A Novel Model for Describing Rail Weld Irregularities and Predicting Wheel-Rail Forces Using a Machine Learning Approach

Rail weld irregularities are one of the primary excitation sources for vehicle-track interaction dynamics in modern high-speed railways.They can cause significant wheel-rail dynamic interactions,leading to wheel-rail noise,component damage,and deterioration.Few researchers have employed the vehicle-track interaction dynamic model to study the dynamic interactions between wheel and rail induced by rail weld geometry irregularities.However,the cosine wave model used to simulate rail weld irregularities mainly focuses on the maximum value and neglects the geometric shape.In this study,novel theoretical models were developed for three categories of rail weld irregularities,based on measurements of the high-speed railway from Beijing to Shanghai.The vertical dynamic forces in the time and frequency domains were compared under different running speeds.These forces generated by the rail weld irregularities that were measured and modeled,respectively,were compared to validate the accuracy of the proposed model.Finally,based on the numerical study,the impact force due to rail weld irrregularity is modeled using an Artificial Neural Network(ANN),and the optimum combination of parameters for this model is found.The results showed that the proposed model provided a more accurate wheel/rail dynamic evaluation caused by rail weld irregularities than that established in the literature.The ANN model used in this paper can effectively predict the impact force due to rail weld irrregularity while reducing the computation time.

Layer-Constrained Triangulated Irregular Network Algorithm Based on Ground Penetrating Radar Data and Its Application

In this paper,a layer-constrained triangulated irregular network（ LC-TIN） algorithm is proposed for three-dimensional（ 3 D） modelling,and applied to construct a 3 D model for geological disease information based on ground penetrating radar（ GPR） data. Compared with the traditional TIN algorithm,the LCTIN algorithm introduced a layer constraint to the discrete data points during the 3 D modelling process,and it can dynamically construct networks from layer to layer and implement 3 D modelling for arbitrary shapes with high precision. The experimental results validated this method,the proposed algorithm not only can maintain the rationality of triangulation network,but also can obtain a good generation speed. In addition,the algorithm is also introduced to our self-developed 3 D visualization platform,which utilized GPR data to model geological diseases. Therefore the feasibility of the algorithm is verified in the practical application.

A Control Simulation Method of High-Speed Trains on Railway Network with Irregular Influence

Based on the discrete time method, an effective movement control model is designed for a group of high- speed trains on a rail network. The purpose of the model is to investigate the specific traffic characteristics of high-speed trains under the interruption of stochastic irregular events. In the model, the high-speed rail traffic system is supposed to be equipped with the moving-block signalling system to guarantee maximum traversing capacity of the railway. To keep the safety of trains＇ movements, some operational strategies are proposed to control the movements of trains in the model, including traction operation, braking operation, and entering-station operation. The numerical simulations show that the designed model can well describe the movements of high-speed trains on the rail network. The research results can provide the useful information not only for investigating the propagation features of relevant delays under the irregular disturbance but also for rerouting and reseheduling trains on the rail network.

The Analysis of Frequency Reuse Irregular Patterns in Mobile Telephone Network

Some frequency reuse irregular patterns in radionetwork design are proposed,the characteristic and applica-tion measures of these patterns are analyzed.Then this paperaccounts that frequency reuse irregular patterns is a usefulway to impove spectrum efficiency and it is significative forartificial intelligence to be applied in this field.

An analysis method for correlation between catenary irregularities and pantograph-catenary contact force

Pantograph-catenary contact force provides the main basis for evaluation of current quality collection; however,the pantograph-catenary contact force is largely affected by the catenary irregularities.To analyze the correlated relationship between catenary irregularities and pantograph-catenary contact force,a method based on nonlinear auto-regressive with exogenous input(NARX) neural networks was developed.First,to collect the test data of catenary irregularities and contact force,the pantograph/catenary dynamics model was established and dynamic simulation was conducted using MATLAB/Simulink.Second,catenary irregularities were used as the input to NARX neural network and the contact force was determined as output of the NARX neural network,in which the neural network was trained by an improved training mechanism based on the regularization algorithm.The simulation results show that the testing error and correlation coefficient are 0.1100 and 0.8029,respectively,and the prediction accuracy is satisfactory.And the comparisons with other algorithms indicate the validity and superiority of the proposed approach.

