Identificação de sinais ruidosos em sistemas de potência e representações na forma de pólos e resíduos

AUTOR(ES)
DATA DE PUBLICAÇÃO

2007

RESUMO

This thesis presents a technique for continuous systems identification from information in z domain transform. The computations allow for the noisy signal action on the output signal. The transfer functions are calculated from time-domain system impulse response, taking into account test system input and output signals. The transfer function is identified in the z domain. However, its structure is presented in the form of poles and their associated residues. The poles are calculated in the s domain by using a mapping transform function which depends on the sampling period and the z poles. The noisy signals are filtered through a denoising process (noise minimization) whose procedure uses wavelets transform. Tests are performed on three power systems to assess the validity of the identification method under different operating conditions and different noise levels. The reduced models identified are used for the design of power system stabilizers (PSSs). These tests are carried out considering a database representing the Brazilian interconnected power grid of the year 2005. In all cases, the identification proved to be quite satisfactory, comparing to the results obtained with a complete and identified model. Linear and non-linear simulations of the identified and the full models show the effectiveness of the system identification technique.

ASSUNTO(S)

denoising poles and residues automacao eletronica de processos eletricos e industriais transformada z estabilizadores de sistemas de potência identificação de sistemas denoising power system stabilizers z transform pólos e resíduos identification systems

ACESSO AO ARTIGO

Documentos Relacionados