A method for optimizing curved surface mapping trajectory and multimodal force control for complex wall grinding

By optimizing the surface mapping trajectory and using multimodal force control methods, the problems of insufficient path accuracy and force control effect in grinding complex curved surfaces are solved, achieving efficient and high-quality grinding results.

CN120287291BActive Publication Date: 2026-11-17SOUTHEAST UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202510443956.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-04-10
Publication Date
2026-11-17
Estimated Expiration
2045-04-10

AI Technical Summary

Technical Problem

Existing grinding robots suffer from insufficient grinding path accuracy, poor force control, and inadequate dynamic performance in complex curved surface environments, resulting in low grinding efficiency and unstable quality.

Method used

By employing surface mapping trajectory optimization and multimodal force control methods, and by constructing a path optimization model, combined with dynamic attitude adjustment of the robotic arm end effector and a hybrid force control strategy, efficient grinding of complex surfaces can be achieved.

Benefits of technology

It significantly improves grinding efficiency and precision, avoids over-cutting or under-cutting, and ensures grinding quality and stability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120287291B_ABST
    Figure CN120287291B_ABST
Patent Text Reader

Abstract

A method for optimizing curved surface mapping trajectory and multimodal force control for complex wall grinding is proposed, which can efficiently adapt to the grinding requirements of complex curved surfaces. By introducing projection technology to determine the grinding trajectory, constructing a path optimization model, and combining it with multi-node force fusion attitude control, interpolation point calculation, attitude change, and position planning matching are achieved, effectively generating a robot grinding path that includes grinding position, attitude, and adjustment parameters. The multimodal force control method combines radial height state switching based on the state classification of convex and concave path segments to dynamically adjust the force control mode. Furthermore, multi-node fusion attitude control is introduced to accurately track the contact force on the sanding disc surface and dynamically adjust the robot's end-effector attitude to follow the curved surface characteristics. This overcomes the problems of irregular curved surfaces and unstable force control in traditional grinding techniques, and can be widely applied to automated and efficient grinding operations in fields such as construction and shipbuilding, achieving efficient and high-precision grinding of complex curved surfaces.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of industrial robot automation technology, specifically a method for optimizing curved surface mapping trajectory and multimodal force control for grinding complex wall surfaces, which is particularly suitable for efficient and high-precision grinding operations on complex curved surfaces with large curvature, such as building walls and ship hulls. Background Technology

[0002] In complex wall surface environments, traditional manual sanding methods are limited by the operator's individual skills and physical strength, resulting in high labor intensity, low efficiency, and poor work consistency, easily leading to substandard surface flatness and smoothness. In recent years, with the rapid development of industrial intelligence, automated robots have gradually been applied to sanding tasks. However, existing sanding robots still have the following shortcomings in structural design and functional implementation:

[0003] (1) In terms of grinding path planning, for complex curved surface operations, existing robots mostly design grinding paths based on fixed trajectories or approximate interpolation schemes. For complex curvature changes in convex and concave areas, there are often problems such as insufficient trajectory accuracy and low efficiency caused by frequent path adjustments.

[0004] (2) In terms of end-effector posture adjustment, existing robots often cannot dynamically adapt to drastic changes in wall curvature and cannot make efficient posture adjustments to complex curved surfaces, which can easily lead to over-cutting or under-cutting in local areas, thus affecting the polishing effect.

[0005] (3) The force control strategy lacks flexibility. Traditional force control methods are difficult to adjust the grinding force in real time to adapt to the curvature complexity and surface material of different areas, which makes it difficult to guarantee the quality and stability of wall surface machining.

[0006] Therefore, to address the problems of low sanding path accuracy, poor force control, and insufficient dynamic performance in complex curved surface environments, a wall sanding control method based on curved surface mapping trajectory optimization and multimodal force control is urgently needed. This invention proposes an efficient sanding solution for complex wall surface environments to address these challenges. Summary of the Invention

[0007] To address the aforementioned technical challenges, this invention proposes a method for surface mapping trajectory optimization and multimodal force control for grinding complex wall surfaces. This system combines trajectory optimization modeling, dynamic attitude adjustment of the robotic arm's end effector, and a hybrid force control strategy, enabling it to efficiently adapt to the grinding requirements of complex curved surfaces. By constructing a path optimization model and introducing multi-node force fusion attitude control, the grinding efficiency and accuracy of the robotic system are further improved.

