基于BP神经网络的自适应模糊半参数时间序列模型

本文介绍了一种自适应模糊半参数时间序列模型,该模型将半参数技术与反向传播神经网络(BPNN)相结合,形成具有LR-型模糊数据的自适应模糊时间序列模型。首先,提出了基于非参数核、加权最小二乘和交叉验证的混合方法,该方法可以同时估计回归参数和光滑函数以及带宽的最优值;其次,基于非线性残差序列建立BP神经网络,通过神经网络的运算得到新的偏差,使得在不确定条件下获得更丰富的信息,提高了预测精度。本文采用一些常见的拟合优度准则来检验所提出的自适应模糊半参数时间序列模型的性能。通过一个模拟仿真的例子,说明了该方法的有效性。最后,对所得结果的统计分析表明,该模型对模糊时间序列数据的预测具有可靠性和有效性,优于其他模糊时间序列预测模型。

自适应模糊ARDL时间序列半参数模型

传统的模糊最小二乘时间序列模型在包含异常值的模糊数据集中表现出较差的性能。通过引入异常值检测策略,提出了一种降低异常值对未来预测影响的方法。为此,本文介绍了一种基于半参数技术的自适应模糊自回归分布滞后(ARDL)时间序列模型,该模型将半参数技术与加权最小二乘结合,根据加权平方距离误差,提出了在存在异常值时确定精确系数的估计程序,然后利用迭代算法对模糊时间序列模型的参数进行了估计。本文还扩展了传统时间序列模型中常用的几种拟合优度准则,以比较所提出的模糊时间序列方法与现有方法的性能。通过实际例子,验证了本文方法的有效性。结果清楚地表明,在模糊数据存在潜在异常值的情况下,本文提出的模型对模糊时间序列数据的预测具有可行性和有效性,优于其他模糊时间序列模型。

基于ARMA模型的结构动力模态参数识别

论述了ARMA时序差分模型与结构振动微分方程之间的等价关系 ,ARMA模型的定阶与参数估计 ,并利用一悬臂梁的振动位移作为时间序列对其模态参数进行了识别 ,其计算结果满足精度要求 。

三维空间过程模拟的时空数据模型研究

现有的地理信息系统软件一般没有考虑时间和空间多维因素 ,难以处理时空多维地学现象。而空间过程模拟是地学现象研究的主要手段。本文对空间过程模拟的数学模型、空间过程模拟模型的特点、现有GIS数据模型的局限性进行了分析 ,提出用快照序列模型和时间参数模型支持过程模拟的方法 ,并结合地震动过程模拟进行实例分析。

非参数技术理论及应用综述

非参数技术中一个最重要的内容是估计方法，常见的非参数估计方法有核估计、局部多项式估计、近邻估计等。光滑参数的选取、“维数灾难”与边界点问题也是与非参数技术有关的重要内容。非参数技术的一个重要的应用是应用在时间序列模型中。

湿热环境作用下HT280双马来酰亚胺树脂的蠕变性能与模型

在80℃/85%RH湿热环境下对HT280双马来酰亚胺树脂进行了环境老化处理,分别测试了湿热老化前后树脂的吸湿率、红外光谱、试样断口形貌特征、动态力学热性能和蠕变性能。结果表明,随着湿热老化时间的延长,树脂的吸湿率先快速增加,后趋于平缓。经历湿热老化前后的树脂蠕变曲线均呈速率衰减变化趋势。树脂蠕变性能的变化归因于树脂网链伸长导致的储能模量增加和湿热物理老化的综合作用。通过3种模型对实验数据进行了拟合,发现三参数时间硬化模型的蠕变拟合效果最好;Poynting-Thomson模型次之;Kevin模型拟合精度较差。三参数时间硬化模型能够较好地预测蠕变曲线,对长时间树脂蠕变的研究有一定的帮助。

高速公路事故清理时间影响因素分析(英文)

交通事故是导致高速公路延误的一个主要因素,同时也降低了道路通行能力、车辆以及人员的安全性,甚至有可能导致二次事故发生.对此,相关部门已开始制定一些事故管理方案以降低或者消除事故影响并最大限度地增加道路通行能力和车辆行驶安全性.应密歇根交通部要求对底特律市高速公路网事故清理时间影响因素进行分析,基于交通流数据、道路几何特征及大量的事故数据,建立了参数化危险持续时间模型.所建立的四个模型分别基于不同的概率分布.结果表明,四个模型的性能大体相似,但是相对于其他分布,log-logistic分布更适合于事故清理时间分析.对事故清理时间影响显著的因素有:事故发生的时间、路段交通特性和几何特性,以及事故特征.

