基于模糊神经网络的短时交通流预测方法研究

为满足交通控制和诱导系统的实时性需求,减少交通拥挤状况,降低交通事故突发频率,需要对短时交通流进行预测。当前的短时交通流预测方法是采用K-近邻的非参数回归对其进行预测,预测过程中没有将预测模型中关键因素对交通流的影响进行详细的说明,导致预测结果不准确,存在短时交通流预测误差较大的问题。为此,提出一种基于模糊神经网络的短时交通流预测方法。该方法首先以历史短时交通流数据样本序列为基础,将提取的关联维数作为短时交通流的混沌特征量,然后以该特征量为依据,对短时交通流数据进行聚类,使相同的短时交通流聚合类样本比不同的交通流聚合类样本更为贴近,采用高斯过程回归对短时交通流预测模型进行建设,建设过程中利用差分方法对短时交通流预测序列进行平稳化操作之后,对短时交通流预测模型进行训练,将GPR模型引入至短时交通流预测过程中,得到交通流预测方差估计值,并确定交通流预测值置信区间,由此实现短时交通流的预测。由此实现短时交通流的预测。实验结果证明,所提方法可以准确地预测交通运输系统的实时状况,为车辆行驶的最佳路线进行了有效引导,减少了自然影响方面和人为因素对短时交通流预测结果的干扰,为交通部门对交通路况的控制管理提供了依据。     

一种新的大气折射率剖面模型构建方法

针对现有大气折射率剖面模型中干项简化模型较多、使用中难以取舍和湿项简化模型精度较低等问题,利用某台站实测探空数据,对现有几种大气折射率剖面模型的精度进行了分析。在此基础上,基于小波聚类和EOF(经验正交函数)分析方法,构建了一个能够考虑天气特征的大气折射率剖面改进模型,并利用实测数据对其精度进行了验证。结果表明:改进模型拟合精度较现有模型精度有明显改进。其中,干项模型能够充分考虑台站本地气候特征,精度更高;湿项模型将天气类型细分为四类,比现有模型精度最大提高了约6个折射率单位。同时,改进模型天气类型划分方法较容易被非气象专业人员理解,便于此模型推广应用。     

混合状态下城市快速路交通流短时预测

建立交通流短时预测状态空间模型, 研究混合状态下城市快速路交通流短时预测. 结合城市快速路自由流状态、拥挤流状态和阻塞流状态下交通流参数的时间和空间分布特性, 基于交通流守恒方程和速度动态模型, 借鉴偏微分方程组求解时空离散的思想, 建立三种状态下交通流短时预测模型; 同时考虑进出口匝道、车道数变更以及道路坡度等因素的影响, 将交通流短时预测模型转化为交通流短时预测状态空间模型, 实现混合状态下交通流短时预测. 研究表明, 该方法能够实现混合状态下道路网内的交通流短时预测, 预测精度可达90.23%. 相同条件下, 经典自回归滑动平均模型的预测精度仅为81%. 关键词： 交通流短时预测 自由流状态 拥挤流状态 阻塞流状态     

含有误差校正的小波神经网络交通流量预测

交通流量的准确预测对于高速路管理者进行决策至关重要。建立了小波神经网络(WNN)交通流量预测模型,并通过预测训练误差和测试误差校正预测结果来提高预测精度。首先构建WNN模型对交通流量进行初步预测,然后利用经验模态分解(EMD)和WNN模型对训练误差和测试误差进行预测。分别用训练误差预测值、测试误差预测值和两种误差预测值的加权对流量初步预测结果进行修正得到最终预测值。采用四川省成灌高速路交通流量数据进行了仿真对比实验,仿真结果表明含有误差校正的小波神经网络模型能有效提高交通流量预测精度,并且利用两种误差加权修正模型的预测精度高于利用测试误差的修正模型和利用训练误差的修正模型。     

