{"id":{"repo_id":"aachen","oai_identifier":"oai:publications.rwth-aachen.de:61654"},"canonical_url":"https://search.dev.ndltd.org/etd/aachen/oai:publications.rwth-aachen.de:61654","repository":{"repo_id":"aachen","name":"RWTH Aachen University","base_url":"https://publications.rwth-aachen.de/oai2d"},"display":{"title":"Kurzfristprognosen von Verkehrszuständen auf Basis von Verfahren der Mustererkennung und von dynamischen Routensuch- und Umlegungsverfahren","abstract":"Increasing traffic congestion (especially in areas of conurbation), the shortening of financial resources, and increasing resistance to further expansion of road networks, force consideration of more efficient usage of existing traffic infrastructure. This task requires an efficient traffic management system with intelligent traffic control measures and information for drivers. This also requires systems for the monitoring and prediction of traffic flow that can provide anticipatory and net-wide information about traffic states. In urban road networks, traffic states can only be measured directly on a limited number of network links. This is due to the limited density of traffic detection devices. For a large part of the network links traffic states have to be estimated. Thus, a net-wide prediction requires, in addition to a temporal extrapolation (prognosis), a spatial extrapolation of measured traffic states. An analysis of available approaches and methods for net-wide prediction of traffic states shows that no general methods are currently available that can solve the problem of spatial and temporal extrapolation of locally measured traffic states, in real-time, for large urban road networks. The integration of traffic assignment and traffic simulation models allows the estimating of traffic states, on network elements, using historical and actual measured data. Due to the capability of assignment and/or simulation models to estimate net-wide traffic states, this group of models is especially suitable for area-wide traffic estimating and prediction. The aim of this thesis is to develop a concept for a prediction system based on an iterative and dynamic traffic assignment procedure. This procedure is to be used as a tool for net-wide reconstruction of actual traffic states and for prediction of future traffic states. The iteration of a dynamic macroscopic route choice and traffic assignment procedure on the one hand, and the adjustment of the traffic demand information by means of a method for the estimation of origin-destination matrices on the other hand, allows three benefits: (1) a modelling of the effects on route choice behaviour, (2) reproduction of traffic control measures, and (3) the development of collective and individual traffic guidance strategies. A prediction system in regard to urban traffic management must particularly deal with special traffic events and situations. The approach presented here is based on the combination of the adaptation of an iterative traffic assignment with methods of pattern recognition. This will identify traffic states and special traffic situations. A hierarchically staged pattern recognition process identifies actual traffic states and special traffic situations by comparison of actual information from detection devices with historical data. It also controls the process of dynamic traffic assignment by providing suitable traffic demand information. The result is a reduced calculation effort for the adaptation of the dynamic traffic assignment procedure and an improved reproduction quality of actual traffic states. This system provides traffic flow data and travel times for all network elements for prediction horizons up to 2 hours, the relevant time span for urban traffic management and traffic control applications. The prediction system presented here is designed as self learning. Traffic events and situations that appear for the first time are stored in a traffic knowledge base for future reference when a similar event or situation occurs. All components of the developed prediction system are specified in detail and alternative solutions of applicable approaches are discussed. During the initial phase of implementing the traffic prediction system, creation and structuring of the knowledge base for the traffic patterns is of particular importance. To support this process, data-mining techniques and approaches to segmentation (clustering) of traffic data are also presented and discussed. Next, input data requirements for the developed prediction system are discussed with specifics. In addition to traffic network model requirements, the creation and management of a time series of \"dynamic\" origin destination matrices (travel demand information that is valid for a short time period) is described in detail. Subsequently, the difficulty of availability and the integration of actually measured traffic data into the prediction system, is discussed. The core components of the developed prediction system are applied and tested on a large urban road network using real online traffic data. The capability of the chosen approach to detect traffic states and special events by pattern matching is demonstrated. Additionally it was verified that the adaptation of a dynamic traffic assignment is a suitable method to reconstruct traffic states.","abstract_html":"Increasing traffic congestion (especially in areas of conurbation), the shortening of financial resources, and increasing resistance to further expansion of road networks, force consideration of more efficient usage of existing traffic infrastructure. This task requires an efficient traffic management system with intelligent traffic control measures and information for drivers. This also requires systems for the monitoring and prediction of traffic flow that can provide anticipatory and net-wide information about traffic states. In urban road networks, traffic states can only be measured directly on a limited number of network links. This is due to the limited density of traffic detection devices. For