随着自动驾驶技术的快速发展，高级别辅助驾驶系统的研究备受广大学者青睐。点云成像作为车辆感知环境的重要手段，对自动驾驶领域的研究具有重要工程价值。相对于激光雷达和摄像头等车载传感器，车载毫米波雷达具有更强的穿透雾、烟、尘的能力，具有全天候全天时的特点。因此，本文以高质量的目标点云为研究对象，针对毫米波雷达波束较宽角分辨率差，多径效应下的信号损失天线孔径自由度以及点云质量有效提高的关键问题，对高分辨的点云成像进行研究。本论文的研究内容主要概括为以下两个部分：

（1）针对实际道路中多径传播和同频干扰产生的大量相干源，导致传统的子空间类的超分辨测向算法性能急剧下降甚至失效的问题，提出了一种基于ESPRIT-LIKE去相干测角算法的改进算法。由于传统去相干ESPRIT-LIKE算法重构的协方差Toeplitz矩阵信息利用率低且在较小角度间隔下性能不佳，本文在其基础上重新构造一个提升天线孔径元素信息利用率的Toeplitz矩阵，然后结合前后平滑的思想，对重新构造的矩阵进行平滑修正，并通过子空间类超分辨算法实现对两维角度的估计。实验结果表明，本文提出的改进ESPRIT-LIKE算法在相干环境下的均方根误差和成功率均大于传统算法，大大提升在相干环境下的角度估计性能。

（2）针对传统3D-FFT点云成像方法生成点云质量不佳的问题，提出了一种基于超分辨测角的点云成像方法。首先引入MIMO虚拟孔径技术来增加虚拟孔径，增加天线自由度，从硬件和信号处理的角度增加对雷达对目标的分辨能力。基于MIMO体制的回波储存形式，从数据处理的层面分析目标三维信息的计算，将改进的ESPRIT-LIKE测角算法应用至点云成像，并给出新的点云成像方法。最后通过TI的AWR-1843毫米波雷达在实际道路上进行实测实验，根据实验结果对比分析，基于改进的ESPRIT-LIKE测角算法生成的点云数目和质量更优，可以有效还原运动目标实时的三维轮廓。

With the rapid development of autonomous driving technology, the research of high-level assisted driving systems has been favored by scholars. As an important means of vehicle perception of the environment, point cloud imaging has important engineering value for the research in the field of autonomous driving. Compared with vehicle-mounted sensors such as lidar and cameras, vehicle-mounted millimeter-wave radar has a stronger ability to penetrate fog, smoke and dust, and has the characteristics of all-weather and all-day. Therefore, this paper takes high-quality target point cloud as the research object, aiming at the key problems of poor resolution of millimeter-wave radar beam, signal loss antenna aperture freedom under multipath effect, and effective improvement of point cloud quality. The research content of this paper is mainly summarized in the following two parts:

(1) Aiming at the problem that a large number of coherent sources generated by multipath propagation and co-channel interference in the actual road, resulting in a sharp decline in the performance or even failure of the traditional super-resolution direction finding algorithm of subspace class, an improved algorithm based on the ESPRIT-LIKE decoherent goniometric algorithm is proposed. Due to the low utilization rate of covariance Toeplitz matrix information reconstructed by the traditional decoherent ESPRIT-LIKE algorithm and its poor performance at small angular intervals, this paper reconstructs a Toeplitz matrix to improve the utilization rate of antenna aperture element information on this basis, and then combines the idea of smoothing before and after to smoothly correct the reconstructed matrix, and realizes the estimation of two-dimensional angles through subspace class super-resolution algorithm. Experimental results show that the root mean square error and success rate of the improved ESPRIT-LIKE algorithm in the coherent environment are greater than those of the traditional algorithm, which greatly improves the performance of angle estimation in the coherent environment.

(2) Aiming at the problem of poor point cloud quality generated by traditional 3D-FFT point cloud imaging methods, a point cloud imaging method based on super-resolution angle measurement is proposed. Firstly, MIMO virtual aperture technology is introduced to increase the virtual aperture, increase the degree of antenna freedom, and increase the resolution ability of radar to target from the perspective of hardware and signal processing. Based on the echo storage form of MIMO system, the calculation of target three-dimensional information is analyzed from the level of data processing, and the improved ESPRIT-LIKE goniometry algorithm is applied to point cloud imaging, and a new point cloud imaging method is proposed. Finally, TI's AWR-1843 millimeter wave radar is used to carry out actual measurement experiments on the actual road, and according to the comparative analysis of the experimental results, the number and quality of point clouds generated based on the improved ESPRIT-LIKE goniometric algorithm are better, which can effectively restore the real-time three-dimensional contour of the moving target.

