Font Size: a A A

Research And Implementation Of Semantic Fusion Model In Dynamic Navigation System

Posted on:2015-08-01Degree:MasterType:Thesis
Country:ChinaCandidate:P LiFull Text:PDF
GTID:2298330467474624Subject:Electronic and communication engineering
Abstract/Summary:PDF Full Text Request
Intelligent Transportation System (ITS) has been developed rapidly in recent years. As thefoundation and guarantee of ITS, traffic information is now a research hotpot and a variety of trafficinformation collection technologies appear, of which the detect performance differs, especiallywhen used for real-time dynamic traffic information. To meet the demand of traffic information forITS applications, an effective traffic information fusion technology is needed to mine morecomprehensive, more accurate and more thorough information for multi-source traffic informationand provide personalized service to the users of ITS applications based on their differentrequirements of traffic information.This paper, after an analysis of the requirement of intelligent fusion of traffic information in thescenario of ITS application and the existing fusion models, proposes a hierarchical,simple-to-operate and easy-to-implement traffic information semantic fusion model. Then thesemantic fusion model in dynamic navigation system including the semantic coding layer, semanticprocessing layer and application service layer is constructed. In the semantic coding layer, asemantic coding method is proposed. In the semantic processing layer, a credibility-based semanticlogic algebra is proposed, through which the level of credibility is set based on the performance oftraffic information detection devices, then a semantic fusion of traffic parameters is made based onthe credibility during semantic processing to improve the accuracy of traffic information fusion. Afeedback mechanism of credibility is also added to the algebra. In the application service layer, afuzzy neutral network based semantic fusion mechanism is proposed, through which a five layerfuzzy neutral network model is constructed according to the preference weights of the trafficinformation of the user of the dynamic navigation system, providing personalized services byweight learning. Finally this paper validates the effectiveness and the correctness of the proposedmechanism by simulation and implement of the dynamic navigation system.
Keywords/Search Tags:dynamic navigation system, semantic fusion, semantic encoding, quality of the trafficinformation, fuzzy neural network
PDF Full Text Request
Related items