Fuzzy logic particle filter for high-performance targets tracking in track-while-scan radar
Date
2006
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The efficiency and accuracy of the particle filter and its extensions depend mainly on two key factors: the propagation function used to re-allocate the particles and the number of particles used to estimate the posterior distribution. Previous studies have failed to consider these two factors together. In this thesis, a new self adaptive fuzzy logic particle filter (FLPF) that uses fuzzy logic systems (FLS) is proposed. It estimates the angular turn rate, which is included as a state component, and tunes dynamically the number of particles used to estimate the posterior distribution. The estimated angular turn rate defines the propagation function used to re-allocate the particles. Also, the number of particles used to estimate the posterior distribution is measured using FLS based on the target's maneuverability. Thus, the FLPF can improve the efficiency and accuracy of the particle filter by estimating the two key factors simultaneously using a fuzzy-logic framework. Also in this thesis, a tracker fusion technique is proposed to reduce the computation load when the target is non-maneuvering by using the unscented Kalman filter (UKF) as it has less computational load compared to the particle filters. The UKF is known to be optimal and is employed for state estimation for linear and Gaussian systems. Data association is also considered in this thesis as it is a crucial element and one of the most important components of any track-while-scan (TWS) radar system. A merged probabilistic data association (MPDA) is proposed. It merges the probabilistic nearest-neighbor filter (PNNF) with the joint probabilistic data association (JPDA) approach. The MPDA cooperates with the UKF to track non-maneuvering targets; meanwhile the independent sample based joint probabilistic data association (ISBJPDA) approach cooperates with the FLPF to track maneuvering targets. Different scenarios are simulated to evaluate and analyze the performance of the FLPF algorithm, the MPDA approach, and the tracker fusion technique. The experimental results verify that the proposed MPDA outperforms other data association techniques when applied to dense clutter environments and higher noise levels. Also, the FLPF show better performance compared to the integrated multiple models (IMM) in tracking high-performance maneuvering targets.
Description
Bibliography: p. 181-189
Keywords
Citation
Kamel, H. Z. (2006). Fuzzy logic particle filter for high-performance targets tracking in track-while-scan radar (Doctoral thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca. doi:10.11575/PRISM/1598