[0008] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0009] A method for optimizing curved surface mapping trajectories and multimodal force control for complex wall surface grinding includes the following steps:

[0010] (1) Utilizing the relative positional relationship between the target surface S(x,y,z) and the reference plane P(x,y,z), the surface area is simplified to a two-dimensional plane S(x,z) through projection technology. The curve is discretized to generate grid points, and the set of feature points {Δ,□,○,◇} is identified.

[0011] (2) Define path segment types based on feature point combinations, classifying path segments into convex segments and concave segments; use simulated annealing algorithm to optimize interpolation points s for convex path segments. * For concave path segments, linear interpolation is used to add interpolation points s. # Generate the globally optimal grinding trajectory Q(x,z) that includes information on interpolation points and path turning points;

[0012] (3) Based on the trajectory optimization results of step (2), a model for continuous adjustment of the posture of the end effector of the robotic arm is designed based on cell division and splicing. Through topology planning, the effective contact between the end effector of the robotic arm and the curved surface is ensured.

[0013] (4) Based on the path segment height state in steps (2) and (3) B A multimodal force control method is adopted in the State B When = 0, adaptive impedance control is enabled, in State B When the force is equal to 1, the constant force control mode is activated, and the movement speed of the robotic arm end effector and the grinding motor speed are coordinated to achieve wall grinding under different force control modes;

[0014] (5) Based on step (4), a multi-node fusion attitude control method with force feedback mechanism is introduced. The fusion attitude control is constructed through three equally spaced multi-node sensing systems to further optimize the contact force distribution on the working plane of the robotic arm end effector.

[0015] As a further improvement of the present invention, in step (1), the curve is discretized to generate grid points, which are divided into N+1 grid points at equal intervals along the curve S(x,z), in the form: S(x,z)={s0,s1,...,s N};

[0016] Feature points {Δ,□,○,◇} are extracted from the grid points and classified. Here, Δ represents the upper tangent point between the surface and the reference plane; □ represents the lower tangent point between the surface and the reference plane; and ○ and ◇ represent the upper and lower intersection points, respectively.

[0017] As a further improvement of the present invention, the convex path segments in the path classification include Δ-◇, Δ-Δ, ○-◇, ○-Δ, and the concave path segments include ◇-□, ◇-○, □-○, □-□;

[0018] The state flag is determined based on the radial height difference Δz of the path segments. B When State B =0, radial height difference greater than 1.5mm; when State B =1, radial height difference less than or equal to 1.5mm.

[0019] As a further improvement of the present invention, in step (2), a secondary interpolation planning is introduced for the raised path segment during the interpolation point optimization process. The optimization objective is to minimize the area removed by the grinding path, and the optimization objective function is:

[0020]

[0021] Where Q(x,z) represents the path function generated by interpolation, and s* represents the interpolation point. The optimal interpolation point s* is solved using the simulated annealing algorithm.

[0022] As a further improvement of the present invention, in step (3), the attitude adjustment of the robotic arm end effector is planned based on the topological configuration of cell division, including the following steps:

[0023] The target surface is decomposed into several cell c. i The attitude transition between cells is planned through the attitude configuration set ζ. Based on the continuity between cells, if the cells are continuous configurations *,*,*,…∈ζ, the current attitude is maintained; if the cells are not continuous configurations *,*,*,…∈ζ, the attitude rotation matrix R(α,β,γ) is calculated for dynamic adjustment.

[0024] As a further improvement of the present invention, in step (4), the hybrid force control strategy includes the following steps:

[0025] When State B When = 0: An adaptive impedance control strategy is adopted to establish the impedance model:

[0026]

[0027] Where: A(q) is the robot Jacobian matrix, C is the Coriolis / centrifugal torque vector, x∈R 3 It is the Cartesian position vector of the robotic arm's end effector, Λ=(AM -1 A T ) -1 It is a (3*3) inertial transformation matrix for the end effector of a robotic arm. F g=A |T τ g ,F f =A ||T τ f ,F a =A ||T τ a It is the force mapping matrix at the end, A || Let A be the dynamic uniform generalized inverse matrix of matrix A;

[0028] When State B When = 1: Force-position hybrid control is enabled, and the control model is as follows:

[0029]

[0030] Where: q∈R n n = 7 is the vector of joint variables, and M(q) is the inertia matrix. is the vector of Coriolis / centrifugal torque, g(q) is the vector of gravitational torque, and τ is the control torque.