[1] 陈垦,杨洋,谭屈山,李伟,何俊,曹骥.基于毫米波雷达的高速公路三急一速自动检测算 法[J].中国交通信息化,2023(02):108-112.

[2] 张皓月,肖乃瑶,王涛.雷视融合设备在高速公路事件分析系统中的应用[J].中国交通 信息化,2022(S1):276-278.

[3] 姜浩浩,曲卫,童菲.近场雷达三维成像技术综述[J].兵器装备工程学报,2021,42(11):25- 30.

[4] Liu J, Zhang K, Sun Z, Wu Q, He W, Wang H. Concealed Object Detection and Recognition System Based on Millimeter Wave FMCW Radar[J]. Applied Sciences, 2021,11(19):8926.

[5] 秦志斌,王慧娟.智慧高速毫米波雷达应用浅析[J].中国交通信息化,2021(03):130-131.

[6] 杨帆. 毫米波雷达点云成像关键技术研究[D].哈尔滨:哈尔滨工业大学,2019.

[7] 李小路, 曾晶晶, 王皓等. 三维扫描激光雷达系统设计及实时成像技术[J].红外与激光工程, 2019: 1-7.

[8] 单文秋. 车载毫米波雷达前视成像系统研究[D].呼和浩特:内蒙古工业大学硕士学位论文, 2018.

[9] Visentin, Sagainov A, Hasch J, et al.Classification of Objects in Polarimetric RadarImages using CNNs at 77 GHz[C]. IEEE Asia Pacific Microwave Conference,2017:356-359.

[10] Tan B, Ma Z, Zhu X, et al. 3D Object Detection for Multi-frame 4D Automotive Millimeter-wave Radar Point Cloud[J]. IEEE Sensors Journal, 2022.

[11] Qian K, He Z, Zhang X. 3D point cloud generation with millimeter-wave radar[J].Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 2020, 4(4): 1-23.

[12] Li X, Wang X, Yang Q, et al. Signal processing for TDM MIMO FMCW millimeter-wave radar sensors[J]. IEEE Access, 2021, 9: 167959-167971.

[13] Kang S, Jang M, Lee S. Autoencoder-Based Target Detection in Automotive MIMO FMCW Radar System[J]. Sensors, 2022, 22(15): 5552.

[14] Zhang G, Geng X, Lin Y J. Comprehensive mpoint: A method for 3d point cloudgeneration of human bodies utilizing fmcw mimo mm-wave radar[J].Sensors, 2021,21(19): 6455.

[15] Chipengo U, Sligar A, Carpenter S. High fidelity physics simulation of 128 channel MIMO sensor for 77GHz automotive radar[J]. IEEE Access, 2020, 8: 160643-160652.

[16] Diab A, Kashef R, Shaker A. Deep Learning for LiDAR Point Cloud Classificationin Remote Sensing[J]. Sensors, 2022, 22(20): 7868.

[17] Lu M, Guo Y, Zhang J,et al. Automatic Markerless Registration Of Mobile LiDARPoint-Clouds[J]. IEEE Geoscience & Remote Sensing Symposium, 2014: 173-176.

[18] Knudde N, Vandersmissen B, Parashar K, et al.Indoor Tracking of Multiple Personswith a 77 GHz MIMO FMCW Radar[C].IEEE 2017 European Radar Conference (EURAD), 2017: 61-64.

[19] Jabłoński P, Iwaniec J, Jabłoński M. Multisensory Testing Framework for AdvancedDriver Assistant Systems Supported by High-Quality 3D Simulation[J].Sensors,2021,21(24):8458-8459.

[20] Li G, Zhou Q, Xu G, et al . Lidar-radar for underwater target detection using a modulated sub-nanosecond Q-switched laser[J]. Optics & Laser Technology,2021,142:107234.

[21] Xiao K, Qian J, Li T, Peng Y. Multispectral LiDAR Point Cloud Segmentation for Land Cover Leveraging Semantic Fusion in Deep Learning Network[J]. Remote Sensing,2022,15(1):243.

