the multi-arm space robot is generated, including:

[0070] Based on operability performance evaluation index M o , Sports energy consumption performance evaluation index M s and joint torque peak performance evaluation index M τ , combined with the joint angle constraints of the multi-arm space robot, a multi-objective optimization model of the grasping configuration of the multi-arm space robot is generated:

[0071] findΘ opt

[0072] min M(θ)=[M o (θ)M s (θ)M τ (Θ)] T

[0073] s.tΘmin ≤Θ≤Θ max

[0074] Among them, Θ opt is the optimal grasping configuration of the multi-arm space robot, and Θ is the angle vector of each joint of the multi-arm space robot.

[0075] The grasping planning method of the multi-arm space robot designed in the embodiment of the present invention effectively solves the problem of poor stability of the grasping points generated by the target object. At the same time, it comprehensively considers the self-coupling characteristics of the multi-arm space robot system and realizes the grasping configuration optimization combined with the multi-objective optimization method, which effectively meets the grasping planning requirements of the multi-arm space robot before grasping the target object.

Brief Description of the Drawings

[0076] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creativity and labor.

[0077] Figure 1 is an operation flow chart provided by an embodiment of the present invention;

[0078] Figure 2 It is a schematic diagram of a multi-arm space robot-target object grasping model provided by an embodiment of the present invention;

[0079] Figure 3 It is a flow chart of a stable grasping point selection algorithm for a multi-arm space robot provided by an embodiment of the present invention;

[0080] Figure 4 is a schematic diagram of a single robot arm DH coordinate system provided by an embodiment of the present invention;

[0081] Figure 5 is a schematic diagram of an equivalent dynamics model of a multi-arm space robot provided in an embodiment of the present invention;

[0082] Figure 6 This is a three-grasp point selection effect diagram of different target objects provided by an embodiment of the present invention;

[0083] Figure 7 is a Pareto diagram of the optimization of the grasping configuration of the multi-arm space robot provided in an embodiment of the present invention;

[0084] Figure 8 is an initial grasping diagram of a multi-arm space robot provided in an embodiment of the present invention;

[0085] Fig. 9This is a diagram showing the grasping effect of the multi-arm space robot provided in an embodiment of the present invention. [Specific implementation method]

[0086] In order to better understand the technical solution of the present invention, the embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0087] It should be clear that the described embodiments are only some embodiments of the present invention, not all 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.

[0088] The embodiment of the present invention provides a grasping planning method for a multi-arm space robot, which includes the following steps. For the operation process, please refer to Figure 1 :

[0089] Step 101, using the three-dimensional model of the target object, according to the stable contact condition between the end of the multi-arm space robot and the surface of the object, a stable contact quality index is determined.

[0090] For details, please refer to Figure 2 , which is a multi-arm space robot-target object grasping model. The target object three-dimensional model is a point cloud model containing the shape information of the target object, and each point contains the corresponding normal vector information.

[0091] The force generated by the end of the i-th robotic arm of a multi-arm space robot At contact point c i The force spiral It can be expressed as:

[0092]

[0093] Among them, B i The contact point c i The force spiral basis at the contact point is determined by the friction contact type at the contact point. When the end of the robot touches the object, consider transforming the grasping force from the local coordinate system to the object's center of mass coordinate system, then the contact point c i The pose relative to the object's center of mass coordinate system is R Si , it can be expressed as:

[0094]

[0095] The end of the i-th arm of the multi-arm robot is at c i Gripping force The torque generated at the center of mass of the object is Grasping force The force spiral generated relative to the center of mass of the object It is expressed as follows:

[0096]

[0097] in,

[0098]

[0099] is the force spiral transformation matrix of the contact position relative to the center of mass of the target object, for The antisymmetric matrix G i is the linear mapping matrix of the contact force at the end of the robot i converted to the object mass center coordinate system. Then the comprehensive force spiral F generated by each contact point at the mass center of the target object S for:

[0100]

