Method for optimizing configuration before capturing of double-branch modular robot

Abstract

The invention provides a pre-capture configuration optimization method for a double-branch modular robot, and the method comprises the steps: constructing a kinematics and dynamics model of the double-branch modular robot through a three-dimensional model of the double-branch modular robot; according to the double-branch modular robot dynamic model, a contact model between the robot and a captured target is determined, and a contact collision dynamic equation of the double-branch modular robot is constructed; according to contact collision analysis before the double-branch modular robot captures the target, key factors before the double-branch modular robot captures the target are considered, and a performance evaluation index and an optimization model are constructed; and through a multi-objective optimization algorithm, solving to obtain a configuration solution set before the double-branch modular robot captures, and according to a good and bad solution distance method, obtaining an optimal configuration before the capturing. According to the technical scheme provided by the embodiment of the invention, the disturbance between the tail end pose and the base pose of the double-branch modular robot before capturing can be effectively reduced, and the contact collision force is reduced.

Description

Technical Field

[0001] The present invention belongs to the field of robot configuration optimization, and relates to a pre-capture configuration optimization method for a double-branch modular robot.

Background Art

[0002] Modular robots are complex systems containing multiple modular units. Through the adjustment of joint arrangements and connection relationships inside them, they can possess different motion performances (such as workspace, dexterity, etc.) and operation capabilities (such as dynamic load capacity, end output force, etc.). In a space environment, modular robots need to perform precise operations in a microgravity environment. This environment leads to inertia-dominated motion, and the robot will be affected by greater dynamic coupling during operation, thus affecting the controllability and stability of its motion. Configuration optimization can help modular robots avoid singularities to the greatest extent, ensure that each joint can maintain smooth and continuous motion during work, and prevent situations such as attitude loss of control or limited operation range. At the same time, based on the particularity of space missions and the high-difficulty operation requirements, the pre-capture configuration optimization method for modular robots can enable them to perform tasks with high precision, high stability, and high efficiency, ensuring sufficient safety and adaptability in complex space missions.

[0003] Currently, the configuration optimization methods for single-branch modular robots are already very sufficient. Compared with single-branch modular robots, the coupling relationship between the manipulator and the base of double-branch modular robots is more complex, with more degrees of freedom, and the singularity problem is more complex. However, at the present stage, regarding the configuration optimization methods for double-branch modular robots, most of the focus is on the operable factors after the robot captures the target. For a freely floating double-branch modular robot, when capturing the target, due to the contact collision between the manipulator and the target, it will cause disturbances at the end and the base of the manipulator while causing end disturbances, and may even lead to capture failure. Therefore, considering the contact collision effects of double-branch modular robots and combining the key factors during the task execution process, the optimal configuration of the double-branch modular robot before capturing the target is selected. Therefore, in-depth research on the configuration optimization method of double-branch modular robots has great theoretical value and practical significance for promoting the wide application of modular robots in the fields of aerospace and so on.

Summary of the Invention

[0004] In view of this, the present invention provides a pre-capture configuration optimization method for a double-branch modular robot for the background of on-orbit operation tasks in space. The proposed method can effectively reduce the end pose and base attitude disturbances of the double-branch modular robot before capture and reduce the contact collision force.

[0005] The present invention provides a pre-capture configuration optimization method for a double-branch modular robot, including:

[0006] Step S1 uses the three-dimensional model of the dual-branch modular robot to construct the kinematic and dynamic models of the dual-branch modular robot;

[0007] Step S2 determines the contact model between the robot and the capture target based on the dynamic model of the dual-branch modular robot, and constructs the contact collision dynamic equation of the dual-branch modular robot;

[0008] Step S3 constructs the performance evaluation index and optimization model by considering the key factors before the dual-branch modular robot captures the target based on the contact collision analysis before the dual-branch modular robot captures the target;

[0009] Step S4 uses the multi-objective optimization algorithm to solve and obtain the solution set of the configuration before the dual-branch modular robot captures, and obtains the optimal configuration before capture according to the distance method of the superior and inferior solutions.

