Dissertation > Excellent graduate degree dissertation topics show

The Research of Multi-sensor Target Tracking Algorithm Based on Information Fusion

Author: WenRuQuan
Tutor: CaoJie
School: Lanzhou University of Technology
Course: Control Theory and Control Engineering
Keywords: Maneuvering Target Tracking Information Fusion Nonlinear filtering algorithm Particle filter algorithm Improved particle filter Interacting multiple model unscented particle filtering algorithm
CLC: TP202
Type: Master's thesis
Year: 2009
Downloads: 311
Quote: 3
Read: Download Dissertation

Abstract


Broad application prospects as target tracking in military and civilian fields, target tracking problem has received extensive attention, maneuvering target tracking has become a hot research topic in the current. Currently, target tracking studies mainly focused on two aspects to improve the algorithm performance of maneuvering target tracking and multi-sensor information fusion. This paper aims to study the maneuvering target tracking algorithm, and the algorithm exist some problems, combined with the information fusion technology to improve the algorithm. Kalman filter linear Gaussian system optimal estimation. Maneuvering target tracking systems are generally nonlinear. Extended Kalman filter (EKF) is the most commonly used algorithms to solve the nonlinear system, the idea of ??the algorithm is to make the system of non-linear linear processing, and then based on the framework of the Kalman filter recursive filtering. EKF system model, the linear approximation introduces error, and the need to calculate the Jacobian matrix is ??not easy to achieve. EKF these problems, the introduction of unscented transform (UT) to improve the Kalman filter, unscented Kalman filter (UKF), the algorithm Sigma Point to describe the state of the system, do not need to make a linearization processing. Through theoretical analysis and simulation experiments show that: compared with the EKF, UKF achieve simple versatility, tracking a high precision, stable performance characteristics. It is entirely possible to replace EKF become a common nonlinear filtering algorithm. The particle filter (PF) is currently the most popular form of maneuvering target tracking algorithm, the algorithm is no limit on non-Gaussian nonlinear system, compared with the EKF and UKF, the application of the particle filter algorithm wider. PF algorithm, particle prone to degradation the importance function plays a crucial role in the selection of the degradation of the particles. In order to overcome this phenomenon, EKF and UKF algorithm introduced in this of importance function improved, better than PF extended particle filter algorithm (EPF) and the unscented particle filter algorithm (UPF). The results showed that the EPF and UPF performance than PF better the highest UPF algorithm tracking accuracy, but its running time is long, therefore less real-time simulation experiments. Finally, the multi-model multi-sensor maneuvering target tracking algorithm. Interacting multiple model algorithm (IMM) is the multi-model algorithm mainstream. In this paper, the IMM algorithm and UPF algorithm has the advantage, a fusion algorithm (IMM-UPF) algorithm - interactive multiple model unscented particle. The establishment of the model set according to the characteristics of the filter used, the direct use of the the three nonlinear model based on Markov transition probability to work in parallel. Through a large number of test simulation proved the effectiveness of the algorithm, and its performance is even better than the UPF.

Related Dissertations

  1. Maneuver Detection and Tracking with the Radar Rate Measurement,TN953
  2. The Maneuvering Target Tracking Research Based on VRPF,TN957.52
  3. Research on Multi-Sensor Embattling and Continious Tracking Technique Applying for Anti-Stealth and Anti-Interference,TN953
  4. Tongue Feature Extraction and Research of Fusion Classification,TP391.41
  5. Multi-Sensor Information Fusion and Its Applications on Wearable Computer,TP202
  6. Study on the Technique of Information Fusionapplied to Enbedded Driver Fatigue Detection,TP368.12
  7. The Research on the Target Localization and Tracking Based on WSN,TN929.5
  8. Research and Implementation of the Key Technologies for Multimeida Sensor Terminal,TP212.9
  9. Research on Wide Area Backup Protection Based on Centralized Decision-making,TM774
  10. Underwater Manipulator information fusion and operations planning studies,TP241
  11. Based on information fusion safety assessment of genetically modified foods,TS201.6
  12. Information Fusion Based on Multi- wheeled mobile robot navigation technology,TP242
  13. Based on Bayesian Networks motor fault diagnosis method,TM307.1
  14. Multi-sensor multi-target tracking and fusion algorithms track,TN953
  15. Multiscale information fusion algorithm,TP202
  16. Based on Information Fusion Fault Diagnosis of Analog Circuits,TN710
  17. Multi-sensor Information Fusion Methods for Integrated Identification of Friend or Foe Based on Evidential Networks,TP202
  18. Study on Freeway Automatic Incident Detection Algorithm Based on Information Fusion,U491
  19. Applied Research of Multi-sensor Information Fusion Technology in the Virtual Chinese Medicine Bone-setting Manipulation,R274
  20. Based on a combination of multi-sensor automated guided vehicle systems research and application,TP242.2
  21. Research on Early-warning and Forecasting System Algorithm for Distributor Fault Arc Based on Information Fusion,TM592

CLC: > Industrial Technology > Automation technology,computer technology > Automation technology and equipment > General issues > Design, performance analysis and synthesis
© 2012 www.DissertationTopic.Net  Mobile