论文标题
动态环境中的轨迹计划：安全人类互动中的触觉感知的应用
Trajectory planning in Dynamics Environment : Application for Haptic Perception in Safe HumanRobot Interaction
论文作者
论文摘要
储层计算是预测湍流的有力工具，其简单的架构具有处理大型系统的计算效率。然而，其实现通常需要完整的状态向量测量和系统非线性知识。我们使用非线性投影函数将系统测量扩展到高维空间，然后将其输入到储层中以获得预测。我们展示了这种储层计算网络在时空混沌系统上的应用，该系统模拟了湍流的若干特征。我们表明，使用径向基函数作为非线性投影器，即使只有部分观测并且不知道控制方程，也能稳健地捕捉复杂的系统非线性。最后，我们表明，当测量稀疏、不完整且带有噪声，甚至控制方程变得不准确时，我们的网络仍然可以产生相当准确的预测，从而为实际湍流系统的无模型预测铺平了道路。
In a human-robot interaction system, the most important thing to consider is the safety of the user. This must be guaranteed in order to implement a reliable system. The main objective of this paper is to generate a safe motion scheme that takes into account the obstacles present in a virtual reality (VR) environment. The work is developed using the MoveIt software in ROS to control an industrial robot UR5. Thanks to this, we will be able to set up the planning group, which is realized by the UR5 robot with a 6-sided prop and the base of the manipulator, in order to plan feasible trajectories that it will be able to execute in the environment. The latter is based on the interior of a vehicle, containing a person (which would be the user in this case) for which the configuration will also be made to be taken into account in the system. To do this, we first investigated the software's capabilities and options for path planning, as well as the different ways to execute the movements. We also compared the different trajectory planning algorithms that the software is capable of using in order to determine which one is best suited for the task. Finally, we proposed different mobility schemes to be executed by the robot depending on the situation it is facing. The first one is used when the robot has to plan trajectories in a safe space, where the only obstacle to avoid is the user's workspace. The second one is used when the robot has to interact with the user, where a dummy model represents the user's position as a function of time, which is the one to be avoided.
