基于人工势场在机械臂避障路径规划中的研究综述

机械臂路径规划避障是目前机器人领域研究的热点,人工势场因算法简单、高效被越来越多地应用到机械臂路径规划避障中。本文首先介绍了人工势场的原理,分析了传统人工势场存在的目标不可达和局部极小值等问题,并对人工势场在机械臂避障规划中已有的算法进行分类综述,最后对人工势场在机械臂避障方面的发展趋势进行展望。     

基于人工势场一遗传算法的机械避障方法研究

人工势场法是机器人避障规划中常用的方法,具有简单、实时性强的特点,但其存在目标不可达和局部极小值问题;文中针对机械臂的避障问题,首先利用人工势场法确定机械臂末端的避障轨迹,使用改进的斥力场克服势场法的目标不可达问题,利用随机逃离和沿等势线逃离相结合的方法解决局部极小值问题;其次针对机械臂末端避障轨迹中的每个点,采用遗传算法计算其对应的机械臂逆运动学解,在确保杆件不发生碰撞的情况下,充分保持关节角变化的柔顺性;仿真结果表明算法能够有效地实现机械臂的避障规划.     

改进RRT-人工势场法的机械臂堆垛运动方法

快速拓展随机树算法(RRT)在机械臂路径规划中存在随机性强、搜索效率低、规划路径长等问题,不能在货柜堆垛场景中取得相对最优的光滑路径.对此,该文提出了一种改进RRT-人工势场法混合算法进行货柜堆垛机械臂运动规划.首先,对传统快速拓展随机树算法进行改进,在传统快速拓展随机树算法的全局搜索的基础上引入目标搜索,增强了随机树...     

面向未知地图的六足机器人路径规划算法

针对移动机器人路径规划中无法准确得知全局地图的问题,提出了一种基于模糊规则和人工势场法的局部路径规划算法。首先,利用测距组与模糊规则,进行障碍物的形状分类,构建局部地图;其次,在人工势场法中引入了一种修正的斥力函数,基于局部地图,利用人工势场法进行局部路径规划;最后,随着机器人的运动,设置时间断点,以减少路径震荡。针对随机障碍物和凹凸障碍物的地图,分别采用传统人工势场法和改进的人工势场法进行仿真,其结果表明：在遇到随机障碍物时,相比传统人工势场法,改进的人工势场法能够显著减少与障碍物的碰撞;在遇到凹凸障碍物时,改进的人工势场法能够很好地完成路径规划的目标。所提算法对地形变化适应能力强,能够实现在未知地图下的六足机器人路径规划。     

足球机器人路径规划的改进型人工势场算法研究

传统人工势场法不能适应复杂动态环境且容易产生局部极小,论文提出了一种改进型的人工势场算法,该算法考虑了机器人和障碍物的速度、加速度等动态特性,对传统人工势场进行了有效的调节,使其能更好地适应动态复杂环境,对局部极小问题进行判定,通过改变斥力场和引力场的影响力来解决局部极小问题,将该优化算法运用到足球机器人仿真比赛中,结果表明基于改进型人工势场优化算法能够在动态对抗性的环境中有效地实现最优路径规划,弥补了传统人工势场的不足。     

基于激光雷达的机器人改进人工势场路径规划研究

人工势场法路径规划需要建立在已知环境下障碍物分布位置的基础之上,而激光雷达传感器可以用于未知环境下障碍物分布位置的探测和获取,因此可以将两者结合,从而解决移动机器人未知环境下路径规划问题。该课题建立在人工势场法理论基础之上,在Matlab中构建路径规划仿真图形用户界面,利用机载激光雷达传感器探测获取障碍物的位置分布信息,通过串口将数据传输至Matlab软件中并显示。在Matlab软件下编写人工势场路径规划的实现算法,进行仿真实验。实验结果表明,传统人工势场法路径规划存在的两个问题,分析原因后给出一种改进的人工势场法,并在之前的图形用户界面下继续进行仿真实验。仿真结果表明：改进的人工势场法有效地实现了路径优化的目标。     

