Data selection using support vector regression

Geophysical data sets are growing at an ever-increasing rate, requiring computationally efficient data selection(thinning)methods to preserve essential information. Satellites, such as Wind Sat, provide large data sets for assessing the accuracy and computational efficiency of data selection techniques. A new data thinning technique, based on support vector regression(SVR), is developed and tested. To manage large on-line satellite data streams, observations from Wind Sat are formed into subsets by Voronoi tessellation and then each is thinned by SVR(TSVR). Three experiments are performed. The first confirms the viability of TSVR for a relatively small sample, comparing it to several commonly used data thinning methods(random selection, averaging and Barnes filtering), producing a 10% thinning rate(90% data reduction), low mean absolute errors(MAE) and large correlations with the original data. A second experiment, using a larger dataset, shows TSVR retrievals with MAE < 1 m s-1and correlations 0.98. TSVR was an order of magnitude faster than the commonly used thinning methods. A third experiment applies a two-stage pipeline to TSVR, to accommodate online data. The pipeline subsets reconstruct the wind field with the same accuracy as the second experiment, is an order of magnitude faster than the nonpipeline TSVR. Therefore, pipeline TSVR is two orders of magnitude faster than commonly used thinning methods that ingest the entire data set. This study demonstrates that TSVR pipeline thinning is an accurate and computationally efficient alternative to commonly used data selection techniques.     

海底地形数据组织与可视化选取技术

海洋测绘数据信息化是构建智慧海洋的基础, 而海底地形数据是海洋测绘数据中的重要内容。为更加便捷和高效地 管理与利用地形数据, 本文结合工程实际需求, 研发了海底地形数据管理平台。首先, 基于海底地形数据的特点, 提出多分 辨率海底地形组织模型, 对不规则区域的海底地形进行多分辨率模型构建, 实现了对多波束实测数据的组织管理; 在此基础 上, 提出了多尺度海底地形分级渲染优化算法与海底地形数据空间截取算法, 利用分级渲染与数据裁剪算法实现对海底地形 数据的可视化表达预览和选取服务; 最后, 以典型的多波束测深数据为例, 基于研发平台完成了数据从存储 、组织管理到可 视化预览选取的一体化管理与服务, 验证了本文算法的可靠性与实用性, 同时可为海底地形数据管理服务提供一定参考。     

An agent-integrated irregular automata model of urban land-use dynamics

Urban growth models are useful tools to understand the patterns and processes of urbanization. In recent years, the bottom-up approach of geo-computation, such as cellular automata and agent-based modeling, is commonly used to simulate urban land-use dynamics. This study has developed an integrated model of urban growth called agent-integrated irregular automata (AIIA) by using vector geographic information system environment (i.e. both the data model and operations). The model was tested for the city of San Marcos, Texas to simulate two scenarios of urban growth. Specifically, the study aimed to answer whether incorporating commercial, industrial and institutional agents in the model and using social theories (e.g. utility functions) improves the conventional urban growth modeling. By validating against empirical land-use data, the results suggest that a holistic framework such as AIIA performs better than the existing irregular-automata-based urban growth modeling.     

基于指标规范值的海水水质评价的SVR模型

为了建立具有普适通用的海水水质评价的支持向量机模型,在设置各指标参照值和指标规范变换式,并对指标进行规范变换的基础上,应用免疫进化优化算法,建立基于指标规范值的海水水质评价的回归支持向量机模型。将优化好的模型用于珠江口海水水质的评价,其评价结果与BP神经网络的评价结果基本一致,从而表明基于指标规范值的支持向量机模型用于海水水质评价是可行的,且该模型较传统的支持向量机评价模型具有较好的普适性和通用性。     

复杂机场环境的多维动态可视化方法

针对飞行器可视导航中复杂机场空地环境多维要素动态交互可视化的难题,基于空地环境要素威胁语义统一的信息场模型,采用体绘制、面绘制与切片绘制混合的方法对机场近区域复杂多样的地形地貌、电磁和气象等威胁态势信息进行交互式动态可视化表达。以某机场的典型模拟数据为例,实现了威胁态势的动态交互可视化,验证了该方法的实用性和有效性。     

