CN-120039424-B - Space non-cooperative target optimal envelope track tracking control method under multi-target constraint

CN120039424BCN 120039424 BCN120039424 BCN 120039424BCN-120039424-B

Abstract

The invention discloses a space non-cooperative target optimal envelope track tracking control method under multi-target constraint, which mainly comprises the steps of designing a full-drive multi-finger space capturing configuration with self-adaptive configuration adjustment and escape prevention capability, generating an effective multi-finger envelope configuration by adopting a bidirectional Hausdorff distance based on finger characteristics and target edge point characteristics, constructing a multi-finger dynamic capturing domain according to target motion characteristics and joint characteristic length, carrying out space expected track parameterization by adopting a continuous high-order Bezier curve, planning a smooth envelope track under the minimum base disturbance based on multi-target constraint and multi-mode particle collaborative learning mechanism, and designing a multi-finger collaborative hierarchical tubular model prediction control strategy based on a disturbance observer through the multi-finger nonlinear coupling characteristic of a reverse dynamic compensation system to realize envelope track tracking control. The method can effectively overcome the interference caused by the complexity of the target shape, the uncertainty of the motion and the nonlinear dynamic coupling, and realize the accurate tracking control of the space non-cooperative target track.

Inventors

CHEN BIN
XI HOUYIN
TIAN JIAN
CHEN TIANWEN
ZHANG YIZHUANG
ZHANG XIAODONG

Assignees

北京邮电大学
北京空间飞行器总体设计部
中国航天科技创新研究院

Dates

Publication Date: 20260508
Application Date: 20250403

Claims (4)

1. The method for tracking and controlling the optimal envelope track of the space non-cooperative targets under the multi-target constraint is characterized by comprising the following steps: S1, designing a full-drive multi-finger space capturing configuration with self-adaptive configuration adjustment and escape prevention capacity, and comprehensively utilizing a Lagrange method and an assumption modal method to establish a multi-finger enveloping capturing coupling dynamics model; S2, generating multi-finger candidate configuration according to the shape characteristics and the joint threshold value of the target, selecting an effective multi-finger enveloping configuration similar to the target contour by adopting a bidirectional Hausdorff distance, and constructing a multi-finger dynamic capturing domain according to the target motion characteristics and the joint characteristic length; S3, establishing a continuous high-order Bezier curve by using time interval mapping, designing an optimal track optimization cost function fused with joint motion restriction, palm posture disturbance and connecting rod self-collision avoidance constraint, and solving a multi-finger smooth envelope track minimizing base disturbance based on a particle collaborative learning mechanism with multi-modal characteristics; s4, separating a single-finger kinetic equation according to a multi-finger system distributed kinetic equation and a coupling relation between fingers, decoupling a nonlinear tracking error system into a linear subsystem based on inverse dynamics, designing a multi-finger collaborative hierarchical tubular model prediction control strategy based on disturbance observer compensation, and realizing multi-finger envelope track tracking control.
2. The method for controlling the tracking of the optimal envelope trajectories of spatially non-cooperative targets under a multi-target constraint according to claim 1, wherein the step S2 comprises: S201, determining a near joint angle of a multi-finger mechanism according to the shape characteristics of a rolling target, and generating multi-finger candidate configuration of a series of finger joints by utilizing joint threshold values and sampling step sizes; S202, adopting position coordinates of central points of finger joints, fingertips and connecting rods as characteristic point sequences, and calculating a bidirectional Hausdorff distance according to a finger feature set and a target edge feature point set: Wherein, the Representing the feature set of the kth finger, Representing a set of target edge feature points, a function Indicating that the maximum value is taken, And All represent unidirectional Hausdorff distance; s203, establishing constraint conditions of escape between the rolling target fingers according to the projection relation between the multi-finger end effector and the palm bottom plane: Wherein, the Represents the finger near joint angle satisfying the constraint condition, Representing an initial multi-assignment projection and perpendicular to The angle between the two is set to be equal, Representing joints And (3) with The distance between the two plates is set to be equal, Representing joints And (3) with The distance between the two plates is set to be equal, the configuration with the smallest sum of the finger joint movement ranges is selected as the optimal configuration of the multi-finger mechanism: Wherein, the Representing the optimal registration of the multiple fingers, Representing a multi-finger candidate configuration, The range of motion of the ith joint is represented, and N represents the number of finger joints.
3. The method for controlling the tracking of the optimal envelope trajectories of spatially non-cooperative targets under a multi-target constraint according to claim 1, wherein the step S3 comprises: S301, mapping a track planning time interval to a standard interval through normalization, and establishing a fifth-order Bezier curve of a joint space expected track according to a starting point, an end point and a shape control point; S302, representing the position relationship of connecting rods of any two fingers in the enveloping process as space geometric vectors, and constructing a repulsive force potential field function according to the minimum collision prevention distance between the connecting rods: Wherein, the Representing a virtual repulsive force, Representing the coefficient of repulsive force, Indicating the effective range of the repulsive force, Indicating the minimum distance between the links, Is a constant; Taking the motion limitation of the multi-finger joints, the disturbance of the palm postures and the constraint of the multi-finger connecting rod for avoiding the self-collision into consideration, constructing an optimal joint track multi-target optimization cost function: Wherein, the The joint angle is indicated by the angle of the joint, A weight matrix representing the end of the multi-finger, A penalty factor representing the palm-pose disturbance, The position and the posture of the tail end of the multi-finger are represented, Representing a drift disturbance of the palm posture, Representing euclidean norms; s303, designing a dynamic collaborative learning mechanism according to fitness, search progress and particle distance, fusing multi-mode characteristics to adjust speed updating strategies of different areas, and designing an updating mechanism of an enhanced particle swarm optimization algorithm: Wherein, the Representing the non-negative inertial weight of the vehicle, And The position and velocity of particle k at the s-th iteration are shown, Indicating the individual optimal positions of the particles k, Representing the co-learned optimal position of the particles k, Representing a velocity adjustment of the particles based on multi-modal identification, Representing dynamic collaborative weights.
4. The method for controlling the tracking of the optimal envelope trajectories of spatially non-cooperative targets under a multi-target constraint according to claim 1, wherein the step S4 comprises: s401, representing inertial coupling and speed coupling among fingers by adopting linear feedforward torque, and separating a single-finger kinetic equation according to a multi-finger system distributed kinetic equation and a coupling relation among fingers: Wherein, the Representing the inertial matrix of the finger, Representation and representation And The non-linear terms of the correlation, Representing an external disturbance such as a wind or a wind, Representing the control moment of the finger(s), Representing the coupling terms between the fingers of a person, Represents joint friction; S402, linearizing coupling moment according to finger cross symmetrical distribution and adjacent finger coupling relation, and reducing a multi-input multi-output nonlinear tracking error system into a single-input single-output linear system based on nonlinear and coupling effects of an inverse dynamics control compensation system; s403, designing an uncertainty in an disturbance observation controller estimation system, and actively compensating external disturbance according to disturbance observance: Wherein, the Represents the auxiliary variable(s), Representing the derivative of the auxiliary variable, The parameters of the observer are represented by a set of parameters, Representing the observed amount of disturbance of the system, The trajectory of the actual state is represented, Representing a matrix of states of the system, Representing an input matrix of the system; S404, indirectly measuring the synergy degree of the multi-finger system by adopting single finger inertia force change, establishing a multi-finger collaborative optimization cost function according to the energy, state tracking error and terminal state constraint of a nominal model in a prediction time domain, and designing a pipe model prediction controller to inhibit residual interference according to optimal control input tracks and deviation feedback of actual state tracks and nominal state tracks: Wherein, the Representing the control inputs of the tube model controller, Representing the trajectory of the optimal control input, Indicating that the feedback gain is to be given, Representing a nominal state trajectory.

