Features of underwater echo extraction based on the stationary wavelet transform and singular value decomposition

A feature extraction method of underwater echo is proposed, which uses the redundancy property of SWT (the Stationary Wavelet Transform) and the steady property of SVD (Singular Value Decomposition). Since using singular values of SWT coefficients matrix as feature vectors, with the K-L transform to multi-subspace signal obtained from SWT, it is a feature compression method essentially. In contrast to the method of sub-bands energy feature based on discrete orthogonal wavelet transform (DWT), better results of lake trial data are acquired with this method: (1) Under the same sample and distance within-class, distance between-class is larger than former; (2) Correct recognition rates are also higher than former, whether the training and testing samples are chosen from the same lake trail or not; (3) Sample sets varying, the variation range of correct recognition rates is far less than the former. Thus this method can obtain more robust, effective features and better correct recognition results.     

天坛圜丘台声学现象的新发现及其解释

１９５３年汤定元院士在科学通报上撰文对天坛声学现象进行了研究，成为我国教科书和各类科普读物关于天坛声学现象的传统解释。我们的研究成果表明传统解释有些与现象相悖。通过现场测试、分析和计算对天坛圜丘台声学现象作新的解释。     

Single—Shot Magnetic Resonance Imaging of a Moving Object Based on the High—Speed Spiral—Scan Echo Planner Technique at 4.7T

The single-shot spiral magnetic resonance imaging (MRI) technique was implemented and optimized on a Bruker Biospec 47/30 scanner.The technique includes the pulse sequence for generation and detection of the k-space MRI signal (free induction decay (FID)),and PC-based programs for the data grid and image reconstruction.The temporal resolution is 70ms (14images per second),which consists of a data acquisition time as short as 13.7ms,a spin-echo time of 13.6ms and a magnetization recovery time of 43ms.This makes it possible to take real-time images of moving objects.The technique is demonstrated using a pendulum (tube) filled with water.     

CS-100型回声测深仪通过鉴定

船舶总公司第七研究院七二六研究所开发的多功能、高精度、全国产化的CS-100型回声测深仪,于1990年12月17—19日由船舶总公司第七研究院主持了鉴定。来自全国有关的科研、使用等27个单位的50多名专家和技术人员代表参加了鉴定会。 CS-100型回声测深仪的主要技术性能为: 工作频率:218kHz。测深范围:0.5—100m。测量精度:记录式:±0.5％±0.05m,数字式:±0.5％±0.1m。声速修正:1400—1560m/s。浅水报警:0—9.9m,可自行设定,声、光同时报警,并可自动消除声报警。工作电源:直流24V±(10—15)％,1.5A。     

储备池状态空间重构与混沌时间序列预测

分析了现有的基于回声状态网络(ESN)的迭代预测方法，指出了该方法在理论上存在的问题以及应用中存在的障碍.提出了一种基于储备池的直接预测方法，该方法利用预测原点和预测时域之间的关系直接构建预测器，因此可以预先对预测器的稳定性施加约束，从而避免了在迭代预测方法中由于网络回路闭合而产生的稳定性问题.在仿真中，首先以Lorenz时间序列为例分析了迭代预测方法在闭合回路前后储备池的变化情况，然后通过Mackey-Glass标杆问题的测试验证了直接预测方法的可行性.     

幼年蝙蝠与母亲之间的超声沟通

众所周知,蝙蝠善于利用超声回声定位来探测环境和捕捉猎物,但鲜为人知的是幼年蝙蝠与成年蝙蝠之间也可以通过超声进行沟通。近期,日本科学家Shizuko Hiryua和Hiroshi Riquimaroux实验研究了一种叫作Pipistrellus Abramus的蝙蝠,通过观     

水声无源材料回声降低测量时反聚焦方法

提出了利用时间反转(时反)聚焦技术的水声无源材料回声降低测量方法。首先实现有界水声环境下有、无试样接收信号的时反聚焦,然后利用聚焦信号测量试样反射系数,最后通过标准试样对反射系数进行修正,从而得到试样的回声降低测量值。在小型水池中进行了铝板试样和钢板试样回声降低的测量,铝板试样的尺寸为1.1m×1.0×0.005m,测量频率范围为3~20kHz,钢板试样尺寸与铝板试样相同,测量频率范围为0.5~20kHz,测量结果经修正后与平面波理论计算值基本一致,和理论计算值的相对误差小于10%,扩展不确定度小于1.5dB。本方法采用时反原理实现了接收信号的空时聚焦,提高了测量信混比,因此适用于非自由场环境下无源材料回声降低的测量试验,尤其适用于低频的回声降低测量。      

心理声学模型及其在MP3编码中的应用

心理声学模型是音频感知编码的核心部分,其直接影响着音频编码的质量及压缩比.基于心理声学的基本原理、绝对听觉阈值、掩蔽效应及临界频带等相关内容,并结合心理声学的数学表达,按照MP3标准编码程序中的各个模块来详细分析算法流程.最后通过相应算法来描述预回声产生机理及其在MP3编码中的抑制方法.     

一种基于误差补偿的多元混沌时间序列混合预测模型

针对多元混沌时间序列的预测问题, 考虑到单纯改进储备池算法无法明显地提高预测精度, 提出一种基于误差补偿的时间序列混合预测模型. 实际观测的数据既包含线性特征又包含非线性特征. 首先利用自回归移动平均模型预测线性特征, 使得残差数据仅含非线性特征; 然后, 建立正则化回声状态网络模型预测; 最后, 将非线性部分的预测值与线性部分的预测值相加, 以实现高精度的多元混沌时间序列预测. 基于Lorenz和太阳黑子-黄河径流量时间序列的仿真实验验证了本文所提模型的有效性. 关键词： 回声状态网络 混沌 多元时间序列预测 误差补偿     

Low-frequency features of the ultrasound echo from an adhesively bonded layer-substrate structure

The low-frequency features of the ultrasound reflection spectra from the structure of a single layer on a substrate bonded by a thin adhesive layer are theoretically studied; the low-frequency here means the frequency of the interrogating ultrasonic wave is less than the quart-wavelength resonance frequency of the adhesive layer. The possibility of the inversion of the thickness and the evaluation of the cohesion strength of the adhesive layer from the resonance frequency shifts of the layered system is indicated. An analytic solution to the nonlinear equation satisfied by the resonance frequency is presented by Taylor expansion method showing satisfactory agreement with the numerical results by Newton iterative method. The results indicate larger range for application than the traditional spring model for the thin adhesive layer. In a much lower frequency range the thin adhesive layer may be regarded to be a spring.