Scannapieco, Antonio Fulvio (2017) Ultralight Radar Sensor for Autonomous Operations by Mini- and Micro-UAS. [Tesi di dottorato]

Preview

Text scannapieco_antoniofulvio_29.pdf Download (120MB) | Preview

Collection description

In recent years the boost in operations by mini- and micro-UAS (Unmanned Aircraft Systems, also known as Remotely Piloted Aircraft Systems - RPAS - or simply drones) and the successful miniaturization of electronic components were experienced. Radar sensors demonstrated to have favorable features for these operations. However, despite their ability to provide meaningful information for navigation, sense-and-avoid, and imaging tasks, currently very few radar sensors are exploited onboard or developed for autonomous operations with mini- and micro-UAS. Exploration of indoor complex, dangerous, and not easily accessible environments represents a possible application for mini-UAS based on radar technology. In this scenario, the objective of the thesis is to develop design strategies and processing approaches for a novel ultralight radar sensor able to provide the miniaturized platform with Simultaneous Localization and Mapping (SLAM) capabilities, mainly but not exclusively indoors. Millimeter-wave Interferometric Synthetic Aperture Radar (mmw InSAR) technology has been identified as a key asset. At the same time, testing of commercial lightweight radar is carried out to assess potentialities towards autonomous navigation, sense-and-avoid, and imaging. The two main research lines can be outlined as follows: - Long-term scenario: Development of very compact and ultralight Synthetic Aperture Radar able to provide mini- or micro-UAS with very accurate 3D awareness in indoor or GPS-denied complex and harsh environments. - Short-term scenario: Assessment of true potentialities of current commercial radar sensors in a UAS-oriented scenario. Within the framework of long-term scenario, after a review of state-of-art SAR sensors, Frequency-Modulated Continuous Wave (FMCW) SAR technology has been selected as preferred candidate. Design procedure tailored to this technology and software simulator for operations have been developed in MATLAB environment. Software simulator accounts for the analysis of ambiguous areas in a three-dimensional environment, different SAR focusing algorithms, and a Ray-Tracing algorithm specifically designed for indoor operations. The simulations provided relevant information on actual feasibility of the sensor, as well as mission design characteristics. Additionally, field tests have been carried out at Fraunhofer Institute FHR with a mmw SAR. Processing approaches developed from simulations proved to be effective when dealing with field tests. A very lightweight FMCW radar sensor manifactured by IMST GmbH has been tested for short-term scenario operations. The codes for data acquisition were developed in Python language both for Windows-based and GNU/Linux-based operative systems. The radar provided information on range and angle of targets in the scene, thus being interesting for radar-aided UAS navigation. Multiple-target tracking and radar odometry algorithms have been developed and tested on actual field data. Radar-only odometry provided to be effective under specific circumstances.

Downloads

Actions (login required)

View Item