[0101] Among them, F S is the comprehensive force spiral generated by each contact point at the center of mass of the target object, G = [G 1 … G k ] is the comprehensive grasping matrix of the multi-arm space robot, is the grasping force vector generated by the ends of each mechanical arm of the multi-arm space robot;

[0102] Firstly, the stable contact conditions between the end of the multi-arm space robot and the surface of the object are analyzed. There exists f that satisfies:

[0103]

[0104] Among them, w ext The force spiral is generated by the external force. The multi-arm space robot and the surface of the target object are considered to have friction point contact. In order to prevent the contact point from sliding, the force at the contact point with the target object needs to meet the following requirements:

[0105]

[0106] Among them, μ f is the Coulomb friction coefficient, f in is the force at the end of the robot arm along the normal direction of the contact surface, f io 、f it is the force along the tangent direction. Specifically, in this embodiment, the Coulomb friction coefficient μ f =0.3.

[0107] Step 102: generating a stable grasping point selection strategy for the multi-arm space robot according to the stable contact quality index, wherein the stable grasping point selection strategy includes the stable grasping points of the multi-arm space robot for the target object.

[0108] Specifically, the stable contact quality index of the multi-arm space robot is first constructed, including two quality indexes: the volume of the force spiral space ellipsoid and the maximum resistance force spiral;

[0109] (1) Ellipsoid volume of force spiral space

[0110] For the comprehensive force spiral F S satisfy:

[0111] ||F S || 2 =F S T F S =f T (G+) T G + f=f T (GG T ) -1 f (9)

[0112] Among them, G + is the generalized inverse matrix of the comprehensive capture matrix G, let ||F S || 2 ≤1, the external force spiral can be equivalent to an ellipsoid. According to the matrix singular value decomposition theory, the singular values ​​of the grasping matrix G correspond to the major and minor axes of the ellipsoid, the minor axis is the minimum singular value, and the major axis is the maximum singular value. The larger the volume of the ellipsoid, the better the grasping stability. The force spiral space ellipsoid volume index is established as follows:

[0113]

[0114] Among them, G is the comprehensive capture matrix, M f is the volume index of the force spiral space ellipsoid

[0115] (2) Maximum resistance spiral

[0116] According to the contact type between the multi-arm space robot and the target object, the friction cone at the contact point is linearized into an m-polygon three-dimensional polyhedron convex cone. i U f in The limit value is

[0117]

[0118] Among them, s i,j Indicates the contact point c iThe j-th edge of , can be constructed based on the above decomposition to form a linearized force spiral ω i,j =(s i,j τ i,j ) T , where τ i,j =p i ×s i,j For the contact point set C between the multi-arm robot and the surface of the object, 1 ,p 2 ,...,p k}, there exists an unlinearized force spiral set F S ={ω 1 ,ω 2 ,...,ω i ,...,ω n}, according to the friction cone at the linearized contact point, the corresponding friction contact linearized force spiral set can be constructed:

[0119] F SL ={ω 1,1 ,ω 1,2 ,...,ω 1,m ,ω 2,1 ,...,ω i,j ,...,ω k,m} (12)

[0120] in, is the force spiral of the jth edge after linearization of contact point i, i = 1, 2, ..., k, j = 1, 2, ..., m, k is the number of manipulators of the multi-arm robot, and m is the number of linearized edges of the friction cone. The linearized force spiral set F constructed by k contact points SL There are k×m single-force helices.

[0121] The necessary and sufficient condition for the closure of the grasping force is that the origin O of the force spiral space is located inside the convex hull of the force spiral, that is, The grasping quality index is defined as the maximum force rotation max(w ext ), which can be equivalent to O as the center in Coh(F SL ), the maximum resistance spiral quality index is constructed as follows:

[0122]

[0123] Among them, r max (a,b) represents the calculation function of the maximum inscribed sphere radius in the convex hull b with a as the center. Represents the convex hull of the linearized force spiral construction for the combination l.

