四足机器人变步态运动控制研究
发布时间：2018-09-14 13:27
【摘要】：相较于轮式机器人,足式机器人具有更好的非连续地形适应性,但在运动速度与能耗效率上,足式机器人尚低于轮式机器人。因而如何在利用足式机器人的优点的情况下,提高足式机器人运动速度与能耗效率成为足式机器人领域研究重点。为了解决机器人的大速度范围、高效运动的问题,研究四足机器人变步态运动控制有着重要的现实意义。本课题以四足机器人平面模型作为研究对象,在建立四足机器人平面动力学模型和能耗模型的基础上,以机器人的稳定性和能耗最优作为优化目标,基于粒子群算法和Matlab数值仿真平台,对机器人多种步态下的运动以及变步态运动控制方法进行了分析。首先,分析了四足机器人变步态的运动特点,建立了四足机器人变步态运动平面动力学模型;结合对平面动力学模型进行稳定性分析和能耗分析的结果,得到了四足机器人变步态平面模型。其次,为了对机器人的变步态运动进行控制,以机器人平面模型的稳定性和能耗经济性作为优化目标,得到了平面模型运动的适应性函数;基于适应性函数和粒子群智能算法,给出了一种机器人平面模型控制参数搜索方法;利用该方法对机器人的变步态运动控制参数进行规划,得到了机器人在多种步态、大速度范围下做匀速及加减速运动的稳定控制解。通过对得到的控制解组成的解集进行分析,根据不同步态的稳定性、能耗等参数随前进速度变化的关系给出了机器人的运动速度区间、各个步态的最适运动速度区间,得到了机器人变步态运动的控制参数规划结果。最后,本课题采用SimMechanics-Simulink仿真建立了四足机器人虚拟样机,设计了基于离线控制解集的虚拟样机控制策略,进行了起步、加速、、减速、渐停等一系列仿真实验。实验实现了机器人虚拟样机大速度范围、高能耗效率的多步态、变步态稳定运动,验证了基于四足机器人平面模型的变步态运动规划结果及机器人变步态控制方法的有效性。
[Abstract]:Compared with wheeled robots, foot robots have better adaptability to discontinuous terrain, but they are still lower than wheeled robots in terms of motion speed and energy efficiency. Therefore, how to improve the speed and energy efficiency of foot robot becomes the focus of research in the field of foot robot under the condition of making use of the advantages of foot robot. In order to solve the problem of high speed and high efficiency of robot, it is very important to study variable gait motion control of quadruped robot. Based on the plane dynamics model and energy consumption model of quadruped robot, the stability and energy consumption optimization of quadruped robot is taken as the optimization object. Based on particle swarm optimization (PSO) and Matlab numerical simulation platform, the motion and variable gait motion control methods of robot in multiple gaits are analyzed. Firstly, the characteristics of variable gait motion of quadruped robot are analyzed, and the plane dynamics model of variable gait motion of quadruped robot is established, and the results of stability analysis and energy consumption analysis of the model are combined. The variable gait plane model of quadruped robot is obtained. Secondly, in order to control the variable gait motion of the robot, the stability and energy efficiency of the robot plane model are taken as the optimization objectives, and the adaptive function of the plane model motion is obtained, which is based on the adaptive function and the particle swarm intelligence algorithm. In this paper, a parameter search method for planar model control of a robot is presented, by which the variable gait motion control parameters of the robot are planned, and the robot in various gait states is obtained. The stable control solution of uniform velocity and acceleration and deceleration motion is given in large range of velocity. By analyzing the solution set composed of the control solution, according to the relation between the stability of different gait and the energy consumption and so on, the moving speed interval of robot and the optimum moving velocity interval of each gait are given. The control parameter planning results of the variable gait motion of the robot are obtained. Finally, the virtual prototype of quadruped robot is established by SimMechanics-Simulink simulation, and the control strategy of virtual prototype based on off-line control solution is designed. A series of simulation experiments, such as starting, accelerating, decelerating and stopping, are carried out. The multigait and variable gait steady motion of the virtual prototype with large speed range and high energy consumption efficiency is realized experimentally. The results of variable gait motion planning based on the planar model of quadruped robot and the effectiveness of the variable gait control method are verified.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242
本文编号：2242839
[Abstract]:Compared with wheeled robots, foot robots have better adaptability to discontinuous terrain, but they are still lower than wheeled robots in terms of motion speed and energy efficiency. Therefore, how to improve the speed and energy efficiency of foot robot becomes the focus of research in the field of foot robot under the condition of making use of the advantages of foot robot. In order to solve the problem of high speed and high efficiency of robot, it is very important to study variable gait motion control of quadruped robot. Based on the plane dynamics model and energy consumption model of quadruped robot, the stability and energy consumption optimization of quadruped robot is taken as the optimization object. Based on particle swarm optimization (PSO) and Matlab numerical simulation platform, the motion and variable gait motion control methods of robot in multiple gaits are analyzed. Firstly, the characteristics of variable gait motion of quadruped robot are analyzed, and the plane dynamics model of variable gait motion of quadruped robot is established, and the results of stability analysis and energy consumption analysis of the model are combined. The variable gait plane model of quadruped robot is obtained. Secondly, in order to control the variable gait motion of the robot, the stability and energy efficiency of the robot plane model are taken as the optimization objectives, and the adaptive function of the plane model motion is obtained, which is based on the adaptive function and the particle swarm intelligence algorithm. In this paper, a parameter search method for planar model control of a robot is presented, by which the variable gait motion control parameters of the robot are planned, and the robot in various gait states is obtained. The stable control solution of uniform velocity and acceleration and deceleration motion is given in large range of velocity. By analyzing the solution set composed of the control solution, according to the relation between the stability of different gait and the energy consumption and so on, the moving speed interval of robot and the optimum moving velocity interval of each gait are given. The control parameter planning results of the variable gait motion of the robot are obtained. Finally, the virtual prototype of quadruped robot is established by SimMechanics-Simulink simulation, and the control strategy of virtual prototype based on off-line control solution is designed. A series of simulation experiments, such as starting, accelerating, decelerating and stopping, are carried out. The multigait and variable gait steady motion of the virtual prototype with large speed range and high energy consumption efficiency is realized experimentally. The results of variable gait motion planning based on the planar model of quadruped robot and the effectiveness of the variable gait control method are verified.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242
【参考文献】
相关期刊论文 前6条
1 张秀丽;梁艳;;一种仿婴儿欠自由度四足爬行机器人[J];机器人;2016年04期
2 胡文皓;陈曙东;辛欣;;求解物流配送问题的混合粒子群算法[J];华侨大学学报(自然科学版);2016年03期
3 王东风;孟丽;赵文杰;;基于自适应搜索中心的骨干粒子群算法[J];计算机学报;2016年12期
4 姜建国;叶华;刘慧敏;张丽媛;孟宏伟;;融合快速信息交流和局部搜索的粒子群算法[J];哈尔滨工程大学学报;2015年05期
5 丁良宏;;BigDog四足机器人关键技术分析[J];机械工程学报;2015年07期
6 王绍治;郭伟;于海涛;李满天;;基于CPG的六足步行机器人运动控制系统研究[J];机械与电子;2010年08期
相关博士学位论文 前1条
1 王鑫;类豹型四足机器人高速运动及其控制方法研究[D];哈尔滨工业大学;2013年
相关硕士学位论文 前1条
1 王浩威;四足机器人跳跃步态参数分析及优化研究[D];哈尔滨工业大学;2011年
，本文编号：2242839
本文链接：https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2242839.html
