随机阵列的时反聚焦方法研究
文献类型：学位论文
| 作者 | 马登永 |
| 学位类别 | 博士 |
| 答辩日期 | 2009-05-27 |
| 授予单位 | 中国科学院声学研究所 |
| 授予地点 | 声学研究所 |
| 关键词 | 随机阵列 时间反转 迭代预滤波 最小二乘 逆滤波聚焦 |
| 其他题名 | Acoustic field focusing of a random array using time-reversal method |
| 学位专业 | 信号与信息处理 |
| 中文摘要 | 时反方法在最近几十年内已经得到了迅速发展，并且被广泛应用于医学上的超声成像和治疗、水下声学、非破坏性检测、房间声学、无线电通信等领域中。由于有限的阵列孔径，常规时反方法在空时域的聚焦性能并不理想，为了解决该问题，许多学者提出了改进的最优时反方法，以改善空时域的聚焦效果。在实际应用中，这些方法都需要进行多次的收发操作，实现麻烦费时、难于实时处理。另外，现有时反方法的声场聚焦实验，都是基于线阵列或弧阵列进行的，对于随机阵列的时反声聚焦实验还没有进行过深入研究。 针对这些问题，本文进行了深入研究，寻求解决问题的有效方法，本文的主要创新点包括以下三个方面： （1） 基于NR-TR方法的物理模型，提出了两种随机阵列的NR-TR聚焦方法：SCNR-TR和MCNR-TR方法，这两种方法都能够在阵元频响存在差异，阵元位置随机布放的情况下，实现期望位置点的声场聚焦。SCNR-TR方法实现简单快速，但聚焦信号中包含阵元之间的交叉干扰分量；MCNR-TR方法通过引入阵元独立发射和同步校准机制，消除了交叉干扰分量的影响。实验结果，表明了两种方法对随机阵列NR-TR聚焦的有效性。 （2） 提出了基于ILS-PF的随机阵列最优NR-TR聚焦方法，以解决SCNR-TR方法时域聚焦信号的拖尾现象和波形展宽问题。该方法在保持SCNR-TR方法空域分辨率的同时，利用迭代预滤波器去均衡时反操作算子的时域响应，以达到逆滤波聚焦的效果，避免了传统聚焦方法因空域均衡引入的重复收发操作。在实际应用中，发射阵列只需要对时反信号进行迭代预滤波处理后再重新发射，就可以在聚焦点位置恢复出期望的聚焦信号波形，实现简单快速。仿真和实验结果表明，所提出方法能够在时域上取得逆滤波聚焦效果，同时保持了较高的空域分辨率。 （3） 针对主动时反方法同样存在的时域聚焦缺陷，提出了基于ILS-PF的随机阵列最优主动时反聚焦方法，即将ILS-PF方法和随机阵列结构引入到主动时反聚焦，利用迭代预滤波器对源信号进行预滤波处理，再按照常规主动时反方法的操作步骤进行声场聚焦，能够在期望位置取得时域逆滤波聚焦效果，实现简单快速，便于实时处理。仿真和实验结果表明，该方法能够取得时域逆滤波聚焦效果，同时其空域分辨率比常规主动时分方法有所提高。 这些理论和实验结果，对于时反方法在传感器网络及声通信领域的应用研究，具有一定的指导意义。 |
| 英文摘要 | In recent decades, the time reversal methods have undergone rapid development and have been widely applied in various fields including the ultrasound based medical imaging and therapy, underwater acoustics, nondestructive testings, room acoustics, and radio communications. Owing to the limited array aperture, the conventional time reversal methods are lack of satisfactory spatial-temporal focusing performance. Many authors have proposed optimal time reversal (TR) methods to improve the focusing quality. However, when applied in practice, all of these methods require repetitive execution of transmit-receive operations which are not only time consuming but are also difficult to implement in real-time processing. Furthermore, the acoustic TR focusing experiments are usually conducted using a single linear or concave array, the TR experiments of using single random array have not yet been studied in depth. This thesis presents a number of effective methods to address these issues. The main novel contributions from this study are summarized as follows: 1) On the base of the Nonreciprocal-Time Reversal (NR-TR) method, Single Channel Nonreciprocal-Time Reversal (SCNR-TR) and Multiple Channels Nonreciprocal- Time Reversal (MCNR-TR) methods are proposed. Using both of the methods, robust acoustic field focusing at the predefined spot can be acquired from a random array with the existence of difference among array element response. The implementation of SCNR-TR method is simple and fast but the focusing signal contains the cross talk between the individual array elements. MCNR-TR method is developed to cancel the cross talk by introducing alternative transmission among the array elements and provides a correction procedure for the synchronization part in order to cancel the cross talk interference. The experimental results have demonstrated that these methods can be used to realize the NR-TR focusing of the random array effectively. 2) Based on the Iterative Least Squares-Prefiltering (ILS-PF) process, the optimal NR-TR focusing method of a single random array is presented which eliminates the broadening and the tailing phenomena associated with the temporal focusing signal. This method employs the iterative pre-filters to equalize the temporal responses of the Time Reversal Operator (TRO) to achieve the temporal inverse filtering focusing while retaining the high spatial resolution. The advantage of this method is that it bypasses the repetitious transmit-receive operations resulting from the spatial response equalization of TRO present in the conventional optimal focusing methods. In practical applications, the expected waveform of the focusing signal can be recovered easily and quickly at the focusing spot which re-emits the time reversed signal pre-filtered iteratively. The methods are validated through simulation and experimental results. 3) The optimal active time reversal (A-TR) focusing method of a single random array is developed in order to address the disadvantages associated with the temporal focusing technique. The new method incorporates the ILS-PF method and the structure of a single random array into the conventional A-TR method. The method acquires the inverse filtering focusing at the expected position through the emission of the iteratively prefiltered source signal followed by the execution of TR operation according to the procedures of conventional A-TR method. The results from the simulation and experiments have shown that this method can be used to obtain the temporal inverse filtering focusing with somewhat higher spatial resolution compared to using the conventional A-TR method. These theories and experimental results are of significant importance for the research works on the application of time reversal methods to sensor networks and acoustic communication. |
| 语种 | 中文 |
| 公开日期 | 2011-05-07 |
| 页码 | 127 |
| 源URL | [http://159.226.59.140/handle/311008/470]  |
| 专题 | 声学研究所_声学所博硕士学位论文_1981-2009博硕士学位论文 |
| 推荐引用方式 GB/T 7714 | 马登永. 随机阵列的时反聚焦方法研究[D]. 声学研究所. 中国科学院声学研究所. 2009. |
入库方式： OAI收割
来源：声学研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。