Behavior Expression And Planning Of Six-Parallel-Legged Robots

The six-legged robot is a complicated legged system due to its multiple inputs,end-effectors and outputs.Considering the topological complexity of the robot,a great variety of motion patterns can be applied.It is an issue to describe and plan the motion of all end-effectors and outputs in practical use.Considering the numerous potential applications of a legged system,various task requirements need to be satisfied.It is another issue to establish the mapping between tasks and motion patterns,in order to realize the collaborative planning of all end-effectors and outputs.By studying the six-parallel-legged robot,this paper establishes the models of task decomposition and behavior expression,proposes three behavior characteristics as well as their expression and planning methods,and finally carries out the experimental validations.The models of task decomposition and behavior expression have instructional significance in practical use of six-legged robots.The detailed contents are as follows.(1)The kinematics and the influence coefficient models of the six-parallel-legged robot are established and experimentally validated.The kinematics of both the single leg and the whole robot in four standing patterns are derived,which are essential for the position planning in task execution.Furthermore,to shorten the task time,the influence coefficient models are established for velocity and acceleration planning.Both the first and second order influence coefficient models are derived to describe the velocity and acceleration relationships between the active joints and the end-effectors respectively.(2)The Topology-Motion-Trajectory(TMT)model is proposed to express the robot behavior.The TMT model defines three types of behavior characteristics to express the robot behavior quantitatively,including characteristics of topology,motion and trajectory.The topology characteristics describe which end-effectors move.The motion characteristics describe which outputs of each end-effector are necessary.And the trajectory characteristics describe how to generate the trajectory of each output.The three behavior characteristics describe all possible motions including the motion patterns and the exact trajectory equations,which provides a general way to express and plan the robot behavior.(3)The time-optimal trajectory planning method for six-legged robots is proposed considering constraints of input power.Due to the time-invariance property of the influence coefficient,the time invariant parameters are extracted from time-varying constraints of the actuators and the gait cycle.As a result,the time-optimal trajectory planning methods of both the single leg and the whole robot are proposed.The Switch Points Method(SPM)is presented to solve the time-optimal trajectory planning problem of the single leg.And based on SPM,the Two-Step Optimization Method(TSOM)is proposed to solve the time-optimal trajectory planning problem of the whole robot.The models and the methods are validated by simulations and experiments.By using TSOM,the velocity ability of the robot walking is analyzed,and the time-optimal trajectory library of the sixparallel-legged robot is established.(4)The task-subtask-behavior model and the collaborative planning method are proposed.According to the time sequence of the subtasks,a task is decomposed into serial or parallel subtasks.Each subtask is mapped to a serial of behaviors.In such way,the collaborative planning of multiple end-effectors and outputs of the robot can be realized.The model is applied to plan the door-opening task and the transportation task under dynamic disturbances.The door-opening task is decomposed into 6 serial subtasks and 26 behaviors to realize the collaborative planning.And the transportation task is decomposed into 2 parallel subtasks and 6 behaviors.Both tasks are experimentally validated.