EXTENDED MONTE CARLO LOCALIZATION ALGORITHM FOR MOBILE SENSOR NETWORKS

A real-world localization system for wireless sensor networks that adapts for mobility and irregular radio propagation model is considered. The traditional range-based techniques and recent range-free localization schemes are not well competent for localization in mobile sensor networks, while the probabilistic approach of Bayesian filtering with particle-based density representations provides a comprehensive solution to such localization problem. Monte Carlo localization is a Bayesian filtering method that approximates the mobile node's location by a set of weighted particles. In this paper, an enhanced Monte Carlo localization algorithm-Extended Monte Carlo Localization (Ext-MCL) is proposed, i.e., the traditional Monte Carlo localization algorithm is improved and extended to make it suitable for the practical wireless network environment where the radio propagation model is irregular. Simulation results show the proposal gets better localization accuracy and higher localizable node number than previously proposed Monte Carlo localization schemes not only for ideal radio model, but also for irregular one.

LDM (Layered Deployment Model): A Novel Framework to Deploy Sensors in an Irregular Terrain

Deployment of sensors in any irregular terrain with 100% coverage and connectivity is a challenging issue in the field of Wireless Sensor Networks. Traditional deployments often assume homogeneous environments, which ignore the effect of terrain profile as well as the in-network obstacles situated randomly like buildings, trees, roads and so on. Proper deployment of sensors in such irregular region and its corresponding routing is one of the most fundamental challenges of Wireless Sensor Networks. In this work, we have considered that the terrain is irregular in shape and there may be obstacles within the terrain in any random position with any random shape, which is the reality in real world. With this novel framework, we have shown that an opti-mum deployment can be achieved in such irregular terrain without compromising coverage as well as con-nectivity between the sensor nodes for effective routing.

Searching for Strange Attractor in Sliver Irregularity Series

The chaotic nonlinear time series method is applied to analyze the sliver irregularity in textile processing.Because it unifies the system's determinacy and randomness,it seems more adaptive to describe the sliver irregularity than conventional methods.Firstly,the chaos character,i.e.fractal dimension,positive Lyapunov exponent,and state space parameters,including time delay and reconstruction dimension,are calculated respectively.As a result,a positive Lyapunov exponent and a fractal dimension are obtained,which demonstrates that the system is chaotic in fact.Secondly,both local linear forecast and global forecast models based on the reconstructed state are adopted to predict a segment part of the sliver irregularity series,which proves the validity of this analysis.Therefore,the sliver irregularity series shows the evidence of chaotic phenomena,and thus laying the theoretical foundation for analyzing and modeling the sliver irregularity series by applying the chaos theory,and providing a new way to understand the complexity of the sliver irregularity much better.

面向电力输电场景概略点云的无人机巡航线自动生成

针对电力输电场景无人机自动巡检技术需求,提出一种基于概略点云的无人机巡航线自动生成方法。首先,采用基于边长约束的不规则三角网(TIN)区域生长方法,将电力输电场景概略点云区分割成不同区域,并在此基础上基于区域面积约束提取电力线走廊点云;其次,利用最小外接矩形方法实现电力线走廊点云分段处理,并根据每段点集的最小外接矩形信息生成初始巡航线位置;最后,采用点集中心进行初始巡航线的位置优化,实现巡航线的快速生成。实验结果表明,所提方法能快速实现电力输电场景点云的巡航线自动生成,对无人机巡航线实时提取及自主巡航具有一定的应用价值。

Road Network Analysis with GIS and GRASS-GIS:A Probabilistic Approach

In this paper we show how it can be useful to the probability of intersections in the determination of a classification rule for raster conversions in Geographical Information System(GIS)and GRASS GIS for the Road Network Analysis(RNA).We use a geometric probabilities approach for irregular path considering these results for transportation planning operations.We study two particular problems with irregular tessellations,in order to have a situation more realistic respect to map GIS and considering also the maximum value of probability to narrow the range of possible probability values.

多源车载数据驱动的地铁轨道不平顺智能识别方法

针对轨道不平顺检测成本高与时效性低等不足,从车辆动态响应与轨道不平顺之间的相关性为切入点,提出一种多源车载数据驱动的轨道不平顺智能识别方法。首先,建立地铁车辆系统动力学模型,获取车辆振动与运动姿态响应数据;其次,通过相关性分析算法,选取强相关性数据,制作网络模型数据集;最后,建立卷积神经网络-长短期记忆网络(CNN-LSTM),通过粒子群算法优化(PSO)神经网络模型参数,建立PSO-CNN-LSTM模型,实现对轨道不平顺的识别拟合。研究结果表明:在车辆动态响应信号中,与横向信号与轨道不平顺之间的相关性相比,垂向信号的更强,同时,车体的运动姿态如车体点头角速度与不平顺有明显的相关性。所提出的PSO-CNN-LSTM模型轨道垂向与横向不平顺识别拟合度分别达0.92和0.76。与经典的全连接神经网络FCNN和支持向量机SVR相比,PSO-CNN-LSTM有更好的识别效果与时效性。