[0031] As a further improvement of the present invention, in step (5), the multi-node sensing system includes three equally spaced force sensors arranged to record the contact force respectively. Each sensor node adopts an impedance control model:

[0032]

[0033] in: Let e ​​be the contact force at node i. i =x i -x i 0 Represents the generalized position of node i. and These are the velocity and acceleration of node i, respectively. and These are the inertia, damping, and stiffness parameters of node i, respectively.

[0034] As a further improvement of the present invention, the multi-node sensing system introduces a virtual node c, and establishes a force coupling relationship through the following generalized variables:

[0035]

[0036] in: It is a generalized nonlinear monotonically decreasing function, w1 and w2 are weighting coefficients. Due to the relationship between the node positions, nodes 1 and 2 can share the attitude offset α, and node 3 shares the offset β with the virtual node.

[0037] The technical solution of the present invention has the following beneficial effects:

[0038] (1) The inventors achieved secondary optimization of path interpolation by discretizing the wall curve and the reference plane and extracting feature points, thereby obtaining the optimal grinding trajectory, which significantly reduced path redundancy and improved grinding efficiency.

[0039] (2) The present invention uses a cell topology model to make fine and continuous adjustments to the posture of the end effector of the robotic arm, which effectively avoids overcutting or undercutting caused by posture mismatch during the grinding process.

[0040] (3) The present invention is based on a hybrid force control mode that dynamically switches force control states. It can select a suitable control mode according to the real-time height difference, and coordinate the motor speed and movement speed at the same time. It can maintain the consistency of the polished surface while ensuring rapid material removal.

[0041] (4) The present invention establishes multi-node fusion attitude control, which can adjust the attitude of the end effector of the robotic arm and control the force feedback in real time in complex curved surface environment, reducing the possibility of over-cutting or under-cutting that may occur during position control grinding. Attached Figure Description

[0042] Figure 1 This is a schematic diagram of the path planning and mesh parameterization of the present invention;

[0043] Figure 2 This is a schematic diagram illustrating the positional relationship between the cells of the curved surface and the reference plane of this invention;

[0044] Figure 3 This is a schematic diagram of the cell splicing state of the robotic arm end effector posture according to the present invention;

[0045] Figure 4 This is a schematic diagram of the sensor arrangement of the multi-node sensing system of the present invention;

[0046] Figure 5 This is a flowchart of the algorithm implementation in an example of the present invention;

[0047] Figure 6 This is a schematic diagram showing the positional relationship of the three force sensors on the end effector of the robotic arm according to the present invention. Detailed Implementation

[0048] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0049] This invention provides a wall polishing method based on surface mapping trajectory optimization and multimodal force control, comprising the following steps:

[0050] Step 1: The grinding trajectory is determined by projection technology, which realizes the calculation of interpolation points, control of attitude change and matching with position planning, and effectively generates a robot grinding path that includes grinding position, attitude and force control strategy adjustment.

[0051] Step 2: A multimodal fusion response hybrid force control algorithm based on radial height state switching is introduced, which introduces multi-node fusion attitude control to accurately track the contact force on the sand table surface and dynamically adjust the robot end posture.

[0052] Step 1, robot polishing path planning, includes the following steps:

[0053] Step 1-1: Perform secondary optimal path planning based on the curved surface modeling trajectory, search for and obtain the position of trajectory interpolation point, determine the concave and convex positions of the wall, and obtain the optimal grinding trajectory.

[0054] Step 1-2: Based on Step 1-1, further model the robot end effector posture, optimize the contact problem between the plane and the curved surface of the robotic arm end effector, and dynamically adjust the robot end effector posture.

[0055] Step 1-1 includes:

[0056] (1): Simplify the target surface: decompose the surface to be polished into several cells, and project the surface to be polished S(x,y,z) and the reference plane P(x,y,z) to a two-dimensional plane of O-XY plane to obtain two-dimensional curves S(x,z) and P(x,z).