[0124] Secondly, select the stable grasping points of the target object according to the following process. Please refer to the algorithm process Figure 3 :

[0125] Step 1: Construct a random combination of n-1 points and construct the corresponding force spiral set G asc ={ω 1 ,ω 2 ,...,ω n-1} and W asc ={ω 1,1 ,ω 1,2 ,...,ω 1,m ,ω 2,1 ,...,ω n-1,m}∪{O}.

[0126] Step 2: Use the multidimensional scaling method (MDS) to calculate the unlinearized force spiral G asc Dimensionality reduction processing is performed to obtain the reduced-dimensional force spiral set Y Gmds , as shown below:

[0127]

[0128] Among them, Θ G3 With Λ G3 is a dual-centered matrix The first three eigenvectors and eigenvalues ​​of D G 2 is the square distance matrix obtained by calculating the square Euclidean distance between pairs of data points; H is the centering matrix;

[0129] Step 3: Similarly, for the force spiral set W asc The linearized force spiral set Y after dimensionality reduction is obtained Wmds , and recombined to form Y Gasc With Y Wasc . Construct Coh(Y Wasc ), calculate the convex hull supporting hyperplane set U passing through point O;

[0130] Step 4: If Then point O is located at Coh(Y Wasc ) inside, for All meet the gripping force closure, constructing G σ ={ω 1 ,ω 2 ,...,ω n-1 ,ω c}, jump to step 8, otherwise skip this step;

[0131] Step 5: If Point O is located at Coh(Y Wasc ) surface, Constructing a force spiral collection

[0132] Step 6: If for Constructed G σ ={ω 1 ,ω 2 ,...,ωω n-1 ,ω t}Satisfy the force closure condition;

[0133] Step 7: If This means that it is impossible to construct a force closure grasping point, so ω is randomly selected. rmds ∈{N 1 ∪N 2}, replace Y Gmds Zhong and ω rmds The closest ω to the Euclidean distance imds , reformed Return to step 3 and rebuild the corresponding With G k , otherwise skip this step;

[0134] Step 8: Based on the obtained G k (or G σ ), calculate the force spiral space ellipsoid volume M corresponding to each group of grasping point combinations f With maximum resistance spiral M r Quality indicators;

[0135] Step 9: After calculating the combined quality indexes of multiple groups of grasping points, a multi-criteria decision operation based on the entropy weight method and the distance comprehensive evaluation method is performed on the force spiral space ellipsoid volume and the maximum resistance spiral to select a stable grasping point set; first, the generated data is standardized, and the standardized result of the quality index j of the i-th group of grasping points is r ij , calculate the entropy value E corresponding to each quality indicator j :

[0136]

[0137] Where ε is the number of generated grasping point combinations, and the weight w corresponding to each quality indicator is calculated based on the entropy value. j , as shown below:

[0138]

[0139] Thus, the relative proximity C corresponding to each group of grasping points can be constructed i , as shown below:

[0140]

[0141] Among them, vij For r ij Multiply by weight w j The result after transformation, v ij =w j × ij , are the positive and negative ideal solutions of quality index j, According to the relative proximity C i A stable grasping point set can be obtained in is the contact position of the end of the i-th robotic arm in the l-th combination, i = 1, 2, ..., k, thereby completing the selection of stable grasping points of the multi-arm space robot for the target object.

[0142] Step 103, using the three-dimensional model of the multi-arm space robot, generating kinematic and dynamic equations of the multi-arm space robot.

[0143] Specifically, construct a single robot arm DH coordinate system such as Figure 4 As shown, the kinematic equations of the multi-arm space robot are constructed:

[0144]

[0145] in, Represents the velocity vector of the end of the multi-arm robot; represents the base velocity vector of the multi-arm robot system; Indicates the speed of each joint of the multi-arm robot; J b , J m Represent the Jacobian matrices related to the base and the multi-manipulator respectively;

[0146] The equivalent dynamic model of the multi-arm space robot is as follows: Figure 5 As shown, the forward dynamics equation of the multi-arm space robot is constructed as follows:

[0147]

[0148] Among them, H(Θ) is the comprehensive generalized inertia matrix of the multi-arm space robot, τ is the comprehensive generalized driving torque of the multi-arm space robot, Divide the system into linear terms.