[0010] In the above method, step S1 includes:

[0011] Step S1.1 constructs the kinematic equation of the dual-branch modular robot.

[0012]

[0013] Among them, and respectively represent the velocity vectors at the ends of the left and right branches; J bL and J bR respectively represent the Jacobian matrices of the velocity transfer from the base to the ends of the left and right branches of the modular robot; represents the velocity vector of the robot base; J θL and J θR respectively represent the Jacobian matrices of the joint angles and the velocity transfer at the ends of the left and right branches of the modular robot, where n l and n r respectively represent the degrees of freedom of the left and right branches of the modular robot; represents the joint angular velocity vector; it can be simplified as:

[0014]

[0015] Among them, represents the end velocity vector of the dual-branch modular robot, J b represents the Jacobian matrix of the velocity transfer from the base to the end of the dual-branch modular robot, J θ represents the Jacobian matrix of the joint angle and the velocity transfer at the end;

[0016] Therefore, based on the conservation of momentum and angular momentum of the dual-branch modular robot in the free-floating state, the kinematic equation is further constructed as:

[0017]

[0018] Among them, J float represents the generalized Jacobian matrix of the double-branch modular robot;

[0019] Step S1.2 constructs the forward dynamics equation of the joint space of the double-branch modular robot.

[0020] In the space environment, the gravitational potential energy of the double-branch modular robot is negligible. The forward dynamics equation of the joint space can be obtained by using the Lagrangian method:

[0021]

[0022] Among them, q = [q b qθ] T represents the base and joint angle vector of the double-branch modular robot; H(q) represents the inertia matrix of the double-branch modular robot; is the system nonlinear term; F s represents the generalized driving force; F e represents the end output force.

[0023] In the above method, the step S2 includes:

[0024] Based on the continuous contact collision dynamics modeling method, a nonlinear damping model is used to describe the contact collision process, where the contact collision force of the nonlinear damping model is:

[0025]

[0026] Among them, i = L, R represents the left and right branches of the modular robot, k represents the Hertz stiffness, ξ represents the collision coefficient, λ represents the damping coefficient, δ represents the relative compression amount along the normal direction of the contact surface, represents the relative compression velocity; combined with the dynamics model of the double-branch modular robot, a contact collision dynamics model is constructed:

[0027]

[0028] Among them, F e = [F eL , F eR T , F ei = n i F i , ||F ei || = F i , i = L, R, n i represents the normal direction vector of the contact surface of the i branch;

[0029] In the above method, the step S3 includes:​

[0030] Step S3.1 Define the comprehensive contact and collision force, end pose perturbation, and base attitude perturbation of the dual-branch modular robot before capture as three performance evaluation indicators according to the contact and collision dynamics model of the dual-branch modular robot;

[0031] Step S3.1.1 Construct a performance evaluation indicator for the comprehensive contact and collision force according to the contact and collision dynamics model of the dual-branch modular robot;

[0032] The comprehensive contact and collision force function is defined as the average of the left and right branch contact and collision forces of the dual-branch modular robot before capturing the target:

[0033] f 1 =mean(F eL ,F eR )

[0034] where the mean(·) function is used by the user to calculate the average of the input data;

[0035] Step S3.1.2 Construct a performance evaluation indicator for the end pose perturbation of the dual-branch modular robot;

[0036] After the space target contacts the end effector of the dual-branch modular robot, a large end pose perturbation of the robot may cause the end to deviate from the target and lead to mission failure. According to the impulse and angular momentum theorem, we can obtain:

[0037]

[0038] where m e is the equivalent mass of the dual-branch modular robot, I e is the equivalent inertia of the dual-branch modular robot, δv e and δω e are the angle and angular velocity change amounts of the dual-branch modular robot after contacting the target, respectively, and P c and M c represent the contact force impulse and torque impulse amplitudes generated at the end after the space target contacts and collides with the dual-branch modular robot;