[0057] (2): As attached Figure 1 As shown, the path along curve S(x,z) is discretized into N+1 grid points as follows:

[0058] 0 = s0, s1, s2...s N-1 ,s N :=s end

[0059] (3): Based on the path turning points of the protrusions and depressions of the wall surface through the mesh search, the path points need to satisfy the following constraints:

[0060] Subject to:

[0061]

[0062] Where: e si The normal to curve S is represented by Δ, which indicates the upper tangent point between the surface and the plane; □ indicates the lower tangent point; ◇ indicates the descending intersection point; and ○ indicates the ascending intersection point.

[0063] (4): Define path segment types based on feature point combinations: convex segments are Δ-◇, Δ-Δ, ○-◇, ○-Δ, and concave segments are ◇-□, ◇-○, □-○, □-□.

[0064] (5): Further, a grid search feature point {□○Δ◇} is used for path planning. The positional information between path points is analyzed to distinguish between raised and recessed areas on the workpiece surface. The height difference of the raised path segment in the z direction is analyzed. Based on the change in height difference, the state of the grid points is further classified.

[0065]

[0066] Where: State B When = 0, the radial distance between the surface to be polished and the reference plane is greater than 1.5mm. B When = 1, the radial distance is less than 1.5 mm, State B Different force control modes are adopted under different conditions to provide grinding signals for subsequent force control;

[0067] (6): It is necessary to perform secondary planning on the original modeling mapping trajectory S(x,z) to reduce the number of path segments and improve the efficiency of the polishing operation. Between the set of turning path points {□○Δ◇}, construct interpolation points s*a (s*a∈s). a (a∈i) to segment the existing path segment, at interpolation points s*a, s i s j Establish a continuously differentiable function Q(x,z) between them. a .

[0068] (7): In step (6), the function Q(x,z) a Divide the domain into m finite intervals, and construct piecewise functions in each interval using different linear expressions. Topologically apply this function to the entire polishing path to obtain the generalized function Q(x,z).

[0069] (8): Based on step (7), the problem of selecting interpolation points is transformed into the problem of minimizing the area removed by grinding. The optimization model is as follows:

[0070]

[0071] (9) The simulated annealing algorithm is used to solve for the optimal interpolation point s between each convex path segment. * For concave path segments, an interpolation point s is added in conjunction with the traditional linear interpolation method. # This ensures that the robotic arm's end effector remains stable during contact with the surface to be ground. Furthermore, the interpolation points s in the convex and concave path segments are integrated. * s # and path turning point s i The information is used to solve for the optimal trajectory planning.

[0072] Step 1-2: Based on the optimal grinding trajectory in Step 1-1, optimize the posture of the robotic arm's end effector. The specific steps are as follows:

[0073] (1): Based on the analysis at the microscopic level of contact, the positional relationship between the curved surface and the reference plane can be divided into eight categories, as shown in the appendix. Figure 2 As shown, based on the meshing approach in step 1-1, the surface is divided into several cells c. i And splicing is performed to meet the grinding requirements, ensuring that the plane of the end effector of the robotic arm makes full contact with the curved surface to be ground or removes material;

[0074] (2): Define ζ as the set of all valid configurations, and M as the set of all reachable elements c on the surface. i The set of elements, where the end-effector pose is also represented by M, has a one-to-one correspondence with surface elements. Configuration D∈ζ can continuously execute tasks without changing the end-effector pose, and there are infinitely many non-overlapping configurations D in the set ζ.

[0075] (3): Further define {*,*,*,…} as an n-unit fusion unit, if If {, *, *, *, ...}∈D, then the units are discontinuous; conversely, if {, *, *, ...}∈D, then the units are continuous.

[0076] (4) Based on step (3), taking a rectangular curved surface as an example, as shown in the attached figure. Figure 3 As shown, considering only the concatenation between two points m1, m2 ∈ M, for example, with c1 as the initial cell, there are 4 cases where the cells are consecutive and 4 cases where the cells are consecutive, as follows:

[0077]

[0078] (5) Further construct the topology configuration, whose elements are cells and configurations are D. Each cell is assigned an index to record possible configuration combinations. Since the number of cell types is finite, the configuration set {*,*,*,…} should also be finite. Therefore, a surface has only two possibilities in a general sense: first, continuous task execution does not require changing the end effector pose; second, during task execution, the end effector pose is continuously adjusted according to the state of configuration D to adapt to surface changes, i.e. The set is ordered as follows:

[0079]

[0080] (6) After establishing the topological configuration, search according to the sorting of the set. The node cell c between {*,*,*,…}∈D iAn index is used for corresponding attitude planning. Furthermore, the attitude index is matched with the position index to map the end-effector position and attitude to the robot arm joint space, thereby achieving trajectory planning and solving.