应用可见-近红外光谱技术进行白醋品牌和pH值的快速检测

提出了一种基于可见-近红外透射光谱技术快速判别白醋品牌和测定pH值的方法。应用可见-近红外透射光谱获取不同品牌白醋的透射光谱曲线,并对获得的原始光谱数据进行平滑、变量标准化以及一阶导数等预处理,然后利用主成分分析对原始光谱数据进行聚类分析,根据主成分的累计贡献率选取主成分数,并将所选取的主成分作为三层BP神经网络的输入。通过定标集样本对BP神经网络进行训练,得到三层优化神经网络结构：5输入层节点,6隐含层节点和2输出层节点,各层传递函数均采用Sigmoid函数。利用该模型对预测集样本进行预测。实验结果表明在阈值设定为±0.1的情况下该模型对预测集样本品牌鉴别准确率达到了100%,pH预测值与实际测量值偏差小于5%,得到了理想的结果。所以利用可见-近红外光谱技术结合主成分分析和神经网络算法能够快速准确的判定白醋品牌和pH值。     

基于模糊树模型的混沌时间序列预测

应用模糊树模型,对混沌时间序列进行建模和预测.该方法可以根据建模数据在空间中的分布信息,基于二叉树结构自适应划分输入空间,得到模糊子空间,在与叶节点对应的子空间上建立线性函数作为模糊规则的后件,用隶属度函数将各分片线性函数光滑连接,最后得到一个精度比较高的非线性映射.通过对Mackey-Glass、Lorenz和Henon混沌时间序列的建模和预测研究,仿真结果表明,该方法具有建模精度高、运行速度快、泛化能力强、预测步数多、适用范围广等优点. 关键词： 模糊树模型 混沌时间序列 预测     

纺织品近红外光谱定性分析的一种新方法

近红外光谱分析技术可用于对样本的快速无损检测,在人们的生产和生活中发挥着越来越重要的作用。支持向量机是建立定性分析模型的常用方法,可通过寻找最优分类超平面将两类样本分开。在小样本情况下,支持向量机方法有其独特的优势。主成分分析是常用的数据降维方法,可将数据降维之后作为支持向量机方法的输入变量,简化模型并提高模型识别的准确性。因此,基于主成分分析的支持向量机(简称PCA-SVM)适合用于建立近红外光谱定性分析模型。多模型方法是人们使用较少的建模方法,用该方法建立的模型一般具有较好的稳定性。将多模型方法与PCA-SVM方法成功结合形成了新方法。以棉锦混合、棉涤混合纺织品为例,用新方法建立了这两类纺织品样本的近红外光谱定性分析模型。建模时将光谱数据按照波长分为4组,用每组光谱数据建立一个子模型,将子模型的输出值进行加权平均便得到最终的预测结果。这样可以更充分地使用光谱数据中所包含的信息。为了便于对比不同的方法,仍使用上述校正集和验证集,又用PCA-SVM方法建立了这两类纺织品样本的近红外光谱定性分析模型。对预测结果做交叉验证,用新方法所建模型判别的正确率的平均值为85.49%,正确率的标准差为0.066 7, 用PCA-SVM方法所建模型判别的正确率的平均值为83.34%,正确率的标准差为0.109 6。研究结果表明用新方法所建模型的分类效果好于用PCA-SVM方法所建模型的分类效果;用新方法建立的模型的稳定性明显高于用PCA-SVM方法建立的模型的稳定性。用PCA-SVM方法所建模型的预测效果受校正集构成情况的影响较大,而用新方法所建模型的预测效果则相对稳定。对废旧纺织品进行分类回收可大量节约纺织原材料,但采用人工分拣方式效率低且成本高。采用近红外光谱分析方法对纺织品进行分类,为废旧纺织品的大规模精细分拣和分级奠定了一定的基础。该新方法有望用于某些其他类型样本的分类。     

基于三维荧光光谱的Krawtchouk图像矩算法在多环芳烃定量分析中的应用

以多环芳烃中的芴和苊为研究对象,提出一种将三维荧光光谱技术与Krawtchouk图像矩、广义回归神经网络相结合的定量分析的方法。利用FS920荧光光谱仪获取样品的三维荧光光谱数据,得到对应的三维光谱灰度图。直接计算三维光谱灰度图的Krawtchouk矩,将得到的Krawtchouk矩经平均影响值筛选后作为广义回归神经网络的输入,建立多环芳烃(PAHs)的定量模型。预测8组混合溶液的测试样本,芴和苊的平均相对误差分别为0.98%和2.15%。研究结果表明,Krawtchouk矩经过筛选后预测结果更为准确,该方法能够有效提取光谱的特征信息,简单、准确的预测PAHs的浓度。     