[0039] According to the end linear velocity direction vector u and the rotation axis direction vector z of the end angular velocity of the dual-branch modular robot, we can obtain

[0040]

[0041] where J v and J ω represent the angular velocity and linear velocity Jacobian matrices of the dual-branch modular robot, respectively. Thus, a performance evaluation indicator for the end pose perturbation can be constructed:

[0042]

[0043] Among them, χ 1 and χ 2 represent the linear velocity and angular velocity weight coefficients of the end perturbation of the double-branch modular robot;

[0044] Step S3.1.3 Construct the evaluation index of the base attitude perturbation performance of the double-branch modular robot;

[0045] According to the dynamic equation of the double-branch modular robot, the functional relationship between the base attitude perturbation and the joint angle and contact collision impulse can be established:

[0046]

[0047] Among them, is the change in the base angular velocity, is the inertia matrix related to the base angular velocity, is the coupling matrix related to the base angular velocity and joint angular velocity, is the inertia matrix related to the joint angle, is the Jacobian matrix related to the joint angular velocity, is the Jacobian matrix related to the joint angle, Φ e represents the contact collision impulse received at the end;

[0048] When the amplitude of the contact collision impulse remains constant, the base attitude perturbation is only related to the joint angles of the double-branch modular robot. To optimize the base attitude perturbation caused by contact collision, the corresponding objective function can be constructed:

[0049]

[0050] Among them, ||·|| is the two-norm of any vector, f 3 The smaller the numerical value, the smaller the base attitude perturbation.

[0051] Step S3.2 Considering the joint limit constraints of the modular robot, the following constraints are established for the pre-capture configuration of the double-branch modular robot:

[0052] θ min ≤θ≤θ max

[0053] Among them, θ min represents the minimum joint angle sequence of the double-branch modular robot, and θ max represents the maximum joint angle sequence of the double-branch modular robot. The joint angles corresponding to the robot capture configuration should satisfy this range.

[0054] Step S3.3 combines Step S3.1 and Step S3.2 to define the pre-capture configuration optimization model of the dual-branch modular robot as:

[0055] findθ best

[0056]

[0057] s.t.θ min ≤θ≤θ max

[0058] where θ best is the optimal pre-capture configuration of the dual-branch modular robot.

[0059] In the above method, the said Step S4 includes:

[0060] Step S4.1 Based on the multi-objective optimization model constructed in S3.3, the non-dominated sorting genetic algorithm (NSGA-II) is used to solve the above optimization model to obtain the Pareto non-dominated solution set;

[0061] Step S4.2 uses the technique for order preference by similarity to ideal solution (TOPSIS) to evaluate the above Pareto solution set, realizes the comprehensive evaluation of the results of the three performance evaluation indexes in S4.1, and thus sorts the solution set of the pre-capture configuration of the dual-branch modular robot to obtain the optimal pre-capture configuration;

[0062] Step S4.2.1 Arrange the results of the three performance evaluation indexes in the Pareto solution set obtained in S4.1 to construct a performance evaluation index solution set matrix

[0063]

[0064] where d ij represents the result of the j-th performance evaluation index of the i-th solution.

[0065] Step S4.2.2 Conduct standardization processing on the performance evaluation index solution set matrix to obtain a standardized performance evaluation index solution set matrix, where each matrix element can be expressed as:

[0066]

[0067] Step S4.2.3 Conduct positive normalization processing on the results of each performance evaluation index to obtain a positive normalized performance evaluation index solution set matrix, where each matrix element is:

[0068]

[0069] where, and They respectively represent the maximum and minimum values of the j-th performance evaluation index in the standardization matrix.

[0070] Step S4.2.4 sets the weight coefficients of each performance evaluation index and calculates the score s of the i-th optimization solution i 。

[0071]

[0072] where w j represents the weight value corresponding to the j-th performance evaluation index function.

