Autonomous Navigation of Unmanned Vehicles: A Fuzzy Logic Perspective

摘要

Unmanned robotic vehicles are capable of performing desired tasks in unstructured, uncertain and potentially hostile environments. They may be remotely-operated or function (semi-) autonomously without human intervention. However, it will long before unmanned robot vehicles function as completely autonomous entities in diverse environments. Current unmanned vehicles adhere to different levels of autonomicity as defined by existing technology limitations and used sensors. Important operational characteristics related to unmanned vehicle functionality (aerial, aquatic or terrestrial), include the following: Perception: Acquire and use knowledge about the environment and itself. This is done by taking measurements using various sensing devices and then extracting meaningful information that should be used in all later tasks (such as localization, planning, collision free motion control, recharging, etc). Intelligence: Operate for a considerable time period without human intervention. This is associated with the learning and inference capabilities, which of the vehicle should have to be able to adapt (its behavior or/and shape) to the environment. Action: Travel from point A to point B. The vehicle should utilize predefined and acquired knowledge to move in dynamic environments without involving humans in the navigation loop. In robotics, autonomy is mainly associated with navigation issues. From a conceptual point of view, autonomous navigation of robotic vehicles may be achieved via continuous interaction between perception, intelligence and action, as shown in Figure Autonomous navigation conceptual loop