Designing Stable Finite State Machine Behaviors Using Phase Plane Analysis and Variable Structure Control

This paper discusses how phase plane analysis can be used to describe the overall behavior of single and multiple autonomous robotic vehicles with finite state machine rules. The importance of this result is that we can begin to design provably stable group behaviors from a set of simple control laws and appropriate switching points with decentralized variable structure control. The ability to prove stable group behavior is especially important for applications such as locating military targets or land mines. In this paper, we demonstrate how phase plane analysis has been used to explain the behavior of a 16 cm³ autonomous line-tracking robot with four finite states. After which, the analysis is extended to include the design of a decentralized variable structure controller that guides multiple vehicles to a goal while avoiding each other.     

On Finding the Relevant Dynamics for Model-Based Controlling Walking Robots

Leg dynamics are often ignored in the real-time control of walking robots because of the high gearing used in leg transmissions. However, the use of a gear reduction high enough to discount Coriolis and centripetal components yields additional non-desired dynamics, which are friction, backlash and elasticity. In such cases, simplifying robot dynamics without considering the effect of gear dynamics leads to unavoidable errors. In order to make dynamic equations reflect the reality of the physical system, it is of paramount importance to model the most significant effects acting on the system. Robot dynamics could then be analyzed and related to trajectory parameters for motion-control purposes. In this paper, a method to derive the dynamics of a robot leg as a function of leg-trajectory parameters is proposed. The method experimentally finds the simplified equations of motion that reflect the reality of the physical system. The resulting model is an accurate and simple representation of the system dynamics, taking into account the most relevant dynamics affecting the system. The simplification of the model allows it to be used in a real-time dynamic-control system.     

基于轨迹引导的移动机器人导航策略优化算法

针对在杂乱、障碍物密集的复杂环境下移动机器人使用深度强化学习进行自主导航所面临的探索困难,进而导致学习效率低下的问题,提出了一种基于轨迹引导的导航策略优化（TGNPO）算法。首先,使用模仿学习的方法为移动机器人训练一个能够同时提供专家示范行为与导航轨迹预测功能的专家策略,旨在全面指导深度强化学习训练;其次,将专家策略预测的导航轨迹与当前时刻移动机器人所感知的实时图像进行融合,并结合坐标注意力机制提取对移动机器人未来导航起引导作用的特征区域,提高导航模型的学习性能;最后,使用专家策略预测的导航轨迹对移动机器人的策略轨迹进行约束,降低导航过程中的无效探索和错误决策。通过在仿真和物理平台上部署所提算法,实验结果表明,相较于现有的先进方法,所提算法在导航的学习效率和轨迹平滑方面取得了显著的优势。这充分证明了该算法能够高效、安全地执行机器人导航任务。     

卫星自主导航中月亮摄动及其仿真研究

卫星自主导航是卫星控制技术发展的必然趋势,是当今世界航天领域研究的前沿热点之一。该文针对基于组合大视场星敏感器星光折射的卫星天文自主导航方法,应用广义卡尔曼滤波理论,建立了月亮引力摄动下的系统状态模型和系统观测模型。对系统模型进行了状态最优估计周围的线性化和采样周期的离散化,给出了干扰方程。在滤波计算过程中,采用了改进的广义卡尔曼滤波算法,以某地球同步轨道通信卫星为背景,进行了计算机仿真研究。分析和讨论了月亮摄动力对卫星自主导航的影响,给出了其影响的变化规律。研究和仿真结果表明,月亮引力是卫星自主导航的主要摄动因素之一,直接影响卫星轨道倾角的变化,并且卫星轨道越高,影响效果越明显。     

Robust dynamic surface control of flexible joint robots using recurrent neural networks

A robust neuro-adaptive controller for uncertain flexible joint robots is presented. This control scheme integrates H-infinity disturbance attenuation design and recurrent neural network adaptive control technique into the dynamic surface control framework. Two recurrent neural networks are used to adaptively learn the uncertain functions in a flexible joint robot. Then, the effects of approximation error and filter error on the tracking performance are attenuated to a prescribed level by the embedded H-infinity controller, so that the desired H-infinity tracking performance can be achieved. Finally, simulation results verify the effectiveness of the proposed control scheme.     

Development of an adaptive fuzzy behavioural control system with experimental and industrial applications

This paper discusses a newly developed Adaptive Fuzzy Behavioural Control System which has been designed for use with an autonomous mobile robot. The merit of the system is the capability to adapt the fuzzy function to suit the environmental conditions encountered without changes to the rule base. This is compared to conventional behavioural control and demonstrated by simulation. The paper also discuses the results of both experimental and industrial applications in which this new control system was applied.     

多动态目标的多机器人协同环航控制

主要研究了多个非完整机器人对多个动态目标的协同环航控制问题;首先,针对多目标护航任务,建立目标扩展圆形构型,以期完成紧密目标保卫任务;其次,针对护航机器人,通过利用自身及相邻节点的位置与方位信息及所包围的动态目标的中心位置及扩展半径设计分布式时变圆形编队控制协议,实现预定几何分布下的机器人环航编队设计;其中,通过引入虚拟信号变量设计期望的速度及角速度控制律,利用反步技术提出了一种新的考虑时变护航半径的分布式控制策略;在对目标节点速度的温和假设下,所提出的编队控制器可以驱动多护航机器人渐近收敛到以多目标中心的圆上,同时维持一个预定的几何编队配置;Lyapunov分析证明了所有的误差都可以渐进稳定到原点,数值仿真核实了所构建的控制方案的可行性。     

一类轮式机器人的重心控制

介绍了一类建筑领域等非结构环境下使用的由6个转动关节和4个直动关节组成的轮式机器人,其特殊的结构加上重心控制,能使机器人在不平整地面上行走而不倾倒。绘制了机器人重心控制原理,推导出使机器人不倾倒的重心控制运动方程。给出了实验用的部分软硬件原理框图,并通过实验验证了方案的可行性。     

无人机自主性概念研究

讨论了与无人机自主性有关的三组概念.介绍了一种无人机自主性的定义和作为其工程概念的自主能力等级.最后分析了无人机自主性,提出了无人机自主性的研究内容.     

Autonomous knowledge acquisition based on artificial curiosity: Application to mobile robots in an indoor environment

This paper describes an autonomous system for knowledge acquisition based on artificial curiosity. The proposed approach allows a humanoid robot to discover, in an indoor environment, the world in which it evolves, and to learn autonomously new knowledge about it. The learning process is accomplished by observation and by interaction with a human tutor, based on a cognitive architecture with two levels. Experimental results of deployment of this system on a humanoid robot in a real office environment are provided. We show that our cognitive system allows a humanoid robot to gain increased autonomy in matters of knowledge acquisition.