[0073] Step S4.2.5 sorts the Pareto solution set according to the scores and obtains the optimal configuration before the capture of the dual-branch modular robot corresponding to the highest score.

BRIEF DESCRIPTION OF THE DRAWINGS

[0074] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those of ordinary skill in the art, without creative and laborious efforts, other drawings can also be obtained based on these drawings.

[0075] Figure 1 is a schematic flow chart of a method for optimizing the pre-capture configuration of a dual-branch modular robot provided by an embodiment of the present invention;

[0076] Figure 2 is a schematic diagram of the structure of the dual-branch modular robot in the simulation experiment of an embodiment of the present invention;

[0077] Figure 3 is a schematic diagram of the multi-objective optimization result in the simulation experiment of an embodiment of the present invention.

DETAILED DESCRIPTION OF THE EMBODIMENTS

[0078] In order to better understand the technical solutions of the present invention, the following will describe the embodiments of the present invention in detail with reference to the drawings.

[0079] It should be clear that the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0080] An embodiment of the present invention provides a method for optimizing the pre-capture configuration of a dual-branch modular robot. Please refer to Figure 1 , which is a schematic flow chart of a method for optimizing the pre-capture configuration of a dual-branch modular robot provided by an example of the present invention. As shown in Figure 1 , the method includes the following steps:

[0081] Step 101: Construct the kinematic and dynamic models of the double-branch modular robot using its three-dimensional model.

[0082] Specifically, in the first step, first construct the kinematic equation of the double-branch modular robot:

[0083]

[0084] Among them, and respectively represent the velocity vectors at the ends of the left and right branches; J bL and J bR respectively represent the Jacobian matrices for the velocity transmission from the base to the ends of the left and right branches of the modular robot; represents the velocity vector of the robot base; J θL and J θR respectively represent the Jacobian matrices for the transmission of the joint angles and end velocities of the left and right branches of the modular robot, where n l and n r respectively represent the degrees of freedom of the left and right branches of the modular robot; represents the joint angular velocity vector.

[0085] Secondly, simplify and represent the above kinematic equation:

[0086]

[0087] Among them, represents the end velocity vector of the double-branch modular robot, J b represents the Jacobian matrix for the velocity transmission from the base to the end of the double-branch modular robot, J θ represents the Jacobian matrix for the transmission of the joint angle and end velocity.

[0088] Based on this, according to the conservation of momentum and angular momentum of the double-branch modular robot in the free-floating state, further construct the kinematic equation as:

[0089]

[0090] Among them, J float represents the generalized Jacobian matrix of the double-branch modular robot.

[0091] Step 2: Construct the forward dynamics equation of the joint space of the double-branch modular robot.

[0092] In the space environment, the gravitational potential energy of the double-branch modular robot is negligible. Using the Lagrangian method, the forward dynamics equation of the joint space can be obtained:

[0093]

[0094] where \(q = [q b _{q\theta}] T represents the base and joint angle vector of the double-branch modular robot; \(H(q)\) represents the inertia matrix of the double-branch modular robot; is the system nonlinear term; \(F s represents the generalized driving force; \(F e represents the end output force.

[0095] Step 102: According to the dynamics model of the double-branch modular robot, determine the contact model between the robot and the captured target, and construct the contact collision dynamics equation of the double-branch modular robot.

[0096] Specifically, based on the continuous contact collision dynamics modeling method, a nonlinear damping model is used to describe the contact collision process, where the contact collision force of the nonlinear damping model is:

[0097]

[0098] where \(i = L, R\) represents the left and right branches of the modular robot, \(k\) represents the Hertz stiffness, \(\xi\) represents the collision coefficient, \(\lambda\) represents the damping coefficient, \(\delta\) represents the relative compression amount along the normal direction of the contact surface, represents the relative compression velocity; combined with the dynamics model of the double-branch modular robot, construct the contact collision dynamics model:

[0099]

[0100] where \(F e = [F eL , F eR T , \(F ei = n i F i , \(\vert\vert F ei \vert\vert = F i , \(i = L, R\), \(n i represents the normal direction vector of the contact surface of the \(i\)-th branch.

