• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>Evolutionary and Bio-inspired Computation: Theory and Applications II >A biologically inspired approach to modeling unmannedvehicle teams
【24h】

A biologically inspired approach to modeling unmannedvehicle teams

机译：从生物学角度启发无人驾驶车辆模型的方法

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Cooperative motion control of teams of agile unmanned vehicles presents modeling challenges at several levels. The "microscopic equations" describing individual vehicle dynamics and their interaction with the environment may be known fairly precisely, but are generally too complicated to yield qualitative insights at the level of multi-vehicle trajectory coordination. Interacting particle models are suitable for coordinating trajectories, but require care to ensure that individual vehicles are not driven in a "costly" manner. From the point of view of the cooperative motion controller, the individual vehicle autopilots serve to "shape" the microscopic equations, and we have been exploring the interplay between autopilots and cooperative motion controllers using a multi-vehicle hardware-in-the-loop simulator. Specifically, we seek refinements to interacting particle models in order to better describe observed behavior, without sacrificing qualitative understanding. A recent analogous example from biology involves introducing a fixed delay into a curvature-control-based feedback law for prey capture by an echolocating bat. This delay captures both neural processing time and the flight-dynamic response of the bat as it uses sensor-driven feedback. We propose a comparable approach for unmanned vehicle modeling; however, in contrast to the bat, with unmanned vehicles we have an additional freedom to modify the autopilot. Simulation results demonstrate the effectiveness of this biologically guided modeling approach.
机译：敏捷无人驾驶车辆团队的协同运动控制在几个层面上都提出了建模挑战。描述各个车辆动力学及其与环境的相互作用的“微观方程式”可能相当精确，但通常过于复杂而无法在多车轨协调的水平上获得定性见解。相互作用的粒子模型适用于协调轨迹，但是需要注意确保不会以“昂贵”的方式驾驶单个车辆。从协同运动控制器的角度来看，单个车辆的自动驾驶仪可以“塑造”微观方程，并且我们一直在使用多车硬件在环仿真器探索自动驾驶仪和协同运动控制器之间的相互作用。 。具体而言，我们在不牺牲定性理解的前提下，寻求对交互粒子模型的改进，以便更好地描述观察到的行为。来自生物学的最新类似示例涉及将固定延迟引入基于曲率控制的反馈定律中，以通过回声定位棒捕获猎物。由于使用传感器驱动的反馈，这种延迟捕获了蝙蝠的神经处理时间和飞行动力响应。我们提出了一种无人驾驶车辆建模的可比方法。但是，与蝙蝠相比，无人驾驶载具具有更大的自由来修改自动驾驶仪。仿真结果证明了这种生物学指导的建模方法的有效性。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号