节点文献

汽车毫米波防撞雷达的研究与实现

Research and Implementation of Automotive Millimeter-wave Collision Avoidance Radar

【作者】 赵爽

【导师】 陈殿仁;

【作者基本信息】 长春理工大学 , 通信与信息系统, 2013, 博士

【摘要】 随着时代的发展和社会的进步,汽车数量日益增多,交通事故时常发生。汽车防撞雷达通过判断车辆前方有无目标,对目标的速度以及距离进行测量,从而提醒驾驶员避免交通事故的发生,具有非常广阔的应用前景。本文主要进行了雷达目标识别的算法研究以及雷达的可实现研究,论文的主要工作集中体现在以下几个方面:雷达目标识别技术研究及高精度测距修正算法研究、毫米波雷达抗干扰研究、高线性度线性调频源以及误差研究。作者在以下几个方面的研究工作中取得了进展及创新:1、针对雷达的发射波形问题,提出了采用脉冲分段对称三角线性调频波形作为雷达的发射波形。当雷达采用对称三角线性调频波形作为发射波形时,经过拉伸处理后的上、下扫频段回波信号中间会包含一小段线性调频信号。当采用FFT作为回波信号的主要处理方法时,这部分线性调频信号并不能提供有效信息,会使发射能量有在这线性调频段的时间内造成浪费。采用脉冲分段对称三角线性调频波形作为雷达的发射波形,可以保证经过拉伸处理后的回波信号的采样数据中不包含线性调频信号,因而提高了信号的信噪比,并提高了目标距离估计算法的精度。2、在目标测距算法中,提出了通过计算上下扫频段回波信号频谱幅度差值,修正目标距离的算法,提高了测距精度。当采用采样点补零、旁瓣抑制加权等算法对雷达采样信号进行处理时,会使目标回波的一维FFT处理结果在频谱主瓣内出现多条谱线,造成目标的模糊,无法分辨是一个目标,还是多个目标。针对此问题,本文采用二维FFT计算出目标的多普勒频移,在已知多普勒频移的情况下,得出一维FFT频谱幅度差值归一化曲线,然后将目标回波的上下扫频段一维FFT频谱幅度差值归一化,在归一化差值曲线上,查找到对应的修正频率,对目标频谱进行修正。通过修正算法,可以将单一目标在主瓣内的多条谱线修正为一条,解决了目标模糊问题,同时提高了测距精度。3、对于距离相近速度不同的目标识别,提出了采用CZT变换提取目标的速度信息,同时提取出上下扫频段频谱幅度差值,再进行距离修正的方法。当雷达前方存在距离相近速度不同的目标时,它们的一维FFT频谱重叠在一起,无法将一维FFT频谱幅度差值归一化,不能应用距离修正算法对目标距离进行修正。但由于目标的速度不同,可以在已知二维FFT的速度结果的情况下应用CZT变换,提取出各个目标的幅度信息,再进行频谱幅度差值归一化,对距离进行修正。此方法可以使雷达对距离相近速度不同的目标进行识别,大幅提高了各个目标的距离测量精度。4、针对不同车辆同时使用防撞雷达造成的互相干扰问题,提出了采用系统跳频的回波处理算法结合低旁瓣微带天线的方法,分别在频域与空间上尽可能降低雷达被干扰的几率。汽车防撞雷达的工作环境比较复杂,经常会遇到多部雷达同时工作的情况。采用系统跳频的方法,使雷达的发射频率各不相同,再通过对雷达的上、下扫频段的目标和干扰信号进行分析,对上、下扫频段目标进行配对,可以使雷达能够区分前方目标和干扰信号。针对雷达可能方向上的干扰,提出了采用微带天线,结合阵列加权以及方向图零点约束的方法,使雷达的发射天线和接收天线的旁瓣尽可能减小,并且发射天线旁瓣与接收天线的零点相对应,尽可能在满足雷达探测角度的前提下,降低雷达天线旁瓣增益,减小雷达的方向上的干扰。5、研究了基于DDS技术的宽带线性调频源系统,通过分析系统的误差模型,提出了根据DDS的相位截断位数选取线性调频步长,以及根据线性调频的频率捷变速度选取接收机A/D采样速度的办法,减小了由于相位截断误差和线性调频不连续所带来的系统误差。

