溢油SAR图像分类中的纹理特征选择

针对海洋SAR图像的特点,采用基于灰度共生矩阵的纹理分析方法,提出适用于海洋溢油SAR图像分类的纹理特征量。并讨论了纹理特征量的筛选和纹理窗口大小的确定等问题。最后采用人工神经网络方法验证了SAR图象分类效果。     

用小波分析与灰色支持向量机预测高铁变形

针对高铁变形监测数据的非线性特征,构建一种基于小波与灰色支持向量机的高铁变形预测组合模型。利用小波分析获取不同时频尺度上的随机序列和近似序列,通过嵌入维数的确定和高低频数据的相关性分析,将重构后的随机序列作为遗传算法优化SVR模型的输入,对近似序列则采用灰色支持向量机来描述其演变特征,最后将二者预测结果进行耦合叠加,得到小波灰色支持向量机的组合模型预测结果。以贵广高铁实测数据为例,将均方差、平均绝对误差、平均绝对相对误差作为评判指标对预测效果进行评价,结果表明该模型较好地拟合了近似分量,同时避免了细节分量的过拟合,为高铁变形预测提供了新途径。     

基于灰色系统GM(1,1)千将坪滑坡变形监测分析

滑坡在世界范围内广泛分布,我国地域辽阔,有着较为复杂的地理环境,是世界上受灾害影响十分严重的国家之一。2003年7月13日在湖北省秭归县发生千将坪滑坡,由于千将坪滑坡是发生在三峡库区蓄水达到135 m后,因此,受到国内外众多研究机构及学者的关注。本文主要利用千将坪滑坡体上安置的大量GPS监测点,依据监测点在滑坡体上的展布情况及数据完整程度选取特征点进行研究,通过编程建立千将坪滑坡灰色系统GM(1,1)位移预测模型,综合分析预测滑坡变形情况并进行精度评价。     

改进不等时距权重的灰色残差组合修正模型及其应用

为提高传统不等时距灰色模型(TUTGM)的预测精度,提出了一种改进不等时距权重的灰色残差组合修正模型(IUTWGMRCC)。首先在传统不等时距灰色模型中引入时距权重分配系数,按照累加生成和累减还原过程的生成序列不同,构建了4种不同的预测模型,并依据相似度准则确定最优拟合序列和预测值;然后采用正弦函数和谐波变化生成的周期序列函数修正残差序列,进一步提高模型的预测精度;最后对建筑物3个观测点的沉降量进行预测。结果表明,累减还原过程引入不等时距权重的灰色模型预测精度最高,经残差组合修正后,预测结果的后验差比分别为0.04、0.11和0.05,精度等级为1级。     

深远海调查作业流程的探索与思考

随着海洋活动的进一步开展、海洋资源的深入研究以及近年来国内外海底观测技术的大力发展,对深远海调查作业规范化、程序化及安全性的要求也逐步提高。文章以"大洋一号"船实践经验为例,分析定点取样类、拖曳类、走航探测类3类作业的区别、特点及安全要素,探索建立安全性更高的深远海调查作业流程及应急处置通则。文章认为应对不同条件下作业流程的制定进行进一步的探索和细化,应根据船舶、海区环境、作业类别、设备特点及人员配备等实际情况制定行之有效的适应本航次要求的作业流程,以支撑调查作业有序、快速、安全、可靠的开展。     

Evaluation and Driving Force Analysis of Marine Sustainable Development based on the Grey Relational Model and Path Analysis

With the rapid development of the marine economy, the demand for marine resources development and the pressure on marine environmental protection are gradually increasing. It is critical to evaluate and analyze the driving forces of marine sustainable development in order to promote the coordinated development of the marine economy, resources and environment. Taking Jiangsu Province of China as an example, this paper constructs an evaluation index system for marine sustainable development from the three aspects of marine economy, resources and environment, and calculates the weight of the variation coefficient for each indicator. Based on the grey relational model, the average value of the relational degree, calculated by the average value method of correlation coefficients and the weighting method, is then used to evaluate the status of marine sustainable development in this province. The comprehensive index model is used to analyze the dynamic trend of the evolution of marine sustainable development. The driving forces of marine sustainable development are analyzed by the path analysis method combined with the average values of the grey relational degree for each indicator. This analysis found that the marine sustainable development in 2016 and 2012 was good, the situation in 2007 was bad, and the remaining years were intermediate. Compared with the previous years, the optimal conditions of 2008 and 2012 were obvious. The main driving factors of marine sustainable development are cargo throughput of coastal ports, economic losses caused by storm surges in coastal areas, the area of marine nature reserves in coastal areas, coastal wind power generation capacity, and marine biodiversity.     

滑坡前兆突变异常识别方法

将滑坡位移观测值时间序列视为非平稳随机过程, 采用建立系统的同态模型的方法提取短期异常 , 并给出了异常识别准则、有序度计算公式; 用一阶差分进行高通滤波的方法提取临滑异常并进行预测; 根据异常时刻组成的灾变日期集用灰色灾变模型进行预测。     

聊城市种植业优势产业选择研究

运用多目标灰色局势决策理论方法,结合聊城市具体实际,按照产值、收入、分业比重等指标,分析和确定聊城市各县市区的优势产业,并提出最佳发展方案。     

System Analysis of Disaster Prevention Design Criteria for Coastal and Estuarine Cities

—In China,estuarine and coastal cities are mostly regional economic development centers.Thedisasters by combined effect of upper reach flood,storm surge and typhoon waves are primary obstaclesto the economic development of such cities.Thus the risk analysis and system analysis of flood-stormsurge-wave disaster,economic loss and flood-storm surge control measures play a very important role inthe sustainable development of coastal cities.There are three types of coastal cities for consideration.Thefirst type of city is like Tianjin.The most significant damage is from the upper reach flood.The effect ofstorm surge is negligible,because in the estuary of the Haihe River,tidal locks are built.The Grey MarkovModel(GMM)is used to forecast the flood peak level.GMM combines the Grey system and the Markovtheory into a high-precision model.The predicted flood peak levels are close to the measured data.A syn-thetic model is established for economic assessment,risk analysis and flood-control benefit estimation.Asa n     

用灰色关联法探讨干旱地区蒸发器的代表性

本文以20m²蒸发池为标准,选定咍地坡站和阿克苏站1982-1988年的部分同期观测资料,研究其它蒸发器的代表性。采用灰色关联法计算各种蒸发器与20m²蒸发器的关联度。结果表明:咍地坡站,ф20cm的关联度最高,其次为ITII-3000;阿克苏站,ITII-3000型关联度最高,其次为E-601。该结果与20m²蒸发量与各型号的线性相关分析结果不同。哈地坡站线性相关系数以ITII-3000最大,其次足E-601;阿克苏站以E-601最大,ITII-3000次之。