汽車防撞雷達與複合式波形設計

Abstract

本論文針對汽車防撞雷達中，頻率調變連續波雷達(Frequency Modulation Continue Wave, FMCW)提出複合式波形設計，並透過觀察速度與距離二維變相，整合出碰撞反應時間的預警機制，再配合演算法過濾多目標資訊，找出危險程度高的物體，據此給予使用者警示。 複合式波形雷達結合了都普勒雷達的連續波形與傳統FMCW雷達中的三角波形，第一部分連續段因為收發訊號雜訊互相抵銷，使得雜訊比能夠比連續頻率調變雷達更好，頻率上還有目標與雜波分離的優點，而連續段偵測上只能獲得速度與角度資訊，缺乏距離資訊。FMCW雷達三角波段，其頻率成分包含速度與距離，可以偵測出距離、速度與角度資訊，但因為頻率包含距離成分，因此受到雜波干擾嚴重，而複合式波段結合兩種波形的優點並改善原有的缺點。 本論文在雷達提供目標物資訊前，透過去雷達演算法預先對資訊做處理，找出對使用者威脅程度高的目標，讓使用者在處理資訊上注意力更加集中，使得思考速度更加迅速。對於目標物的威脅程度，本論文提出同時根據距離與速度二維變相，整合出車輛發生碰撞時間(Time To Collision, TTC)的概念作為危險評估標準，並配合後視雷達的實際應用，藉由模擬實際道路雜波模型與目標物資訊來驗證雷達演算法的功能。

Abstract In the thesis, we designed a multi-waveform for vehicle avoidance collision radar. By using Time To Collision (TTC) alarm system and paring radar algorithm, dangerous targets could be found and alarmed the driver . The multi-waveform is composed of Doppler radar's waveform and general Frequency Modulation Continue Wave (FMCW) triangle waveform. The first part is Continue wave (CW) chirp. In the CW chirp, the intermediate frequency (IF) signal is always worked in same frequency so that the transmit signal's noise and receiver's noise may cancel each other, resulted in the better Signal to Noise Ratio (SNR) could also achieved due to the separation of the target's frequency from clutter. However, we can only get distance and angle information in the CW chirp. By triangle chirp, we can more information, like distance , velocity, and angle information. Combined two waveforms, multi-waveform radar has both of their advantages. Before radar send targets' messages to the user, we prior to find the dangerous targets and remove the unnecessary information by radar algorithm so that the user can think easily and concentrate on process the targets' messages .At last, in order to verify the radar algorithm work well, we simulate the target's signal model, clutter model, and noise in real lane environment .