[22] Munir N,Awrangjeb M,Stantic B.Power Line Extraction and Reconstruction Methodsfrom Laser Scanning Data: A Literature Review[J]. Remote Sensing,2023,15(4):973-977.

[23] Hojatimalekshah A,Gongora J ,Enterkine J,Glenn N F, Caughlin T T, Marshall H P,

Hiemstra C A.Lidar and deep learning reveal forest structural controls on snowpack[J]. Frontiers in Ecology and the Environment,2023,21(1):49-54.

[24] Gao H, Cheng B, Wang J, et al. Object classification using CNN-based fusion of vision and LIDAR in autonomous vehicle environment[J]. IEEE Transactions on Industrial Informatics, 2018, 14(9): 4224-4231.

[25] Małgorzata M,Pradeep V W S, Lorenz H,Dario F. Sanchez, Nicola C, Federica M,Gu-dial B,Jürgen B,Helena V S. Cracks, porosity and microstructure of Ti modifiedpolymer-derived SiOC revealed by absorption, XRD- and XRF-contrast 2D and 3D

imaging[J]. Acta Materialia,2020,198:23-27.

[26] Yan B,Antonio A,Sara P,Aldo D C,Celine N,Jean-Jacques P,Laurent H. Aging effectin interface-engineered perovskite solar cells with 2D nanomaterials:A depth profileanalysis[J]. Materials Today Energy,2018,9:1-10.

[27] 安霸发布用于自动驾驶的集中式4D成像毫米波雷达架构[J].单片机与嵌入式系统应用,2023,23(02):96.

[28] Dr.Wang.4D成像毫米波雷达系统的发展与趋势[J].汽车与配件,2022(21):50-53.

[29] Wang Z,Zhang Y.Estimation of ship berthing parameter based on Multi-LiDAR andMMW radar data fusion[J]. Ocean Engineering,2022,266(P5).

[30] Xiao Y, Yin S ,Cui G, Yao L, Fang Z ,Zhang W. A Near-Field Area Object Detection Method for Intelligent Vehicles Based on Multi-Sensor Information Fusion[J]. World Electric Vehicle Journal,2022,13(9):160-166.

[31] Liu H, Pan W,Hu Y, Li C,Yuan X, Long T.A Detection and Tracking Method Basedon Heterogeneous Multi-Sensor Fusion for Unmanned Mining Trucks[J]. Sensors,2022,22(16):5989.

[32] Sun R, Suzuki K, Owada Y, Takeda S, Umehira asahiro, Wang X, Kuroda H. A millimeter-wave automotive radar with high angular resolution for identification of closely spaced on-road obstacles.[J]. Scientific reports,2023,13(1):3233-3237.

[33] Kim Eunhee,Kim Ilkyu, Kim Wansik. Non-Uniform MIMO Array Design for RadarSystems Using Multi-Channel Transceivers[J]. Remote Sensing,2022,15(1):78-91.

[34] Shu F, Qin Y, Liu T, et al.Low-complexity and high-resolution DOA estimation forhybrid analog and digital massive MIMO receive array [J]. IEEE Transactions on Communications, 2018, 66(6): 2487-2501.

[35] 晋良念,朱皓,王燃.一种毫米波雷达快速高分辨三维点云成像方法[J].仪器仪表学报,2022,43(12):158-167.

[36] 晋良念,王燃.毫米波FMCW MIMO雷达三维点云成像方法[J].雷达科学与技术,2022,20(05):497-506.

[37] 黄李波, 徐嘉. 车用成像毫米波雷达点云目标分类[J].电子技术与软件工程,2021(11):100-103.

[38] 封泽运, 张慧,林凤泰等. 车载MIMO毫米波雷达高分辨点云成像[C]//中国电子学会数字信号处理专家委员会. 第十四届全国DSP应用技术学术会议论文集, 2022: 76-82

[39] Vallet P, Mestre X, Loubaton P.Performance analysis of an improved MUSIC DOA estimator[J]. IEEE Transactions on Signal Processing, 2015, 63(23): 6407-6422.

[40] Wang G,Zheng G,Wang Hn,Chen C.Meter Wave Polarization-Sensitive Array Radar for Height Measurement Based on MUSIC Algorithm[J]. Sensors,2022,22(19):7298-7280.

[41] Bilal A,Cho C S.Localization of Dielectric Anomalies with Multi-Monostatic Using 2D MUSIC Algorithm with Spatial Smoothing[J]. Sensors,2022,22(14):6872-6880.