滚动球变换支持下的TIN-DDM地形特征线自动提取方法

针对传统规则格网数字高程模型地形特征线提取方法中存在阈值难以定量调控、连接方式无法自适应调整以及地形特征线类型不完整的问题,将滚动球变换模型应用于不规则三角网数字水深模型(triangulated irregular network digital depth model,TIN-DDM)地形特征线的自动提取,在构建临界滚动球半径关联的地形特征点定量识别判定准则基础上,引入地形形态边界点概念,采用逆向工程的建模思路,建立了以剖分单元为基础的地形特征线自动提取模型,结合地形类型判定准则的多尺度表达特性及顾及水深数值的地形特征优化模型,提出了可多尺度表达且类型完整的地形特征线自动提取方法。试验结果表明:相比于经典地表流水模拟方法,该方法可实现完整、连续且细分的TIN-DDM地形特征线自动提取及多尺度表达,且提取的地形特征线具有更高的地形重构精度。

麻家边水库库容测量与变化分析

以麻家边水库为例,根据中小型水库库容测量的工作内容,探讨无人机载激光雷达测量、无人船自动测量、等高线法和不规则三角网法等先进方法在水库库容测量中的应用,并通过实地调查和对比不同时间的卫星影像资料合理解释水库库容变化,可为类似项目提供有益参考。

基于ArcEngine的等高线质量检查方法研究

等高线的精准描绘是地图绘制的关键问题。本文针对地图生产中等高线与高程注记点,研究可能出现的矛盾类型,引入基于等高线和高程注记点的不规则三角网检查方法与放射检查方法,以期为生产实践提供参考。基于ArcEngine组件库和C#语言,开发了等高线质量检查程序,并用实例数据对其检查结果的准确度进行了验证。研究结果表明:等高线节点构建不规则三角网进行点线矛盾的检查方法具有较高的精度,可满足项目生产需求。

基于改进SURF的增强现实图像匹配方法

针对传统增强现实图像匹配算法鲁棒性不强且效率不高的问题,提出一种改进的SURF匹配算法。首先,使用SURF算法进行特征点检测,并通过Haar小波模板确定特征主方向,在得到特征主方向后构建特征描述符;由于传统SURF算法采用高达64维的矩形描述符,导致算法的计算量非常大,并且鲁棒性不强。因此,该文使用DAISY圆形描述符替代原始算法中的矩形描述符,DAISY是三层同心圆结构,每层包含8个采样点,可以得到25个维度的描述符,这种结构使得算法的鲁棒性大大增强并且降低了计算复杂度;接着,使用特征描述符计算欧氏距离进行特征点匹配;最后,对得到的匹配点集使用随机抽样一致(RANSAC)与三角不规则网络(TIN)算法进行优化,剔除误匹配点。实验结果表明,该算法虽然略微增加了时间复杂度,但鲁棒性变得更强,并且算法的效率和匹配精度也大大提高,平均精度达到了95%以上。

非常规油气藏不规则复杂裂缝表征方法

非常规油气资源储量大,开发难度高。储层压裂改造是非常规油气资源开发的关键技术手段。天然裂缝和压裂裂缝具有不规则性和复杂性。针对现有裂缝表征方法难以准确刻画裂缝真实形状和宽度变化等不规则特性这一问题,提出了基于非结构PEBI网格的不规则复杂裂缝表征方法。首先,建立基于PEBI网格的天然裂缝表征流程,实现对任意区域或限定区域天然裂缝的准确表征;其次,建立基于Delaunay三角形网格和PEBI网格的压裂裂缝表征及优化方法,分析网格尺寸及优化次数对裂缝表征精度的影响;第三,建立基于非结构网格的非平面裂缝表征方法,实现对弯曲裂缝的刻画和表征,裂缝形态和分布与实际情况更相符;第四,提出非均匀裂缝宽度表征方法,实现对变宽度即同一条裂缝宽度和导流能力不均匀分布裂缝的精细表征;第五,在全区域和限定区域内,实现耦合不规则压裂裂缝和不规则天然裂缝的复杂缝网表征。针对大规模天然裂缝与压裂裂缝相交、裂缝宽度非均匀分布、非平面裂缝等复杂条件下的裂缝网络表征,通过调整网格优化次数,能够提高缝网表征质量。利用PEBI网格能够灵活准确逼近裂缝复杂边界条件的优势,实现快速、准确地显示处理大量不规则天然裂缝和压裂裂缝。形成的不规则复杂裂缝表征方法,有助于提高非常规油气藏裂缝网络的表征精度和数值模拟计算准确性。