基于可见-近红外光谱的可乐品牌鉴别方法研究

提出了一种采用可见-近红外光谱分析技术快速鉴别可乐品牌的新方法。采用美国ASD公司的便携式光谱仪对三种不同品牌的可乐进行光谱分析,各获取55个样本数据。将样本随机分成150个建模样本和15个预测样本,采用平均平滑法和标准归一化方法对样本数据进行预处理,再用主成分分析法对光谱数据进行聚类分析并获得各主成分数据。将建模样本的主成分数据作为BP网络的输入变量,可乐品牌作为输出变量,建立三层人工神经网络鉴别模型,并用模型对15个预测样本进行预测。结果表明,预测准确率为100%,实现了可乐品牌快速、准确的鉴别。     

基于递阶模糊聚类的混沌时间序列预测

提出一种新的基于递阶模糊聚类系统的模糊建模方法.目的在于通过一系列的步骤优化T-S模糊模型结构，实现非线性系统的建模和预测.首先利用最近邻聚类法初始划分输入空间，得到规则数及初始聚类中心，用模糊C均值算法(FCM)进一步优化聚类中心；然后利用加权最小二乘法估计模糊模型的初始参数，进一步利用带遗忘因子的递推最小二乘法优化结论参数.采用该方法对Mackey-Glass混沌时间序列进行预测实验，结果表明可以对Mackey-Glass混沌时间序列进行准确建模和预测，证明了本方法的有效性. 关键词： 递阶模糊聚类 模糊建模 混沌时间序列 最小二乘     

Mesoscale models and their application to air quality studies

Summary The use of mesoscale models in conjunction with air quality models enables better prediction of pollution transport and dispersion in the atmosphere. The accuracy of the prediction depends upon the physical processes included in the models, on the mathematical representation, and on the initial meteorological and other external data. In most situations large power plants are located near the coast or in hilly terrain. Under these situations the meteorological fields which are relevant to air quality studies change tremendously with time and space (both vertically and horizontally). So far most of the air quality studies and consequently the decisions regarding sitting and management of power plants in coastal or hilly sites are made according to simple Gaussian models which are based on limited observational data, and, therefore, are of doubtful meanings. To obtain better prediction of pollutant concentration dense meteorological observations have to be made. Since such observations are very expensive, it is unrealistic to require them. A reliable mesoscal model, however, can supply the necessary meteorological fields with relatively little expenses with any needed resolution, providing that appropriate computers are available. Model results can be very useful in prediction of maximum concentration that can be expected from routine simultaneous release from several sources during critical weather conditions. Since the late seventies, with the huge development of large computers, it has been made possible also to carry out three-dimensional simulations. Lately with the development of advanced microcomputers and appropriate software it is even possible to perform two- and even three-dimensional experiments on such computers within a reasonable time frame.     

Predicting ramps by integrating different sorts of information

Although predicting sudden rapid changes of renewable energy outputs is useful for maintaining the stability of power grids with many renewable energy resources, the prediction is difficult so far. Here we list causes for the uncertainty for our prediction, quantify them, and forecast whether such sudden rapid changes are likely to happen or not by integrating their quantifications with a method of machine learning. We test the proposed forecast using a toy model and real datasets of solar irradiance and wind speed.     

