Principal Investigator: Jonathan M. Palma
Associated Researchers: Karina Barbosa, Hicham El Aiss, Márcio Júnior Lacerda, Cecília Freitas Morais, André Marcorin de Oliveira, Gabriela W. Gabriel, Leonardo de Paula Carvalho etc
Principal Investigator: André Marcorin de Oliveira
Supervisor: Oswaldo L. V. Costa
Abstract: The purpose of this research plan for the Post-Doctorate Fellowship with the National Institute of Science and Technology for Cooperative Autonomous Systems Applied to Security and Environment is to further develop the theory of control systems of Markov jump linear systems with partial observation of the Markov chain, by deriving new conditions for the design of dynamic controllers and filters that do not depend explicitly on the mode of operation but instead rely on an estimation provided by a fault detector. The model that will be considered is more general and encompasses the so-called detector approach already studied in the literature. The proposed techniques will be validated by means of relevant applications in Active Fault-Tolerant Control Systems.
Grantee: André Marcorin de Oliveira
Supervisors: Oswaldo L. V. Costa, Jamaal Daafouz
Abstract: This document presents a research project to be performed by the PhD applicant André Marcorin de Oliveira from the University of São Paulo - USP, Brazil, under the supervision of Prof. Dr. Jamal Daafouz (Université de Lorraine - France) and Prof. Dr. Oswaldo Luiz do Valle Costa (University of São Paulo - Brazil). The proposed research activities will take place at the Université de Lorraine in Nancy, France, from September, 1st 2017 to August, 31th 2018. The main goal is to bring both groups of research closer in order to share their experiences and to extend the theory of partial observation of the Markov mode by developing conditions for the synthesis of filters and dynamical controllers that do not depend explicitly on the mode.
Grantee: André Marcorin de Oliveira
Supervisor: Oswaldo L. V. Costa
Abstract: This research plan concerns the analysis of methods for obtaining parameters which governs the dynamics of stochastic systems. Specifically, we aim to study the Markov jump linear systems with partial observation of the Markov mode in order to estimate the state and the mode itself through information of filters and detectors. In addition, the behaviour and performance of estimated mode dependent controllers applied to Markov linear systems shall be studied.
Grantee: André Marcorin de Oliveira
Supervisor: Alim Gonçalves
Abstract: This research plan concerns the project of controllers for mechanical systems in which the communication between the controller and the plant, both the measurements and the controlling signal, occurs through a non-ideal communication channel. Particularly, we aim to study the effects of packet loss through the network. The project is going to be implemented in a rotary inverted pendulum whose modeling, identification and discretization are included in the project objectives. In addition, different models for the communication channel are going to be tested, e.g. the Gilbert-Elliot, Fritchman and McCullough models, including different probabilities for the events associated with the channel modeling.
Principal Investigator: André Marcorin de Oliveira
Supervisor: Josué Ramos
Principal Investigator: André Marcorin de Oliveira
Supervisor: Josué Ramos