Sensing a complex world: Infinite dimensional observer theory for robots

感知复杂的世界：机器人的无限维观察者理论

• 批准号: DP160100783
• 负责人: Prof Jochen Trumpf
• 金额: $ 22.04万
• 依托单位: The Australian National University
• 依托单位国家: 澳大利亚
• 项目类别: Discovery Projects
• 财政年份: 2016
• 资助国家: 澳大利亚
• 起止时间: 2016-04-01 至 2020-12-31
• 项目状态: 已结题
• 来源: https://dataportal.arc.gov.au/RGS/Web/Grants/DP160100783
• 关键词: Sensing; complex; world; Infinite; dimensional

This project aims to develop the foundational theory and design paradigms to support the new generation of sensor systems crucial to enabling widespread robotic automation in unstructured environments such as mining, agriculture and urban transport. Modern dense robotic sensor modalities such as CMOS (complementary metal oxide semiconductor) sensors, LIDAR and dense acoustic arrays are best modelled as a fine grid of measurements from an infinite dimensional dynamical system. The project plans to develop infinite dimensional invariant observer theory to formulate implementable algorithms that run in real-time on embedded hardware, providing detailed information that enables robots to undertake tasks that are presently impossible with state-of-the-art sparse sensing paradigms.

该项目旨在开发基础理论和设计范式，以支持新一代传感器系统，这些系统对于在采矿、农业和城市交通等非结构化环境中实现广泛的机器人自动化至关重要。现代密集的机器人传感器模式，如CMOS（互补金属氧化物半导体）传感器，激光雷达和密集的声学阵列最好建模为一个精细的网格测量从一个无限维的动态系统。该项目计划开发无限维不变观测器理论，以制定可在嵌入式硬件上实时运行的可实现算法，提供详细信息，使机器人能够承担目前最先进的稀疏传感范例不可能完成的任务。