基于随机森林特征选择的城市绿化乔木树种分类

城市绿化在改善空气、水和土壤质量,吸收和减少二氧化碳及各种污染物,缓解城市热岛和减少雨水径流等方面发挥着重要作用。及时准确地获取树种信息是城市规划与绿化管理的先决条件,对进一步改善城市生态环境也具有重要意义。基于遥感技术,使用高空间分辨率的WorldView-2卫星影像,采用光谱、纹理、指数以及几何等多种特征相结合的面向对象方法,并通过随机森林进行特征选择,对福州大学旗山校区北部的榕树、杧果、香樟、重阳木、羊蹄甲、垂叶榕以及木棉7种主要绿化乔木进行树种分类。实地验证结果表明：通过特征选择可以减少或规避数据冗余以及休斯效应的产生,该方法可以提高现有同类型树种分类的精度,当淘汰全部特征的20%,利用34个特征（包括15个光谱特征、6个纹理特征、8个指数特征和5个几何特征）进行分类时,总精度最高,可达74.95%,Kappa系数为0.67。其中,光谱平均值的特征重要性最高,而各波段的标准差的重要性较低。WorldView-2卫星影像的4个新增波段,特别是黄光和红边波段及其构建的指数特征重要性较高,也说明这些波段在植被遥感,特别是树种分类中极具应用前景。     

Automatic Sounding Generalization in Nautical Chart Considering Bathymetry Complexity Variations

Sounding acts as the main feature in a digital nautical chart as it describes the concerned marine topography for the safety of navigation. Unlike the geometry-oriented selection of point feature, the generalization of soundings for chart compiling is expected to be context-oriented, which means bathymetry complexity variations across the study region should be preserved in the sounding selection process. However, such variations are not explicitly accessible to automated systems. This paper proposes an approach that effectively analyzes and measures bathymetry complexity from sounding data, with a focus on topography variations among different regions. The presented approach first divides the exploring region into several subregions, by adopting techniques of computational geometry and graph theory. Then, the approach quantitatively measures the bathymetry complexity of the subregions from grid-based digital terrain model. Finally, a composite bathymetry complexity index integrating aspects of steepness and depth variation is developed to guide the operation of sounding selection in different subregions. Generally, when seafloor is rugged with steep slopes, the number of soundings is high. While in flatter areas, a smaller amount of soundings is retained. The potential of our approach is demonstrated by an application to a real data set.     

结合POI数据的道路自动选取方法

人们对道路重要性的认知与道路周边设施有重要关系,针对现有地图综合道路选取方法中对语义特征考虑不足的问题,将POI数据引入到道路的语义特征分析中,提出一种综合考虑道路空间特征和语义特征的道路综合自动选取方法。首先,结合POI位置数据构造了道路语义特征度量的3个新参量：设施点密度、重要设施比率和专题设施比率;然后,与道路长度、连接值、总深度值、平均线密度等反映道路几何、拓扑和分布特征的度量参量一起,通过归一化和熵值法赋权进行整合计算,得到道路重要性值;最后,综合考虑道路重要性值、道路stroke构成和stroke连通度的约束条件进行道路的分步选取。实验结果表明,该方法在保留主要道路、保持道路分布疏密特征和道路连通性的同时,较好地顾及了道路的语义特征信息。     

基于机器学习的格点气温预报订正方法

使用2017年9月至2021年3月国家级业务化运行的智能网格实况分析产品和欧洲中期天气预报中心全球模式(EC)产品,根据湖北省的地理分布特征构建6个分区,采用基于LightGBM机器学习算法建立的气温预报方法,生成湖北省0.05°×0.05°格点气温预报产品。利用2021年4—9月的预报产品和格点实况资料进行检验,结果表明:基于机器学习的气温预报方法(MLT)取得了较好的预报效果,其在0～72 h时效内优于中央气象台下发的气温精细化指导预报(SCMOC)和EC产品;MLT在山区的误差较平原大,但山区的订正幅度大于平原,日最高气温的订正幅度大于日最低气温的订正幅度;4—9月MLT、SCMOC、EC产品的平均绝对误差(MAE)日变化都呈现了白天偏高、夜间偏低、午后凸起的单峰特征,MLT的MAE值较SCMOC和EC产品的更低,并且在转折性天气中仍具有优势;站点检验与格点检验结论一致,基于格点建模的气温预报产品对站点预报同样得到了订正。机器学习在格点气温的模式订正方面可以作为一个行之有效的手段。     

A generalized conditional intensity measure approach and holistic ground‐motion selection