[0149] Step 104, determining the performance evaluation index of the grasping configuration of the multi-arm space robot based on the stable grasping points of the multi-arm space robot for the target object and the kinematic and dynamic equations of the multi-arm space robot; and, based on the performance evaluation index, generating a multi-objective optimization model of the grasping configuration of the multi-arm space robot.

[0150] Specifically, based on the kinematic equation of the multi-arm space robot, the operability performance evaluation index M is constructed.o ,include:

[0151] According to the speed of each joint of the multi-arm space robot during its motion Limit value, construct the joint angular velocity constraint weight matrix:

[0152]

[0153] in, is the joint angular velocity constraint weight matrix, is the joint angular velocity of joint j of robot arm i, is the joint angular velocity constraint value of joint j of robot arm i, satisfying

[0154] According to the joint angular velocity constraint weight matrix, the operability performance evaluation index is constructed:

[0155]

[0156] Among them, M o is the operability performance evaluation index, G is the comprehensive grasping matrix of the multi-arm space robot, is the pseudo-inverse matrix of the Jacobian matrix J; under a fixed basis, J = J m , J m is the Jacobian matrix of the robot arm.

[0157] According to the dynamic equation of the multi-arm space robot, the motion energy consumption performance evaluation index M is constructed. s ,include:

[0158] According to the initial state configuration of the multi-arm space robot Configuration with target state Using quintic polynomial interpolation for joint space trajectory planning:

[0159]

[0160] Assume t 0 ,t f is the initial moment and the final moment, based on the configuration of the multi-arm space robot at the initial moment and the final moment (Θ 0 , Θ f ), joint angular velocity Joint angular acceleration The boundary of the joint can be solved and the joint angular velocity at each moment can be obtained

[0161] According to the forward dynamics equation of the multi-arm space robot, the time-varying function τ(t) of the torque of each joint of the multi-arm space robot is obtained, and the motion energy consumption performance evaluation index of the multi-arm space robot is constructed according to the changing parameters:

[0162]

[0163] Among them, M s is the motion energy consumption performance evaluation index, λ is the multi-arm space robot from Θ init to Θ end The number of segments in joint space trajectory planning. t i-1 represents the start time of the i-th segment, t i Represents the end time of the i-th segment;

[0164] According to the change of joint torque in the joint space planning process of the multi-arm space robot, the joint torque peak performance evaluation index M is constructed. τ ,include:

[0165] According to the forward dynamics equation of the multi-arm space robot and the time-varying function τ(t) of the joint torque of the multi-arm space robot, the joint torque peak performance evaluation index is constructed:

[0166]

[0167] Among them, M τ is the joint torque peak performance evaluation index, and for any time, Represents the joint torque of joint j of robot arm i.

[0168] The motion constraint conditions of the multi-arm space robot are the joint angle constraint conditions of the multi-arm space robot. According to the joint angle restrictions of the multi-arm space robot, the joint angle constraint conditions of the multi-arm space robot are established as follows:

[0169] Θ min ≤Θ≤Θ max (25)

[0170] in, is the vector composed of the joint angles of the multi-arm space robot, represents the j-th joint angle of robot arm i, is the vector of the maximum value of each joint angle of the multi-arm space robot, The vector of the minimum limit values ​​of each joint angle of the multi-arm space robot;

[0171] The operability performance evaluation index M o , Sports energy consumption performance evaluation index M s and joint torque peak performance evaluation index Mτ , and the motion constraints of the multi-arm space robot, a multi-objective optimization model of the grasping configuration of the multi-arm space robot is generated, including:

[0172] Based on operability performance evaluation index M o , Sports energy consumption performance evaluation index M s and joint torque peak performance evaluation index M τ , and the motion constraints of the multi-arm space robot, a multi-objective optimization model of the grasping configuration of the multi-arm space robot is generated:

[0173]

[0174] Among them, Θ opt is the optimal grasping configuration of the multi-arm space robot, and Θ is the angle vector of each joint of the multi-arm space robot.

