부산대학교 도서관

A method is developed and tested for extending the range of low-cost radar chipsets for use in urban air mobility (UAM) vehicles. The method employs weak-signal correlation techniques and long measurement intervals to achieve a 1 km range. Low-cost radar is an enabling technology for vertical take-off and landing (VTOL) aircraft envisioned for large-scale deployment in urban areas. These aircraft must be autonomously piloted to make them economically feasible, but autonomous systems have yet to match a human pilot's ability to detect and avoid (DAA) obstacles. Visible light cameras are useful for this application, but cameras alone are insufficient, as they are fundamentally unable to resolve range. Existing commercial radar units would suffice for DAA, but their large size weight, power, and cost (SWaP-C) militates against their application to UAM. The technique detailed in this paper is a fused camera-radar solution that exploits the camera's excellent angular resolution to guide radar signal processing so that signals arriving from a camera-detected target are combined constructively. Such guided processing significantly extends the range of low SWaP-C radar chipsets, making them useful for DAA. An analysis of the fused technique's robustness to target velocity uncertainty is presented, along with experimental results indicating that a typically-sized VTOL aircraft would be detectable at a range of 1 km.