Aerosol Type Identification Using PARASOL Multichannel Polarized Data

PARASOL （Polarization ＆ Anisotropy of Reflectances for Atmospheric Sciences coupled with Ob- servations from a Lidar） multi-channel and mul- ti-directional polarized data for different aerosol types were compared. The PARASOL polarized radiance data at 490 nm, 670 nm, and 865 nm increased with aerosol optical thickness （AOT） for fine-mode aerosols; however, the polarized radiances at 490 nm and 670 nm decreased as AOT increased for coarse dust aerosols. Thus, the vari- ation of the polarized radiance with AOT can be used to identify fine or coarse particle-dominated aerosols. Polar- ized radiances at three wavelengths for fine- and coarse-mode aerosols were analyzed and fitted by linear regression. The slope of the line for 670 nm and 490 nm wavelength pairs is less than 0.35 for dust aerosols. However, the value for fine-mode aerosols is greater than 0.60. The Support Vector Machine method （SVM） based on 12 vector features was used to discriminate clear sky, coarse dust aerosols, fine-mode aerosols, and cloud. Two cases were given and validated by AErosol RObotic NETwork （AERONET） measurements, MODIS （Mod- erate Resolution Imaging Spectroradiometer） FMF （Fine Mode Fraction at 550 nm） images, PARASOL RGB （Red Green Blue） images, and CALIOP （Cloud-Aerosol Lidar with Orthogonal Polarization） VFM （Vertical Feature Mask） data.     

海底双相随机介质声反射的SVM神经网络分类识别研究

考虑海底沉积介质为双相介质,为了更好地模拟实际海底底质的不均匀性,将随机介质理论引入双相介质理论。首先,通过基于随机-双相介质理论的高阶有限差分数值技术模拟计算海底底质分别为泥质砂、泥、泥质砾时的地震反射波信号。然后利用小波变换分别求取不同底质的一次反射波的包络作为其特征向量,最后利用基于粒子群智能算法优化的支持向量机神经网络对这些反射波信号进行分类识别。为了进一步考察所用方法的抗噪能力,对正演得到的海底底质反射波信号分别加入10%、30%、50%的高斯白噪音之后再进行分类,支持向量机仍然取得了较好的分类预测效果。基于上述正演模拟及分类识别方法的论证,提出了一套行之有效的微机软件模拟海底沉积物分类识别的一般化流程,这将有利于开展海底沉积物反射特征的进一步研究。     

Nonlinear optimization method of ship floating condition calculation in wave based on vector

Ship floating condition in regular waves is calculated. New equations controlling any ship＇s floating condition are proposed by use of the vector operation. This form is a nonlinear optimization problem which can be solved using the penalty function method with constant coefficients. And the solving process is accelerated by dichotomy. During the solving process, the ship＇s displacement and buoyant centre have been calculated by the integration of the ship surface according to the waterline. The ship surface is described using an accumulative chord length theory in order to determine the displacement, the buoyancy center and the waterline. The draught forming the waterline at each station can be found out by calculating the intersection of the ship surface and the wave surface. The results of an example indicate that this method is exact and efficient. It can calculate the ship floating condition in regular waves as well as simplify the calculation and improve the computational efficiency and the precision of results.     

正二十面体全球剖分模型的Geometry Clipmap球面绘制

地形绘制一直是图形学研究的热点问题,尤其是球面地形绘制,其在形状和数据组织方面比平面地形绘制更加复杂。在已有球面地形绘制算法的基础上,提出一种基于Geometry Clipmap的球面地形剖分与绘制方法。该方法以构建正二十面体球面网格为基础,将正二十面体划分为十个菱形区域,采用球面菱形网格的剖分,针对每个菱形区域的周边网格进行重新剖分和组合,形成一个虚拟的3×3的大菱形区域,扩大了Clipmap的活动范围,并在一定程度上解决了Clipmap的跨边界问题。实验结果表明了本文方法的可行性和有效性。