[0101] Step 103: According to the contact collision analysis of the double-branch modular robot before capturing the target, consider the key factors before the double-branch modular robot captures the target, and construct the performance evaluation index and optimization model.

[0102] Specifically, first, according to the contact collision dynamics model of the double-branch modular robot, define the comprehensive contact collision force, end pose perturbation, and base pose perturbation before the double-branch modular robot captures the target as three performance evaluation indexes.

[0103] ​Step 1: Based on the contact and collision dynamics model of the dual-branch modular robot, construct a comprehensive evaluation index for the performance of the contact and collision force.

[0104] The comprehensive contact and collision force function is defined as the mean of the left and right branch contact and collision forces before the dual-branch modular robot captures the target:

[0105] f 1 = mean(F eL , F eR )

[0106] where the mean(·) function is used by the user to calculate the mean of the input data.

[0107] Step 2: Construct an evaluation index for the pose perturbation performance of the end of the dual-branch modular robot; after the spatial target contacts the end effector of the dual-branch modular robot, a large pose perturbation at the end of the robot may cause the end to deviate from the target and lead to task failure. According to the impulse and angular momentum theorem, we can obtain:

[0108]

[0109] where m e is the equivalent mass of the dual-branch modular robot, I e is the equivalent inertia of the dual-branch modular robot, δv e and δω e are the changes in angle and angular velocity of the dual-branch modular robot after contacting the target, respectively, and P c and M c represent the impulse of the contact force and the amplitude of the torque impulse generated at the end after the spatial target contacts and collides with the dual-branch modular robot.

[0110] According to the direction vector u of the linear velocity at the end of the dual-branch modular robot and the direction vector z of the rotation axis of the angular velocity at the end, we can obtain:

[0111]

[0112] where J v and J ω represent the angular velocity and linear velocity Jacobian matrices of the dual-branch modular robot, respectively. From this, we can construct an evaluation index for the pose perturbation performance at the end:

[0113]

[0114] where χ 1 and χ 2 represent the weight coefficients of the linear velocity and angular velocity of the pose perturbation at the end of the dual-branch modular robot.

[0115] Step 3: Construct the evaluation index for the base attitude disturbance performance of the double-branch modular robot.

[0116] Based on the dynamic equation of the double-branch modular robot, a functional relationship between the base attitude disturbance, joint angles, and contact collision impulses can be established:

[0117]

[0118] where, is the change in the base angular velocity, is the inertia matrix related to the base angular velocity, is the coupling matrix related to the base angular velocity and joint angular velocities, is the inertia matrix related to joint angles, is the Jacobian matrix related to joint angular velocities, is the Jacobian matrix related to joint angles, Φ e represents the contact collision impulse received at the end effector.

[0119] When the amplitude of the contact collision impulse remains constant, the base attitude disturbance is only related to the joint angles of the double-branch modular robot. To optimize the base attitude disturbance caused by contact collision, a corresponding objective function can be constructed:

[0120]

[0121] where, ||·|| is the two-norm of any vector, and f 3 The smaller the value, the smaller the base attitude disturbance.

[0122] Secondly, considering the joint limit constraints of the modular robot, the following constraints are established for the pre-capture configuration of the double-branch modular robot:

[0123] θ min ≤ θ ≤ θ max

[0124] where, θ min represents the minimum joint angle sequence of the double-branch modular robot, and θ max represents the maximum joint angle sequence of the double-branch modular robot. The joint angles corresponding to the robot's capture configuration should satisfy this range.

