Design and testing of a UAV-based, through-rubble vital sign detection RADAR using metaheuristic topology optimization

Abstract

This thesis forms part of a larger effort at MIT Lincoln Laboratory to develop a micro-UAV based platform, capable of detecting survivors through rubble, under the funding of the New Technology Initiative (NTI) program. In support of this goal the thesis makes three distinct contributions. First, the operating environment of a disaster scenario is characterized. To do so, the electrical properties of different residential construction materials are determined and an analytical model for the behavior of a radar operating in this environment is developed. Second, preliminary efforts were made towards the miniaturization of the radar back-end by designing circuitry that generates the transmit waveform. Additionally an analog filter was developed to attenuate unwanted signals on the receive end. Lastly metaheuristic topology optimization was applied towards the design of antennas for the radar front-end. A novel, algorithmic extension to an established metaheuristic algorithm is demonstrated and tested. Additionally a new antenna parametrization based on Bezier curves is developed and evaluated against the established pixel-based parametrization.