预测矿井涌水量的计算级别与精度评述

根据水利部2005年发布的《建设项目水资源论证导则（试行）》技术文件中的有关规定与要求,参照《供水水文地质勘察规范》以及《固体矿产资源/储量分类,》认为地下水可开采量与预测矿井涌水量的计算可按勘查研究程度,将其可分为A、B、C、D、E等5个精度等级。参照各级别的允许误差,论证了目前矿井涌水量预测计算公式及方法的可靠程度。     

基于人工神经网络的地震经济损失评估

地震经济损失估计是涉及多方面、多级次的复杂指标体系的非线性动力问题。其指标体系的选取以具有代表性、可靠、易获取、易于定量化为原则。应用MATLAB6.5神经网络工具箱,建立了适用于震后经济损失快速评估的三层BP神经网络地震灾害经济评估模型。     

加权支持向量回归机及其在水质预测中的应用

支持向量机是一种基于结构风险最小化原理的学习技术，也是一种新的具有很好泛化性能的回归方法。本文对用于回归估计的标准支持向量机加以改进，提出了一种新的用于回归估计的支持向量机学习算法，针对各样本重要性的差异，给各个样本的惩罚系数和误差要求赋予不同权重，并利用加权支持向量回归机的理论及其算法构建水质预测模型。实验结果表明，该方法对水质具有较好的预测效果。     

A level-set corrector to an adaptive multiscale permeability prediction

We consider the inverse problem of permeability estimation for two-phase porous-media flow. The novel approach is based on regularization by zonation, where the geometry and size of the regions are chosen adaptively during the optimization procedure. To achieve this, we have utilized level-set functions to represent the permeability. The available data are sparsely distributed in space; hence, it is reasonable to confine the estimation to coarse-scale structures. The level-set approach is able to alter the boundaries between regions of different permeability without strict restrictions on their shape; however, when the data are sparse, a reasonable initial guess for the permeability is required. For this task, we use adaptive multiscale permeability estimation, which has the potential of identifying main permeability variations. These are described by a piecewise constant function, where the constant values are attained on rectangular zones. In the current work, we develop a level-set corrector strategy, assuming adaptive multiscale permeability estimation as a predictor.     

Best prediction in linear models with mixed integer/real unknowns: theory and application

In this contribution, we extend the existing theory of minimum mean squared error prediction (best prediction). This extention is motivated by the desire to be able to deal with models in which the parameter vectors have real-valued and/or integer-valued entries. New classes of predictors are introduced, based on the principle of equivariance. Equivariant prediction is developed for the real-parameter case, the integer-parameter case, and for the mixed integer/real case. The best predictors within these classes are identified, and they are shown to have a better performance than best linear (unbiased) prediction. This holds true for the mean squared error performance, as well as for the error variance performance. We show that, in the context of linear model prediction, best predictors and best estimators come in pairs. We take advantage of this property by also identifying the corresponding best estimators. All of the best equivariant estimators are shown to have a better precision than the best linear unbiased estimator. Although no restrictions are placed on the probability distributions of the random vectors, the Gaussian case is derived separately. The best predictors are also compared with least-squares predictors, in particular with the integer-based least-squares predictor introduced in Teunissen (J Geodesy, in press, 2006).     

Gauss-Markov模型的岭型GM估计

针对目前文献中提出的各种抗差有偏估计没有对设计空间的粗差进行控制的问题,将岭估计与GM估计结合起来提出了一种新的抗差有偏估计———岭型GM估计,理论分析和计算结果均表明岭型GM估计不仅能消除法矩阵病态性的不良影响,而且能同时具有观测空间和设计空间的抗差性,是一种性能良好的估计。     

Application of statistical method in pollution episode analyses

Observed pollutants are of both primary and secondary origins, influenced by local topography and meteorology. Identifying sources and relative contribution often require vast detailed data collection and complicated models. This study applied a statistical time series analysis to two selected pollution events, spring and fall, at two sites in northern Taiwan. Vector moving average representations were used to quantitatively examine relationships among chemical pollutants and estimate their lifetimes. Results from impulse responses show that wind direction change alters the characteristic of pollution observed in opposite sites of Taipei City, from chemical dominant system to transport dominant one and vice versa. Chemicals are clearly separated into photochemical pollutants and primary pollutants. Results pointed out that Taipei City is the major source of photochemical smog, but not these primary pollutants. Derived chemical lifetimes at same location vary from 20% to four times under different meteorological condition. Estimated concentrations of hydroxyl radical range between 2 to 8 × 10⁷ cm⁻³. Photochemical pollutants are responsible for parts of PM10 collected in both station observed. Oxidation of SO₂ is only important in PM10 observed at one station. This study provides a simpler tool to derive information usually from complex models, therefore, is suitable as complement in decision-making process.     

一种简化的Helmert方差分量估计算法

Ｈｅｌｍｅｒｔ方法是通过求出两类观测值的单位方差,来修正其先验权,用迭代法最终求得正确的权。本文提出,如果已知两类观测值的单位方差之比,即可定出正确的权,从而使计算简单化。     

方向观测后方交会的点位精度

以虚拟误差方程,经严密推证,给出了以方向作为独立观测值的后方交会点位精度估算公式。以大桥施工控制网为例,介绍了３个方向与４个方向后方交会的精度状况及其应用。     

L1估计的分布及统计性质

推导了Ｌ１估计的密度函数,讨论了Ｌ１估计的统计性质,证明Ｌ１估计是无偏估计,同时还证明Ｌ１估计优于ＬＳ估计。