A particle swarm optimization-based state estimation scheme for moving objects

Abstract

The α − β and α − β − γ filters based on the Kalman-like estimation scheme have been recognized as outstanding tools for estimating position, velocity and acceleration signals of moving objects. Nevertheless, the performance of estimation heavily depends on the parameters α, β, and γ. In general, the choice of parameters is a trade-off multi-objective optimization problem between the tracking accuracy and noise suppression capability. In this paper, a weighted particle swarm optimization (PSO) method is proposed to help search globally optimal parameter values in parameter space to meet some specifications. A numerical example is employed to illustrate the developed optimum estimation scheme, which is capable of tracking the desired signals accurately and, at the same time, reducing the noise disturbance remarkably. In order to test practically the effectiveness of the estimation scheme, an experimental apparatus, named twin rotor system, will be used as a test bed to demonstrate its performance. The experimental results strongly suggest that the PSO-based state estimators are encouraging for practical applications.