Share Email Print

Proceedings Paper

Performance analysis of fuzzy logic particle filter compared to fuzzy IMM in tracking high-performance targets
Author(s): Hazem Kamel; Wael Badawy
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

A high-performance target may accelerate at non-uniform rates, complete sharp turns within short time periods, thrust, roll, and pitch; which may not follow a linear model. Even though the interacting multiple model (IMM) can be considered as a multimodal approach, it still requires prior knowledge about the target model. To overcome this weakness, a fuzzy logic particle filter (FLPF) is used. It is comprised of single-input single-output; which is presented by fuzzy relational equations. A canonical-rule based form is used to express each of these fuzzy relational equations. The dynamics of the high-performance target are modeled by multiple switching (jump Markov) systems. The target may follow one-out of-seven dynamic behavior model at any time in the observation period under assumption of coordinate turn model. The FLPF has the advantage that it does not require any prior knowledge of statistical models of process as in IMM. Moreover, it does not need any maneuver detector even when tracking a high performance target; which results in less computational complexities. By using an appropriate fuzzy overlap set, only a subset of the total number of models need to be evaluated, and these will be conditioned on acceleration values close to the estimate. This reduces the computational load compared to the fuzzy IMM (FIMM) algorithm. To achieve the whole range of maneuver variables, more models can be added without increasing the computational load as the number of models evaluated is determined only by the overlap. An example is included for visualizing the effectiveness of the proposed algorithm. Simulation results showed that the FLPF has good tracking performance and less computational load compared to the FIMM when applied to systems characterized by large scan periods.

Paper Details

Date Published: 2 May 2006
PDF: 12 pages
Proc. SPIE 6238, Acquisition, Tracking, and Pointing XX, 62380H (2 May 2006); doi: 10.1117/12.664688
Show Author Affiliations
Hazem Kamel, Univ. of Calgary (Canada)
Wael Badawy, Univ. of Calgary (Canada)

Published in SPIE Proceedings Vol. 6238:
Acquisition, Tracking, and Pointing XX
Michael K. Masten; Larry A. Stockum; William E. Thompson, Editor(s)

© SPIE. Terms of Use
Back to Top