[0175] Step 105 , based on the stable grasping points of the multi-arm space robot on the target object and the multi-objective optimization model, and with stable contact as the goal, grasping planning is performed between the multi-arm space robot and the target object.

[0176] Specifically, the stable grasping planning of the multi-arm space robot for the target object is realized according to the following process:

[0177] Step 1: Generate stable grasping points of the target object according to the three-dimensional model of the target object and the grasping quality index and grasping point selection strategy;

[0178] Step 2: Based on the generated stable grasping points, the optimal grasping configuration of the multi-arm space robot is solved through multi-objective optimization;

[0179] Step 3: Combined with the multi-arm space robot's stable grasping points and optimal grasping configuration for the target object, solve the joint trajectory of the multi-arm space robot from the initial state to the target state to achieve stable grasping planning for the multi-arm space robot.

[0180] According to the above method provided by the embodiment of the present invention, a simulation experiment is carried out to study the grasping planning method of a multi-arm space robot.

[0181] In this embodiment, it is assumed that the target objects to be grasped are spheres, cubes, cylinders, and space station cargo ship models, and the stable grasping point generation strategy is verified. The sizes and data contained are shown in Table 1.

[0182] Table 1 Target object corresponding parameters

[0183]

[0184] Assume that the friction coefficient between the end of the multi-arm space robot and the surface of the target object is μ f= 0.2, the number of linearized edges of the friction cone is m = 10, and the contact force is f in =[00-1.0] T , the negative sign indicates that the normal direction points to the inside of the target object. The above parameter values ​​and the target object are used to carry out the simulation verification of grasping point generation with three contact points, four contact points and five contact points.

[0185] Please refer to the simulation results Figure 6 , which are the grasping point generation results of the three contact points of four target objects. Each target object contains three different stable grasping qualities, where the C value represents the grasping quality corresponding to the result. As can be seen from the figure, Figure 6 (a), (d), (g) and (j) are all stable grasping points, and higher quality output corresponds to more stable grasping points, which proves the effectiveness of the proposed grasping point generation strategy for multi-arm space robots. The grasping point pose information corresponding to the four target objects in the simulation results (relative to the center of mass of each target object) is shown in Table 2.

[0186] Table 2 Simulation results of the stable grasping point pose of the target object

[0187]

[0188] Assuming that the structure and parameters of each arm of the multi-arm space robot are the same, the dynamic parameters and initial configuration of the robot arms are shown in Tables 3 and 4.

[0189] Table 3 DH and dynamic parameters of single manipulator

[0190]

[0191]

[0192] Table 4 Initial parameter configuration of multi-arm space robot manipulator

[0193]

[0194] According to the data shown in Tables 2 to 4 as input, the operability, motion energy consumption and joint peak torque performance evaluation indicators are taken as the optimization objectives of the multi-arm space robot grasping configuration. Based on the MATLAB language, a multi-objective grasping configuration optimization simulation experiment based on MOPSO is carried out. The algorithm program setting parameters are shown in Table 5.

[0195] Table 5 MOPSO multi-objective optimization parameter setting table

[0196]

[0197]

[0198] After the simulation program is initialized, the simulation program is run to optimize the multi-objective grasping configuration of the multi-arm space robot using the MOPSO algorithm. After 3000 iterations, the multi-objective optimization simulation results of the simple model of the space station cargo ship are as follows: Figure 7 shown.

[0199] Figure 7 The three-dimensional surface formed by the scatter plot in is the Pareto frontier of the multi-objective optimization result. The three coordinate axes in the figure represent the three objective functions of grasping configuration optimization. The values ​​of decision variables and optimization objectives corresponding to points A, B, and C are shown in Table 6.

