一种基于BP算法学习的小波神经网络   总被引：2，自引：1，他引：2  

陈哲  冯天瑾  陈刚 《中国海洋大学学报(自然科学版)》2001,31(1)

为发展 Szu的基于信号表示的小波神经网络 ,提出一种多输入多输出的小波网络模型 ,网络隐层采用框架小波函数、输出层采用 Sigmoid激励函数 ,并选用“熵误差函数”以加速网络的学习速度。奇偶判别和混沌时间序列预测例子的实验结果表明了它具有良好的函数逼近能力和推广能力 ,收敛速度和均方误差均优于相同结构的多层感知器模型。  相似文献   

两类双结构矩阵的标准型及其应用     

余多  刘新国 《中国海洋大学学报(自然科学版)》2007,37(1):47-52

给出2类双结构矩阵的标准型。作为应用,还利用新得出的标准型对这2类双结构矩阵特征值问题发展了结构算法,其中对第一类双结构矩阵特征值问题发展了只用实运算的结构算法。  相似文献   

基于ArcEngine的数字高程模型更新方法研究   总被引：1，自引：0，他引：1  

施永胜  邹峥嵘  孙鸿睿 《海洋测绘》2006,26(1):50-53

采用COM技术,以ARCENGINE所提供的类库和凸包算法,实现了数字高程模型(DEM)数据更新的边界检测和利用矢量地形数据采集更新数字高程模型数据的方法。保证了数字高程模型数据的精度,可以满足数字高程模型的现势性需求。  相似文献   

海洋调查数据的可视化建模及实现方法     

孙菁  杨冠杰 《中国海洋大学学报(自然科学版)》2001,31(4):579-585

科学计算可视化作为新一代的分析工具 ,正在进入物理海洋学领域。文中介绍对海洋调查数据进行可视化的关键问题——数据建模 ,并从分析海洋调查数据的特点出发提出针对海洋研究中可视化建模的原则和有效途径 ,给出相关的实现方法。其中 ,重点讨论局部体样条建模中必须面对的区域分割优化问题 ,引入遗传算法并给出较好的解决方案。  相似文献   

基于模拟退火算法的饲料配方设计     

陈世希  李翠华  王渊源 《台湾海峡》2001,20(Z1):142-147

本文在饲料配方设计中首次引入新颖、高效的组合优化算法--模拟退火算法，并获得理想的数值结果，说明了在传统的饲料配方设计中高技术的应用可以获得更高的效益.  相似文献   

琼州海峡铁路轮渡新海湾港区泥沙淤积问题     

沈莹  李大山  张镜潮  杨正己 《海洋工程》1997,(3)

海南省拟建琼州海峡铁路轮渡。本文通过对海峡南岸新海湾轮渡港址附近自然条件、岸滩演变和输沙特征的分析研究，对不同类型港区布置方案港内开挖区和航道的淤积强度、淤积量等进行了计算，并对不同防护建筑物的有效拦沙时间等进行了探讨和计算，提出了港区类型推荐方案。  相似文献   

应用遗传神经网格方法分析赤潮监测数据   总被引：1，自引：1，他引：1  

杨建强  高振会  孙培艳  石强  李钦亮 《海洋科学进展》2002,20(2):77-82

简要介绍了遗传神经网络方法的原理 ,探讨了应用遗传神经网络方法研究辽东湾海域丹麦细柱藻 (Leptocylindrusdanicus)赤潮与其环境因子之间的关系 ,计算了各环境因子对丹麦细柱藻赤潮的贡献。结果表明 ,温度、盐度、DIN的变化对研究海域丹麦细柱藻种群密度的增长有比较重要的影响 ,DIN是营养限制因子。遗传神经网络是分析赤潮监测数据的有效方法 ,值得进一步探索  相似文献   

Geochemical and hydrodynamic constraints on the distribution of trace metal concentrations in the lagoon of Nouméa, New Caledonia   总被引：1，自引：0，他引：1  

Christophe Migon  Sylvain Ouillon  Xavier Mari  Emmanuel Nicolas 《Estuarine, Coastal and Shelf Science》2007,74(4):756

Seawater samples were collected in the lagoon of Nouméa (southwest New Caledonia) along two transects from eutrophic coastal bays to the oligotrophic barrier reef. Land-based emissions to the lagoon were measured with dissolved and particulate concentrations of chromium (Cr) and nickel (Ni), used as tracers of both terrigenous and industrial (Ni ore treatment) activities, as well as dissolved and particulate concentrations of zinc (Zn), used as a tracer of urban effluents. The spatial variability of metal concentrations was related to geochemical and hydrodynamic conditions, i.e., respectively: (1) natural and anthropogenic emission sources, and chemical processes occurring in the water column; and (2) water residence times. The parameter used to describe the residence time of water masses was the local e-flushing time, i.e. the time required for a tracer mass contained within a given station to be reduced by a factor 1/e. High metal concentrations were found in coastal areas (up to 9000 ng dissolved Ni L⁻¹), and steeply decreased with distance from the coast (down to 101 ng dissolved Ni L⁻¹ near the barrier reef) to reach levels similar to those found in remote Pacific waters, suggesting a rapid renewal of waters close to the barrier. Distributions of metals in the lagoon are controlled upstream by land-based emission sources and later chemical processes. Then hydrodynamics constrain metal distributions, as shown by the observed relationship between local e-flushing times and the spatial variability of metal concentrations. In addition, a change in the direction of prevailing winds yielded a decrease of dissolved metal concentrations at the same site by a factor of 2.5 (Cr and Ni) and 2.9 (Zn). It is suggested that the residence time is a key parameter in the control of elemental concentrations in the lagoon waters, as much as land-based emission sources.  相似文献   

一种新型混沌遗传优化算法的研究     

张海燕  潘成欣 《中国海洋大学学报(自然科学版)》2006,36(4):555-558

在对遗传算法交叉、变异后结果的处理方法进行改进并在将进化过程分段的基础上,提出1种新的混沌遗传算法。该算法利用混沌运动的遍历性择优产生初始群体,对每一轮遗传操作所得到的部分当前最优个体进行变尺度混沌搜索。仿真结果表明,该算法具有较高的优化效率,并能求得全局最优解。  相似文献   

Neural network for wave forecasting among multi-stations     

Ching-Piao Tsai  Chang Lin  Jia-N Shen 《Ocean Engineering》2002,29(13):1685-1695

Unlike in the open sea, the use of wind information for forecasting waves may encounter more ambiguous uncertainties in the coastal or harbor area due to the influence of complicated geometric configurations. Thus this paper attempts to forecast the waves based on learning the characteristics of observed waves, rather than the use of the wind information. This is reported in this paper by the application of the artificial neural network (ANN), in which the back-propagation algorithm is employed in the learning process for obtaining the desired results. This model evaluated the interconnection weights among multi-stations based on the previous short-term data, from which a time series of waves at a station can be generated for forecasting or data supplement based on using the neighbor stations data. Field data are used for testing the applicability of the ANN model. The results show that the ANN model performs well for both wave forecasting and data supplement when using a short-term observed wave data.  相似文献