海员避碰行动各指标相关性的多元回归分析

在获取大量海员避碰行动样本的基础上，通过建立多元线性回归的数学模型，分析了转向时机、转向幅度、目标船舷角、会遇两船的船速比等4个指标与目标船最后通过距离之间的相关性大小，结论对海员避碰行动决策有一定的参考价值。由于样本是在对海员进行航海模拟器训练中获得的，从而说明了用航海模拟器来研究避碰行为的可行性以及STCW公约强化模拟器训练的必要性。     

高炉锰铁生产要素的多元回归分析

采用数理统计方法中的多元线性回归方法分析了高炉锰铁实际生产数据(模拟),得到了高炉利用系数、综合焦比与其主要影响因素综合冶炼强度、休风率、焦炭负荷、入炉锰矿品位、锰回收率的一组多元线性回归方程式,确定了上述各因素对高炉利用系数、综合焦比影响的重要性顺序,为今后高炉锰铁生产预测、决策和调控提供了依据。     

Ground truthing sediment profile imagery with traditional benthic survey data along an established disturbance gradient

Macrobenthic surveys are an expensive, slow and labour intensive means to establish the health of benthic communities. Sediment profile imagery (SPI) is a means of rapid reconnaissance for monitoring large areas of the benthos. SPI has often been used to monitor gross anthropogenic disturbance. The aim of this study is to determine if SPI can be used as a tool to reliably map change in communities along natural estuarine gradients. Macrobenthic sampling was carried out at five stations along an established estuarine gradient. This faunal data was analysed using standard multivariate techniques and to ground truth a concurrent SPI survey. Faunal analysis showed that habitat quality in Inner Galway Bay was generally good, with some localised disturbance from the River Corrib and the sewage out flow exterior to the city dock. Four distinct groups were identified with a degree of overlap occurring between stations 3 and 4. While existing SPI indices mapped habitat quality in the same manner as the faunal data for end member stations, the level of distinction between the habitats of an intermediate staging was found to be poor. This lack of distinction amongst the stage 2 and 3 stations was overcome by developing a tailored index, the Galway Bay index of habitat quality (GBHQ). This index was derived from the 5 observed variables in the SPI data that were determined to best match the faunal distribution by permutative mantel testing. The 5 observed variables from the SPI data were the depth of the apparent redox potential discontinuity (aRPD), the depth of penetration by the prism, and the presence/absence of infauna, surface faecal pellet layer and biogenic mounding. The GBHQ was able to distinguish between the 5 stations to a greater extent than previously described indices, showing clearly the separate groupings defined by the faunal data. The index was tested on a follow up SPI survey and shown to be applicable in mapping a broader range of habitats in Galway Bay. Indices generated for localised mapping of estuarine gradients should be derived from observed features and be ground truthed using faunal data. Some aspects of the GBHQ should be generally applicable to fine grained boreal estuarine sediments (aRPD/penetration), while others may be of limited utility in other locations depending on the digging characteristics of the particular SPI camera, and local factors influencing the persistence of biogenic features in the profile. This derivation technique provides a simple way to optimise SPI to particular studies and localities.     

试析语流音变对词语发音和词形的影响

音位与音位结合的时候,一个音位由于受邻音的影响,或者由于说话的快慢、高低、强弱的不同,可能发生不同的变化．文章从音的同化、异化、强化（插入音或加音）、弱化（弱化和脱落）的角度深入分析了语流音变对词语发音和构形的影响．对于长期困扰语言学习者的一些常见语音变化规律和发音现象作出了较详尽的阐释,同时发音的改变最终必然反映到词形变化上．     

基于岩体指标和掘进参数的TBM净掘进速率预测模型

净掘进速率是TBM施工速度的主要评价指标,与围岩物理力学性质、TBM掘进参数之间存在一定相关性。文章以兰州水源地建设工程输水隧洞双护盾TBM施工为背景,基于现场实测数据,选择岩石单轴抗压强度、抗拉强度、变形模量、泊松比、岩石耐磨性CAI值等岩体指标,以及刀盘推力和刀盘转速等掘进参数,进行TBM净掘进速率与有关影响参数之间的单因素相关性分析,得到相应拟合公式;基于TBM净掘进速率与岩体指标、掘进参数之间的相关性,利用多元非线性回归方法建立了TBM净掘进速率预测模型。通过将兰州水源地建设工程输水隧洞实测TBM净掘进速率和预测结果进行对比,验证了TBM净掘进速率预测模型的合理性。研究结果表明:(1)在复杂的多种地质条件下,TBM净掘进速率与岩石单轴抗压强度、抗拉强度、变形模量、岩石耐磨性CAI值、刀盘推力以及刀盘转速呈负相关关系,与泊松比呈正相关关系;(2)干湿状态对岩石耐磨性CAI值有一定影响,饱和状态下岩石耐磨性CAI值与TBM净掘进速率之间的相关性更显著;(3)建立的多元非线性回归预测模型,预测精度较高,可为相似地质条件下TBM净掘进速率估算提供参考。     

A generic data assimilation framework for vehicle trajectory reconstruction on signalized urban arterials using particle filters