深度神经网络在不规则弥漫大B细胞淋巴瘤时间序列数据分类预测中的应用

目的探讨深度神经网络在不规则时间序列数据中的分类效果,并对山西某医院2014-2020年362例弥漫大B细胞淋巴瘤(diffuse large B-cell lymphoma,DLBCL)患者进行复发预测。方法回顾性地收集了确诊且治疗后达到完全缓解的362例DLBCL患者的病例资料,并预测其两年内的复发。先利用LASSO回归进行变量的筛选,再构建基于GRU-ODE-Bayes(gated recurrent unirt-ordinary differential equation-Bayes)的不规则时间序列深度神经网络模型,并与传统模型及其他深度神经网络模型进行比较。结果在本文的所有模型中,传统模型的分类性能不及深度神经网络模型。其中GRU-ODE-Bayes模型最优,其AUC为0.85,灵敏度为0.84,特异度为0.71,G-means为0.77。结论关于不规则DLBCL时间序列数据,与本文其他模型相比,GRU-ODE-Bayes模型可以更精准地预测DLBCL患者的复发情况,可为患者个性化治疗和医生决策提供参考。

基于GWO-CNN-LSTM的铁路轨道高低不平顺值反演模型研究

为利用列车的车体振动加速度来准确反演铁路的轨道高低不平顺值,提出一种基于灰狼优化算法(GWO)、卷积神经网络(CNN)和长短期记忆网络(LSTM)构建车体振动加速度与轨道高低不平顺值的关系模型(GWO-CNN-LSTM)。首先,将轨检车的实测数据根据轨道检测数据特点,采用莱因达准则进行异常值剔除的预处理;然后,利用处理后的数据以车体振动加速度作为模型的输入,以轨道高低不平顺值作为模型的输出,利用CNN学习车体振动加速度的波形信息,将CNN学习到的特征输入LSTM;最后,再用GWO优化LSTM模型的关键参数,进而反演出轨道高低不平顺值,为突出该模型的适应性,又随机选取了其他区段进行了反演。文中采用3个性能指标来评价模型,并与其他经典的方法进行对比分析。结果表明,GWO-CNN-LSTM模型的均方根误差、平均绝对误差和拟合优度分别为0.141、0.107和0.977,与单一的LSTM相比,GWO-CNN结合LSTM可以使拟合优度提高20.1%;与循环神经网络、BP神经网络和支持向量回归相比,所提出的GWO-CNN-LSTM模型其均方根误差降低68.1%~79.1%,平均绝对误差降低60.4%~68.3%,拟合优度提高27.7%~44.9%,验证了GWO-CNN-LSTM模型用于轨道高低不平顺值反演的有效性。GWO-CNN-LSTM模型为铁路轨道高低不平顺反演研究提供了新的思路。

卫星斜视影像地形遮蔽区的自动检测研究

针对卫星斜视影像普遍存在的地形遮挡问题,文章提出了基于有理多项式系数(Rational Polynomial Coefficients,RPC)模型的高分辨率卫星影像遮挡检测方法。利用不规则三角网(Triangulated Irregular Network,TIN)表达三维地形表面,并采用光线追踪法精确探测入射光线与三角网的相交情况以确定交点间的遮挡关系。为提高遮挡检测效率,设计了一种极小化地形搜索区的方法,并利用斜视成像的真实卫星影像及模拟卫星影像进行了实验验证,结果表明该方法对遮挡区域的检测精度高于97%,能有效地识别出受地形遮挡影响的区域,自动生成准确的地形遮蔽区域掩膜影像。

基于道路矢量数据的路口地理实体面图元提取

地理实体数据是实景三维数据中的重要组成部分,路口实体在城市实景三维中有着重要作用,当前路口实体面图元的提取是通过人工查找路口位置并在路口位置处手动构建路口面,这种提取方式容易漏掉路口位置且有人工编辑工作量较多的问题。为准确定位路口位置且在路口位置处高效构建路口面图元,本文设计了一种基于道路矢量数据的路口地理实体面图元的提取方法,通过对道路面数据构建不规则三角网,分析不规则三角网(TIN)在路口处的特点,准确定位出路口位置,并在路口处自动构建路口面图元。本文基于湖南省基础测绘道路矢量数据,对基于道路矢量数据的路口地理实体面图元提取方法的有效性进行验证,实验结果表明,该方法路口定位准确率达97.5%,路口面构建较为准确,可以有效缩短路口实体生产周期,研究结果可以应用于其他条带状交叉口实体类的定位和构建。