共查询到19条相似文献，搜索用时 69 毫秒
1.
应用经验模式分解（EMD）将恒电量瞬态响应信号分解为不同时间尺度的内在模函数（IMF）分量。去除其中的小时间尺度的干扰噪声分量。然后经过拉普拉斯变换获得恒电量频谱以研究电化学腐蚀过程。 相似文献
2.
针对快速滤波分解信号为本征模态函数(IMF)所产生的边界效应问题,提出了一种抑制这种边界效应的方法。即利用快速滤波先将原信号分解为本征模态函数,然后在信号内部截取适当的两段分别延拓到原信号两端,经快速滤波得到分解结果后,再截去延拓的部分,保留原信号长度的分解结果。通过实例验证了该方法的有效性。 相似文献
3.
异常事件对EMD方法的影响及其解决方法研究 总被引：4，自引：1，他引：4
赵进平 《中国海洋大学学报(自然科学版)》2001,31(6):805-814
作者指出异常事件在数据中形成局部的高频信号 ,运用经验模态分解 (EMD)方法分析这种存在异常事件干扰的数据 ,就会产生本征模函数 (IMF)的频率混叠现象 ,而造成物理过程的重叠 ,使得难以用时间过程曲线表现特定的物理过程。这一问题是 EMD方法中尚未妥善解决的问题。为解决这一问题 ,作者利用干扰信号极值及其两边的极大与极小值位置与原始数据有明显对应关系的特征 ,将相关 IMF中的异常信息直接滤除 ,再用 Spline插值方法弥补滤除时段的数据 ,得到重新拟合的该 IMF数据。采用这种方法可以提取出异常信号 ,提取的精度与异常信号的时段长度有关。而且 ,拟合结果消除了异常干扰 ,可以将该 IMF与其余 IMF一起叠加成没有异常干扰的数据。将滤除了异常干扰的数据再次进行 EMD分解 ,可以得到新的 IMF系列 ,而它与不加校正的分解结果有相当大的差别 ,可靠地反映了真实物理过程。结果表明 ,只有在有效滤除异常干扰的情况下才能获得可靠的 IMF系列 ,并准确地描述各种尺度的现象 ;消除了异常干扰的 IMF可以任意单独或组合使用 ,表现各种时间尺度的变化与过程 ;所讨论的方法只适合异常时段较小的情形。对于异常时段接近或大于正常变化周期的干扰还需要探讨其他方法 相似文献
4.
基于粗糙集与人工神经网络的变压器故障诊断 总被引：2，自引：0，他引：2
根据电力变压器故障诊断问题,提出了基于粗糙集与人工神经网络的变压器故障诊断模型,分析了该模型的实现步骤.采用Kohonen网络对连续属性值进行离散化,应用粗糙集理论对特征参数进行属性约简,并把约简结果生成规则作为BP网络的输入.仿真结果表明,把经过粗糙集理论预处理过的数据送入BP网络训练,提高了学习速度和故障诊断正确率,减少了训练时间. 相似文献
5.
本文采用经验模式分解 (EMD)提取信号的内在模函数 (IMF) ,并利用希尔伯特变换对所得IMF进行包络分析 ,提取机械故障特征。与直接对原信号进行包络分析相比较 ,该方法提取的机械故障特征更明显。数值模拟和对故障轴承振动信号分析表明了该方法的有效性。 相似文献
6.
为提高非线性和非平稳海水温度时间序列的预测能力,提出了一种基于经验模态分解(Empirical Mode Decomposition.简称EMD)的BP神经网络预测方法.该方法首先对原始序列进行经验模态分解,将其分解为多个平稳性得到很大改善的本征模态函数(Intrinsic Mode Function,简称IMF)之和,然后时每个本征模态函数进行预测,最后再根据EMD方法的完备性把预测结果相加得出原始序列的预测结果.预测试验结果表明.基于EMD的BP神经网络预测的精度比单纯用BP神经网络预测有很大提高. 相似文献
7.
8.
针对海洋地震资料高分辨率宽频处理的要求,也为了克服常规谱白化、反Q滤波等方法无法同时增强时域和频域局部细节的不足,基于希尔伯特黄变换(HHT)的谱白化方法对海洋地震资料进行高分辨率处理应用研究。对地震记录进行固有模态分解(EMD),得到不同尺度的IMF分量;再利用常规谱白化方法对每个分量根据瞬时频率进行合理的振幅均衡;将均衡后的IMF分量进行反变换重构地震记录,从而得到高分辨地震数据。通过理论模型和实际地震资料实验,与常规谱白化方法的对比分析,表明该方法信号局部时频刻画能力以及相对振幅保真性优于常规方法,同时表明此方法能够有效增强地震信号时域和频域的分辨率,使地震剖面更为连续和清晰,并具有较高的信噪比。 相似文献
9.