[0081] Step 2: A multimodal force control method based on radial height state switching includes the following steps:

[0082] Step 2-1: Based on the radial height State from Step 1-1 B Depending on the condition, adjust the corresponding force control strategy and integrate the grinding motor speed to quickly remove surface material;

[0083] Since the attitude adjustment of the robotic arm end effector in steps 1-2 is based solely on position control and lacks flexibility, and due to the complexity of the wall surface, the local plane B fitted by the robotic arm end effector may spatially intersect with the reference plane A. (When in the grinding state) B When = 1, if the working plane of the robotic arm end effector is parallel to the reference plane, it may cause problems such as jitter and undercut.

[0084] Step 2-2: Based on the attitude control of the robotic arm end effector in Step 1-2, a force feedback mechanism is introduced to construct a multi-node sensing system with three equally spaced nodes for multi-node force fusion attitude control.

[0085] The specific steps of step 2-1 are as follows:

[0086] (1): Force control strategy deployment mechanism:

[0087] According to the state flag in step 1-1 B Switch control strategy. When z > 1.5mm, i.e., State B =0. The system switches to adaptive impedance control to optimize grinding by adjusting contact impedance. Simultaneously, a coupling model between motor speed and radial height difference is established to achieve rapid removal of surface material. When z < 1.5 mm, i.e., State B =0. The system switches to adaptive constant force control to maintain the force applied to the wall within the desired range, ensuring stable and efficient sanding results;

[0088] (2): When State B =0, adaptive impedance control is performed. The translational motion model of the robot radial grinding is as follows:

[0089]

[0090] Where: A(q) is the robot Jacobian matrix, C is the Coriolis / centrifugal torque vector, x∈R 3 It is the Cartesian position vector of the robotic arm's end effector, Λ=(AM-1 A T ) -1 It is a (3*3) inertial transformation matrix for the end effector of a robotic arm. F g =A |T τ g ,F f =A ||T τ f ,F a =A ||T τ a It is the force mapping matrix at the end, A || Let A be the dynamic uniform generalized inverse matrix of matrix A;

[0091] (3): Based on the dynamic response between the force applied to the end effector of the robotic arm and the position of the robotic arm, the robotic arm smoothly grinds to the desired position. The real-time displacement of the end effector of the robotic arm is as follows:

[0092]

[0093] in: K d and B d These are appropriate inertia and damping matrices, which are positive definite and are usually set as constant diagonal matrices, x d ,x e These represent the desired position and the current position, respectively. The desired speed and the current speed.

[0094] (4): When State B When the value is 1, during the contact process between the end effector of the robotic arm and the wall, horizontal position control and radial translation force control are required. The force-position hybrid control model is established as follows:

[0095]

[0096] Where: q∈R n n = 7 is the vector of joint variables, and M(q) is the inertia matrix. is the vector of Coriolis / centrifugal torque, g(q) is the vector of gravitational torque, and τ is the control torque.

[0097] (5) Based on the radial height state switching of the trajectory planning in step 1-1, switch the force control strategy in real time, integrate the grinding motor speed, quickly remove surface materials, and meet the requirements of wall flatness.

[0098] Step 2-1 can maintain the stability of the radial grinding force of the robotic arm end effector. However, since the working surface of the robotic arm end effector is large, the grinding quality cannot be met by only considering the control of the radial force. Therefore, based on step 2-1, the multi-node force fusion attitude control in step 2-2 is carried out to optimize the contact force of the working plane of the robotic arm end effector.

[0099] The specific steps of step 2-1 are as follows:

[0100] (1): As attached Figure 4 As shown, a multi-node sensing system is constructed using three equally spaced force sensors. Each sensor node employs an impedance-controlled model and is described as a second-order linear system. The force-position relationship at node i can be expressed as:

[0101]

[0102] in: Let e ​​be the contact force at node i. i =x i -x i 0 Represents the generalized position of node i. and These are the velocity and acceleration of node i, respectively. and These are the inertia, damping, and stiffness parameters of node i, respectively.

