Centre de documentation
scientifique

The Swiss trolley / Ralph Glaus (2006)  

Public Titre : The Swiss trolley : a modular system for track surveying Type de document : Rapport Auteurs : Ralph Glaus, Auteur Editeur : Zurich : Schweizerischen Geodatischen Kommission / Commission Géodésique Suisse Année de publication : 2006 Collection : Geodätisch-Geophysikalische Arbeiten in der Schweiz, ISSN 0257-1722 num. 70 Importance : 184 p. Format : 21 x 30 cm ISBN/ISSN/EAN : 978-3-908440-13-0 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Navigation et positionnement [Termes IGN] capteur imageur [Termes IGN] capteur non-imageur [Termes IGN] données localisées [Termes IGN] filtre de Kalman [Termes IGN] fonction spline [Termes IGN] fusion de données [Termes IGN] GPS en mode cinématique [Termes IGN] GPS en mode différentiel [Termes IGN] GPS-INS [Termes IGN] lissage de données [Termes IGN] navigation à l'estime [Termes IGN] odomètre [Termes IGN] positionnement absolu [Termes IGN] précision millimétrique [Termes IGN] prise de vue terrestre [Termes IGN] réseau ferroviaire [Termes IGN] surveillance d'ouvrage [Termes IGN] tachéomètre électronique [Termes IGN] transport ferroviaire Index. décimale : 30.70 Navigation et positionnement Résumé : (Auteur) Modern railway infrastructure requires accurate, absolute referenced spatial data for project planning, construction and maintenance. On the one hand, passenger safety and travel comfort depend to a large extent on accurate tracks. On the other hand, absolute referenced coordinates of railway assets facilitate data exchange between railway operators and third parties. In addition, time slots for maintenance are short, due to the high volumes of traffic on major railway lines. Thus, flexible surveying systems are required yielding accurate data within a short time. The multi-sensor platform Swiss Trolley, which offers such a flexible system, copes with absolute referenced spatial data. The platform is mounted on a track vehicle. This allows for a complete description of the track environment in kinematic mode with a minimum of interference time with regular traffic. The Swiss Trolley features a modular design. A basic module for assessing track key parameters such as chainage, cant, twist, gradients and track gauge covers monitoring tasks on construction sites. A positioning module integrating GPS or total stations allows for the determination of the track axis. A further scan module can be used to generate absolute referenced point clouds in the track environment. This work compiles the development steps of the Swiss Trolley. Relevant side conditions re-garding track surveying, coming from track geometry and the railway operators are summarised and state-of-the-art systems are reviewed. Based on these premises, a niche for Swiss Trolley applications is defined. Sensors providing geometric data in the track environment are evaluated in regard to their suitability and error behaviour. The key problem of the trolley positioning consists in determining the six degrees of freedom of the multi-sensor platform at any point in time. The chosen kinematic approach asks for a careful treatment of time constraints. Each data string coming from a specific sensor must own an accurate time tag. Kinematic surveys at walking speed with subcentimetric accuracy require time tags with millisecond accuracy. The incorporated sensors were investigated regarding their error behaviour. Calibration issues are addressed and approaches for the bias determination are presented. Models for correcting collimation errors and nuisance accelearations are given for the pendulum inclination sensors used. Moreover, emphasis was placed on biases emerging at kinematic surveys for the particular optical total station used. Reduction models for the laser scanner data are proposed and calibration procedures providing intrinsic orientation and latency parameters are given. A kinematic model for Swiss Trolley surveys based on the Frenet base system and its canonical representation was developed. Explicit formulae are given for runs on geometric elements dominating in the railway track environment. For the mutual data processing, a loosely coupled filter concept is proposed consisting of data pre-processing, synchronisation and filtering steps. The core of data processing is a Kalman filter, estimating vehicle and track states in an absolute or a relative reference frame. By means of the filter approach, the observations of the involved sensors can be integrated in a spatial model. Individual filter runs can be assembled by an additional merge step. Merged runs in up and down direction allow for a quality assessment and also allow for the monitoring of eventually remaining biases such as a boresight misalignment or inclination sensor zero point offsets. Positioning accuracies for the static and kinematic case were assessed on the one hand by the comparison of up and down runs. On the other hand, comparisons were carried out with independently measured reference data. The static error behaviour of the Swiss Trolley could be evaluated by using a slab track alignment. Submillimetric positioning accuracies were obtained in combination with high-precision total stations. Kinematic positioning accuracy mainly depends on the positioning sensor