Hybrid Particle Swarm Optimization and Unscented Filtering Technique for Estimation of Non-stationary Signal Parameters

Abstract

This paper proposes an adaptive unscented Kalman filter for parameter estimation of non-stationary signals, like amplitude and frequency, in the presence of significant noise and harmonics. This paper proposes an iterative update equation for model and measurement error covariances Q and R to improve tracking of the filter in the presence of high noise. The initial choice of the model and measurement error covariances Q and R, along with the UKF parameters, are crucial in noise rejection. This paper utilizes a modified particle swarm optimization (MPSO) algorithm for the initial choice of the error covariances and UKF parameters. Various simulation results for time varying signals reveal significant improvement in noise rejection and accuracy in obtaining the frequency and amplitude of the signal.

Keywords:

• Extended Kalman filter
• Modified particle swarm optimization
• Unscented Kalman filter

Additional information

Notes on contributors

P. K. Dash

P. K. Dash (SM’1990) holds D.Sc., Ph.D., M.E., and B.E.. Degrees in Electrical Engineering and had his Post-Doctoral education at the University of Calgary, Canada. He is working as a Director (PG & Research), Silicon Institute of Technology, Bhubaneswar, India. Earlier he was a Professor in the Faculty of Engineering, Multimedia University, and Cyberjaya, Malaysia. He also served as a Professor of Electrical Engineering & Chairman, Center for Intelligent Systems, National Institute of Technology, Rourkela, India for more than 25 years. His research interests are in the area of Power Quality, FACTS, Soft Computing, Deregulation and Energy Markets, signal processing, and Data mining and Control. He had several visiting appointments in Canada, USA, Switzerland, and Singapore. To his credit he has published more than 200 International Journal papers and nearly 100 in International conferences. He is a Fellow of the Indian National Academy of Engineering and senior member of the IEEE, and Fellow of Institution of Engineers, India. E-mail: [email protected]

B. K. Panigrahi

B. K. Panigrahi is working as Assistant Professor in the Department of Electrical Engineering IIT Delhi. Prior to joining IIT Delhi he has served as Lecturer in University collage of Engg, Burla, Sambalpur, Orissa for 13 years. He is a senior member of IEEE. His research interests are application of Evolutionary computation techniques to power system pianning, operation and control. He has published more than 100 technical papers in IEEE, Elsevier and other international reputed journals on soft computing and its applications. E-mail: [email protected]

Shazia Hasan

Shazia Hasan is working as Lecturer in the Department of Electronics Engineering, Silicon Institute of Technology, Bhubaneswar, Orissa. She received her BE degree in electronics and telecommunication engineering from University collage of Engg, Burla,Sambalpur,Orissa, in the year 2002. Her area of interest include Digital signal processing and its application to power system. E-mail: [email protected]