基于大气激光通信系统的实验测量研究

文中介绍了激光通过大气随机信道的光强分布测试系统的结构、工作原理和关键技术,在设计激光光斑测量系统基础上,进行了不同气象条件下的视距实验测量,给出了不同天气情况下的测量结果并进行了分析和讨论。用此系统可以方便、准确、形象地观测激光通过大气随机信道后光强的变化情况,为进一步建立大气随机信道模型提供了有力的依据。     

基于跨平台红外高光谱观测的对流层三维风场测量

精确的风场数据对提高数值天气预报准确性具有重要意义,对流层风是改进天气预报的要素之一。虽然利用气象卫星成像仪对连续云图追踪特征目标进行导风是一种有效的风场观测方法,且在区域和全球尺度上改善了数值天气预报,但仍存在风场高度分配模糊问题而产生误差。星基红外高光谱探测仪具备大气温湿度廓线垂直探测能力,通过分析各个垂直分层内的大气参数运动得到三维风场,能够提升风场垂直高度的准确性,改进风场高度分配模糊问题。提出了利用跨平台极轨气象卫星FY-3D星红外高光谱大气探测仪HIRAS和NOAA-20星跨轨红外探测仪CrIS交叉观测对流层三维风场的创新方法,根据两仪器近重叠轨道星下点交叉观测辐射数据匹配水汽通道图像,通过稠密光流法分析目标运动变化并计算风场,对风矢量进行质量控制后同ERA-Interim再分析资料作定量化比较,分析风速均值绝对偏差、均方根误差和风向均值绝对偏差。分别对2019年2月20日UTC世界时00:00,06:00,12:00的HIRAS和CrIS交叉数据计算200,300,400,600,650和1 000 hPa六组垂直高度风场,结果表明,风速范围的变化趋势与再分析资料表现一致,风速范围随高度降低而减小,高层对20 m·s-1以上风速更敏感,地表附近测得风速集中在10 m·s-1以内。风速均值绝对偏差多数小于3 m·s-1,最大不超过4 m·s-1,风速均方根误差多数小于3.5 m·s-1,最大不超过4.5 m·s-1,风向均值绝对偏差多数小于30°,最大不超过40°。风场误差主要来自仪器自身设计参数不同引入辐射数据的观测偏差,以及因数据空间分辨率不同导致在图像重投影处理过程中引入的定位偏差。     

一种基于支持向量机回归的旋转遮光带日射表散射辐照度修正新算法

通过比较旋转遮光带日射表(RSP)和参考标准表所测散射辐照度之间的数值差异,分析了太阳总辐照度、环境温度、相对湿度和太阳光谱等气象要素对RSP散射辐照度测量误差的影响关系,提出了一种修正RSP散射辐照度测量值的新算法。该算法从支持向量机回归预测角度,建立了对RSP散射误差修正值的一次预测模型,然后根据误差修正值最优预测模型推导出RSP散射辐照度修正算法模型。利用该算法对美国国家太阳辐射研究实验室和劳里观测站采集的RSP散射辐照度数据进行修正,修正后两观测站数据的平均偏差和均方根误差分别降低到-0.2W/m2,3.3W/m2和1.9W/m2,8.5W/m2,显示算法具有良好的修正性能和适用性。该算法能够有效避免Vignola算法中存在的欠修正和Vignola and Augustyn(VA)算法中存在的过修正现象。     

A Fault Diagnosis Method Considering Meteorological Factors for Transmission Networks Based on P Systems

Bad meteorological conditions may reduce the reliability of power communication equipment, which can increase the distortion possibility of fault information in the communication process, hence raising its uncertainty and incompleteness. To address the issue, this paper proposes a fault diagnosis method for transmission networks considering meteorological factors. Firstly, a spiking neural P system considering a meteorological living environment and its matrix reasoning algorithm are designed. Secondly, based on the topology structure of the target power transmission network and the action logic of its protection devices, a diagnosis model based on the spiking neural P system considering the meteorological living environment is built for each suspicious fault transmission line. Following this, the action messages of protection devices and corresponding temporal order information are used to obtain initial pulse values of input neurons of the diagnosis model, which are then modified with the gray fuzzy theory. Finally, the matrix reasoning algorithm of each model is executed in a parallel manner to obtain diagnosis results. Experiment results achieved out on IEEE 39-bus system show the feasibility and effectiveness of the proposed method.     