[0103] (2) Treating the end effector of the robotic arm as a single-degree-of-freedom system interacting with the environment, establish the mass motion equation:

[0104] (3) As attached Figure 4 As shown, a virtual node c is introduced on the x-axis. To achieve multi-node fusion control, force coupling relationships are established between the nodes:

[0105]

[0106] in: It is a generalized nonlinear monotonically decreasing function, w1 and w2 are weighting coefficients. Due to the relationship between the node positions, nodes 1 and 2 can share the attitude offset α, and node 3 shares the offset β with the virtual node.

[0107] (4) Replacing the generalized variable x with attitude offsets α and β, and combining the impedance models of each node, the entire surface contact dynamics system is:

[0108]

[0109] in: These represent the required inertia, damping, and stiffness matrices for each node, respectively. Internal matrix relationships and Similarly, and There is a certain coupling relationship, and the initial expected rotation angle deviation around x and y is e. α ,e β , represented as e α =α d -α0、e β =β d -β0.

[0110] Step 2-2's control strategy enables multi-node force fusion control of the robotic arm's end effector, ensuring precise attitude adjustment and force feedback control in complex environments. When each node in the dynamic system is subjected to an external force F... ext When they are equal, the attitude adjustment process of the robotic arm's end effector ends.

[0111] The following is in conjunction with the appendix Figure 5 The present invention will be further described in detail with reference to specific embodiments:

[0112] This example provides a wall sanding method based on surface mapping trajectory optimization and multimodal force control. First, a smooth and efficient trajectory is obtained through interpolation point optimization. Then, the end effector attitude is dynamically adjusted based on this trajectory, and the force control mode is dynamically switched. The specific implementation steps are as follows:

[0113] Phase 1: Constructing the optimal polishing path, including searching for trajectory interpolation points and marking concave and convex features, etc.

[0114] Step 1: Surface Data Acquisition and Preprocessing

[0115] (1): A 3D laser scanner of model LMS-Z420i was used to scan the wall surface at a spatial resolution of 1mm to obtain a point cloud dataset P={p1(x1,y1,z1),p2(x2,y2,z2),...,p n (x n ,y n ,z n )};

[0116] (2): The ORPEF algorithm was used to perform plane search on the point cloud data to identify the reference plane P(x,y,z), and the fitting accuracy was better than 0.1mm;

[0117] (3): Project the three-dimensional point cloud data onto the O-XZ plane to generate a two-dimensional curve S(x,z) and a reference curve P(x,z);

[0118] Step 2: Marking and classifying curve feature points

[0119] (1): Discretize the curve S(x,z) into 100 segments at equal intervals, forming 101 sampling points {s0,s1,...,s...} 100};

[0120] (2): Calculate the tangent vector V at each point si and normal vector e si And mark four types of feature points: upper tangent point (Δ), lower tangent point (□), upper intersection point (○), and lower intersection point (◇);

[0121] (3): Based on the combination of feature points, divide the segment into convex segments (Δ-◇,Δ-Δ,○-◇,○-Δ) and concave segments (◇-□,◇-○,□-○,□-□);

[0122] (4): Calculate the maximum radial height difference Δz_max for each path segment and set the status flags:

[0123] State B =1, when Δz_max≤1.5mm;

[0124] State B =0, when Δz_max>1.5mm;

[0125] Step 3: Optimization of trajectory interpolation points

[0126] (1) Optimization of protruding segments (State) B =0): Use the simulated annealing algorithm (initial temperature 1000, cooling coefficient 0.95) to search for the optimal interpolation point s*;

[0127] (2) The optimization objective is to minimize the function f(s*)=∫|Q(x,z;s*)-P(x,z)| 2 dx;

[0128] (3) Record the optimal interpolation point s*_opt and generate the cubic spline interpolation curve;

[0129] (4): Depression segment optimization - set 3 equidistant interpolation points s # _j=s_p+j·(s_q-s_p) / 4,j=1,2,3, construct a smooth path using Catmull-Rom spline interpolation;

[0130] Step 4: Trajectory Integrity – Verify the curvature continuity analysis to ensure that the rate of curvature change at feature points is <0.02m. -1 ·s -1 ;

[0131] Phase Two: End-effector posture adjustment based on the optimal polishing trajectory