a large part of the network links traffic states have to be estimated. Thus, a net-wide prediction requires, in addition to a temporal extrapolation (prognosis), a spatial extrapolation of measured traffic states. An analysis of available approaches and methods for net-wide prediction of traffic states shows that no general methods are currently available that can solve the problem of spatial and temporal extrapolation of locally measured traffic states, in real-time, for large urban road networks. The integration of traffic assignment and traffic simulation models allows the estimating of traffic states, on network elements, using historical and actual measured data. Due to the capability of assignment and/or simulation models to estimate net-wide traffic states, this group of models is especially suitable for area-wide traffic estimating and prediction. The aim of this thesis is to develop a concept for a prediction system based on an iterative and dynamic traffic assignment procedure. This procedure is to be used as a tool for net-wide reconstruction of actual traffic states and for prediction of future traffic states. The iteration of a dynamic macroscopic route choice and traffic assignment procedure on the one hand, and the adjustment of the traffic demand information by means of a method for the estimation of origin-destination matrices on the other hand, allows three benefits: (1) a modelling of the effects on route choice behaviour, (2) reproduction of traffic control measures, and (3) the development of collective and individual traffic guidance strategies. A prediction system in regard to urban traffic management must particularly deal with special traffic events and situations. The approach presented here is based on the combination of the adaptation of an iterative traffic assignment with methods of pattern recognition. This will identify traffic states and special traffic situations. A hierarchically staged pattern recognition process identifies actual traffic states and special traffic situations by comparison of actual information from detection devices with historical data. It also controls the process of dynamic traffic assignment by providing suitable traffic demand information. The result is a reduced calculation effort for the adaptation of the dynamic traffic assignment procedure and an improved reproduction quality of actual traffic states. This system provides traffic flow data and travel times for all network elements for prediction horizons up to 2 hours, the relevant time span for urban traffic management and traffic control applications. The prediction system presented here is designed as self learning. Traffic events and situations that appear for the first time are stored in a traffic knowledge base for future reference when a similar event or situation occurs. All components of the developed prediction system are specified in detail and alternative solutions of applicable approaches are discussed. During the initial phase of implementing the traffic prediction system, creation and structuring of the knowledge base for the traffic patterns is of particular importance. To support this process, data-mining techniques and approaches to segmentation (clustering) of traffic data are also presented and discussed. Next, input data requirements for the developed prediction system are discussed with specifics. In addition to traffic network model requirements, the creation and management of a time series of "dynamic" origin destination matrices (travel demand information that is valid for a short time period) is described in detail. Subsequently, the difficulty of availability and the integration of actually measured traffic data into the prediction system, is discussed. The core components of the developed prediction system are applied and tested on a large urban road network using real online traffic data. The capability of the chosen approach to detect traffic states and special events by pattern matching is demonstrated. Additionally it was verified that the adaptation of a dynamic traffic assignment is a suitable method to reconstruct traffic states.","abstract_has_math":false,"creators":["von der Ruhren, Stefan"],"institution":"Publikationsserver der RWTH Aachen University","degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":["Beckmann, Klaus J."],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2006,"date_issued":"2006","date_published":"2006","updated_at":"2026-07-30T19:43:10Z","subjects":["info:eu-repo/classification/ddc/620","Ingenieurwissenschaften","Verkehrsmodell","Verkehrstechnik","Mustererkennung","Verkehrsleitsystem","Verkehrsmanagement","Kurzfristprognose","dynamische Routensuche und Umlegung","traffic model","pattern recognition","traffic management","short term prediction","dynamic traffic assignment"],"languages":["ger"],"rights":["info:eu-repo/semantics/openAccess"],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier","label":"Identifier","values":["https://publications.rwth-aachen.de/search?p=id:%22RWTH-CONV-123293%22"],"render_values":[{"text":"https://publications.rwth-aachen.de/search?p=id:%22RWTH-CONV-123293%22","href":"https://publications.rwth-aachen.de/search?p=id:%22RWTH-CONV-123293%22","code":true}]}]},"links":{"outbound_url":"https://publications.rwth-aachen.de/record/61654","outbound_label":"Repository record","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Beckmann, Klaus J."]