RailwayTrackIrregularityMeasuringbyGNSSINSIntegration.pdf资源-CSDN文库

需积分: 3 195 浏览量 2020-04-07 10:13:56 上传评论收藏 1.29MB PDF 举报

资源推荐

资源详情

资源评论

Railway Track Irregularity Measuring by GNSS/

INS Integration

QIJIN CHEN, XIAOJI NIU, QUAN ZHANG, and YAHAO CHENG

GNSS Research Center, Wuhan University

Received November 2013; Revised August 2014

ABSTRACT: Railway track irregularity measuring is a task of fundamental importance to guarantee operating

safety and arrange proper maintenance, particularly for the high-speed lines. Conventional measuring methods

cannot satisfy the requirements of accuracy and time-efficiency simultaneously. A GNSS/INS integrated technique

is proposed based on the fact that railway track irregularity detection is essentially an issue of relative surveying. Key

technologies of the integration algorithm aiming at track irregularity measuring are proposed to improve the

performance of the GNSS/INS system. Results show that the proposed method can fulfill 1 mm relative accuracy

in identifying track irregularities in the kinematic surveying mode, which means this method can satisfy the

accuracy requirement for a high-speed line and is ten times faster than the conventional method based on total

station. Copyright # 2015 Institute of Navigation

INTRODUCTION

The railway track condition tends to deteriorate

due to some external factors, such as frequent

passage of heavy trains and deformation of the track

bed. These factors make the railway track drift away

from its designed geometry thus result in track

irregularity [1]. Track irregularity, i.e., track

deformation, is one of the most important factors

that cause safety problems and further track

deterioration [2]. So detecting and ﬁxing the track

irregularity is a task of fundamental importance to

guarantee high operating safety, especially for the

high-speed lines.

The requirement of track smoothness for the high-

speed line is quite demanding, since even track

deformations with small magnitude (e.g., 2 mm/5 m

for 30-m chord [3]) would lead to considerable

dynamic vehicle response forces under high speed

operation. To detect such small deformations, the

track position should be determined with sub-

millimeter accuracy. In addition, time slots (i.e.,

skylight time) permitted for railway surveying for

the existing line are limited due to high trafﬁc

volumes. Thus ﬂexible surveying systems are

required yielding accurate data within a short time.

Nowadays, railway track irregularities are

primarily measured dynamically by the special track

inspection locomotive, which can be operated at high

speed but does not satisfy the accuracy requirement

for track correction. Lightweight track trolleys

combined with high-precision total station are widely

deployed as a supplement of the dynamic inspection.

The well-known products of the high-precision track

surveying trolley are Amberg GRP1000 by Amberg

Technologies and Trimble GEDO CE. They have

been widely deployed in China for railway cons-

truction and are ofﬁcially recommended to survey

the unloaded track geometry for track correction

following the dynamic inspection. These types of

track trolleys equipped with motorized total station

claim to measure the absolute deviation of the

railway track from its designed geometry with the

accuracy of 1 millimeter in the stop-and-go mode.

This method has two shortcomings: a) not fast

enough, e.g., we can only survey 150-m of track per

hour using this method; b) surveying accuracy is

determinedly affected by the condition of the railway

construction control network, which is not well

maintained in China for the existing line due to the

huge maintenance cost.

It can be summarized that the conventional

methods cann ot satisfy the requirements of accuracy

and time-efﬁciency simultaneously. So we propose

to use a GNSS/INS technique to meet these two

requirements.

In previous work, the inertial technique has

been widely utilized for railway track surveying and

track condition monitoring. Inertial sensors, e.g.,

NAVIGATION: Journal of The Institute of Navigation

Vol. 62, No. 1, Spring 2015

Printed in the U.S.A.

accelerometers and gyroscopes, are usually equipped

in the railway track surveying equipment to detect

the track irregularity and assess the track condition

by sensing the acceleration and angular rate change

of the corresponding platform. Typically, the inertial

sensors are mounted on the rigid structures of the

train, such as the axle-box, bogie, and car-body to

detect the vertical and lateral track irregularities

[4–11]. The accelerations are ﬁrst converted to

displacement by double integration; then the track

irregularities are detected with the displacement.

But in this case, the solutions suffer from two issues:

a) the inertial sensor drift is difﬁcult to compensate

for; b) the double integration of sensor noise and

bias causes uncontrolled error in the resulting

displacement, so band-pass ﬁlters have to be applied,

which may harm the real motion signal when the

equipment moves at low speed along the track.