[0125] Finally, combining the above steps, the optimization model for the pre-capture configuration of the double-branch modular robot is defined as:

[0126] find θ best

[0127]

[0128] s.t. θ min≤θ≤θ max

[0129] Among them, θ best is the optimal configuration before the double-branch modular robot captures.

[0130] Step 104: Solve the solution set of the configuration before the double-branch modular robot captures through a multi-objective optimization algorithm, and obtain the optimal configuration before capture according to the distance method of the superior and inferior solutions.

[0131] Specifically, first, based on the multi-objective optimization model constructed in Step 103, use the non-dominated sorting genetic algorithm (NSGA-II) to solve the above optimization model to obtain the Pareto non-dominated solution set.

[0132] Secondly, evaluate the above Pareto solution set using the distance method of the superior and inferior solutions, realize the comprehensive evaluation of the results of the three performance evaluation indicators in S4.1, so as to sort the solution set of the configuration before the double-branch modular robot captures, and obtain the optimal configuration before capture.

[0133] Step 1: Arrange the results of the three performance evaluation indicators in the Pareto solution set obtained in S4.1 to construct a performance evaluation index solution set matrix D N :

[0134]

[0135] Among them, d ij represents the result of the jth performance evaluation indicator of the ith solution.

[0136] Step 2: Conduct standardization processing on the performance evaluation index solution set matrix to obtain a standardized performance evaluation index solution set matrix, where each matrix element can be expressed as:

[0137]

[0138] Step 3: Conduct positive normalization processing on the results of each performance evaluation indicator to obtain a positive normalized performance evaluation index solution set matrix, where each matrix element is:

[0139]

[0140] Among them, and respectively represent the maximum and minimum values of the jth performance evaluation indicator in the standardized matrix.

[0141] Step 4: Set the weight coefficients of each performance evaluation indicator, and calculate the score s i .

[0142]

[0143] Among them, w j represents the weight value corresponding to the j-th performance evaluation index function.

[0144] Step 5: Sort the Pareto solution set according to the scores, and obtain the optimal configuration before the capture of the dual-branch modular robot corresponding to the highest score.

[0145] Based on the above method provided by the embodiment of the present invention, a simulation experiment is carried out with the dual-branch modular robot as the object. The configurations of the two branches are the same, and a single branch is composed of three modular motion units, as Figure 2 shown.

[0146] According to the NSGA-II multi-objective optimization algorithm, the optimization results are as Figure 2 shown. According to the scores, the Pareto solution set can be sorted, as shown in Table 1, and the optimal configuration before the capture of the dual-branch modular robot in the ideal case is obtained, that is, the second group of data. Among them, the comprehensive contact collision force data of the optimal configuration is -1.602, which is 14.15% higher than the average level of the Pareto solution set; the end pose perturbation data of the optimal configuration is 2.076, which is 29.56% higher than the average level of the Pareto solution set; the base pose perturbation data of the optimal configuration is 0.445, which is 49.28% higher than the average level of the Pareto solution set, which can prove the effectiveness of this method.

[0147] Table 1 Pareto solution set of optimization results and processing results

[0148]

[0149] The technical solution of the embodiment of the present invention has the following beneficial effects:

[0150] Characterize the contact collision dynamics problem of the dual-branch modular robot in the free-floating state, and realize the contact collision analysis before capture; use the NSGA-II multi-objective optimization algorithm to optimize multiple indexes of the dual-branch modular robot in terms of contact collision force, end pose perturbation, and base pose perturbation, and realize the optimal selection of the configuration before capture through the distance method of excellent and poor solutions, effectively reducing the task failure rate caused by contact collision when the dual-branch modular robot captures the target.

[0151] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0152] The content not detailed in the specification of the present invention belongs to the well-known technology of those skilled in the art.

