Filtragem robusta : uma abordagem por desigualdades matriciais lineares

AUTOR(ES)
DATA DE PUBLICAÇÃO

1998

RESUMO

: This work deals with the problem of robust filtering for both continuous and discrete-time dynamic linear systems. More precisely, the problem of full order linear filtering for uncertain systems with Hoo norm performance criterion is analyzed. Firstly, the results from the literature based on Riccati like equations and (for norm bounded uncertainty) on coupled Linear Matrix Inequalities - [ MIs are presented and discussed. Then, an [ MIs based solution is proposed for systems with polytope type parameter uncertainties. This solution, based on necessary (in the sense that the filtering error is quadratically stable) and sufficient conditions, allows the optimal 1-íooguaranteed cost to be obtained through convex optimization procedures, with convergence assured. The extension to the case of robust mixed H2/Hoo filtering is also presented. Moreover, the problem of singular filtering (that is, the filteringproblem for systems in which the output is not completely corrupted by the noise signal) subject to unknown inputs is also addressed, and a solution based on a decomposition algorithm is proposed, in such a way that a particular structure is imposed to the filter, combining the robustness of the 1-íoofilter with unknown input observer standard properties... Note: The complete abstract is available with the full electronic digital thesis or dissertations

ASSUNTO(S)

riccati sistemas lineares kalman teoria de estimativa equação de otimização matematica filtragem de

ACESSO AO ARTIGO

Documentos Relacionados