arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 0810.2665cs.RO
收藏

基于多智能体系统或运动捕捉的物体、机器人和人体模型路径规划器

Path Planner for Objects, Robots and Mannequins by Multi-Agents Systems or Motion Captures

Institut de Recherche en Communications et Cybernétique de Nantes(南特通信与控制研究所);École Centrale de Nantes(南特中央理工学院);CEA(法国原子能委员会);EADS/CCR(欧洲航空防务与航天公司/核心研究中心)

Damien Chablat

摘要:

为了优化新产品的设计成本和时间,同时提高其质量,并行工程是基于这些产品的数字模型。然而,为了能够避免明确的物理模型而不丢失信息,必须提供新的工具。特别是,一个工具,使其能够简单,快速地检查复杂的机械装置的可维护性使用的数值模型是必要的。十年来,IRCCyN的MCM团队致力于创建用于生成和分析虚拟人体模型轨迹的工具。人类任务的模拟可以通过类似机器人的模拟或通过运动捕捉的模拟来进行。本文给出了这两种方法的一些结果。第一种方法基于多智能体系统和数字模型技术,在考虑人体工程学约束或关节限制的情况下,评估人体模型或机器人的访问和可见性任务的有效路径规划器。人工操作员被集成到过程优化中,以促进对环境的全局感知。该操作员与几个自动本地基本代理实时合作。在第二种方法中,我们与CEA和EADS/CCR解决相关的约束条件的演变的虚拟人在其环境中的基础上,从运动捕捉系统产生的数据。开发了一种使用虚拟指南的方法,以允许用户在没有力反馈的情况下实现精确的轨迹。

英文摘要:

In order to optimise the costs and time of design of the new products while improving their quality, concurrent engineering is based on the digital model of these products. However, in order to be able to avoid definitively physical model without loss of information, new tools must be available. Especially, a tool making it possible to check simply and quickly the maintainability of complex mechanical sets using the numerical model is necessary. Since one decade, the MCM team of IRCCyN works on the creation of tools for the generation and the analysis of trajectories of virtual mannequins. The simulation of human tasks can be carried out either by robot-like simulation or by simulation by motion capture. This paper presents some results on the both two methods. The first method is based on a multi-agent system and on a digital mock-up technology, to assess an efficient path planner for a manikin or a robot for access and visibility task taking into account ergonomic constraints or joint limits. The human operator is integrated in the process optimisation to contribute to a global perception of the environment. This operator cooperates, in real-time, with several automatic local elementary agents. In the second method, we worked with the CEA and EADS/CCR to solve the constraints related to the evolution of human virtual in its environment on the basis of data resulting from motion capture system. An approach using of the virtual guides was developed to allow to the user the realization of precise trajectory in absence of force feedback.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