基于模糊神经网络(FNN)的赤潮预警预测研究 总被引：1，自引：0，他引：1
为研究各种理化因子与赤潮藻类浓度间的非线性对应规律和有效预测赤潮藻类浓度,构建了基于BP 算法的一个四层模糊神经网络模型。将模糊神经网络(FNN)技术引入赤潮预测研究,并与普通 BP 网络、RBF 网络的结果作比较,结果表明,该模型能够较好地反演出各种理化因子与夜光藻密度的非线性对应变化规律,有更好的预测功能。 相似文献
10.
基于BP网络对模拟声呐信号分类 总被引：1，自引：0，他引：1
针对常规的主动声呐调查设备,在简单海洋分层模型的基础上,模拟了多波束类单频信号、侧扫类单频信号、Ch irp调频信号和混合信号4类声呐接收信号,并针对接收信号特征构造了3层BP网络模型,将隐藏层神经元数目设为可调节;利用时间域脉冲宽度和水深与频率域功率谱密度相结合的特征参量,成功地对模拟信号进行了分类。采用改进的BP网络模型,用训练成功的BP网络对102个检测信号进行了分类测试,结果表明,分类成功率较高,可达76%~84.6%,因而利用BP网络可以对不同类别设备的模拟声呐接收信号进行分类。 相似文献
11.
1 .IntroductionTheartificialneuralnetwork(ANN)hasbeenwidelyusedinmanyscientificfieldsinrecentyears .Itisakindofinformationmanagementsystemthatresemblesthehumanbraininworkpattern .Comparedwiththetraditionalmethodsofnumericalsimulation ,ANNhastheadvantagesofrelativein dependenceofphysicalmodel,uniformandsimplewayofrealization ,quicknessofcomputing ,andsoon .Sincethemodelofartificialneuronswasfirstlyintroducedin 1 943,ithasbeendevelopedthroughseveralstages.TheapplicationofANNhadnotbeenpopular… 相似文献
12.
在快速滤波分解信号为最宽带通本征模态函数方法（简称FFDSI方法）的基础上，引入了主控模态函数的概念。将其应用于Chichijima水文站水位信号，从中分解出3个96．93％-主控模态函数，分别代表着不同于调和分潮，且与天体运行规律更加吻合的3种模态潮型。由此可见，主控模态函数在信号分析中的应用价值。 相似文献
13.
In this study,an advanced probabilistic neural network(APNN)method is proposed to reflect the global probability density function(PDF)by summing up the heterogeneous local PDF which is automatically determined in the individual standard deviation of variables.The APNN is applied to predict the stability number of armor blocks of breakwaters using the experimental data of van der Meer,and the estimated results of the APNN are compared with those of an empirical formula and a previous artificial neural network(ANN)model.The APNN shows better results in predicting the stability number of armor blocks of breakwater and it provided the promising probabilistic viewpoints by using the individual standard deviation in a variable. 相似文献
14.
Tsunami run-up height is a significant parameter for dimemsions of coastal structures.In the present study,tsunami run-up heights are estimated by three different Artificial Neural Network (ANN) models,i.e.Feed Forward Back Propagation (FFBP),Radial Basis Functions (RBF) and Generalized Regression Neural Network (GRNN).As the input for the ANN configuration,the wave height (H) values are employed.It is shown that the tsunami run-up height values are closely approximated with all of the applied ANN methods.The ANN estimations are slightly superior to those of the empirical equation.It can he seen that the ANN applications are especially significant in the absence of adequate number of laboratory experiments.The restdts also prove that the available experiment data set can he extended with ANN simulations.This may be helpful to decrease the burden of the experimental studies and to supply results for comparisons. 相似文献
15.
An artificial neural network (ANN) was applied to predict seasonal beach profile evolution at various locations along the Tremadoc Bay, eastern Irish Sea. The beach profile variations in 19 stations for a period of about 7 years were studied using ANN. The model results were compared with field data. The most critical part of constructing ANN was the selection of minimum effective input data and the choice of proper activation function. Accordingly, some numerical techniques such as principal component analysis and correlation analysis were employed to detect the proper dataset. The geometric properties of the beach, wind data, local wave climate, and the corresponding beach level changes were fed to a feedforward backpropagation ANN. The performance of less than 0.0007 (mean square error) was achieved. The trained ANN model results had very good agreement with the beach profile surveys for the test data. Results of this study show that ANN can predict seasonal beach profile changes effectively, and the ANN results are generally more accurate when compared with computationally expensive mathematical model of the same study region. The ANN model results can be improved by the addition of more data, but the applicability of this method is limited to the range of the training data. 相似文献
16.