[42] Li J, Jiang D, Zhang X. DOA estimation based on combined unitary ESPRIT for coprime MIMO radar[J]. IEEE Communications Letters, 2016, 21(1): 96-99.

[43] 朱国辉,汪洋,米胜男.加权最小二乘法和MUSIC法联合测向技术研究[J].雷达科学与技术,2019,17(03):319-323+334.

[44] Zhang Y, Ng B P. MUSIC-like DOA estimation without estimating the number of sources[J]. IEEE Transactions on Signal Processing, 2009, 58(3): 1668-1676.

[45] Sun F, Gao B, Chen L, et al . A low-complexity ESPRIT-based DOA estimation method for co-prime linear arrays[J]. Sensors, 2016, 16(9): 1367.

[46] 康晓涛,晁文兰,王昊,石要武.均匀线阵双基地MIMO雷达目标角度估计[J].吉林大学学报(工学版),2014,44(05):1460-1465.

[47] Wang R, Wang Y,Li Y,Cao W,Yan Y. Geometric Algebra-Based ESPRIT Algorithm for DOA Estimation[J]. Sensors,2021,21(17):5933.

[48] Du R, Wang J, Liu F. Unitary-ESPRIT Algorithm for Carrier Frequency Offset Estimation for Interleaved OFDMA Uplink Systems[J]. Wireless personal communications, 2013, 69(4).

[49] Liu D, Zhao Y, Cao C, Pang X . A novel reduced-dimensional beamspace unitary ESPRIT algorithm for monostatic MIMO radar[J]. Digital Signal Processing,2021,114:103027.

[50] 窦慧晶,梁霄,张文倩.基于压缩感知理论的二维DOA估计[J].北京工业大学学报,2021,47(03):231-238.

[51] Zhang Y,Zhang LY,Han J F,Ban Z,Yang Y. A new DOA estimation algorithm basedon compressed sensing[J]. Cluster Computing,2019,22(1):895-903.

[52] 石要武, 陈淼, 单泽涛, 等 .基于特征空间MUSIC算法的相干信号波达方向空间平滑估计[J]. 吉林大学学报: 工学版, 2017, 47(1): 268-273.

[53] Gao Y, Chang W, Pei Z, et al. An improved music algorithm for doa estimation of coherent signals[J]. Sensors and Transducers, 2014, 175(7): 75-82.

[54] Gong J, Guo Y, Wan Q. An improved spatial difference smoothing method based on multistage wiener filtering[J]. Mathematical Problems in Engineering, 2019, 2019: 1-8.

[55] Qi B, Liu D. An enhanced spatial smoothing algorithm for coherent signals DOA estimation[J]. Engineering Computations, 2022, 39(2): 574-586.

[56] Zhagypar R, Zhagyparova K, Akhtar M T. Spatially smoothed TF-root-MUSIC for DOA estimation of coherent and non-stationary sources under noisy conditions [J]. IEEE Access, 2021, 9: 95754-95766.

[57] Zhang W, Han Y, Jin M, et al. Multiple-Toeplitz matrices reconstruction algorithm for DOA estimation of coherent signals[J]. IEEE Access, 2019, 7: 49504-49512.

[58] Zhang D, Zhang Y, Zheng G, et al. ESPRIT-like two-dimensional DOA estimation for monostatic MIMO radar with electromagnetic vector received sensors under the condition of gain and phase uncertainties and mutual coupling[J]. Sensors, 2017, 17(11): 2457.

[59] Chen H, Hou C, Zhu W , et al. ESPRIT-like two-dimensional direction finding formixed circular and strictly noncircular sources based on joint diagonalization[J].Signal Processing, 2017, 141: 48-56.

[60] 张薇. 基于矩阵重构的相干信源波达方向估计算法研究[D].哈尔滨:哈尔滨工业大学,2021.

[61] 王凌,李国林,刘坚强,毛维平.一种基于数据矩阵重构的相干信源二维测向新方法[J].西安电子科技大学学报,2013,40(02):130-137.

[62] Zhang W, Han Y, Jin M, Li X.An improved ESPRIT-like algorithm for coherent signals DOA estimation[J]. IEEE Communications Letters, 2020,24(2): 339-343.

[63] Chalabi I. Application of CFAR detection to multiple pulses for gamma distributedclutter[J]. Remote Sensing Letters,2022,13(10):678-688.