The GNSS/INS integrated technique can over-

come the two shortcomings mentioned above by

fusing the GNSS update information and the inertial

outputs in a well-designed Kalman ﬁlter. But few

attempts have been made to use GNSS/INS to

detect the track irregularity in the previous work:

Lück discussed the design of track measurement

system based on INS-GPS and integration technique

[12, 13]. This system is designed for the dynamic

inspection locomotive and experimental results from

the laboratory test vehicle have been presented.

Applanix Corp. in partnership with Plasser proposed

a track surveying solution based on a GPS/INS

technique called POS/TG system [14]. POS TG is

also a position and orientation system designed

for a dynam ic inspection locomotive, whose accuracy

for 10 meter mid-chord offset measurement is

claimed to be 1mm (RMS). But we have little

knowledge on the inner technology and algorithm

of POS/TG.

Our research differs from the previous work in

the fact that the proposed method would combine

advantages of the dynamic method and the

lightweight track trolley based on total station.

The GNSS/INS integrated system is mounted on a

lightweight track trolley rather than a locomotive

to realize kinematic surveying and improve time

efﬁciency. The measurement then would be

accurate enough for track correction/adjustment in

the railway construction and maintenance project.

Challenges for using the GNSS/INS integrated

system to realize sub-millimeter positioning ac-

curacy in this application lie in the fact that the

lightweight track trolley can only move at low speed

compared to the dynamic inspection locomotive,

thus the maneuver experienced by GNSS/INS is

very weak. In this case, the accuracy of the GNSS/

INS system tends to diverge due to the poor

signal-to-noise ratio of the gyro and accelerometer

outputs.

METHODOLOGY

Track Irregularity and Assessment

Track geometry can be regarded as a three-

dimensional curve, which is described in terms of

track curvature, alignment, elevation (cant), and

gauge as functions of distance along the track [5].

Track irregularity refers to the deviation of railway

track from its designed geometry and smoothness,

usually expressed in two separate layouts for

horizontal and vertical. Five typical track irre-

gularities are deﬁned: alignment irregularity,

vertical irregularity, cant, i.e., cross level or super-

elevation, twist, and gauge deviation. Alignment

and vertical irregularity refer to the dev iation of

consecutive measuring points of the rail in the lateral

and vertical direction, respectively, expressed as

excursions from the reference line, i.e., the mean

horizontal and vertical position [15]. Cant, i.e.,

super-elevation, is the difference betwe en the

elevations of the two rails in a curved section, which

is arranged to compensate for part of the lateral

accelerations. Cant irregularity refers to the

difference of the real cant from the nominal value.

Twist refers to the algebraic difference between two

cross levels taken at a deﬁned distance apart, usually

expressed as a gradient between the two points of

measurement [15]. Gauge irregularity refers to the

deviation of the gauge measurement from their

nominal value. In this paper, the alignment and cant

irregularity will be evaluated as examples to

demonstrate the measuring accuracy. The mea-

surement method for vertical irregularity is similar

to that for alignment, and twist can be computed

from consecutive measurements of cant, so these will

not be discussed in detail.

Since rail can be regarded as a three-dimensional

curve, then track irregularities should be expressed

as the relative deviation from their nominal

smoothness. Thus railway track irregularity

detection is inherently an issue of spatial-relative

measurement of the track position and angles. In

other word, once the track position and orientation

are determined, the 3-D track geometry can be

reconstructed based on which track irregularities

could be identiﬁed.

The differential versine, illustrated in Figure 1, is

recommended by the China Ministry of Railway as a

standard method of identifying shortwave alignment

and vertical irregularities and evaluating track

smoothness [3]. As shown in Figure 1, the differential

versines for a 30-m chord are consecutively computed

by subtracting two versines 5 meter apart, for

example p25 and p33, by applying equation (1).

Δh ¼ h

25S

 h

33S

ðÞh

25D

 h

33D

ðÞ

≤ 2mm (1)

where, h

25D

, h

33D

refer to the nominal/designed

versine at p25 and p33, which would be zero for the

84 Navigation Spring 2015

straight line section and constant for the curved

section. h

25S

, h

33S

are the corresponding real

versines computed based on the track position

determination, respectively. The tolerance of the

differential versines Δh for the shortwave track

irregularities, i.e., 30-m chord, is 2 millimeters as

shown in equati on (1). This tolerance is quite

demanding, and implies that a 0.6 mm one- σ error

is required for translating the corresponding

tolerance on a one-dimensional 98.8% interval to

standard deviation for a single point position

determination.