[0132] Step 5: Divide the curved area traversed by the optimal grinding trajectory into 10×10 regular grid cells, calculate the normal vector n_ij at the center point of each cell c_ij; assign a unique index D_ij=10·(i-1)+j to each cell to facilitate subsequent posture matching;

[0133] Step 6: Define the attitude configuration set ζ = {D1, D2, ..., D_m}, where each configuration represents a set of allowed attitudes, and calculate the angle θ_ij = arccos(n_i·n_j) between the normal vectors of adjacent cells;

[0134] Step 7: Based on steps 1-2, establish continuity judgment criteria, and further construct the topological relationship matrix T and the set of continuous regions {C1,C2,...,C_p};

[0135] Step 8: Trajectory-Attitude Mapping and Switching Point Determination. Map the optimized trajectory Q(x,z) onto the surface cells to obtain the cell sequence {c_1,c_2,...,c_k}. Analyze the continuous region sequence S=[C_s1,C_s2,...,C_s_q] that the trajectory passes through.

[0136] Step 9: For each pair of adjacent regions (C_s_j, C_s_{j+1}), determine the optimal attitude transition point, and further use the quaternion spherical linear interpolation (SLERP) algorithm to perform continuous attitude adjustment;

[0137] Phase 3: Multimodal hybrid force control and multi-node fusion attitude adjustment

[0138] Step 10: Based on the first and second stages, and according to the State obtained in trajectory planning... B Logo, design of the force control mode of the robotic arm end effector: when State B When the state changes from 0 to 1: impedance control → force-position hybrid control, when State B When the value changes from 1 to 0: force-position hybrid control → impedance control;

[0139] Step 11: Based on State B State switching enables real-time adjustment of the radial force control of the robotic arm's end effector, while simultaneously constructing a multi-node force sensing system for the end effector.

[0140] (1): Three force sensors (nodes 1-b, 2-r, and 3-l) are evenly installed at the bottom of the end effector of the robotic arm, as shown in the attached figure. Figure 6 The three nodes form an equilateral triangle, and each sensor node is responsible for collecting the contact force signal at that point.

[0141] (2): A control strategy based on multi-node force fusion is established. A virtual node c is introduced on the x-axis of the grinding direction to coordinate the force coupling control of the three actual sensor nodes. The force coupling relationship between the nodes is modeled as follows:

[0142]

[0143] (3): During the control process, the generalized variable x is replaced with attitude angle offsets α and β, and the overall surface contact dynamics model is established by combining the impedance model of each sensor node:

[0144]

[0145] This invention proposes a multi-sensor fusion hybrid force control algorithm that optimizes the grinding trajectory through surface modeling and geometric mapping, combined with radial height state switching. On the one hand, it realizes the calculation of interpolation points, control of posture changes and matching with position planning, effectively generating a robot grinding path that includes grinding position, posture and force control strategy adjustment. On the other hand, it accurately tracks the contact force on the sanding disc surface, dynamically adjusts the robot end posture, and adjusts the grinding force in real time to ensure uniform effect, thereby significantly improving grinding efficiency and quality.

[0146] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any other way. Any modifications or equivalent changes made based on the technical essence of the present invention shall still fall within the scope of protection claimed by the present invention.

Claims

1. A method for optimizing curved surface mapping trajectories and multimodal force control for grinding complex wall surfaces, characterized in that, Includes the following steps: (1) Utilizing the relative positional relationship between the target surface S(x,y,z) and the reference plane P(x,y,z), the surface area is simplified to a two-dimensional plane S(x,z) through projection technology. The curve is discretized to generate grid points, and the set of feature points {∆,□,○,◊} is identified. (2) Define path segment types based on feature point combinations, classifying path segments into convex segments and concave segments; use simulated annealing algorithm to optimize interpolation points s for convex path segments. ∗ For concave path segments, linear interpolation is used to add interpolation points s. # Generate the globally optimal grinding trajectory Q(x,z) that includes information on interpolation points and path turning points; (3) Based on the trajectory optimization results of step (2), a model for continuous adjustment of the posture of the end effector of the robotic arm is designed based on cell division and splicing. Through topology planning, the effective contact between the end effector of the robotic arm and the curved surface is ensured. (4) Based on the path segment height status in steps (2) and (3) Using a multimodal force control method, in When adaptive impedance control is enabled, The constant force control mode is activated, and the movement speed of the robotic arm end effector and the speed of the grinding motor are coordinated to achieve wall grinding under different force control modes. (5) Based on step (4), a multi-node fusion attitude control method with force feedback mechanism is introduced. The fusion attitude control is constructed through three equally spaced multi-node sensing systems to further optimize the contact force distribution on the working plane of the robotic arm end effector. In step (5), the multi-node sensing system includes three equally spaced force sensors that record the contact force respectively. , , Each sensor node adopts an impedance control model: ; in: The contact force at node i, Represents the generalized position of node i. and These are the velocity and acceleration of node i, respectively. , and These are the inertia, damping, and stiffness parameters of node i, respectively. The multi-node sensing system introduces a virtual node c and establishes a force coupling relationship through the following generalized variables: ; Where: φ is a generalized nonlinear monotonically decreasing function, w1 and w2 are weighting coefficients. Due to the relationship between the node positions, nodes 1 and 2 can share the attitude offset α, and node 3 shares the offset β with the virtual node.