基于改进人工势场法的机械臂避障路径规划

在对机械臂进行避障路径规划时,传统人工势场法依赖周围环境信息进行避障,针对障碍物精确信息较难获取,且形状不一,提出了采用椭球包围盒对障碍物进行建模,兼顾了包围盒的简单性和紧密性。同时,针对路径规划中容易陷入局部最优的问题,则提出了使用RRT（rapidly-exploring random tree）方法来解决这种问题,当路径规划陷入局部最优时,使用RRT算法自动生成并添加虚拟目标点,通过改变机械臂周围的环境从而达到跳出局部最优的目的。多次仿真实验表明,采用椭球包围盒建模并采用上述改进算法时,可以快速并且有效地规划出一条无障碍的路径。     

复杂环境中基于人工势场优化算法的最优路径规划

本文提出一种基于人工势场优化的路径规划方法.把人工势场的路径规划结果作为先验知识，对蚁群算法进行初始化，提高了蚁群算法的优化效率;另一方面，机器人的路径也同时得到优化，克服了人工势场法的局部极小问题．仿真实验结果表明，该方法在复杂环境中能有效地实现最优路径规划；并提供了一种把传统规划方法和统计优化相结合、提高规划效率的可行思路．     

基于人工势场法的虚拟夹具建模技术研究

虚拟夹具是在计算机仿真技术、力反馈技术、人工智能等基础上发展起来的一种新兴机器人遥操作辅助技术.基于人工势场法的虚拟夹具建模技术研究对提高机械臂末端抓手自动防撞具有重要意义,特别是在非结构环境和一些工作空间环境难以建模的情况下,应用虚拟夹具法来限制机械臂末端的运动空间,可防止机械臂与周围环境发生碰撞.为此,使用OpenGL与C++相结合的编程方法设计并实现了三角片元法构建虚拟管道的几何模型,采用人工势场法计算实时虚拟力的虚拟夹具建模算法,通过仿真实验验证该算法能有效限制机械臂末端的运动空间及正确计算虚拟力.     

基于改进势场法的机械臂动态避障规划

本文针对刚性机械臂提出了一种基于改进势场法的动态避障规划算法.势场产生只作用于机械臂末端的引力和作用于机械臂与障碍物最近点的排斥力,并依据动力学定律分别产生吸引速度和排斥速度.本文直接在笛卡尔空间中构造基于目标速度的吸引速度,使机械臂能够追踪动态目标;构造方向依据障碍物速度变化的排斥速度,使机械臂连杆也具备避开动态障碍的能力.最终两种速度在机械臂关节空间合成,为机械臂规划一条无碰路径并控制机械臂追踪动态目标.仿真结果证明了该方法的有效性.     

事件空间中非完整力学系统的Herglotz-d′Alembert原理与守恒律

研究事件空间中非完整力学系统的Herglotz型守恒律.给出事件空间中Herglotz型广义变分原理,引入非完整约束并采用交换关系的H9lder定义,导出事件空间中非完整力学系统的新型微分变分原理—Herglotz-d′Alembert原理.引进事件空间中的空间生成元和参数生成元,建立Herglotz-d′Alembert原理不变性条件的变换.基于该原理构建了事件空间中非完整非保守力学系统的Herglotz型守恒定理及其逆定理.作为特例,给出了位形空间的Herglotz型守恒量和事件空间中完整力学系统的Herglotz型守恒量.文末还给出了一个算例.     

Asynchronous multi-domain variational integrators for nonlinear hyperelastic solids

We present the asynchronous multi-domain variational time integrators with a dual domain decomposition method for the initial hyperbolic boundary-value problem in hyperelasticity. Variational time integration schemes, based on the principle of minimal action within the Lagrangian framework, are constructed for the equation of motion and implemented into a variational finite element framework, which is systematically derived from the three-field de Veubeke-Hu-Washizu variational principle to accommodate the incompressibility constraint present in an analysis of nearly-incompressible materials. For efficient parallel computing, we use the dual domain decomposition method with local Lagrange multipliers to ensure the continuity of the displacement field at the interface between subdomains. The α-method for time discretization and the multi-domain spatial decomposition enable us to use different types of integrators (explicit vs. implicit) and different time steps on different parts of a computational domain, and thus efficiently capture the underlying physics with less computational effort. The energy conservation of our nonlinear, midpoint, asynchronous integration scheme is investigated using the Energy method, and both local and the global energy error estimates are derived. We illustrate the performance of proposed variational multi-domain time integrators by means of three examples. First, the method of manufactured solutions is used to examine the consistency of the formulation. In the second example, we investigate energy conservation and stability. Finally, we apply the method to the motion of a heterogeneous plane domain, where different integrators and time discretization steps are used accordingly with disparate material data of individual parts.     