GNSS/INS Measuring Principle

The inertial navigation system (INS) is well

known to be capable of providing position and

attitude measurements with extremely high relati ve

accuracy in a short time. The GNSS/INS inte grated

system can take advantage of GNSS position and

velocity measurement as update information to

compensate the error accumulation of INS, so as to

provide continuous high-precision solutions. For the

railway track irregularity measuring, if the wheels

of the track trolley without suspension can keep

continuous contact with the railway track, then the

3-D track geometry can be uniquely determined by

the position and attitude sequences provided by the

GNSS/INS system as shown in Figure 2. The track

trolley structure is designed rigidly without

suspension, thus the GNSS/INS position and

attitude variations can strictly reﬂect the variation

of the track smoothness, i.e., track irregularity. For

example, the roll angle reﬂects the difference

between the elevations of the two rails, according to

which the cant deviation can be detected. Similarly,

the pitch and heading angle reﬂect the vertical and

alignment variations of the track to some extent. In

this case, the spatial-relative property of the track

irregularity is converted to the time-relative

property through the moving trolley determined by

the time series of position and attitude provided by

the GNSS/INS system. With the GNSS/INS

smoothed position and attitude, we can reconstruct

the 3-D track geometry in the computer software,

based on which the track parameters can be

calculated and track irregularities can be identiﬁed.

Figure 3 illustrates the track irregularity measuring

principle of the track trolley based on the GNSS/INS

integrated system.

GNSS/INS Algorithm Description

In this section we will discuss the main point of

the GNSS/INS algorithm, including the Kalman

ﬁlter design, the aided method for the accuracy

improvement. Since the radius of curvature of the

high speed railway track is usually large and when

the track trolley runs on the railway at low speed

(i.e., < 10 m/s), the maneuver experienced by the

GNSS/INS system is very weak. In this case, the

accuracy of the GNSS/INS solutions tends to diverge

due to poor signal-to-noise ratio of the gyro and

accelerometers outputs, especially for the heading

angle measurement. To improve the accuracy of

GNSS/INS solution, non-holonomic constraints

(NHC) and smoothing algorithm are implemented

in the GNSS/INS integration algorithm, i.e., Kalman

ﬁlter.

Fig. 2–Illustration of the railway track trolley based on the GNSS/

INS system. The trolley is designed rigidly without suspension and

the wheels can keep continuous contact with the rails.

Fig. 1–Differential vesine method of identifying the shortwave (30-chord) irregularity for the

slab track (refer to [3]). The supporting points, e.g., sleepers of the rails are consecutively

arranged with 0.625 meter intervals, and differential versines are obtained by subtracting

two versines 5 meters apart.

85Chen et al.: Railway Track Irregularity by GNSS/INSVol. 62, No.1

剩余10页未读，继续阅读

评论收藏

内容反馈

Stone_89217

粉丝: 5
资源: 9

Railway Track Irregularity Measuring by GNSS INS Integration.pdf

最新资源

Railway Track Irregularity Measuring by GNSS INS Integration.pdf

irregulation.rar_PSD function_irregulation_track irregularity

LARGE TRACK IRREGULARITY DIVIATIONS OF TRACK QUALITY INDEX

The effect of track alignment irregularity on wheel-rail contact geometry relationship in a turnout zone (2012年)

A new track inspection car based on a laser camera system

Average irregularity representation of a rough surface for ray reflection

Shigley-s-Mechanical-Engineering-Design

Optical System Design

机器创作的象征音乐的评价指标_matlab音乐

Wave Propagation Phenomena of Single Mode Pulse Laser in a Thick Ytterbium Vapor Column

Designing A Common Sense Approach to Web & Mobile Application Design

边缘强度相似度用于图像质量评估

图新地球的自定义图源lrc

AUTOSAR架构详解（讲得好）.pdf

各工况测试曲线参数.xlsx

智能网联汽车技术路线图 2.0.pdf

全国道路网SHP数据.zip

MAX96712 FDS.pdf

蔚来汽车汽车以太网网络架构.pdf

ISO11898全套资料

Matlab/Simulink两种智能小车的建模与仿真

Trimble DataTransfer157.exe

ISO14229中文+英文原版.rar

中国九大流域shp.zip

ymodbus在windows下的修改版，能够被vs2019编译通过。

ISO26262汽车功能安全学习笔记(纯手绘版-思维导图).pdf

ISO14229 UDS中文翻译版-542页-无水印.pdf

VCU控制策略+MATLAB模型

坐标转换软件（百度、高德火星坐标与CGCS2000、WGS84之间的互转）

最新资源