},{"key":"dc:creator","label":"Author","values":["von der Ruhren, Stefan"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:coverage","label":"Dc Coverage","values":["DE"]},{"key":"dc:date","label":"Dc Date","values":["2006"]},{"key":"dc:publisher","label":"Institution","values":["Publikationsserver der RWTH Aachen University"]},{"key":"dc:relation","label":"Dc Relation","values":["info:eu-repo/semantics/altIdentifier/urn/urn:nbn:de:hbz:82-opus-17837"]},{"key":"dc:type","label":"Dc Type","values":["info:eu-repo/semantics/doctoralThesis","info:eu-repo/semantics/publishedVersion"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["info:eu-repo/classification/ddc/620","Ingenieurwissenschaften","Verkehrsmodell","Verkehrstechnik","Mustererkennung","Verkehrsleitsystem","Verkehrsmanagement","Kurzfristprognose","dynamische Routensuche und Umlegung","traffic model","pattern recognition","traffic management","short term prediction","dynamic traffic assignment"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["ger"]},{"key":"dc:rights","label":"Dc Rights","values":["info:eu-repo/semantics/openAccess"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://publications.rwth-aachen.de/record/61654","https://publications.rwth-aachen.de/search?p=id:%22RWTH-CONV-123293%22"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Increasing traffic congestion (especially in areas of conurbation), the shortening of financial resources, and increasing resistance to further expansion of road networks, force consideration of more efficient usage of existing traffic infrastructure. This task requires an efficient traffic management system with intelligent traffic control measures and information for drivers. This also requires systems for the monitoring and prediction of traffic flow that can provide anticipatory and net-wide information about traffic states. In urban road networks, traffic states can only be measured directly on a limited number of network links. This is due to the limited density of traffic detection devices. For a large part of the network links traffic states have to be estimated. Thus, a net-wide prediction requires, in addition to a temporal extrapolation (prognosis), a spatial extrapolation of measured traffic states. An analysis of available approaches and methods for net-wide prediction of traffic states shows that no general methods are currently available that can solve the problem of spatial and temporal extrapolation of locally measured traffic states, in real-time, for large urban road networks. The integration of traffic assignment and traffic simulation models allows the estimating of traffic states, on network elements, using historical and actual measured data. Due to the capability of assignment and/or simulation models to estimate net-wide traffic states, this group of models is especially suitable for area-wide traffic estimating and prediction. The aim of this thesis is to develop a concept for a prediction system based on an iterative and dynamic traffic assignment procedure. This procedure is to be used as a tool for net-wide reconstruction of actual traffic states and for prediction of future traffic states. The iteration of a dynamic macroscopic route choice and traffic assignment procedure on the one hand, and the adjustment of the traffic demand information by means of a method for the estimation of origin-destination matrices on the other hand, allows three benefits: (1) a modelling of the effects on route choice behaviour, (2) reproduction of traffic control measures, and (3) the development of collective and individual traffic guidance strategies. A prediction system in regard to urban traffic management must particularly deal with special traffic events and situations. The approach presented here is based on the combination of the adaptation of an iterative traffic assignment with methods of pattern recognition. This will identify traffic states and special traffic situations. A hierarchically staged pattern recognition process identifies actual traffic states and special traffic situations by comparison of actual information from detection devices with historical data. It also controls the process of dynamic traffic assignment by providing suitable traffic demand information. The result is a reduced calculation effort for the adaptation of the dynamic traffic assignment procedure and an improved reproduction quality of actual traffic states. This system provides traffic flow data and travel times for all network elements for prediction horizons up to 2 hours, the relevant time span for urban traffic management and traffic control applications. The prediction system presented here is designed as self learning. Traffic events and situations that appear for the first time are stored in a traffic knowledge base for future reference when a similar event or situation occurs. All components of the developed prediction system are specified in detail and alternative solutions of applicable approaches are discussed. During the initial phase of implementing the traffic prediction system, creation and structuring of the knowledge base for the traffic patterns is of particular importance. To support this process, data-mining techniques and approaches to segmentation (clustering) of traffic data are also presented and discussed. Next, input data requirements for the developed prediction system are discussed with specifics. In addition to traffic network model requirements, the creation and management of a time series of \"dynamic\" origin destination matrices (travel demand information that is valid for a short time period) is described in detail. Subsequently, the difficulty of availability and the integration of actually measured traffic data into the prediction system, is discussed. The core components of the developed prediction system are applied and tested on a large urban road network using real online traffic data. The capability of the chosen approach to detect traffic states and special events by pattern matching is demonstrated. Additionally it was verified that the adaptation of a dynamic traffic assignment is a suitable method to reconstruct traffic states."]