| The tensegrity structure is a type of hybrid soft-rigid structure,characterized by its deployable,lightweight,and the ability to resist external disturbance.The structure is particularly suitable for building the body of robots,and in recent years,robots based on this type of structure have received widespread attention from scholars around the world,resulting in some of research achievements.Due to the flexibility of this robot,its body is prone to vibration or shaking during high-speed movement.Vibration suppression of tensegrity robots using traditional control methods,which leads to the slow motion of robots.For this reason,taking advantage of the characteristics that the tensegrity structure is prone to vibration,research on vibration-driven tensegrity robots has attracted the attention of scholars in this field.However,currently developed vibration-driven tensegrity robots are mostly tailored for specific environments and tasks,and are trained using machine learning algorithms to obtain optimal driving frequencies and gaits.This method has relatively weak anti-disturbance ability,making it difficult to lock specific frequencies in the structure and to fully utilize the passive dynamics of tensegrity robots for motion control.In view of this,this paper studies on the modal analysis and self-excitation of tensegrity robots,a self-excitation method can lock at specific natural frequencies of the robot is proposed,and then utilizing the passive dynamics of tensegrity structures for motion control.The paper will research on form-finding analysis,modal analysis and tests,analysis and establishment of the self-excitation controller,and self-excitation motion control of tensegrity structures,etc.To determine the equilibrium configuration of a tensegrity structure under different levels of prestress,this study analyzes the equilibrium conditions of the tensegrity structure firstly,further researches on the form-finding algorithm.This paper analyzes the existing problems of the node-space dynamic relaxation form-finding algorithm.In order to solve problems of node-space form-finding algorithm,this paper introduces the concept of "task space" in the field of robots into form-finding algorithms,and proposes a taskspace dynamic relaxation algorithm for tensegrity structures based on substructure.In this algorithm,the high stiffness part of the structure is treated as a whole,and virtual damping and mass are added in its six-degree-of-freedom space to construct the task-space pseudo-dynamic equation,then iterative form-finding for the tensegrity structure.This algorithm solves problems that the node-space form-finding algorithm cannot maintain the relative position constraints of the bar elements in each high stiffness substructure and hard to converge.In this paper,a number of different models of tensegrity structures are established,and multiple simulations are performed to verify the effectiveness of the task-space form-finding algorithm.To investigate the vibration characteristics of tensegrity structures,this paper derives the node-space stiffness and mass matrices of tensegrity structures.It is found that the high dimensionality of these matrices makes modal analysis relatively complex.On the other hand,the node-space stiffness and mass matrices are difficult to apply effectively to control of robotic motion.To address these issues,this paper proposes a substructure-based modal analysis method,derives the transformation matrix,and transforms stiffness and mass matrices from node space into task space.For the tensegrity structures with different configurations,the vibration equations are established,modes of tensegrity structures are analyzed in different spaces,and analysis results in different spaces are compared.In addition,this paper also performs modal tests on different tensegrity structures.To address the issue of modal self-locking excitation in tensegrity robots,this study establishes the linear dynamics of a tensegrity robot,derives the mapping relationship between cable-space force and task-space force,proposes a modal self-excitation method of tensegrity robots,it can self-excite the specific order mode of robots.By using the self-excitation method,the robot is controlled to realize the limit cycle oscillation motion at the natural frequency.In order to judge the stability of the limit cycle oscillation,the stability of the limit cycle is analyzed by using Nyquist diagram and numerical calculation.The self-excitation simulation of the tensegrity robot is performed,and the influence of prestress and mass on the robotic self-excitation motion is explored.The self-excitation experiment platform of the tensegrity robot is built,the modal self-excitation experiments are performed,and robotic modes switching and multi-modal self-excitation experiments are realized.The self-excitation motion control of the tensegrity robot is studied,and the nonlinear dynamics of the tensegrity robot is analyzed.The dynamics model of the tensegrity robotic arm is established,the self-excitation method is applied to the nonlinear dynamics,and the tensegrity robotic arm is controlled to form a rhythmic motion.A physical prototype of the tensegrity robotic arm is built,and the self-excitation method is used to control the tensegrity robotic arm to form a rhythmic motion.The effects of disturbance,prestress and mass on the self-excitation motion of the robot are explored by experiments.For the locomotion of the tensegrity robotic snake on the ground,the dynamics and friction models of the tensegrity robotic snake are established,and a cross-feedback self-excitation control method is proposed,the robotic snake is controlled to produce traveling wave propulsion motion on the ground. |