【Abstract】 With the development of society, the number of cars is increasing and the traffic accidents occur frequently. Automotive collision avoidance radar can detect whether the target exists in front of the car and measure the speed and the distance, thus reminds the driver to avoid traffic accidents. It has very broad application prospect. This paper mainly did research on the radar target recognition algorithms and the realization of the radar. The work of this paper epitomized in the following areas:the radar target recognition technology、 precision ranging correction algorithm、Millimeter-wave radar anti-jamming and the correction of the chirp source of high linearity and error. The paper made progress and innovation in the following areas:1、 For radar transmitted waveform, proposed the pulse segmented LFMCW as the radar transmitter waveform. When the radar uses the LFMCW waveform as the transmitted waveform, the upper and lower sweep band echo contain a short linear FM signal after the stretching process. When using FFT as a main process method of the echo signal, this part of the chirp signal and can not provide valid information. This causes the emission energy wasted in the chirp segment time. While using the pulse segmented LFMCW as the radar transmitted waveform, it can be guaranteed the echo signal does not contain the linear FM signal after the stretching treatment. It improves the SNR of the signal and the target distance estimation algorithm accuracy.2、In the target distance algorithm, by calculating the spectral amplitude difference of the upper and lower sweep band echo signal, corrected the target distance algorithm, which increased ranging precision. When plus zero to the simple point and use the side lobe suppression weighting algorithm as radar signal processing, multi-spectral lines will appear within the main lobe of the spectrum when using FFT process the data of the radar. It will cause the target fuzzy, unable to distinguish whether it is a target or multiple targets. To solve this problem, this paper used two-dimensional FFT calculate target Doppler shift. In the case of the known Doppler frequency shift, drawn a one-dimensional FFT spectrum amplitude difference normalized curve, and then normalized one-dimensional FFT spectrum amplitude difference of the target echo up and down sweep band. In the normalized difference curve, the corresponding correction frequency to correct the target spectrum has been funded. Though the correction algorithm, the multiple spectral lines of the main lobe could be corrected to one. At the same time, it improved the ranging accuracy.3、For the targets of similar distance but different speed,proposed use CZT transform extract the speed information of the target and the amplitude difference of the band spectrum.Then use the distance correction method to correct the target distance. When there are similar distance but different speed targets in front of the the radar, Their one-dimensional FFT spectrum overlap so can not normalize the one-dimensional FFT Spectral amplitude. However, due to the different speed of the target, the amplitude information of the respective target can be extracted by using CZT to process the results of the two-dimensional FFT. Then the distance of the target can be corrected by using amplitude spectrum difference normalized. So that the radar can identify different target with similar distance but different speed, and increases the distance measurement accuracy of the various target.4、For multiple radar work simultaneously interference problems, the the system hopping echo processing algorithms combined with low sidelobe microstrip antenna have been proposed. The work environment of the automotive collision avoidance radar is complex, it often encounters multiple radar work at the same time. Using system hopping, each radar has a different transmitter frequency. Through analyzing the upper and lower sweep band of the radar and the interference signal, pairing of upper and lower sweep band target, allows the radar to be able to distinguish between the front of the target and the interference signal. In the millimeter wave automotive anti-collision radar antenna, the paper designed microstrip array antenna. Used the weighted algorithm combined with the zero constraints to reduce the antenna sidelobes. Then reduce the interference caused by the possible directions.5、Broadband chirp source systems based on DDS technology have been researched. By analysis system error model, proposed according to the DDS phase truncation digit to select chirp step and according to Chirp frequency agility speed to select the A/D sampling speed of the receiver. Reduces the system error caused by the discontinuous phase truncation and linear FM.

节点文献中: