Remedying the effects of array shape distortion on the spatialfiltering of acoustic data from a line array of hydrophones

The effects of both small perturbations and large deformations to the array's shape on both conventional and adaptive beamformers are shown for two frequencies: the spatial Nyquist frequency (or design frequency) of the array and a frequency about three times greater. Large shape deformations lead to a decrease in the conventional beamformer's output power for a beam steered in the direction of the signal source, together with an increase in the sidelobe levels (or secondary maxima), while small perturbations in the array shape have little effect. Signal suppression is observed to be far greater for the adaptive beamformer because it is very sensitive to system errors. The imposition of a weight norm constraint on the adaptive beamformer reduces the signal suppression only for small shape perturbations array shape estimation techniques are needed to reduce signal suppression for large shape deformations. The adverse effects of a nonlinear array shape on both conventional and adaptive beamforming are shown to be substantially reduced by applying techniques that estimate the coordinates of the hydrophones prior to beamforming     

基于RBF神经网络的多系统钟差预报算法

针对海上条件下,对于实时定位应用,实时数据流无法下载的情况,文中提出一种基于RBF神经网络的卫星钟差预报算法,给出基函数的中心、方差以及隐含层到输出层的权值的计算方法,采用滑动窗口的方法,用样本数据训练后的网络预测下一个历元的钟差值,依次往后训练网络直到预测完整个时间段,通过实验验证了算法的可用性。短期预报中,GPS预报精度在1 ns以下,BDS和GLONASS在2～3 ns左右;长期预报中,GPS预报精度在几十纳秒左右,而BDS和GLONASS在几百纳秒左右,文中给出了相应的结果分析。     

基于权重调整的BP神经网络在Nino区海温预报中的应用

传统BP神经网络在训练完之后,其权重是固定不变的,加上神经网络的样本的标准化处理,将使得网络不易描绘样本峰值.因此,本文考虑变权的方法,以调节训练后的BP网络权重,基于变权次数,建立不同网络模型,并利用不同网络输出值与相应实测值进行比较.结果表明:变权BP网络预报效果有较大提升,同时,降低了对因子相关性的要求.     

Adaptive beamforming of a towed array during a turn

During maneuvering, towed array beamforming degrades if a straight array is assumed. This is especially true for high-resolution adaptive beamforming. It is experimentally demonstrated that adaptive beamforming is feasible on a turning array, provided that array shape is estimated. The array shape can be inferred solely from the coordinates of the tow vessel's Global Positioning System (GPS) without any instrumentation in the array. Based on estimated array shape from the GPS, both the conventional beamformer and the white noise constrained (WNC) adaptive beamformer are shown to track the source well during a turn. When calculating the weight vector in the WNC approach, a matrix inversion of the cross-spectral density matrix is involved. This matrix inversion can be stabilized by averaging the cross-spectral density matrix over neighboring frequencies. The proposed algorithms have been tested on real data with the tow-vessel making 45/spl deg/ turns with a 500-m curvature radius. While turning, the improvement in performance over the assumption of a straight array geometry was up to 5 dB for the conventional beamformer and considerably larger for the WNC adaptive beamformer.     

Neural adaptive sensory processing for undersea sonar

Neural adaptive beamformers (NABFs) utilize neural paradigms to accomplish desired adaptations that are associated with sensory-field-responsive partitioning and selection processes. Kohonen-type organization and Hopfield-type optimization have been formulated as NABF mechanisms and have been applied to test data. Formulations and results are included. NABFs are also used in conjunction with a learning network for interpretation of weight sets as population codings of direction. An example is included. Desirable qualities of human auditory response are being interpreted in the context of neural adaptive beamforming for the purpose of creating an integrated processing structure that incorporates NABFs, a cochlear model, and an associative memory as part of a total spatiotemporal processing scheme for selective attention     

基于粒子群算法优化的BP神经网络在海水水质评价中的应用

针对目前存在的海水水质受多因素影响、评价难的现状,提出了一种基于粒子群算法（PSO）优化误差反向传播（BP）神经网络的海水水质评价模型。该模型通过PSO得到BP神经网络最优的权值和阈值,结合青岛东部海域10个监测站点的数据得到水质评价结果。实验证明,该模型和单因子评价、传统的BP神经网络评价相比较,具有训练时间短、预测精度高的特点,在海水水质评价中具有良好的应用价值。     

基于GA-BP神经网络的临洪河口湿地土地覆盖分类算法研究

针对传统分类方法易受到"同物异谱"和"同谱异物"影响，致使河口湿地覆盖分类精度较低的问题，提出一种基于遗传算法优化BP神经网络分类算法。以江苏省临洪河口湿地为研究区，选用哨兵Sentinel-2影像，经辐射校正、大气校正和图像裁剪等预处理后，构建基于自适应遗传算法优化的BP神经网络算法开展临洪河口湿地土地覆盖分类研究，并与传统BP神经网络、支持向量机和随机森林算法进行精度比较。研究结果表明：遗传算法优化后的BP神经网络算法开展河口湿地土地覆盖分类的总精度为96.162 7%，Kappa系数为0.952 0；与传统BP神经网络、支持向量机和随机森林分类算法的分类总精度相比，分别提高了7.359 7%、11.677 9%和6.042 4%；对应的Kappa系数也相应提高了0.090 8、0.118 0和0.074 8；有效解决了河口湿地土地覆盖分类精度低的问题。遗传算法优化后的BP神经网络可实现河口湿地土地覆盖的高精度分类，促进湿地资源的合理开发和保护，为实现海洋生态文明建设提供技术支撑。     

基于头脑风暴优化算法与BP神经网络的海水水质评价模型研究

针对基于传统BP神经网络的海水水质评价模型存在易陷入局部极小等问题,提出了一种新的利用头脑风暴优化算法(BSO)优化BP神经网络的海水水质评价模型(BSO-BP)。该模型引入具有全局寻优特点的头脑风暴优化算法,用于模拟人类提出创造性思维解决问题的过程,具有强大的全局搜索和局部搜索的能力,同时利用BP神经网络所具有良好的非线性映射能力、学习适应能力和容错性,最大程度上考虑到海洋水质评价因素的非线性和非平稳的关系,得到BP神经网络的各层权值、阈值的最优解,使得海水水质评价结果准确合理。并以胶州湾海域的12个监测站位的监测数据作为评价样本进行水质评价,实验结果表明该评价模型能够克服局部极小问题,评价结果准确性较高,并具有一定的实用性。     

深度学习的BP神经网络在GNSS水准拟合中的应用

针对误差逆向传播BP (back propagation)神经网络在GNSS水准拟合中存在梯度消失、陷于局部最小点的问题,通过使用深度学习中的分段线性整流函数Relu(rectified linear units)作为神经元激活函数,自适应矩估计Adam (adaptive moment estimation)算法作为网络优化函数,提出了一种基于深度学习的BP神经网络模型。研究结果表明:改进后的BP神经网络内外符合精度分别提高近50%和25%,可达0.9 cm和2.4 cm,为GNSS水准拟合提供了新的思路。     

Adaptive separation of mixed broadband sound sources with delays bya beamforming Herault-Jutten network

The Herault-Jutten network has been used to separate independent sound sources that have been linearly mixed. The problem of separating a mixture of several independent signals in free-field conditions or a signal and echoes in confined spaces is compounded by propagation time delays between the source(s) and the microphones because the conventional Herault-Jutten network cannot tolerate time delays. In this paper, we combine a symmetrically balanced beamforming array with the conventional Herault-Jutten network. The resulting system can adaptively separate signals that include delays introduced by the propagation medium. The proposed algorithm has been simulated in digital communication multipath channels where intersymbol interference exists. The simulation results show two clear advantages of the proposed method over the conventional adaptive equalization: (1) there is no penalty for very long impulse responses caused by long delays, and (2) no training signals are needed for equalization. The design of a multibeamformer to handle the source separation of multiple broad-band signals is also presented     

时空联合估计权向量的MVDR自适应波束形成方法

为实现短数据条件下权向量的稳定优化估计,提出1种时空联合估计权向量的MVDR自适应波束形成方法。该方法结合时域解析信号的MVDR自适应波束形成算法中构造时域解析信号的方法和直接数据域算法中空间滑动的方法,以减少一半权向量为代价,实现了在更短的数据长度下稳定优化地估计协方差矩阵和权向量。数值仿真实验和海上实验数据处理结果表明:与常规波束形成和直接数据域相比,该方法具有更好的稳定性和更好的波束性能,即主瓣更窄,旁瓣更低,阵增益更高。     

Improved time-delay estimates of underwater acoustic signals usingbeamforming and prefiltering techniques

Passive sonar systems that localize broadband sources of acoustic energy estimate the difference in arrival times (or time delays) of an acoustic wavefront at spatially separated hydrophones, The output amplitudes from a given pair of hydrophones are cross-correlated, and an estimate of the time delay is given by the time lag that maximizes the cross correlation function. Often the time-delay estimates are corrupted by the presence of noise. By replacing each of the omnidirectional hydrophones with an array of hydrophones, and then cross-correlating the beamformed outputs of the arrays, the author shows that the effect of noise on the time-delay estimation process is reduced greatly. Both conventional and adaptive beamforming methods are implemented in the frequency domain and the advantages of array beamforming (prior to cross-correlation) are highlighted using both simulated and real noise-field data. Further improvement in the performance of the broadband cross-correlation processor occurs when various prefiltering algorithms are invoked     

Application of probabilistic neural network to design breakwater armor blocks

This study presents a probabilistic neural network (PNN) technique for predicting the stability number of armor blocks of breakwaters. The PNN is prepared using the experimental data of Van der Meer. The predicted stability numbers of the PNN are compared with those of previous studies, i.e. by an empirical formula and a previous neural network model. The agreement index between the measured and predicted stability numbers by PNN are better than those by the previous studies. The PNN offers a way to interpret the network's structure in the form of a probability density function and it is easy to implement. Therefore, it can be an effective tool for designers of rubble mound breakwaters.     

基于粒子群优化神经网络的水深反演模型

针对直接采用BP神经网络反演水深收敛速度慢,且易陷入局部最优的问题,提出了一种基于粒子群(PSO)优化BP神经网络的水深遥感新模型。该模型首先利用粒子群算法对BP神经网络的权重和阈值进行优化,然后将该优化值作为BP神经网络的初始值,最后再将PSO优化后的模型用于测试海区的反演精度评估。实验结果表明,该模型的网络收敛速度明显加快,水深反演的精度也得到提高。     

Robust LS-SVM regression for ore grade estimation in a seafloor hydrothermal sulphide deposit

Due to the geological complexities of ore body formation and limited borehole sampling, this paper proposes a robust weighted least square support vectormachine (LS-SVM) regression model to solve the ore grade estimation for a seafloor hydrothermal sulphide deposit in Solwara 1, which consists of a large proportion of incomplete samples without ore types and grade values. The standard LS-SVM classification model is applied to identify the ore type for each in complete sample. Then, a weighted K-nearest neighbor (WKNN) algorithm is proposed to interpolate the missing values. Prior to modeling, the particle swarm optimization (PSO) algorithm is used to obtain an appropriate splitting for the training and test data sets so as to eliminate the large discrepancies caused by randomdivision. Coupled simulated annealing (CSA) and grid search using 10-fold cross validation techniques are adopted to determine the optimal tuning parameters in the LS-SVM models. The effectiveness of the proposed model by comparing with other well-known techniques such as inverse distance weight (IDW), ordinary kriging (OK), and back propagation (BP) neural network is demonstrated. The experimental results show that the robust weighted LS-SVM outperforms the othermethods, and has strong predictive and generalization ability.     

双通道实时数据采集处理系统的设计与实现

本文提出了一种基于 TMS32 0 C31串行接口的双通道实时数据采集处理系统的设计与实现方案 ,该设计以 TMS32 0 C31和 TL C32 0 AD5 0 C为核心器件 ,具有两个独立的 A/D,D/A通道 ,能够实现 32位浮点计算和 16位数据采集与回放。应用该系统进行归一化最小均方误差 (Normal-ized L east Mean Square,NLMS)算法实时自适应噪声抵消实验 ,实验结果表明 ,该系统能够实现实时的自适应噪声抵消 ,可广泛应用于实时语音信号处理等领域。     

Prediction of breaking waves with neural networks

The height of a wave at the time of its breaking, as well as the depth of water in which it breaks, are the two basic parameters that are required as input in design exercises involving wave breaking. Currently the designers obtain these values with the help of graphical procedures and empirical equations. An alternative to this in the form of a neural network is presented in this paper. The networks were trained by combining the existing deterministic relations with a random component. The trained network was validated with the help of fresh laboratory observations. The validation results confirmed usefulness of the neural network approach for this application. The predicted breaking height and water depth were more accurate than those obtained traditionally through empirical schemes. Introduction of a random component in network training was found to yield better forecasts in some validation cases.     

Development of wavelet network model for accurate water levels prediction with meteorological effects

Accurate water levels modeling and prediction is essential for safety of coastal navigation and other maritime applications. Water levels modeling and prediction is traditionally developed using the least-squares-based harmonic analysis method that estimates the harmonic constituents from the measured water levels. If long water level measurements are not obtained from the tide gauge, accurate water levels prediction cannot be estimated. To overcome the above limitations, the current state-of-the-art artificial neural network has recently been developed for water levels prediction from short water level measurements. However, a highly nonlinear and efficient wavelet network model is proposed and developed in this paper for water levels modeling and prediction using short water level measurements. Water level measurements (about one month and a week) from six different tide gauges are employed to develop the proposed model and investigate the atmospheric changes effect. It is shown that the majority of error values, the differences between water level measurements and the modeled and predicted values, fall within the −5 cm and +5 cm range and root-mean-squared (RMS) errors fall within 1–6 cm range. A comparison between the developed highly nonlinear wavelet network model and the harmonic analysis method and the artificial neural networks shows that the RMS of the developed wavelet network model when compared with the RMS of the harmonic analysis method is reduced by about 70% and when compared with the RMS of the artificial neural networks is reduced by about 22%. It is also worth noting that if the atmospheric changes effect (meteorological effect) of the air pressure, the air temperature, the relative humidity, wind speed and wind direction are considered, the performance accuracy of the developed wavelet network model is improved by about 20% (based on the estimated RMS values).     

On the ability to estimate narrow-band signal parameters usingtowed arrays

Using the Cramer-Rao lower bound (CRLB) as an indicator of potential performance, the limits on the estimation and resolution capabilities of a towed line array of uniformly spaced hydrophones to provide frequency and bearing information about narrowband signals are examined. It is assumed that a monochromatic plane wave arrives at the array for each source. Several versions of the bounds are computed using different assumptions about which parameters have known values and about the way in which the samples are taken in space and in time. It is shown that the CRLB values for different situations can be compared to provide information about the effective use of a moving aperture for estimation of the parameters of narrowband signals arriving at the array. It is also shown that adding at least one hydrophone occupying a fixed position in space can improve the bearing estimates of a towed array by supplying additional frequency information if both the bearings and frequencies of the sources are unknown     

Adaptive port-starboard beamforming of triplet sonar arrays

For a low-frequency active sonar (LFAS) with a triplet receiver array, it is not clear in advance which signal processing techniques optimize its performance. Here, several advanced beamformers are analyzed theoretically, and the results are compared to experimental data obtained in sea trials. Triplet arrays are single line arrays with three hydrophones on a circular section of the array. The triplet structure provides the ability to solve the notorious port-starboard (PS) ambiguity problem of ordinary single-array receivers. More importantly, the PS rejection can be so strong that it allows to unmask targets in the presence of strong coastal reverberation or traffic noise. The theoretical and experimental performance of triplet array beamformers is determined in terms of two performance indicators: array gain and PS rejection. Results are obtained under several typical acoustic environments: sea noise, flow noise, coastal reverberation, and mixtures of these. A new algorithm for (beam space) adaptive triplet beamforming is implemented and tuned. Its results are compared to those of other triplet beamforming techniques (optimum and cardioid beamforming). These beamformers optimize for only one performance indicator, whereas in theory, the adaptive beamformer gives the best overall performance (in any given environment). The different beamformers are applied to data obtained with an LFAS at sea. Analysis shows that adaptive triplet beamforming outperforms conventional beamforming algorithms. Adaptive triplet beamforming provides strong PS rejection, allowing the unmasking of targets in the presence of strong directional reverberation (e.g., from a coast) and at the same time provides positive array gain in most environments.