With trajectory data, a complete microscopic and macroscopic picture of traffic flow operations can be obtained. However, trajectory data are difficult to observe over large spatiotemporal regions—particularly in urban contexts—due to practical, technical and financial constraints. The next best thing is to estimate plausible trajectories from whatever data are available. This paper presents a generic data assimilation framework to reconstruct such plausible trajectories on signalized urban arterials using microscopic traffic flow models and data from loops (individual vehicle passages and thus vehicle counts); traffic control data; and (sparse) travel time measurements from whatever source available. The key problem we address is that loops suffer from miss- and over-counts, which result in unbounded errors in vehicle accumulations, rendering trajectory reconstruction highly problematic. Our framework solves this problem in two ways. First, we correct the systematic error in vehicle accumulation by fusing the counts with sparsely available travel times. Second, the proposed framework uses particle filtering and an innovative hierarchical resampling scheme, which effectively integrates over the remaining error distribution, resulting in plausible trajectories. The proposed data assimilation framework is tested and validated using simulated data. Experiments and an extensive sensitivity analysis show that the proposed method is robust to errors both in the model and in the measurements, and provides good estimations for vehicle accumulation and vehicle trajectories with moderate sensor quality. The framework does not impose restrictions on the type of microscopic models used and can be naturally extended to include and estimate additional trajectory attributes such as destination and path, given data are available for assimilation.     

基于混沌理论的短时交通流量预测

随着智能交通的发展,实时动态交通分配成为当前研究热门问题。短时交通流预测是实时动态交通分配的关键技术之一,在当今交通控制以及车辆导航中具有不可替代的地位。通过对交通流数据进行分析,得出交通系统具有耗散系统特性,并且存在混沌。在此基础上,运用混沌理论对交通流数据进行相空间重构,并用多元局域预测法对时间序列进行预测。通过分析预测数据,得出基于混沌理论的短时交通流量预测在2～5 min内具有较高的预测精度。     

A Comparative Study of Three Multivariate Short-Term Freeway Traffic Flow Forecasting Methods With Missing Data

Short-term traffic flow forecasting is a critical function in advanced traffic management systems (ATMS) and advanced traveler information systems (ATIS). Accurate forecasting results are useful to indicate future traffic conditions and assist traffic managers in seeking solutions to congestion problems on urban freeways and surface streets. There is new research interest in short-term traffic flow forecasting due to recent developments in intelligent transportation systems (ITS) technologies. Previous research involves technologies in multiple areas, and a significant number of forecasting methods exist in the literature. However, most studies used univariate forecasting methods, and they have limited forecasting abilities when part of the data is missing or erroneous. While the historical average (HA) method is often applied to deal with this issue, the forecasting accuracy cannot be guaranteed. This article makes use of the spatial relationship of traffic flow at nearby locations and builds up two multivariate forecasting approaches: the vector autoregression (VAR) and the general regression neural network (GRNN) based forecasting models. Traffic data collected from U.S. Highway 290 in Houston, TX, were used to test the model performance. Comparison of performances of the three models (HA, VAR, and GRNN) in different missing ratios and forecasting time intervals indicates that the accuracy of the VAR model is more sensitive to the missing ratio, while on average the GRNN model gives more robust and accurate forecasting with missing data, particularly when the missing data ratio is high.     

Estimates of vertical mixing during a Lagrangian experiment off the Galician coast

An adjoint 1-D model was used to determine vertical diffusivity coefficients from temperature profiles collected within a filament escaping from the Galician coast following an upwelling event. The optimisation scheme ended with relatively high diffusivity values within the thermocline (9×10⁻⁵ m² s⁻¹). Such high values are relevant for biogeochemical exchanges between surface and deep waters in stratified areas.The optimised values were several orders of magnitude higher than the bulk of diffusivity measurements recorded with a free-falling device; however, the optimisation solution was consistent with the arithmetic mean of the measurements in the thermocline (7.7×10⁻⁵ m² s⁻¹), giving more weight to the few largest values. Below the thermocline, the data assimilation method failed because of the three-dimensional nature of the advective field of the upwelling system. Ignoring this advective forcing in the model led to estimates that were two orders of magnitude too high.The results suggest that turbulent mixing is a random process where a few intense events determine the average mixing that drives the long-term evolution of the water column structure. This statistical property is very important when one wants to use instantaneous diffusivity measurements for modelling purposes.     

Quantifying uncertainty in high-resolution coupled hydrodynamic-ecosystem models

Marine ecosystem models are becoming increasingly complex and sophisticated, and are being used to estimate the effects of future changes in the earth system with a view to informing important policy decisions. Despite their potential importance, far too little attention has been, and is generally, paid to model errors and the extent to which model outputs actually relate to real-world processes. With the increasing complexity of the models themselves comes an increasing complexity among model results. If we are to develop useful modelling tools for the marine environment we need to be able to understand and quantify the uncertainties inherent in the simulations. Analysing errors within highly multivariate model outputs, and relating them to even more complex and multivariate observational data, are not trivial tasks. Here we describe the application of a series of techniques, including a 2-stage self-organising map (SOM), non-parametric multivariate analysis, and error statistics, to a complex spatio-temporal model run for the period 1988–1989 in the Southern North Sea, coinciding with the North Sea Project which collected a wealth of observational data. We use model output, large spatio-temporally resolved data sets and a combination of methodologies (SOM, MDS, uncertainty metrics) to simplify the problem and to provide tractable information on model performance. The use of a SOM as a clustering tool allows us to simplify the dimensions of the problem while the use of MDS on independent data grouped according to the SOM classification allows us to validate the SOM. The combination of classification and uncertainty metrics allows us to pinpoint the variables and associated processes which require attention in each region. We recommend the use of this combination of techniques for simplifying complex comparisons of model outputs with real data, and analysis of error distributions.