[0200] Table 6 Multi-objective optimization results of multi-arm space robot (four points A, B, C, and D)

[0201]

[0202]

[0203] Depend on Figure 7 It can be seen that the closer to point A, the better the global operability of the multi-arm space robot; the closer to point C, the smaller the energy consumption during movement; the closer to point D, the smaller the peak value of the joint torque; according to the Pareto optimal solution set, the vertex B of the three-dimensional frontier surface is selected as the optimization result, so that the values ​​of the three optimization objectives are at a relatively reasonable level, thereby obtaining the optimal grasping configuration of the multi-arm space robot.

[0204] Combining the stable grasping point of the target object with the optimal grasping configuration, the grasping planning of the target object by the multi-arm space robot can be realized. The final simulation results before and after are referenced. Figure 8 , Fig. 9 . Figure 8 is the grasping state of the multi-arm space robot before grasping the target object. Fig. 9 For the actual crawling status, combined with Figure 8 , Fig. 9 It can be concluded from Table 6 that the grasping configuration optimization method can achieve low motion energy consumption and joint peak torque while ensuring high operability, and effectively realize the grasping operation of the target object.

[0205] The technical solution of the embodiment of the present invention has the following beneficial effects:

[0206] In the technical solution of the embodiment of the present invention, a grasping planning method for a multi-arm space robot is designed, and the three-dimensional model of a known target object is used to analyze the stable contact conditions between the end of the multi-arm space robot and the surface of the object, and a stable contact quality index is constructed based on the stable contact conditions; based on the stable contact quality index of the multi-arm space robot, a stable grasping point selection strategy for the multi-arm space robot is designed to generate a stable grasping point of the multi-arm space robot for the target object; the three-dimensional model of the multi-arm space robot is obtained, and the kinematic and dynamic equations of the multi-arm space robot are constructed based on the three-dimensional model; based on the stable grasping points of the multi-arm space robot for the target object and the kinematic and dynamic equations of the multi-arm space robot; a performance evaluation index of the grasping configuration of the multi-arm space robot is constructed, and based on the performance evaluation index, a multi-objective optimization model of the grasping configuration of the multi-arm space robot is constructed, and the MOPSO algorithm is used to perform multi-objective optimization on the grasping configuration of the multi-arm space robot; based on the stable grasping points of the multi-arm space robot for the target object and the optimal grasping configuration, a grasping plan for achieving stable contact between the multi-arm space robot and the target object is performed.

[0207] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

[0208] The contents not described in detail in the specification of the present invention belong to the common knowledge of those skilled in the art.

Claims

1. A grasping planning method for a multi-arm space robot, characterized in that: The method comprises: Using the three-dimensional model of the target object, the stable contact quality index is determined according to the stable contact condition between the end of the multi-arm space robot and the surface of the object; According to the stable contact quality index, a stable grasping point selection strategy of the multi-arm space robot is generated, wherein the stable grasping point selection strategy includes a stable grasping point of the multi-arm space robot for the target object; Using the three-dimensional model of the multi-arm space robot, the kinematic and dynamic equations of the multi-arm space robot are generated; Determine the performance evaluation index of the grasping configuration of the multi-arm space robot according to the stable grasping point of the multi-arm space robot for the target object and the kinematic and dynamic equations of the multi-arm space robot; and generate a multi-objective optimization model of the grasping configuration of the multi-arm space robot according to the performance evaluation index; According to the stable grasping points of the multi-arm space robot for the target object and the multi-objective optimization model, and with stable contact as the goal, grasping planning is performed between the multi-arm space robot and the target object.

2. The method according to claim 1, characterized in that The three-dimensional model of the target object is a point cloud model containing shape information of the target object, and each point contains corresponding normal vector information; The stable contact conditions between the end of the multi-arm space robot and the surface of the object include: for There exists f that satisfies: Among them, w ext is the force spiral generated by the external force, F S is the comprehensive force spiral generated by each contact point at the center of mass of the target object, G = [G1 … G k ] is the comprehensive grasping matrix of the multi-arm space robot, is the grasping force vector generated by the end of each manipulator of the multi-arm space robot, and k is the number of contact points; The force at the contact point between the multi-arm space robot and the target object needs to meet the following requirements: Among them, μ f is the Coulomb friction coefficient, f in is the force at the end of the robot arm along the normal direction of the contact surface, f io 、f it is the force along the tangential direction.