},{"key":"dc:source","label":"Dc Source","values":["Aachen : Publikationsserver der RWTH Aachen University 185 S. : graph. Darst., Kt. (2006). = Aachen, Techn. Hochsch., Diss., 2006"]},{"key":"dc:title","label":"Title","values":["Kurzfristprognosen von Verkehrszuständen auf Basis von Verfahren der Mustererkennung und von dynamischen Routensuch- und Umlegungsverfahren"]}]}],"canonical_facts":{"dc:contributor":["Beckmann, Klaus J."],"dc:coverage":["DE"],"dc:creator":["von der Ruhren, Stefan"],"dc:date":["2006"],"dc:description":["Increasing traffic congestion (especially in areas of conurbation), the shortening of financial resources, and increasing resistance to further expansion of road networks, force consideration of more efficient usage of existing traffic infrastructure. This task requires an efficient traffic management system with intelligent traffic control measures and information for drivers. This also requires systems for the monitoring and prediction of traffic flow that can provide anticipatory and net-wide information about traffic states. In urban road networks, traffic states can only be measured directly on a limited number of network links. This is due to the limited density of traffic detection devices. For a large part of the network links traffic states have to be estimated. Thus, a net-wide prediction requires, in addition to a temporal extrapolation (prognosis), a spatial extrapolation of measured traffic states. An analysis of available approaches and methods for net-wide prediction of traffic states shows that no general methods are currently available that can solve the problem of spatial and temporal extrapolation of locally measured traffic states, in real-time, for large urban road networks. The integration of traffic assignment and traffic simulation models allows the estimating of traffic states, on network elements, using historical and actual measured data. Due to the capability of assignment and/or simulation models to estimate net-wide traffic states, this group of models is especially suitable for area-wide traffic estimating and prediction. The aim of this thesis is to develop a concept for a prediction system based on an iterative and dynamic traffic assignment procedure. This procedure is to be used as a tool for net-wide reconstruction of actual traffic states and for prediction of future traffic states. The iteration of a dynamic macroscopic route choice and traffic assignment procedure on the one hand, and the adjustment of the traffic demand information by means of a method for the estimation of origin-destination matrices on the other hand, allows three benefits: (1) a modelling of the effects on route choice behaviour, (2) reproduction of traffic control measures, and (3) the development of collective and individual traffic guidance strategies. A prediction system in regard to urban traffic management must particularly deal with special traffic events and situations. The approach presented here is based on the combination of the adaptation of an iterative traffic assignment with methods of pattern recognition. This will identify traffic states and special traffic situations. A hierarchically staged pattern recognition process identifies actual traffic states and special traffic situations by comparison of actual information from detection devices with historical data. It also controls the process of dynamic traffic assignment by providing suitable traffic demand information. The result is a reduced calculation effort for the adaptation of the dynamic traffic assignment procedure and an improved reproduction quality of actual traffic states. This system provides traffic flow data and travel times for all network elements for prediction horizons up to 2 hours, the relevant time span for urban traffic management and traffic control applications. The prediction system presented here is designed as self learning. Traffic events and situations that appear for the first time are stored in a traffic knowledge base for future reference when a similar event or situation occurs. All components of the developed prediction system are specified in detail and alternative solutions of applicable approaches are discussed. During the initial phase of implementing the traffic prediction system, creation and structuring of the knowledge base for the traffic patterns is of particular importance. To support this process, data-mining techniques and approaches to segmentation (clustering) of traffic data are also presented and discussed. Next, input data requirements for the developed prediction system are discussed with specifics. In addition to traffic network model requirements, the creation and management of a time series of \"dynamic\" origin destination matrices (travel demand information that is valid for a short time period) is described in detail. Subsequently, the difficulty of availability and the integration of actually measured traffic data into the prediction system, is discussed. The core components of the developed prediction system are applied and tested on a large urban road network using real online traffic data. The capability of the chosen approach to detect traffic states and special events by pattern matching is demonstrated. Additionally it was verified that the adaptation of a dynamic traffic assignment is a suitable method to reconstruct traffic states."],"dc:identifier":["https://publications.rwth-aachen.de/record/61654","https://publications.rwth-aachen.de/search?p=id:%22RWTH-CONV-123293%22"],"dc:language":["ger"],"dc:publisher":["Publikationsserver der RWTH Aachen University"],"dc:relation":["info:eu-repo/semantics/altIdentifier/urn/urn:nbn:de:hbz:82-opus-17837"],"dc:rights":["info:eu-repo/semantics/openAccess"],"dc:source":["Aachen : Publikationsserver der RWTH Aachen University 185 S. : graph. Darst., Kt. (2006). = Aachen, Techn. Hochsch., Diss., 2006"],"dc:subject":["info:eu-repo/classification/ddc/620","Ingenieurwissenschaften","Verkehrsmodell","Verkehrstechnik","Mustererkennung","Verkehrsleitsystem","Verkehrsmanagement","Kurzfristprognose","dynamische Routensuche und Umlegung","traffic model","pattern recognition","traffic management","short term prediction","dynamic traffic assignment"],"dc:title":["Kurzfristprognosen von Verkehrszuständen auf Basis von Verfahren der Mustererkennung und von dynamischen Routensuch- und Umlegungsverfahren"],"dc:type":["info:eu-repo/semantics/doctoralThesis","info:eu-repo/semantics/publishedVersion"]},"updated_at":"2026-07-30T19:43:10Z"}