Investigations on the effect of process parameters on the growth of vertically oriented graphene sheet in plasma-enhanced chemical vapour deposition system

A multistage numerical model comprising the plasma kinetics and surface deposition sub-models is developed to study the influence of process parameters, namely, total gas pressure and input plasma power on the plasma chemistry and growth characteristics of vertically oriented graphene sheets (VOGS) grown in the plasma-enhanced chemical vapour deposition system containing the Ar + H₂ + C₂H₂ reactive gas mixture. The spectral and spatial distributions of temperature and number densities, respectively, of plasma species, that is, charged and neutral species in the plasma reactor, are examined using inductively coupled plasma module of COMSOL Multiphysics 5.2 modelling suite. The numerical data from the computational plasma model are fed as the input parameters for the surface deposition model, and from the simulation results, it is found that there is a significant drop in the densities of various plasma species as one goes from the bulk plasma region to the substrate surface. The significant loss of the energetic electrons is observed in the plasma region at high pressure (for constant input power) and low input power (for constant gas pressure). At low pressure, the carbon species generate at higher rates on the catalyst nanoislands surface, thus enhancing the growth and surface density of VOGS. However, it is found that VOGS growth rate increases when input plasma power is raised from 100 to 300 W and decreases with further increase in the plasma power. A good comparison of the model outcomes with the available experimental results confirms the adequacy of the present model.     

Impact of aerosols on the forecast accuracy of solar irradiance calculated by a numerical weather prediction model

The impact of aerosols on the forecast accuracy of solar irradiance calculated by a fine-scale, one day-ahead, and operational numerical weather prediction model (NWP) is investigated in this study. In order to investigate the impact of aerosols only, the clear sky period is chosen, which is defined as when there are no clouds in the observation data and in the forecast data at the same time. The evaluation of the forecast accuracy of the solar irradiance is done at a single observation point that is sometimes affected by aerosol events. The analysis period is one year from April 2010 to March 2011. During the clear sky period, the root mean square errors (RMSE) of the global horizontal irradiance (GHI), direct normal irradiance (DNI), and diffuse horizontal irradiance (DHI) are 40.0?W?m^?2, 84.0?Wm^?2, and 47.9?W?m^?2, respectively. During one extreme event, the RMSEs of the GHI, DNI, and DHI are 70.1?W?m^?2, 211.6?W?m^?2, and 141.7?W?m^?2, respectively. It is revealed that the extreme events were caused by aerosols such as dust or haze. In order to investigate the impact of the aerosols, the sensitivity experiments of the aerosol optical depth (AOD) for the extreme events are executed. The best result is obtained by changing the AOD to 2.5 times the original AOD. This changed AOD is consistent with the satellite observation. Thus, it is our conclusion that an accurate aerosol forecast is important for the forecast accuracy of the solar irradiance.     

Application of the Transmission Line Matrix method for outdoor sound propagation modelling – Part 2: Experimental validation using meteorological data derived from the meso-scale model Meso-NH

Outdoor sound prediction is both a societal concern and a scientific issue. This paper deals with numerical simulations of micrometeorological (temperature and wind) fields for environmental acoustics. These simulations are carried out using the reference meso-scale meteorological model at the Meteo-France weather agency (Meso-NH). Meso-NH predictions at very fine scales (up to 3 m), including new developments (drag force approach), are validated both numerically and experimentally under stable, unstable and neutral conditions. Then, this information can be used as input data for the acoustic propagation model. The time-domain acoustic model is based on the Transmission Line Matrix method. Its development has also been promoted for application to outdoor sound propagation, i.e. to take into account topography, ground impedance, meteorological conditions, etc. In part 1, the presentation and evaluation of the Transmission Line Matrix method showed the relevance of this method’s use in the context of environmental acoustics. Finally, simulated noise levels under different propagation conditions were compared to in situ measurements. Satisfactory results were obtained regarding the variability of the observed phenomena.     

Forecasting the daily electricity consumption in the Moscow region using artificial neural networks

In [1] we demonstrated the possibility in principle for short-term forecasting of daily volumes of passenger traffic in the Moscow metro with the help of artificial neural networks. During training and predicting, a set of the factors that affect the daily passenger traffic in the subway is passed to the input of the neural network. One of these factors is the daily power consumption in the Moscow region. Therefore, to predict the volume of the passenger traffic in the subway, we must first to solve the problem of forecasting the daily energy consumption in the Moscow region.