3. The method according to claim 1, characterized in that The stable contact quality index includes the force spiral space ellipsoid volume index and the maximum resistance force spiral; Among them, the force spiral space ellipsoid volume index is: Among them, G is the comprehensive capture matrix, M f It is the volume index of the ellipsoid in the force spiral space; Maximum resistance spiral quality index: Among them, M r is the maximum resistance spiral quality index, r max (a,b) represents the calculation function of the maximum inscribed sphere radius in the convex hull b with a as the center. Represents the convex hull of the linearized force spiral construction for the combination l.

4. The method according to claim 1, characterized in that: The step of generating a stable grasping point selection strategy for a multi-arm space robot according to the stable contact quality index comprises: Construct a random combination of n-1 points and construct the corresponding force spiral set G asc ={ω1,ω2,...,ω n-1 } and W asc ={ω 1,1 ,ω 1,2 ,...,ω 1,m ,ω 2,1 ,...,ω n-1,m }∪{O}, the force spiral set G asc Dimensionality reduction processing is performed to obtain the reduced-dimensional force spiral set Y Gmds : Among them, Θ G3 With Λ G3 is a dual-centered matrix The first three eigenvectors and eigenvalues ​​of D G 2 is the square distance matrix obtained by calculating the square Euclidean distance between pairs of data points; H is the centering matrix; Force spiral set W asc Dimensionality reduction processing is performed to obtain the reduced-dimensional force spiral set Y Wmds , for Y Gmds With Y Wmds Recombined to form Y Gasc With Y Wasc , and reconstruct Coh(Y Wasc ), calculate the convex hull supporting hyperplane set U passing through point O; like Then point O is located at Coh(Y Wasc ) inside, for All meet the gripping force closure, constructing G σ ={ω1,ω2,...,ω n-1 ,ω c }; like Point O is located at Coh(Y Wasc ) surface, Constructing a force spiral collection like for Constructed G σ ={ω1,ω2,...,ω n-1 ,ω t }Satisfy the force closure condition; like Randomly select ω rmds ∈{N1∪N2}, replace Y Gmds Zhong and ω rmds The closest ω to the Euclidean distance imds , reformed And rebuild the corresponding and Restart the algorithm calculation; Based on the obtained G σ , a multi-criteria decision operation based on entropy weight method and distance comprehensive evaluation method is performed on the force spiral space ellipsoid volume and the maximum resistance spiral to select a stable grasping point set: in, is the stable grasping point at the end of the i-th robot arm in the l-th combination, i = 1, 2, …, k.

5. The method according to claim 1, characterized in that The three-dimensional model of the multi-arm space robot is composed of a plurality of 7-DOF mechanical arms. The three-dimensional model of the multi-arm space robot is used to generate the kinematic and dynamic equations of the multi-arm space robot, including: Construct the kinematic equations of the multi-arm space robot: in, Represents the velocity vector of the end of the multi-arm robot; represents the base velocity vector of the multi-arm robot system; Indicates the speed of each joint of the multi-arm robot; J b , J m Represent the Jacobian matrices related to the base and the multi-manipulator respectively; Construct the forward dynamics equations of the multi-arm space robot: Among them, H(Θ) is the comprehensive generalized inertia matrix of the multi-arm space robot, τ is the comprehensive generalized driving torque of the multi-arm space robot, Divide the system into linear terms.

