航空激光测深仪海面和海底激光入射点分布

依据卵形扫描系统的原理,以严格的数学推导为基础,推导了海平面激光入射点坐标的计算公式及相应的优化算法。建立起了海底激光入射点坐标及该点垂直水深的严密计算模型。通过仿真试验,得出了激光镜面反射点X轴的修正量随驱动电机转角的变化趋势,单个周期的海面激光入射点分布及30个周期的激光脉冲海面重叠图和试验脉冲对应的定位角度数据。同时给出了每个测量数据对应扫描镜中心点的位置和每个激光发射时对应XYZ坐标系中坐标原点的经纬度的计算方法。     

基于多普勒频移的单频载波相位周跳探测

阐述了多普勒探测周跳的原理,利用了静态和车载动态高频数据的L1上的载波和多普勒观测值,试验探测了其周跳。     

时变参数PGM（1,1）模型在变形监测中的应用研究

给出了时变参数PGM（1,1）模型的数值解法,比较了其与GM（1,1）、PGM（1,1）模型的预测精度,分析了灰区间作为预测结果的可靠性。     

基于非等时距数据序列的高层建筑沉降预测

将泊松曲线引入到高层建筑的沉降预测分析中。采用牛顿插值平均法将非等时距数据转化成等时距数据,通过三段计算法求解模型系数,并结合高层建筑沉降监测工程实例计算分析,结果表明:泊松曲线预测与实际沉降数据吻合较好,拟合曲线精度较高,预测效果明显好于传统的双曲线法和指数曲线法,具有较高的实际应用价值。     

面向单脉冲信号分类的集成特征选择与评价

受大量射频干扰信号影响, 快速从海量观测数据中准确识别出单脉冲信号已成为天文数据处理的一项重要任务, 而设计和提取有效数据特征, 是利用机器学习进行单脉冲信号高效识别的决定因素. 针对如何选择最优特征, 进而提升单脉冲信号的分类精度这一关键问题, 设计了面向单脉冲信号分类的集成特征选择方法. 方法首先混合单脉冲信号的参数特征、统计特征和抽象特征, 然后分别利用5种单一特征选择方法选出各自的最优特征集, 最后利用贪心策略对5种单一方法获取的最优特征集进行集成筛选, 获取最优集成特征集. 实验表明, 最优特征集合既包含统计特征也包含抽象特征. 在相同特征数量下, 利用集成特征选择比单一特征选择能获得更高的模型精度, 可使F1值最高提升1.8%. 在海量数据背景下, 集成特征选择对减少特征数量、提升分类性能和加快数据处理速度具有重要作用.     

Bankfull Discharge Recurrence Interval in Gravel-bed Rivers

Bankfull discharge was identified in some 30 gravel-bed rivers representing in total c. 40 gauging stations. The catchment sizes cary from 4km² to nearly 2700km². Bankfull discharge value increases with basin size. In the case of gravel-bed rivers developed on an impermeable substratum, the following equation emerges: Q_b=0·087 A^1·044. Bankfull discharge recurrence interval was determined by fitting maximum annual floods (T_a) into Gumbel's distribution and then using the partial duration series (T_p) in this same distribution. Recurrence interval is below 0·7 years (T_p) for small pebble-bed rivers developed on an impermeable substratum; it reaches 1·1 to 1·5 years when the catchment size of these rivers exceeds 250km². Rivers incised in the soft schists of the Famenne show larger channel capacity at bankfull stage, a small width/depth ratio and thus higher recurrence intervals (1·4–5·3 years with T_a and 1–4·4 years with T_p). Baseflow-dominated gravel-bed streams and sandy or silty rivers experience less frequent bankfull discharges, with a recurrence interval higher than 2 or even 3 years (T_p). © 1997 John Wiley & Sons, Ltd.     