used. Optical total stations providing synchronised angle and distance data allow for subcentimetric positioning. High-precision DGPS position-ing yields subcentimetric accuracy for the horizontal component. The typical vertical accuracy is better than two centimetres. The integrated longitudinally mounted inclination sensor slightly augments the mere GPS solution. The attitude determination of the platform is a result of the combined data treatment. For GPS surveys, the typical pitch angle accuracy is two mrad. Yaw angles essentially correspond to the derivation of the trajectory with respect to the covered path and are determined with one mrad accuracy. Roll angle accuracy is dominated by the inclination sensor measurements across the track. The typical accuracy is 0.3 mrad. For the scan module, laser dots in the absolute reference frame are degraded by the uncertainty of the trajectory and the platform attitude amplified by a geometry-depending lever. The absolute accuracy of such a dot is three centimetres using a time-of-flight laser scanner. Relative accuracy between two adjacent dots amounts to five millimetres. The Swiss Trolley was successfully applied on numerous assignments. Adaptations for the multi-sensor platform exist for tunnel site locomotives and road-vehicles. Note de contenu : 1 Introduction 2 Track Geometry 2.1 Nominal Geometries 2.1.1 Introduction 2.1.2 Horizontal Layout 2.1.3 Vertical Layout 2.2 Rules and Standards of Different Countries 2.2.1 Horizontal Layout 2.2.2 Vertical Layout 2.2.3 Cant 2.3 Kinematic Model of Motion 2.3.1 Kinematics in the Frenet System 2.3.2 Canonical Representation of the Most Common Track Curves 2.4 Remarks on Track Accuracy 2.4.1 General Remarks 2.4.2 Relative and Absolute Accuracy of a Track 2.5 Methods for Track Surveying 2.5.1 Overview 2.5.2 Relative Track Surveying 2.5.3 Absolute Track Surveying 2.5.4 Selected Track-Surveying Systems 2.5.5 The Swiss Trolley - Finding the Niche 3 Potentials and Limitations of a Kinematic Track-Surveying System 3.1 Kinematic Surveying 3.2 Absolute Position Fixing 3.2.1 GNSS 3.2.2 Tracking Total Stations 3.3 Dead Reckoning 3.3.1 Inertial Navigation Systems (INS) 3.3.2 Yaw Rates by Chord Techniques 3.3.3 Odometers 3.3.4 Height Determination by an Inclination Sensor 3.4 Attitude Determination 3.5 Kinematic Surveys of the Railway Inventory 3.5.1 Track Gauge Measuring Systems 3.5.2 Laser Scanners 3.5.3 3D Cameras 3.5.4 Ground Penetration Radar (GPR) 3.6 Synchronisation 3.7 Modelling 3.8 Transformation 4 The Track-Surveying Trolley 4.1 Introduction 4.1.1 Development 4.1.2 Concept 4.2 Data Acquisition 4.2.1 Electronic Box 4.2.2 A/D Conversion 4.2.3 Data Synchronisation 4.3 Reconstruction 4.4 Inclination Sensors 4.4.1 Sensor Characteristics 4.4.2 Calibration of Characteristic Curve 4.4.3 Temperature Influences 4.4.4 Corrections for Non-Orthogonalities (Collimation Error) 4.4.5 Dynamic Behaviour of the Inclination Sensor 4.4.6 Transformation of the Inclination Angles into the Body-System 4.5 Track Gauge Measuring System 4.5.1 Characteristics and Measuring Principle of the Track Gauge Measuring System 4.5.2 Calibration 4.6 Odometers 4.6.1 Characteristics and Calibration 4.7 Integration of Tracking Total Stations 4.7.1 Characteristics 4.7.2 Common Total Station Biases 4.7.3 Deflections of the Vertical 4.7.4 Surveys in Canted Sections 4.7.5 Synchronisation of Distances and Angles 4.7.6 Internal Tacheometer and Radio Latencies 4.8 Integration of GPS 4.8.1 Characteristics 4.8.2 NMEA Data 4.9 Boresight Calibration of Prism and Antenna Phase Centre 4.10 Laser Scanners 4.10.1 Characteristics 4.10.2 Model 4.10.3 Yaw Angle Correction 4.10.4 Evaluation of the Laser Scanner Precision 4.10.5 Variance Propagation for a Given Scanner Arrangement 4.10.6 Kinematic Calibration of Rmb, xmb and the Latency 5 Data Processing 5.1 Introduction 5.2 Post-Processing Software Concept 5.3 Data Preprocessing 5.3.1 Blunder Labelling 5.3.2 Reduction, Model 5.3.3 Linear Filters 5.3.4 Synchronisation 5.3.5 Reduction to the Centre Line of the Track 5.4 Trajectory Smoothing by a Kalman Filter 5.4.1 Discrete Kalman Filter 5.4.2 Backward Filter and Smoother 5.4.3 Absolute Model 5.4.4 Relative Model 5.5 Smoothing Splines 5.5.1 Smoothing Splines with First Derivatives 5.5.2 Comparison between Kalman Filter and Smoothing Splines 5.6 Merging Trajectories 5.6.1 Strategies for Merging 5.6.2 Chaining the Pieces 5.6.3 Merging 5.6.4 Linking Scans to Merged Trajectories 6 Applications 6.1 Slab Track Alignment 6.2 Kinematic Track Axis Surveys 6.2.1 Comparison between Forward Filter, Backward Filter and Smoother 6.2.2 Filter Tuning 6.2.3 Comparison between Absolute and Relative Model 6.2.4 The Influence of Inclinometer Measurement on GPS Heights 6.2.5 The Smoother in Action - GPS Example 6.2.6 The Smoother in Action - Total Station Example 6.3 Kinematic Scanning 7 Conclusions Numéro de notice : 15261 Affiliation des auteurs : non IGN Thématique : POSITIONNEMENT Nature : Rapport de recherche En ligne : https://www.sgc.ethz.ch/sgc-volumes/sgk-70.pdf Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=55115

Exemplaires(2)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
15261-01	30.70	Livre	Centre de documentation	Géodésie	Disponible
15261-02	30.70	Livre	Centre de documentation	Géodésie	Disponible