IDENTIFICATION OF NONLINEAR SYSTEMS USING FIXED BUDGET KERNEL LMS ALGORITHM

Abstract

In this paper a new version of kernel normalized least mean squares algorithm is applied to identification of nonlinear system. To maintain a fixed amount of support vectors, requisite for practical kernel-based algorithm, a pruning criterion is used. After admitting a new input vector to the dictionary, a least important entry is selected and discarder. A case of nonlinear system identification is presented, proving that algorithm performs well and it can maintain a performance comparable to state-of-the-art algorithms, using smaller number of support vectors.

Authors and Affiliations

Dominik Rzepka, Piotr Otfinowski

Keywords

Related Articles

NUMERICAL COMPUTATIONS OF THE FRACTIONAL DERIVATIVE IN IVPS, EXAMPLES IN MATLAB AND MATHEMATICA

The paper concerns a numerical method that deals with the computations of the fractional derivative in Caputo and Riemann-Liouville definitions. The method can be applied in time stepping processes of initial value probl...

DETERMINATION OF ELECTROMAGNETIC RADIATION EMITTED BY TELECOMMUNICATION EQUIPMENT OF TRANSMISSION CENTERS

Determining and forecasting of environmental influence of telecommunication equipment that is a source of electromagnetic radiation is still an actual challenge. The approach proposed enables to assess an impact of wide...

RESEARCH OF DC/DC CONVERTER WITH ISOLATION TRANSFORMER TO SUPPLY THE NEEDS OF TRACTION VEHICLES

This paper presents research results of DC/DC converter with isolation transformer designed to supply the needs of traction vehicles. It also presents comparative research of isolation transformer with high output curren...

SZTUCZNE SIECI NEURONOWE JAKO MOTORYCZNA PAMIĘĆ ASOCJACYJNA RĘKI ROBOTA HUMANOIDALNEGO

Opisane badania dotyczą użycia sztucznych sieci neuronowych przy realizacji asocjacyjnej pamięci motorycznej zarządzającej ręką robota humanoidalnego. Zaproponowano model kognitywnego sterowania odwołujący się do struktu...

ALGORITHM OF TIME STAMPS DISTRIBUTION FOR DATA SYNCHRONIZATION OF VARIOUS MEASUREMENT SYSTEMS

The described in this paper research refer to the finished project which aimed to build the measurement system for phase distribution calculation and two phase flow structure identification based on 3D ECT system. In the...

Download PDF file
  • EP ID EP59394
  • DOI -
  • Views 119
  • Downloads 0

How To Cite

Dominik Rzepka, Piotr Otfinowski (2012). IDENTIFICATION OF NONLINEAR SYSTEMS USING FIXED BUDGET KERNEL LMS ALGORITHM . Informatyka Automatyka Pomiary w Gospodarce i Ochronie Środowiska, 2(4), 10-13. https://europub.co.uk./articles/-A-59394