考虑间歇充填和浓度效应的充填体长期强度试验研究

为探究不同充填间隔时间(FTS)和料浆浓度对胶结充填体长期强度影响机制,配制70%、72%、75%三个浓度、充填间隔时间为12、24、36、48 h的两分层胶结充填体试件,开展单轴抗压强度(UCS)试验并探究其力学特性及其破坏形式。试验结果表明,（1）胶结充填体峰值抗压强度随充填间隔时间增大而呈递减趋势,充填间隔时间一定时胶结充填体抗压强度随料浆浓度增大而增大,且峰值抗压强度与充填间隔时间呈多项式函数规律;（2）胶结充填体试件加载过程中表现为压密阶段、线弹性阶段、裂纹扩展阶段和破坏发展4个阶段,随充填间隔时间延长,胶结充填体的破坏形式可能表现为张拉破坏–拉剪破坏过渡–拉剪混合破坏的损伤模式。研究结论能够为后期充填体强度设计和稳定性控制提供有益参考。     

analyses on the triggering facrors of large quantities of landslides in the upper reaches of the minjiang river,sichuan province

Over the past geological and historical period, tens of thousands of landslides occurred in the upper reaches of the Minjiang River, an area which is characterized by alpine valleys and has been densely populated over the past several hundreds of years. Discussing the triggering factor of these landslides is of great significance to geological hazard mitigation and prevention in this region. In this paper, we focus on four aspects of regional rainfall, shape features of landslide slopes, the corresponding relationship between landslide area and earthquake magnitude, and the recurring features of the reconstructed palaeoearthquake record at Diexi. Compared with those in Nepal, both mean seasonal rainfall accumulation and mean daily rainfall for the past 30 years are too low to reach the threshold values triggering landslides in the upper reaches of the Minjiang River. Secondly, landslides in the study area are usually absent of inner gorges(canyon topography)on the hillslope toes, which are confirmed in previous studies as typical features of landslides triggered by storms. Thirdly, wide distribution of the landslides in the study area supports our notion of earthquake-triggering because the landslides triggered by storms commonly distribute locally. Fourthly, periodicity analysis of the reconstructed palaeoearthquake record at Diexi provides a few cycles of twenty to thirty years, possibly corresponding to the earthquakes of magnitudes>5.0 or 5.5 which are believed to have caused soft-sediment deformation in the study area. In contrast, like the 2008 M_S8.0 Wenchuan earthquake, the average recurrence interval of the large earthquakes in the study area is 2.6ka. They caused tens of thousands of landslides and provided more coarse silt particles for the nearby lake sediments at least in 330 years for each time. This is consistent with exponential increase of earthquake magnitude from large to medium and of the landslide area with the increased earthquake magnitude. To sum up, we suggest that tens of thousands of landslides in the upper reaches of the Minjiang River were most likely triggered by earthquakes instead of storms. This preliminary viewpoint needs further examination in the future.     

上海地区不同重现期的风速估算研究

基于上海气象站历史风速观测资料,采用极值I型和皮尔逊III型分布估算了上海市不同重现期最大风速的时间变化以及各区（县）不同重现期最大和极大风速的空间分布。结果表明,1901~2011年,上海市10、30、50和100 a重现期的最大风速分别为21.0、24.9、26.7和29.2 m/s。1974~2011年期间,上海各区（县）10 m高度10、30、50和100 a重现期的最大风速都是以南部沿海地区南汇或金山最大,分别为19.0、21.4、22.6和24.1 m/s;各重现期极大风速也是以南汇或金山最大,分别为32.3、36.4、38.4和41.0 m/s。中心城区各重现期的最大和极大风速都最小。     

卫星钟差采样间隔对高频PPP-ZTD精度的影响

高频对流层延迟（ZTD）的提取对于反映水汽含量的高时间分辨率瞬时变化及其在暴雨短临预报中的应用具有重要意义。基于精密单点定位技术（PPP）分析了不同采样间隔的卫星钟差对PPP-ZTD精度的影响。结果表明,卫星钟差的时间间隔小于30 s时,所获得的PPP-ZTD(RMS<4 mm)比5 min间隔的（RMS<6 mm）精度要高;而5 s与30 s采样间隔的卫星钟差所获得的ZTD精度相当。