摘要
从生存理论这种新视角出发能有效结合机器人动力学模型来分析其避障问题。针对轮式机器人的欠驱动特性,提出了一种生存区域构造新方法,并结合机器人模型将障碍约束就转化为一组生存不等式;为了保证机器人能够规避附近障碍物,给出了该区域全局生存的充分性证明;同时为满足基本导航要求,设计了该要求所需的方向控制器;最后生存不等式和方向控制器输出会作为一组约束条件使轮式机器人避障转化为一组线性规划问题。机器人可以根据导航需要优化特定指标,如最大化控制输出以实现激进的高速避障,并对机器人在未知环境条件下算法的使用进行了描述。此外,通过给出不同控制器参数的仿真结果验证了该生存控制器的有效性并展示了控制参数对运动性能影响。
From a new view of viability theory, obstacle avoidance problems can be effectively analyzed combined with robotic dynamics. To solve the underactuated characteristic of wheeled robots, a new viability zone construction approach is introduced and combined with the robot model, the obstacle constrains can be transferred into a set of viability inequations. For the assurance of robotic obstacle avoidance capability, the sufficient globe viability proof for the zone is given. Meanwhile, to meet the demand of basic navigation, a direction controller is designed for this function. At last, the viability inequations and the output of the direction controller are turned into a set of constrains and the wheeled robot obstacle avoidance problem can be transferred into a linear programming problem. The robot can optimize a specific index according to the navigation demand, such as maximize the control output to achieve aggressive high speed obstacle avoidance and the usage of the algorithm has been described in the unknown environment case. Moreover, the simulation results of different controller parameters prove the validity of the viability of the controller and indicate the influence of the control parameters on the movement performance.
出处
《控制工程》
CSCD
北大核心
2016年第11期1707-1713,共7页
Control Engineering of China
基金
国家自然科学基金(11171221)
国家教育部博士点基金资助项目(20123120110004)
上海市一流学科建设项目(XTKX2012)
上海理工大学博士启动经费(1D-10-303-002)
关键词
生存理论
仿射非线性系统
自主避障
线性规划
Viability theory
affine nonlinear system
obstacle avoidance
linear programming