Discrete-time stabilization of integrators in cascade: real-time stabilization of a mini-rotorcraft

This paper deals with the problem of stabilizing n integrators in cascade with bounded input using a discrete-time controller. We propose here a control algorithm using saturation functions for the case of 3 integrators and then we extend the development to the case of n integrators. The control methodology developed is simple and is explicitly given. We also present a discrete-time algorithm to control n integrators with a bounded input including a delay and we design an observer by considering that we only measure the output of the integrators chain. We present a real-time application of the proposed control algorithm for the stabilization of the orientation of a mini-rotorcraft.     

A meshless method for solving the free boundary problem associated with unilateral obstacle

In this paper, we discuss the numerical method for solving the first kind of elliptic variational inequality. We first use the fundamental solution as the basis function to approximate the solution of variational inequality, then we employ Uzawa's algorithm to determine the free boundary and the solution. Numerical examples are given to verify the efficiency of the method.     

自适应轮廓的变分水平集复杂背景多目标检测

无需重新初始化的变分水平集模型能够避免经典水平集模型的重复初始化步骤,进而简化计算,降低检测所需时间,同时能够有效利用图像的边缘梯度信息,从而准确定位图像的局部结构.但该模型不能自适应获得初始化曲线,水平集的拓扑结构也无法改变,不能解决多个目标的检测问题.针对以上问题,本文提出了一种基于自适应轮廓的变分水平集复杂背景多目标检测方法,该方法采用帧间差分算法与K-means聚类算法相结合,以获得多个运动目标的初始化曲线,通过形态学方法来降低图像噪声的干扰,从而快速自适应的估计复杂背景下运动目标的位置和轮廓大小.该算法进一步对无需初始化的变分水平集进行改进,将其由单目标检测模型扩展为多目标检测模型,并修正原模型难以处理图像灰度不均匀的问题,最终实现对复杂背景下多个目标的检测.在标准数据库和实际数据集上的测试结果表明,本文所提方法能够准确的定位不同尺度和灰度目标的轮廓,从而提高算法的演化迭代效率及准确性.     

Mixed interpretation and extensions of the equivalent mass matrix approach for elastodynamics with contact

The present paper highlights the idea that the mass redistribution technique introduced by Khenous and co-workers (2006) [17], [18] for elastodynamics with impact, can be reinterpreted as a mixed formulation in displacements and velocities in which a special compatibility condition is enforced. Such a formulation opens the route to various extensions and to variational integrators for impact problems, potentially with variational time adaption. Those ideas are exemplified in the design of such integrators and various mass redistribution schemes.     

The maximum principle for partially observed optimal control problems of mean-field FBSDEs

This paper is concerned with a partially observed optimal control problem described by mean-field forward and backward stochastic differential equations. Moreover, the control variable enters the diffusion coefficient and the control domain is non-convex. Utilising Girsanov's theorem as well as extended Ekeland's variational principle, a maximum principle is established in the form of Pontryagin's type. As an application, a linear-quadratic control problem is studied in terms of the stochastic filtering.     

Removal of Contrived Chaos in Finite-difference Methods

A chaos-free numerical method will be developed for the solution of a system of non-linear initial-value problems (IVP's) associated with the transmission dynamics of two HIV subtypes. It will be shown that integration of this four-dimensional system with some standard numerical integrators like the fourth-order Runge-Kutta algorithm leads to scheme-dependent numerical instabilities.     

Efficient numerical schemes for Chan-Vese active contour models in image segmentation

Multimedia Tools and Applications - In this paper, we introduce multi-symplectic Lagrangian variational integrators for solving Chan-Vese active contour models in image segmentation. Energy...     

Regularity of null-additive fuzzy measure on metric spaces

In this paper, we investigate the regularity of null-additive fuzzy measures on metric spaces and show that Lusin's theorem on fuzzy measure space remains valid for those fuzzy measures with null-additivity. Egoroff's theorem for fuzzy measures on metric spaces is presented. These are generalizations and improvements of the previous results obtained by others.