具有一阶自回归误差的半参数可加模型的估计

研究具有一阶自回归误差的半参数可加时间序列模型的估计问题.假设回归函数来源于某个参数分布族,采用局部二次拟合准则,结合非参数核函数方法,给出参数向量的估计量和回归函数的半参数估计量.在一定正则条件下,证明估计量具有相合性.通过模拟研究与实证分析检验方法的有效性和可行性.

中国经济增长质量的地区差异研究——基于半参数个体时间异质模型的检验

随着中国经济步入以中高速、优结构、新动力为主要特征的新常态,提高经济增长质量成为经济发展中迫切需要解决的重要问题。采用主成分分析法测算中国30个省(市、自治区)1998-2012年间经济增长质量指数值,运用半参数个体时间异质模型对各地区经济增长质量的差距进行检验,结果表明:中国30个省(市、自治区)的经济增长质量水平在1998-2012年间表现出收敛的趋势,且各地区经济增长质量差距的水平收敛假设被拒绝,地区经济增长质量的增长率随着时间收敛。分地区再检验结果显示,中部地区和西部地区两个分组的经济增长质量差距呈条件收敛趋势,而东部地区的经济增长质量差距呈发散趋势。进一步,采用聚类方法检验发现30个省(市、自治区)的经济增长质量水平存在一个趋同俱乐部。

Numerical and experimental evaluation on methods for parameter identification of thermal response tests

Several parameter identification methods of thermal response test were evaluated through numerical and experimental study.A three-dimensional finite-volume numerical model was established under the assumption that the soil thermal conductivity had been known in the simulation of thermal response test.The thermal response curve was firstly obtained through numerical calculation.Then,the accuracy of the numerical model was verified with measured data obtained through a thermal response test.Based on the numerical and experimental thermal response curves,the thermal conductivity of the soil was calculated by different parameter identification methods.The calculated results were compared with the assumed value and then the accuracy of these methods was evaluated.Furthermore,the effects of test time,variable data quality,borehole radius,initial ground temperature,and heat injection rate were analyzed.The results show that the method based on cylinder-source model has a low precision and the identified thermal conductivity decreases with an increase in borehole radius.For parameter estimation,the measuring accuracy of the initial temperature of the deep ground soil has greater effect on identified thermal conductivity.

Cluster of Strong Earthquakes and Aftershocks in Ahar-Varzeghan Area on August 11, 2012

Seismic events are very complex spatial-temporal phenomena. Seismic catalogues, reporting information about spatial-temporal distribution of the main shocks, are nowadays available for many seismic areas in the world, very often major events mark the beginning of a series of earthquakes （aflershocks） whose frequency and energy are meanly decreasing in time down to the background level of activity. Azerbaijan is one of the most active segments of the Alpine-Himalayan seismic belt and marks the junction between the African-Arabian and Indian plate to the south, and Eurasian plate to the north. The cluster of earthquakes that struck near Varzeghan-Ahar was centered near the Gosha-Dagh fault, but preliminary data suggested that the fault was not responsible for the temblor. On the late afternoon of Saturday, August 11, 2012, the northwest of Iran was shaken by two of the strong earthquakes in Iranian history. First was hit by Mw （moment magnitude scale） = 6.4 Richter at local time 16：54 （12：23 GMT （Greenwich Mean Time））, and about 11 min later, an Mw = 6.3 struck 10 km to the west. The spatial-temporal clustering of micro earthquakes （aftershocks） near Varzeghan, is parameterized by means of a generalized passion model. The region has known faults but numerous smaller or deeply buried faults remain undated, according to the Geological Survey of Iran.

Estimation of Number Of Small Cattle Through ARIMA Models in Turkey

In this study, the number of sheep and goats in Turkey were analysed by time series analysis method, and the number of great cattle for next years predicted through the most appropriate time series model.Time series was formed using the data on the number of sheep and goats belonging to the period between 1930 and 2014 in Turkey It was determined through autocorrelation function graphic that the series weren＇t stationary at first, but they became stationary after their first difference were calculated. A stagnancy test was performed through extended Dickey-Fuller test. So as to determine the suitability of the model, it was reviewed if autocorrelation and partial autocorrelation graphs were white noise series and also the results of Box-Ljung test were reviwed. Through the ＂tested models, the model estimations, of which parameter estimates were significant and Akaike information criterion （AIC） was the smallest, were performed. The most appropriate model in terms of both the number of sheep and goats is first-level integrated moving average model stated as ARIMA（0,1,1）. In this model, it was estimated that there would be an increase in the number of sheep and goats in Turkey between the years of 2015 and 2020, however, the increase in the number of sheep would be more than the increase in the number of goats.