Patterns in the spatial distribution of Peruvian anchovy (Engraulis ringens) revealed by spatially explicit fishing data 总被引：1，自引：0，他引：1
Peruvian anchovy (Engraulis ringens) stock abundance is tightly driven by the high and unpredictable variability of the Humboldt Current Ecosystem. Management of the fishery therefore cannot rely on mid- or long-term management policy alone but needs to be adaptive at relatively short time scales. Regular acoustic surveys are performed on the stock at intervals of 2 to 4 times a year, but there is a need for more time continuous monitoring indicators to ensure that management can respond at suitable time scales. Existing literature suggests that spatially explicit data on the location of fishing activities could be used as a proxy for target stock distribution. Spatially explicit commercial fishing data could therefore guide adaptive management decisions at shorter time scales than is possible through scientific stock surveys. In this study we therefore aim to (1) estimate the position of fishing operations for the entire fleet of Peruvian anchovy purse–seiners using the Peruvian satellite vessel monitoring system (VMS), and (2) quantify the extent to which the distribution of purse–seine sets describes anchovy distribution. To estimate fishing set positions from vessel tracks derived from VMS data we developed a methodology based on artificial neural networks (ANN) trained on a sample of fishing trips with known fishing set positions (exact fishing positions are known for approximately 1.5% of the fleet from an at-sea observer program). The ANN correctly identified 83% of the real fishing sets and largely outperformed comparative linear models. This network is then used to forecast fishing operations for those trips where no observers were onboard. To quantify the extent to which fishing set distribution was correlated to stock distribution we compared three metrics describing features of the distributions (the mean distance to the coast, the total area of distribution, and a clustering index) for concomitant acoustic survey observations and fishing set positions identified from VMS. For two of these metrics (mean distance to the coast and clustering index), fishing and survey data were significantly correlated. We conclude that the location of purse–seine fishing sets yields significant and valuable information on the distribution of the Peruvian anchovy stock and ultimately on its vulnerability to the fishery. For example, a high concentration of sets in the near coastal zone could potentially be used as a warning signal of high levels of stock vulnerability and trigger appropriate management measures aimed at reducing fishing effort. 相似文献
17.
《Coastal Engineering》2007,54(9):643-656
This paper aims at improving the prediction of wave transmission behind low-crested breakwaters by means of a numerical model based on Artificial Neural Networks (ANNs). The data here used are those gathered within the European research project DELOS.Firstly, the motivations that lead to employ an ANN numerical model to forecast the wave transmission behind low-crested structures are discussed. Then, the ANN model is tested and its architecture is optimized with a test targeted on assessing both the accuracy and the robustness of the method. A study is devoted to investigate the ANN model capability in reproducing some physical relationships among the involved parameters. Finally, comparisons of ANN results with those from experimental formulations based on the classic regression approach demonstrate a considerable improvement in the forecast accuracy.The ANN forecasting tool is available as a user-friendly Internet applet at: http://w3.uniroma1.it/cmar/wave_transm_kt.htm. 相似文献
18.
由于地理信息系统的一个重要特点就是针对不同的使用对象,提供了不同的功能模块,因此开发者就需要一个可以方便管理各个功能模块的系统内核对模块进行统一的管理。介绍了一种面向模块对象设计的G IS系统内核,通过运用多种设计模式,使系统可以方便的裁减各个模块,并在模块之间"低耦合"的条件下完成了它们的相互通信。 相似文献
19.
赤潮预测的人工神经网络方法初步研究 总被引：13，自引：0，他引：13
赤潮是一种由多因素综合作用引发的生态异常现象,具有突发性及非线性等特点。对其进行预测预报一直是海洋科学研究的热点。探讨了应用人工神经网络原理进行赤潮预测的方法,简要介绍了BP和RBF算法的基本原理,用2种算法对不同海域赤潮生物与环境因子之间非线性和不确定性的复杂关系进行学习训练和预测检验,并与传统的统计方法进行了比较。结果表明：人工神经网络方法在模拟和预测方面优于传统的统计回归模型,具有较强的模拟预测能力及实用性,值得进一步探索。 相似文献