6. The method according to claim 1, characterized in that The performance evaluation index of the grasping configuration of the multi-arm space robot is determined according to the stable grasping point of the multi-arm space robot on the target object and the kinematic and dynamic equations of the multi-arm space robot, including: According to the kinematic equation of the multi-arm space robot, the operability performance evaluation index M is constructed. o ; According to the dynamic equation of the multi-arm space robot, the motion energy consumption performance evaluation index M is constructed. s ; According to the change of joint torque in the joint space planning process of the multi-arm space robot, the joint torque peak performance evaluation index M is constructed. τ ; According to the performance evaluation index, a multi-objective optimization model of the grasping configuration of the multi-arm space robot is generated, including: according to the operability performance evaluation index M o , Sports energy consumption performance evaluation index M s and joint torque peak performance evaluation index M τ , as well as the motion constraints of the multi-arm space robot, a multi-objective optimization model of the grasping configuration of the multi-arm space robot is generated.

7. The method according to claim 6, characterized in that According to the kinematic equation of the multi-arm space robot, the operability performance evaluation index M is constructed. o ,include: According to the joint speed of the multi-arm space robot during movement Limit value, construct the joint angular velocity constraint weight matrix: in, is the joint angular velocity constraint weight matrix, is the joint angular velocity of joint j of robot arm i, is the joint angular velocity constraint value of joint j of robot arm i, satisfying i=1,…,7,j=1,…,k; According to the joint angular velocity constraint weight matrix, the operability performance evaluation index is constructed: Among them, M o is the operability performance evaluation index, G is the comprehensive grasping matrix of the multi-arm space robot, is the pseudo-inverse matrix of the Jacobian matrix J; under a fixed basis, J = J m , J m is the Jacobian matrix of the robot arm.

8. The method according to claim 6, characterized in that According to the dynamic equation of the multi-arm space robot, the motion energy consumption performance evaluation index M is constructed. s ,include: According to the initial state configuration of the multi-arm space robot Configuration with target state Use quintic polynomial difference to plan joint space and get the joint angular velocity at each moment According to the forward dynamics equation of the multi-arm space robot, the time-varying function τ(t) of the torque of each joint of the multi-arm space robot is obtained, and the motion energy consumption performance evaluation index of the multi-arm space robot is constructed according to the changing parameters: Among them, M s is the motion energy consumption performance evaluation index, λ is the multi-arm space robot from Θ init to Θ end The number of segments in joint space trajectory planning, t i-1 represents the start time of the i-th segment, t i Represents the end time of the i-th segment.

9. The method according to claim 6, characterized in that According to the change of joint torque in the joint space planning process of the multi-arm space robot, the joint torque peak performance evaluation index M is constructed. τ ,include: According to the forward dynamics equation of the multi-arm space robot and the time-varying function τ(t) of the joint torque of the multi-arm space robot, the joint torque peak performance evaluation index is constructed: Among them, M τ is the joint torque peak performance evaluation index, and for any time, Represents the joint torque of joint j of robot arm i.

10. The method according to claim 6, characterized in that The motion constraint conditions of the multi-arm space robot are the joint angle constraint conditions of the multi-arm space robot; the method further comprises: According to the joint angle restrictions of the multi-arm space robot, the joint angle constraints of the multi-arm space robot are established: I min ≤Θ≤Θ max in, is the vector composed of the joint angles of the multi-arm space robot, is the vector of the maximum value of each joint angle of the multi-arm space robot, It is a vector composed of the minimum limit values ​​of each joint angle of the multi-arm space robot.

11. The method according to claim 6, characterized in that The operability performance evaluation index M o , Sports energy consumption performance evaluation index M s and joint torque peak performance evaluation index M τ , and the motion constraints of the multi-arm space robot, a multi-objective optimization model of the grasping configuration of the multi-arm space robot is generated, including: Based on operability performance evaluation index M o , Sports energy consumption performance evaluation index M s and joint torque peak performance evaluation index M τ , and the joint angle constraints of the multi-arm space robot, a multi-objective optimization model of the grasping configuration of the multi-arm space robot is generated: findΘ opt min M(Θ)=[M o (W)M s (W)M τ (I)] T s.tΘ min ≤Θ≤Θ max Among them, Θ opt is the optimal grasping configuration of the multi-arm space robot, and Θ is the angle vector of each joint of the multi-arm space robot.

Images