Claims

1. A method for pre-capture configuration optimization of a dual-branch modular robot, characterized in that: The method comprises: Step S1 uses the three-dimensional model of the double-branch modular robot to construct a kinematic and dynamic model of the double-branch modular robot; Step S2 determines the contact model between the robot and the capture target based on the dynamics model of the dual-branch modular robot, and constructs the contact and collision dynamics equation of the dual-branch modular robot; Step S3 is based on the contact collision analysis before the dual-branch modular robot captures the target, considering the key factors before the dual-branch modular robot captures the target, and constructing the performance evaluation index and optimization model; In step S4, a multi-objective optimization algorithm is used to solve the pre-capture configuration solution set of the dual-branch modular robot, and the optimal configuration before capture is obtained according to the superior and inferior solution distance method.

2. The method according to claim 1, characterized in that The step S1 comprises: Step S1.1 constructs the kinematic equations of the dual-branch modular robot: in, and Respectively represent the velocity vectors of the left and right branch ends; J bL With J bR The Jacobian matrices representing the velocity transfer between the base and the left and right branch ends of the modular robot respectively; represents the velocity vector of the robot base; J θL With J θR Respectively represent the Jacobian matrix of the joint angle and terminal velocity transmission of the left and right branches of the modular robot, where n l With n r They represent the number of degrees of freedom of the left and right branches of the modular robot respectively; Represents the joint angular velocity vector; it can be simplified as: in, represents the terminal velocity vector of the dual-branch modular robot, J b The Jacobian matrix representing the velocity transfer between the base and the terminal of the dual-branch modular robot, J θ The Jacobian matrix representing the transfer of joint angles and terminal velocities; Based on the conservation of momentum and angular momentum of the dual-branch modular robot in the free-floating state, the kinematic equation is further constructed as follows: Among them, J float Represents the generalized Jacobian matrix of a two-branch modular robot; Step S1.2 constructs the joint space forward dynamics equation of the dual-branch modular robot; The gravitational potential energy of the dual-branch modular robot in the space environment is negligible, and the forward dynamics equation of the joint space can be obtained by using the Lagrangian method: Where q=[q b q θ ] T represents the base and joint angle vector of the dual-branch modular robot; H(q) represents the inertia matrix of the dual-branch modular robot; is the nonlinear term of the system; F s represents the generalized driving force; F e Indicates the end output force.

3. The method according to claim 1, characterized in that The step S2 comprises: Based on the continuous contact collision dynamics modeling method, a nonlinear damping model is used to describe the contact collision process, where the contact collision force of the nonlinear damping model is: Where i = L, R represents the left branch and the right branch of the modular robot, k represents the Hertzian stiffness, ξ represents the collision coefficient, λ represents the damping coefficient, and δ represents the relative compression along the normal direction of the contact surface. Represents the relative compression velocity; combined with the dual-branch modular robot dynamics model, a contact collision dynamics model is constructed: Among them, F e =[F eL ,F eR ] T , F ei =n i F i , ‖F ei ‖=F i , i=L,R,n i Represents the normal direction vector of the contact surface of branch i.

