基于LSTM的海洋表面短期风速预测研究

为实现对海面风速精确的短期预测,提出了一种基于长短期记忆(LSTM,longshort-termmemory)神经网络的短期风速预测模型,选取OceanSITES数据库中单个浮标站点采集的风速历史数据作为模型输入,经过训练设置最佳参数等步骤,实现了以LSTM方法,对该站点所在海区海面风速在各季节性代表月份海面风速的24 h短期预测。同时通过不同预测时长的实验以及与BP(back propagation)神经网络神经网络和径向基函数神经网络(radialbasisfunctionneuralnetwork,RBF)的预测效果对比实验,证明了LSTM预测方法相比上述两种神经网络预测方法,在海表面风速预测应用中的优越性。最后通过多个海域对应的站点风速数据预测实验,证明了LSTM神经网络模型的普遍适用性,由相关系数和预测误差的分析可知该方法具备应对急剧变化数据的预测稳定性,可以作为海洋表面风速短期预测的一种可靠方法。     

基于GF5-AHSI遥感数据的横沙浅海水深反演

高光谱遥感水深反演是一种对传统水深测量方法的补充,具有方便、快捷、经济等突出优势。本文研究区位于上海横沙,属于典型滩涂浅水区,研究数据包括GF5-AHSI高光谱遥感数据和同时期的水深数据。通过数据变换和相关分析等方法提取建模参数,利用单波段比值模型、多元线性回归模型、最优标度回归模型和BP神经网络模型实现该区域水深反演,并对4种模型反演结果的准确性进行了验证和比较。研究发现：最优标度回归模型优于其他3种模型,R2达到了0.972,RMSE为0.47 m,适用于横沙浅海水深反演。     

Distribution and invasion of Spartina alterniflora within the Jiaozhou Bay monitored by remote sensing image

Spartina alterniflora as an alien invasive plant, poses a serious threat to the ecological functions of the coastal wetland of the Jiaozhou Bay. As of 2019, the distribution area of S. alterniflora in the Jiaozhou Bay has reached more than 500 hm². For this reason, combined with field surveys, remote sensing monitoring of the invasion S.alterniflora in the Jiaozhou Bay has been carried out. To accurately identify S. alterniflora within the Jiaozhou Bay coastal wetland, we used a new m...     

一种基于异步串行口的IBM-PC局部网

在综合的科学考察船上以异步通讯口实现的IBM-PC局部网具有较好的性能价格比。本文介绍了该网的联网环境、网络系统的组成及其拓扑结构、网络的服务功能以及网络软、硬件的安装和配置。     

面向高分遥感影像道路提取的轻量级双注意力和特征补偿残差网络模型

针对高分辨率遥感影像背景复杂,道路提取容易受阴影、建筑物和铁路等背景信息干扰的问题,提出一种带有轻量级双注意力和特征补偿机制的DAFCResUnet模型。该模型在ResUnet的基础上,通过增加轻量级的双注意力和特征补偿模块实现模型在性能和时空复杂度上的平衡。其中,双注意力模块可以增强模型的特征提取能力,特征补偿模块可以融合网络中来自深浅层的道路特征。在DeepGlobe和GF-2道路数据集上的实验结果表明,DAFCResUnet模型的IoU和F1-score可以达到0.6713、0.8033和0.7402、0.8507,模型的整体精度优于U-Net、ResUnet和VNet模型。与U-Net和ResUnet模型相比,DAFCResUnet模型仅增加了少量的计算量和参数量,但IoU和F1-score均有较大幅度的提高;与VNet模型相比,DAFCResUnet模型在计算量和参数量远低于VNet的情况下取得了更高的精度,模型在精度和时空复杂度两方面均有优势。相比其他对比模型,DAFCResUnet模型具有更强的特征提取和抗干扰能力,能更好解决道路上的干扰物、与道路特征相似地物、树荫或阴影...     

基于Quanser实验平台的带有输出约束单连杆柔性机械臂的神经网络控制

机械臂在航空航天、服务等领域的应用越来越广泛,其研究也越来越深入.相比于刚性机械臂,柔性机械臂质量轻、能耗小,具有更好的性能.但是,由于柔性机械臂本身的结构与材料具有特殊性,其在运动过程中会产生弹性形变与振动,这就给机械臂的定位、轨迹跟踪带来了困难,因此对其振动抑制的研究具有重要意义.本文利用假设模态法对单连杆柔性机械臂系统进行建模,通过李雅普诺夫直接法实现了闭环系统的稳定性.由于一些实际问题对控制系统的状态量有特殊要求,因此采用正切函数形式的障碍李雅普诺夫策略来处理输出约束问题,之后利用神经网络控制方法来逼近系统的不确定性,通过李雅普诺夫法对闭环系统的稳定性进行了分析,并基于Matlab平台设计仿真、基于Quanser实验平台进行实验,对控制器的控制性能进行了验证.     

Structural-stratigraphic control on the Umitaka Spur gas hydrates of Joetsu Basin in the eastern margin of Japan Sea

Integrated geological, geochemical, and geophysical exploration since 2004 has identified massive accumulation of gas hydrate associated with active methane seeps on the Umitaka Spur, located in the Joetsu Basin on the eastern margin of Japan Sea. Umitaka Spur is an asymmetric anticline formed along an incipient subduction zone that extends throughout the western side of the Japanese island-arc system. Seismic surveys recognized chimney structures that seem strongly controlled by a complex anticlinal axial fault system, and exhibit high seismic amplitudes with apparent pull-up structures, probably due to massive and dense accumulation of gas hydrate. Bottom simulating reflectors are widely developed, in particular within gas chimneys and in the gently dipping eastern flank of the anticline, where debris can store gas hydrates that may represent a potential natural gas resource. The axial fault system, the shape of the anticline, and the carrier beds induce thermogenic gas migration to the top of the structure, and supply gas to the gas hydrate stability zone. Gas reaching the seafloor produces strong seepages and giant plumes in the sea water column.     

海洋测绘成果在港口与航道维护保障中的应用

详细介绍了海洋测绘成果在港口与航道疏浚工程水下挖槽量、回淤量和航道边坡土方量及超宽量计算中的应用,探讨了维护疏浚工程效益分析方法,提出了扩大海洋测绘成果在港口与航道通航保障应用范围的建议.     

广利河口拦门沙发育动态和河口航道的选择

根据现场调查和历史资料,获取了广利河口拦门沙水动力特征、海底地形和底质特征,对拦门沙的动态发育和波浪作用下拦门沙运动状态进行了分析,探讨了广利河口的航道选择方案。分析结果认为,广利河槽外航道宜从东偏南向入海。     

Neural network for wave forecasting among multi-stations

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.