非参数理论综述

对非参数理论进行了系统地综述.非参数理论中一个比较重要的内容是估计方法,常见的非参数估计方法有核估计、局部多项式估计、近邻估计等.光滑参数的选取、"维数灾难"与边界点问题也是与非参数理论有关的重要内容,也对这些方面进行综述.最后,文章还综述了非参数技术在时间序列模型中的有关应用问题.

Navigation using binary pulsars

The issue of navigation using binary pulsars is investigated in this paper.We show how the effect of the binary motion is incorporated in the navigation measurement equation.Necessary reference systems and time scales are introduced and the transformation between different time scales is derived.Based mainly on Damour and Deruelle's binary model and the TEMPO2 software,the timing model resolving the time delay from emission to arrival is established with the solar system,interstellar and binary system delays included,which retains all the terms exceeding 1 ns.The navigation measurement equation is thus built and the parameters needed are listed.A comparison is made between Sheikh's measurement equation and that in this paper.It is found that it is not necessary to introduce a fictitious SSB arrival time as Sheikh does.Near-Earth navigation for the Keplerian orbit in the case of observing one pulsar is investigated.The uncertainties of the orbit parameters are derived using the least square fitting method,which reveals that observing one pulsar enables one to navigate near-Earth spacecraft.

Determination of epidemic parameters from early phase fatality data： A case study of the 2009 A（H1N1） pandemic in Europe

This paper demonstrates that the susceptible-infected-removed （SIR） model applied to the early phase of an epidemic can be used to determine epidemic parameters reliably. As a case study, the SIR model is applied to the fatality data of the 2009 fall wave cycle of the A（H1N1） pandemic in 12 European countries. It is observed that the best estimates of the basic reproduction number R0 and the mean duration of the infection period l/r/ lie on a curve in the scatterplots, indicating the existence of a nearly-invariant quantity which corresponds to the duration of the epidemic. Spline interpolation applied to the early phase of the epidemic, an approximately 10-week period, together with a future control point in the stabilization region, is sufficient to estimate model parameters. The SIR model is run over a wide range of parameters and estimates of R0 in the range 1.2- 2.0 match the values in the literature. The duration of the infection period, 1/η is estimated to be in the range 2.0-7.0 days. Longer infection periods are tied to spatial characteristics of the spread of the epidemic.

Modelling envelope and temporal fine structure components of frequency-following responses in rat inferior colliculus

In studies of auditory perception, a dichotomy between envelope and temporal fine structure(TFS) has been emphasized. It has been shown that frequency-following responses(FFRs) in the rat inferior colliculus can be divided into the envelope component(FFREnv)and the temporal fine structure component(FFRTFS). However, the existing FFR models cannot successfully separate FFREnv and FFRTFS. This study was to develop a new FFR model to effectively distinguish FFREnv from FFRTFS by both combining the advantages of the two existing FFR models and simultaneously adding cellular properties of inferior colliculus neurons. To evaluate the validity of the present model, correlations between simulated FFRs and experimental data from the rat inferior colliculus were calculated. Different model parameters were tested, FFRs were calculated, and the parameters with highest prediction were chosen to establish an ideal FFR model. The results indicate that the new FFR model can provide reliable predictions for experimentally obtained FFREnv and FFRTFS.

Time-varying Performance Prediction and System Identification of Internal Combustion Engines

Design for life-time performance and proper maintenance measures are usually needed to prolong the mean-time-between-failures of complex equipments such as internal combustion engines.To reach this,it is important to obtain the information of time-varying system performance in design stage and to identify the structural change at each moment.So a multidisciplinary model based method is studied in this paper to unify the time-varying performance(TVP) prediction and system identification(SI) of equipments.The related multidisciplinary model in this paper should be not only precise to give simulation results but also sensitive to the variation of system parameters.So the varying history of system performance along with the structural change can be obtained from the model.Then the value of system parameters can be identified by seeking roots with given detected responding data and relationship between system responding data and system parameters.A case study on a low power gasoline engine shows that the method presented in this paper can provide useful information for the development and maintenance of complex equipments.