Dissertation > Excellent graduate degree dissertation topics show

Short-term Urban Traffic Forecasting Based on Multi-kernel SVM Model

Author: OuYangJun
Tutor: LiuXingQuan；LuFeng
School: Central South University
Course: Cartography and Geographic Information Systems
Keywords: Floating Car Exploratory data analysis Traffic anomaly detection Short-term traffic forecasts Support Vector Machine Multicore Particle swarm optimization
CLC: U491.14
Type: Master's thesis
Year: 2011
Downloads: 108
Quote: 1
Read: Download Dissertation

Abstract


Currently, the countries of the world are faced with the growing problem of urban traffic problems. Traffic congestion, traffic accidents, traffic pollution caused great distress to the lives of urban residents, but also a huge loss to the socio-economic development, has become a major problem for all mankind solved. Intelligent Transportation Systems (ITS) is recognized as the only way to solve the problem of urban traffic. Dynamic traffic information platform is the hub of the various modules in the ITS, and real-time traffic information data base and blood of dynamic traffic information platform. Traffic information collection technologies and the Internet, the development and maturation of the wireless network technology to make real-time traffic information acquisition and release it possible, and laid the material foundation for the analysis of real-time traffic conditions. From the dynamic city in the real-time traffic flow data quickly and accurately discover and predict future traffic conditions, and published an intuitive, easy to understand form, supporting dynamic navigation, real-time route guidance and signal lights automatically with service in the city transportation planning and management departments and the majority of traffic travel by participants, to ease traffic pressure and improve traffic safety, improve operational efficiency and reduce air pollution, is of great significance. This article is based on the dynamic traffic information platform, from the characteristics of urban traffic flow, intelligent analysis framework based on the floating car data (Floating Car Data, FCD). The main innovation of this paper is as follows: an urban traffic flow parameters visualization methods, design a sub-sites based on the normal distribution assumption traffic anomaly detection exploratory analysis method. The standard deviation detection method is currently considered the preferred transport abnormality detecting method. In this paper, the mean and variance of the sample center itself vulnerable to data pollution, the standard deviation of the detection method can not reflect the differences of the sample center and extreme abnormal value, its improvements, a traffic based on the sub-sites anomaly detection algorithm. 2 based multicore hybrid support vector machine urban short-term traffic forecasting model. Uncertainty, nonlinear, and time and space of urban road traffic makes the parameter description and knowledge of the transportation system to get extremely difficult, so difficult to obtain satisfactory results of short-term traffic forecasts. In this paper, a hybrid multi-core support vector machine's ability to identify and deal with different categories of input data, a traffic forecast method based on the multicore the mixed support vector machine city short time. The method is based on statistical analysis of traffic state data samples, inherited the good generalization ability of support vector machine, the global optimum and strong self-adaptive characteristics, and improved particle swarm optimization parameters of support vector machine optimized choice. The same time, strong linear correlation for real-time road traffic state historical average traffic status, real-time road traffic status of non-linear correlation with the state of the traffic state and the upstream and downstream sections of the first few periods of real-time traffic, were designed linear kernel mapping function and nonlinear kernel function of urban traffic state and fitting. This method not only consider the significance of the predicted traffic state laws of history, but also take into account the time-varying characteristics of the traffic, to fully extract the relevant parameters of the transport system of knowledge and information.

Related Dissertations

  1. Research on Automatic Detection Algorithm for Substructure Distress of Highway Pavement Based on SVM,U418.6
  2. Research on Autamatic Music Structrue Analysis,TN912.3
  3. Research on Transductive Support Vector Machine and Its Application in Image Retrieval,TP391.41
  4. Research on Feature Extraction and Classification of Tongue Shape and Tooth-Marked Tongue in TCM Tongue Diagnosis,TP391.41
  5. Fault Diagnosis Method Based on Support Vector Machine,TP18
  6. Process Support Vector Machine and Its Application to Satellite Thermal Equilibrium Temperature Prediction,TP183
  7. Computing Minimum Distance between Curves/Surfaces Based on PSO Algorithm,O182
  8. Active Power Filter and Its Application in Distribution Network,TN713.8
  9. Research for Infrared Image Target Identification and Tracking Technology,TP391.41
  10. Research on Fuzzy C-Mean Clustering Algorithm Based on Particle Swarm Optimization and Shuffled Frog Leaping Algorithm,TP18
  11. Study on the Road Condition Monitoring Based on Vehicular 3D Acceleration Sensor,TP274
  12. Research of Diagnosing Cucumber Diseases Based on Hyperspectral Imaging,S436.421
  13. Research on the Improvements and Applications of Particle Swarm Optimization,TP18
  14. Segmentation of cDNA Microarray Image Using Fuzzy C-means Algorithm Optimized by Particle Swarm,TP391.41
  15. The Research on Intrusion Detection System Based on Machine Learning,TP393.08
  16. Research on Modification and Application of Particle Swarm Optimization Algorithm Based on Control Methods,TP301.6
  17. Based on RFID Prison Intelligent Management System Research and Implementation,TP315
  18. Research on Improved K Neighbor Support Vector Machine Algorithm Faced Text Classification,TP391.1
  19. Research of Path Planning for Small-Size Intelligent Soccer Robot in Complex and Dynamic Environment,TP242
  20. Research on Face Recognition Based on AdaBoost Algorithm,TP391.41
  21. The Research on the Target Localization and Tracking Based on WSN,TN929.5

CLC: > Transportation > Road transport > Technical management of traffic engineering and road transport > Traffic engineering and traffic management > Traffic Survey and Planning > Traffic forecasts
© 2012 www.DissertationTopic.Net  Mobile