A generalized conditional intensity measure (GCIM) approach is proposed for use in the holistic selection of ground motions for any form of seismic response analysis. The essence of the method is the construction of the multivariate distribution of any set of ground‐motion intensity measures conditioned on the occurrence of a specific ground‐motion intensity measure (commonly obtained from probabilistic seismic hazard analysis). The approach therefore allows any number of ground‐motion intensity measures identified as important in a particular seismic response problem to be considered. A holistic method of ground‐motion selection is also proposed based on the statistical comparison, for each intensity measure, of the empirical distribution of the ground‐motion suite with the ‘target’ GCIM distribution. A simple procedure to estimate the magnitude of potential bias in the results of seismic response analyses when the ground‐motion suite does not conform to the GCIM distribution is also demonstrated. The combination of these three features of the approach make it entirely holistic in that: any level of complexity in ground‐motion selection for any seismic response analysis can be exercised; users explicitly understand the simplifications made in the selected suite of ground motions; and an approximate estimate of any bias associated with such simplifications is obtained. Copyright © 2010 John Wiley & Sons, Ltd.     

MULTIVARIATE SELECTION OF VARIABLES IN INDUSTRIAL QUALITY CONTROL:OPTIMIZING AVIATION FUEL FINAL CONTROL

This paper deals with a typical question encountered in all industrial analytical laboratories:are all thetests performed in the laboratory to characterize the final product really necessary?In this work the cross-validation technique,Procrustes rotation techniques and statistical variable selection have been used toreduce the 26 initial British Petroleum and ASTM kerosene specification tests to ten'essential'ones.Statistical as well as chemical considerations were used to select the optimum subset of original variablesto be retained from all the possible ones.     

近代回归分析在交通调查分析建模中的应用

本文以出行发生量模型的建立为例,较为系统地讨论了近代回归分析中的自变量选择及回归诊断方法在交通调查分析建模中的应用。分析了在获取调查数据后,如何借助回归自变量选择方法来选择最佳自变量子集,以确定简捷的回归模型;文中应用回归诊断方法,讨论了修正回归模型、探测错误的调查数据的方法,从而为建立简捷、高精度的交通模型打下了基础。最后,作者提出了应用自变量选择及回归诊断方法建立出行发行量模型的一般步骤。     

甘肃西部地区高放废物处置库场址预选调查

本文根据高放废物处置库场址选择的基本要求，对区域地质及深部构造特征、新构造运动及地震分布规律进行了调查，确定北山南带地区为稳定区。在稳定区内根据处置库有利围岩的出露情况，结合岩体工程地质特征分析，初步选出了４个以花岗岩为主体的场址预选区。经综合分析，认为在４个场址预选区中，以Ｉ号预选区的条件最优，其次是Ⅱ号预选区，再次是Ⅲ号和Ⅳ号预选区。场址区的最后确定尚待进一步调查和论证。     

桉树腐殖酸(HA)专用肥区域试验与配方的优选

采用随机区组设计,以等价投入的常规化肥为对照,开展专用肥多点施肥试验研究.通过专用肥对林木生长量、林相特征的影响分析,结果表明:与等价投入(肥料成本)的常规化肥相比,施用系列专用肥的林分生长量均显著或极显著地高于常规化肥处理,且叶片浓绿、树势强壮,施后19个月内,树高、胸径、材积分别提高了7.8%~51.8%、9.8%~53.0%和30%以上,叶片叶绿素含量提高26.3%~94.2%,叶面积指数提高41.0%~172.8%.经对HA桉树专用肥的增产性、适应性和稳定性的综合分析与评定,选出基肥专用的优化配方为:有机质≥30.0%、N P2O5 K2O≥6.0%(2.5-2.0-1.5)、腐殖酸总量≥8.0%,并加上适量的"绿滴"复合微生物菌剂(于施用时和基肥按每50 kg∶1 kg的量掺入);追肥专用的优化配方为:有机质≥20.0%、N P2O5 K2O≥20.0%(9-8-3)、腐殖酸总量≥6.0%.     

Non‐linear system identification of the versatile‐typed structures by a novel signal processing technique

Non‐linear structural identification problems have raised considerable research efforts since decades, in which the Bouc–Wen model is generally utilized to simulate non‐linear structural constitutive characteristic. Support vector regression (SVR), a promising data processing method, is studied for versatile‐typed structural identification. First, a model selection strategy is utilized to determine the unknown power parameter of the Bouc–Wen model. Meanwhile, optimum SVR parameters are selected automatically, instead of tuning manually. Consequently, the non‐linear structural equation is rewritten in linear form, and is solved by the SVR technique. A five‐floor versatile‐type structure is studied to show the effectiveness of the proposed method, in which both power parameter known and unknown cases are investigated. Copyright © 2007 John Wiley & Sons, Ltd.