2. The method for optimizing curved surface mapping trajectory and multimodal force control for grinding complex wall surfaces according to claim 1, characterized in that, In step (1), the curve is discretized to generate grid points, which are then divided into N+1 grid points at equal intervals along the curve S(x,z), in the form: S(x,z)={s0,s1,...,s...} N }; Feature points {∆,□,○,◊} are extracted from the grid points and classified. Here, ∆ represents the upper tangent point between the surface and the reference plane; □ represents the lower tangent point between the surface and the reference plane; and ○ and ◊ represent the upper and lower intersection points, respectively.

3. The method for optimizing curved surface mapping trajectory and multimodal force control for grinding complex wall surfaces according to claim 2, characterized in that, In the path classification, convex path segments include ∆−◊,∆−∆,○−◊,○−∆, and concave path segments include ◊−□,◊−○,□−○,□−□; Determine the status flag based on the radial height difference ∆z of the path segment. :when The radial height difference is greater than 1.5mm; when The radial height difference is less than or equal to 1.5 mm.

4. The method for optimizing curved surface mapping trajectory and multimodal force control for grinding complex wall surfaces according to claim 1, characterized in that, In step (2), a secondary interpolation plan is introduced for the raised path segment during the interpolation point optimization process. The optimization objective is to minimize the area removed by the grinding path, and the objective function is: ; Where Q(x,z) represents the path function generated by interpolation. The interpolation point is represented by a simulated annealing algorithm to find the optimal interpolation point. .

5. The method for optimizing curved surface mapping trajectory and multimodal force control for grinding complex wall surfaces according to claim 1, characterized in that, In step (3), the attitude adjustment of the robotic arm end effector is planned based on the topological configuration of cell division, including the following steps: The target surface is decomposed into several cell c. i The attitude transition between cells is planned through the attitude configuration set ζ. Based on the continuity between cells, if the cells are continuous configurations ∗,∗,∗,…∈ζ, the current attitude is maintained; if the cells are not continuous ∗,∗,∗,…∉ζ, the attitude rotation matrix R(α,β,γ) is calculated for dynamic adjustment.

6. The method for optimizing curved surface mapping trajectory and multimodal force control for grinding complex wall surfaces according to claim 1, characterized in that, In step (4), the hybrid force control strategy of the multimodal force control method includes the following steps: when At that time: An adaptive impedance control strategy was adopted to establish an impedance model. ; Where: A(q) is the robot Jacobian matrix, C is the Coriolis / centrifugal torque vector, x∈R 3 It is the Cartesian position vector of the robotic arm end effector. It is a (3*3) inertial transformation matrix for the end effector of a robotic arm. It is the force mapping matrix at the end, A || Let A be the dynamic uniform generalized inverse matrix of matrix A; when At this time: Force-position hybrid control is enabled, and the control model is as follows: ; Where: q∊R n n = 7 is the vector of joint variables, and M(q) is the inertia matrix. is the vector of Coriolis / centrifugal torque, g(q) is the vector of gravitational torque, and τ is the control torque.

Citation Information

Patent Citations

  • Real-time force control-based robot polishing system and method for complex curved surface constant resection rate

    CN110524371A

  • Industrial robot high-precision constant-force grinding method based on curved surface self-adaption

    CN111055293A