Parameter Estimation of Bilinear Systems Based on an Adaptive Particle Swarm Optimization

Abstract

Bilinear models can approximate a large class of nonlinear systems adequately and usually with considerable parsimony in the number of coefficients required. This paper presents the application of Particle Swarm Optimization (PSO) algorithm to solve both offline and online parameter estimation problem for bilinear systems. First, an Adaptive Particle Swarm Optimization (APSO) is proposed to increase the convergence speed and accuracy of the basic particle swarm optimization to save tremendous computation time. An illustrative example for the modeling of bilinear systems is provided to confirm the validity, as compared with the Genetic Algorithm (GA), Linearly Decreasing Inertia Weight PSO (LDW-PSO), Nonlinear Inertia Weight PSO (NDW-PSO) and Dynamic Inertia Weight PSO (DIW-PSO) in terms of parameter accuracy and convergence speed. Second, APSO is also improved to detect and determine varying parameters. In this case, a sentry particle is introduced to detect any changes in system parameters. Simulation results confirm that the proposed algorithm is a good promising particle swarm optimization algorithm for online parameter estimation.

Department(s)

Electrical and Computer Engineering

Keywords and Phrases

Adaptive Particle Swarm Optimizations; Bilinear Models; Bilinear System; Bilinear Systems; Computation Time; Convergence Speed; Dynamic Inertia; Illustrative Examples; Inertia Weight; Large Class; Linearly Decreasing Inertia Weight PSO; Nonlinear Inertia Weight; Offline; Online Parameter Estimation; Optimization Algorithms; Particle Swarm Optimization Algorithm; Sentry Particles; Simulation Result; Varying Parameters; Convergence of Numerical Methods; Genetic Algorithms; Metal Analysis; Online Systems; Parameter Estimation; Particle Swarm Optimization (PSO); Adaptive Particle Swarm Optimization

International Standard Serial Number (ISSN)

0952-1976

Document Type

Article - Journal

Document Version

Citation

File Type

text

Language(s)

English

Rights

© 2010 Elsevier, All rights reserved.

Publication Date

01 Oct 2010

Share

 
COinS