New advances in complex motion control for single robot systems and multi-agent systems
New advances in complex motion control for single robot systems and multi-agent systems
复制标题
DOI：
10.1007/s11431-016-0779-x
复制
发表时间：
2016-11
期刊：
影响因子：
--
通讯作者：
H. Fang;Shao-lei Lu;Jie Chen
中科院分区：
文献类型：
--
作者：
H. Fang;Shao-lei Lu;Jie Chen
运动控制是一个经典的问题，一直受到控制界和机器人界的关注。这个问题过去常常在理想化的环境中解决，或者使用简化的模型来完成复杂的任务。现在，在现实环境中控制机器人的愿望促使研究人员考虑越来越复杂的系统和场景。这给运动控制问题带来了新的挑战，其中一个可能使运动控制问题复杂化的问题就是环境的可观测性。在完全可观察的环境中，机器人的观察可以揭示环境的当前状态，然后可以用于设计控制律[1-3]。在部分可观测的环境中，由于信息不完整，也称为不完美信息，机器人可能会对环境的一种状态产生不同的观测结果。在多智能体系统中，信息不完全的情况经常出现，其中智能体的通信是有限的，交互是复杂的[4-8]。不完美的信息使控制设计问题复杂化，并可能降低整个系统的性能。通过与人类的合作来提高复杂环境中多智能体系统的性能[9-10]。为了开发实际的机器人应用，需要考虑环境，并且环境越复杂，在设计控制律时就越困难。复杂的环境导致复杂的机制，这使得设计问题变得复杂。最近的研究集中在复杂环境中完整和非完整机器人系统的控制器设计问题，其中不同的方法，
Motion control is a classic problem, which has constantly attracted attentions from the control and the robotics communities. The problem used to be tackled in idealistic settings or using simplified models to complete complex tasks. Now the wish to control robots in realistic environments has driven the researchers to consider more and more complex systems and scenarios. This brings new challenges to the motion control problems.One that may complicate the motion control problem is the observability of the environment. In a fully observable environment, the observation of the robot can reveals the current state of the environment, which can then be used to design control laws [1–3]. In a partially observable environment, the robot may produce different observations for one state of the environment, because of incomplete information, also called imperfect information. The situation of imperfect information is often seen in multi-agent systems, where agent’s communications are limited and interactions are complicated [4–8]. Imperfect information complicates the control design problem and may degrade the performance of the overall system. Cooperation with human was adopted to improve the performance of a multi-agent system in complex environments [9–10]. To develop practical robot applications, it is necessary to take the environment into consideration, and the more complex the environment is, the more difficult it will be when designing the control law. Complex environments lead to complicated mechanisms, which makes the design problem complex. Recent studies focused on the controller design problem in complex environments for holonomic and nonholonomic robot systems, where different approaches to