Description

Space non-cooperative target optimal envelope track tracking control method under multi-target constraint Technical Field The invention relates to the field of space target capture control, in particular to a space non-cooperative target optimal envelope track tracking control method under multi-target constraint. Background With the ever-increasing demand for space exploration and activity, the number of satellites that are successfully launched worldwide has risen year by year. Due to the lack of effective recovery means for the failed satellites, there is a significant increase in non-cooperative targets in earth orbit. Meanwhile, the targets generally lack power control and communication capability and cannot effectively interact with the ground, so that the complexity of the space environment is further increased by on-orbit disassembly and collision, and the safety of on-orbit spacecrafts, space stations and astronauts is seriously threatened. Compared with non-cooperative target capturing modes such as butt-joint type recovery, flying net capturing, rope claw traction, laser pushing and the like, the rigid capturing mode based on the mechanical arm and the tail end mechanism has the advantages of high connection rigidity, reusability, accurate force control and the like, and is widely studied and applied in the industry. However, in the on-orbit capturing process of the actual non-cooperative targets, the targets are influenced by external disturbance forces such as earth attraction, air resistance, solar radiation pressure and the like, the targets usually have large size span, shape diversity and motion uncertainty and roll, spin or nutation motion, meanwhile, the multi-finger system can face constraint problems such as output limitation of an actuator, target motion boundary, joint speed limitation and the like in the track tracking process, and the complex space kinematic relationship and nonlinear dynamics result in dynamic coupling phenomenon of the system, so that the track tracking control effect of the existing method in the capturing of the space non-cooperative targets is poor. Aiming at the problems, the invention focuses on the design of a full-drive multi-finger capturing configuration, the positioning of a dynamic capturing domain under escape constraint, the planning of a minimum base disturbance envelope track and the design of a disturbance compensation multi-finger coordination control strategy, and discloses a space non-cooperative target optimal envelope track tracking control method under multi-target constraint. Disclosure of Invention Aiming at the problems and challenges of the space non-cooperative target track tracking control, the invention provides a space non-cooperative target optimal envelope track tracking control method under multi-target constraint, and the technical scheme and steps of the invention are as follows: step 1, comprehensively considering the overall dimension, mass inertia and motion characteristics of a target, constructing a full-drive multi-finger space capturing configuration with self-adaptive configuration adjustment and escape prevention capability, and constructing a multi-finger enveloping capturing coupling dynamics model based on Lagrange method and hypothesis mode method; step2, generating multi-finger candidate configuration according to the shape characteristic of the target and the joint threshold value, selecting an effective multi-finger envelope configuration similar to the target contour by adopting a bidirectional Hausdorff distance, and constructing a multi-finger dynamic capture domain according to the target motion characteristic and the joint characteristic length, wherein the method comprises the following specific steps: (1) Determining a near joint angle of a multi-finger mechanism according to the shape characteristics of the rolling target, and generating multi-finger candidate configuration of a series of finger joints by utilizing joint threshold values and sampling step sizes; (2) The position coordinates of the finger joints, fingertips and connecting rod center points are used as characteristic point sequences, and the bidirectional Hausdorff distance is calculated according to the finger feature set and the target edge feature point set; (3) And establishing constraint conditions of escape between the rolling target fingers according to the projection relation between the multi-finger end effector and the palm bottom plane, and selecting the configuration with the minimum sum of the finger joint movement ranges as the optimal configuration of the multi-finger mechanism. Step 3, determining a continuous high-order Bezier curve through time interval mapping, establishing an optimal track optimization cost function by fusing multi-objective constraint, and solving a multi-finger smooth envelope track for minimizing base disturbance by adopting a particle collaborative learning mechanism, wherein the method comprises the