Title
Parameter Identification using PSO under measurement noise conditions
Abstract
This work presents an experimental study on the performance of the Particle Swarm Optimization (PSO) algorithm to solve the parameter estimation problem for a DC servomechanism. The study considers the case when the measurements exhibit high levels of noise. The parameters estimates obtained with the PSO algorithm are compared with those obtained using a standard Least Square (LS) algorithm. Two different data sets are applied to the PSO algorithm, the first set considers position measurements from the DC servomechanism without further processing. The second set corresponds to filtered position measurements using a first order filter with several cut-off frequencies. All the experimental results are obtained in a laboratory test platform where the position measurement are obtained through a potentiometer. This sensor produces measurements with high levels of noise, which is evaluated using a Signal-to-Noise Ratio (SNR) index. The experiments show that positive values of the SNR index translate into parameter estimates obtained with the PSO algorithm closer to those produced by the Least Squares algorithm. On the other hand, negative values of the SNR index correspond to significant discrepancies between these estimates.
Year
DOI
Venue
2019
10.1109/CoDIT.2019.8820564
2019 6th International Conference on Control, Decision and Information Technologies (CoDIT)
Keywords
Field
DocType
measurement noise conditions,parameter estimation problem,DC servomechanism,PSO algorithm,position measurement,filtered position measurements,parameter identification,particle swarm optimization algorithm,signal-to-noise ratio,standard least square algorithm
Least squares,Particle swarm optimization,Data set,Algorithm,Least mean square algorithm,Low-pass filter,Servomechanism,Estimation theory,Potentiometer,Mathematics
Conference
ISSN
ISBN
Citations 
2576-3547
978-1-7281-0522-2
0
PageRank 
References 
Authors
0.34
7
3
Name
Order
Citations
PageRank
R. Cortez-Vega100.34
Jéssica J. Maldonado200.34
Rubén Garrido300.34