4. The method according to claim 1, characterized in that The step S3 comprises: Step S3.1 defines the comprehensive contact collision force, terminal posture disturbance, and base posture disturbance of the double-branch modular robot before capture as three performance evaluation indicators based on the contact collision dynamics model of the double-branch modular robot; Step S3.1.1 constructs a comprehensive contact collision force performance evaluation index based on the contact collision dynamics model of the dual-branch modular robot; The comprehensive contact collision force performance evaluation index is defined as the average contact collision force of the left and right branches before the dual-branch module robot captures the target: f1=mean(F eL ,F eR ) Among them, the mean(·) function user calculates the mean of the input data; Step S3.1.2 constructs the evaluation index of the terminal posture disturbance performance of the dual-branch modular robot; After the space target comes into contact with the end effector of the dual-branch modular robot, the large disturbance of the robot's end posture may cause the end to deviate from the target and thus cause the mission failure. According to the impulse and impulse moment theorem, it can be obtained: Among them, m e is the equivalent mass of the dual-branch modular robot, I e is the equivalent inertia of the dual-branch modular robot, δv e With δω e are the angle and angular velocity changes of the dual-branch modular robot after it contacts the target, P c With M c They respectively represent the contact force impulse and torque impulse amplitudes generated on the end after the space target and the dual-branch modular robot come into contact and collide; According to the terminal linear velocity direction vector u and the terminal angular velocity rotation axis direction vector z of the dual-branch modular robot, we can get in, J v With J ω Respectively represent the angular velocity and linear velocity Jacobian matrices of the dual-branch modular robot, from which the terminal posture disturbance performance evaluation index can be constructed: Among them, χ1 and χ2 represent the linear velocity and angular velocity weight coefficients of the end disturbance of the dual-branch modular robot, respectively; Step S3.1.3 constructs a base attitude disturbance performance evaluation index for the dual-branch modular robot; According to the dynamic equation of the dual-branch modular robot, the functional relationship between the base posture disturbance and the joint angle and contact collision pulse can be established: in, is the change in the angular velocity of the base, is the inertia matrix related to the angular velocity of the base, is the coupling matrix between the base angular velocity and the joint angular velocity, is the inertia matrix associated with the joint angle, is the Jacobian matrix related to the joint angular velocity, is the Jacobian matrix associated with the joint angle, Φ e Indicates the contact collision pulse received by the end; When the contact collision pulse amplitude remains constant, the base posture disturbance is only related to the joint angle of the dual-branch modular robot. In order to optimize the base posture disturbance caused by the contact collision, the corresponding objective function can be constructed: Among them, ||·|| is the second norm of any vector, and the smaller the value of f3 is, the smaller the disturbance of the base attitude is; Step S3.2 considers the joint limit constraints of the modular robot and establishes the constraints for the pre-capture configuration of the dual-branch modular robot as follows: i min ≤θ≤θ max Among them, θ min represents the minimum joint angle sequence of the dual-branch modular robot, θ max Represents the maximum joint angle sequence of the dual-branch modular robot. The joint angles corresponding to the robot's capture configuration should be within this range. Step S3.3 combines steps S3.1 and S3.2 to define the pre-capture configuration optimization model of the dual-branch modular robot as follows: findθ best stθ min ≤θ≤θ max Among them, θ best Capture the optimal configuration before the dual-branch modular robot.

5. The method according to claim 1, characterized in that The step S4 comprises: Step S4.1: Based on the pre-capture configuration optimization model constructed in step S3.3, a non-dominated sorting genetic algorithm (NSGA-II) is used to solve the optimization model to obtain a Pareto non-dominated solution set; Step S4.2 uses the superior and inferior solution distance method to evaluate the Pareto solution set in step S4.1, realizes a comprehensive evaluation of the optimization model results, and thus sorts the pre-capture configuration solution set of the dual-branch modular robot to obtain the optimal configuration before capture; Step S4.2.1 Arrange the three performance evaluation index results in the Pareto solution set obtained in S4.1 to construct a performance evaluation index solution set matrix Among them, d ij represents the jth performance evaluation index result of the i-th solution; Step S4.2.2 performs standardization processing on the performance evaluation index solution matrix to obtain a standardized performance evaluation index solution matrix, where each matrix element It can be expressed as Step S4.2.3 performs data forward processing on the results of each performance evaluation index to obtain the forward performance evaluation index solution matrix, where each matrix element for in, and Respectively represent the maximum and minimum values ​​of the jth performance evaluation index in the standardized matrix; Step S4.2.4 sets the weight coefficients of each performance evaluation index and calculates the score s of the i-th optimization solution i Among them, w j Represents the weight value corresponding to the jth performance evaluation index function; Step S4.2.5 sorts the Pareto solution set according to the scores to obtain the optimal configuration of the dual-branch modular robot before capture corresponding to the highest score.

Images