EVENTO
Control and Filtering for Continuous-time Markov Jump Linear Systems with Partial Mode Information
Tipo de evento: Seminário de Avaliação - Série A
A Markov jump linear system (MJLS for short) is a family of linear systems with the transition between each scenario described by a Markov chain. In general, this class of models is used to represent dynamical systems which are subject to abrupt changes in their structures. Over the past decades, MJLS have attracted a significant research effort due to the wide application in several areas, for instance, in fault-tolerant control systems, manufacturing process, economics, flight systems, power systems, communication systems and networked control systems. As is well known, the body of work surrounding the subject of control theory with reference to MJLS is by now fairly extensive, however it is more scarce when it comes to the setting with partial information, i.e., the scenario where one or some parameters are unknown. Thus, problems regarding partially observable MJLS may be related either to the state variable, the Markov chain or both variables. Recently, discrete-time MJLS with perfect measurement of the state and partial information of the Markov chain were treated in [Costa, O. L. V., Fragoso, M. D., & Todorov, M. G., A Detector-Based Approach for the H-2 Control of Markov Jump Linear Systems with Partial Information. IEEE Transactions on Automatic Control, v. 60, p. 1219-1234, 2015]. In this mentioned paper, the H-2 control design is based on the information coming from a suitable detector which provides measurements of the Markov chain. The detector-based formulation generalizes some of the better cases previously considered in the MJLS literature, i.e, complete observation, mode-independent and cluster observation. In opening up this field, in this work we have developed new results regarding H-infinity control and filtering for continuous-time MJLS with partial mode information.
Data Início: 29/11/2016 Hora: 10:00 Data Fim: 29/11/2016 Hora: 13:00
Local: LNCC - Laboratório Nacional de Computação Ciêntifica - Sala de Reuniões da CSC - 1A56
Aluno: Caio Cesar Graciani Rodrigues - LNCC -
Co-Orientador: Marcelo Dutra Fragoso - Laboratório Nacional de Computação Científica - LNCC
Orientador: Marcos Garcia Todorov - Laboratório Nacional de Computação Científica - LNCC
Participante Banca Examinadora: Jack Baczynski - Laboratório Nacional de Computação Científica - LNCC Marcos Garcia Todorov - Laboratório Nacional de Computação Científica - LNCC Nei Carlos dos Santos Rocha - - UFRJ
