Leonardo de Paula Carvalho

Dr. Leonardo de Paula Carvalho

Universidade de São Paulo, São Paulo, Brazil
orcid

Research Focus & Metrics

Academic Career Length

12Years (2015-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
6 January 2020 in IEEE Access
Verified
Journal Article
Open access
LCLeonardo de Paula Carvalho
Leonardo de Paula Carvalho
AOAndre Marcorin De Oliveira
Andre Marcorin De Oliveira
OCOswaldo Luiz V. Costa
Oswaldo Luiz V. Costa
We focus on the Simultaneous Fault Detection and Control (SFDC) in the context of Markov Jump Linear Systems (MJLS). The main novelty of the paper is the design of H∞ and H2 SFDC under the MJLS framework considering partial observation of the Markov chain. Both designs are obtained via Bilinear Matrix Inequalities optimization problem. As secondary results we provide a Mixed H2/H∞ SFDC under the same set up, as well as the implementation of a coordinated descent algorithm to solve the optimization problem formulated as Bilinear Matrix Inequalities (BMI). To illustrate the viability of the proposed solution a numerical example is provided.

1 July 2015
Verified
Conference Paper
Closed access
JJonathan Matias Palma Olate
Jonathan Matias Palma Olate
LCLeonardo de Paula Carvalho
Leonardo de Paula Carvalho
AGAlim P.C. Goncalves
Alim P.C. Goncalves
Christian GalarzaChristian Galarza
Christian Galarza
AOAndre M. De Oliveira
Andre M. De Oliveira
The classical Robust Control Theory applied to network environments does not consider usual network effects like packet losses in the communication between sensors and actuators. In order to design optimal controllers that consider such conditions, one must rely on the class of models dubbed Markovian Jump Linear System (MJLS). In this paper, we study the H-infinity norm degradation in the presence of a limitation on the allowed number of retransmissions in a multi-hop semi-reliable network. Specifically on this study, we implemented the Hop by Hop transport model and also limited the maximum number of retransmissions per packet. For a particular case, it was possible to design a filter with a global number of transmitted packets of only 17.93% of the packets that would be necessary in full reliable communication, yet guaranteeing the estimated error stability.

23 September 2022 in Mathematics
Verified
Journal Article
Open access
PDF file
RFRoberto Matias Fuentes
Roberto Matias Fuentes
JJonathan Matias Palma Olate
Jonathan Matias Palma Olate
HJHildo Guillardi Júnior
Hildo Guillardi Júnior
MLMárcio Lacerda
Márcio Lacerda
LCLeonardo de Paula Carvalho
Leonardo de Paula Carvalho
RORicardo Oliveira
Ricardo Oliveira
This paper investigates the problem of control design for dc–dc converters, where the solution is especially suitable to address variations in the input voltage, a frequent situation in photovoltaic systems, and the problem of constant power load, where a nonlinear load is connected to the output of the converter. The proposed approach models the converters in terms of Linear Parameter-Varying (LPV) models, which are used to compute gain-scheduled robust gains. The synthesis conditions provide stabilizing controllers with an attenuation level of disturbances in terms of the H∞ norm. Moreover, the design conditions can also overcome pole locations to comply with physical application restrictions when ensuring transient performance. The validation of the controllers is made via simulation of the classical converters (buck, boost and buck-boost), showing that the proposed method is a viable and generalized control solution that works for all three converters, with guarantees of closed-loop stability and good performance.

30 September 2020 in International Journal of Robust and Nonlinear Control
Verified
Journal Article
Open access
PDF file
LCLeonardo de Paula Carvalho
Leonardo de Paula Carvalho
TRT. E. Rosa
T. E. Rosa
BJBayu Jayawardhana
Bayu Jayawardhana
OCOswaldo Luiz V. Costa
Oswaldo Luiz V. Costa
We tackle the fault accommodation control (FAC) in the Markovian jump linear system (MJLS) framework for the discrete‐time domain, under the assumption that it is not possible to access the Markov chain mode. This premise brings some challenges since the controllers are no longer allowed to depend on the Markov chain, meaning that there is an asynchronism between the system and the controller modes. To tackle this issue, a hidden Markov chain (, ) is used where denotes the Markov chain mode, and denotes the estimated mode. The main novelty of this work is the design of and FAC under the MJLS framework considering partial observation of the Markov chain. Both designs are obtained via bilinear matrix inequalities optimization problems, which are solved using coordinate descent algorithm. As secondary results, we present simulations using a two‐degree of freedom serial flexible joint robot to illustrate the viability of the proposed approach.

7 August 2018 in Sensors
Verified
Journal Article
Open access
PDF file
JJonathan Matias Palma Olate
Jonathan Matias Palma Olate
CDCristian Duran-Faundez
Cristian Duran-Faundez
LCLeonardo de Paula Carvalho
Leonardo de Paula Carvalho
CCecília Morais
Cecília Morais
RORicardo Oliveira
Ricardo Oliveira
KHKrzysztof Herman
Krzysztof Herman
This paper proposes a new communication protocol for output-feedback control through multi-hop Wireless Sensor Network (WSN). The protocol is based on a Hop-by-Hop transport scheme and is especially devised to simultaneously fulfill two conflicting criteria: the network energy consumption and the stability/performance (in terms of H∞ norm) of the closed-loop system. The proposed protocol can be implemented by means of three heuristics, basically using distinct rules to control the maximum number of retransmissions allowed in terms of the voltage level of the batteries of the network nodes. As another contribution, a Markov jump based representation is proposed to model the packet loss in the communication channel, giving rise to a systematic procedure to determine the transition probability matrix and the Markov chain operation modes of a network with multiple information sources. The synthesis of the output-feedback controller is made in two steps (observer filter plus a state-feedback controller) for the Markov model assuming partial availability of the operation modes. The efficiency and applicability of the communication protocol is illustrated by means of a numerical experiment, based on a physical model of a coupled tanks plant. The features of each heuristic of implementation of the proposed protocol are presented in the numerical comparisons.

10 October 2016
Verified
Conference Paper
Closed access
LCLeonardo de Paula Carvalho
Leonardo de Paula Carvalho
JJonathan Matias Palma Olate
Jonathan Matias Palma Olate
AGAlim P. C. Goncalves
Alim P. C. Goncalves
In the last two decades the car pursuit problem has been heavily studied, mainly because it is a viable solution to minimize the occurrence of accidents on the road. The controller design usually considers that all measured signals are transmitted by a perfect network and that sensors and actuators do not fail. In a real implementation, these equipments may present failure and the network quality of transmission may not be fully reliable. Using Network Control System theory, it is possible to design a controller that takes into account the probabilities of failure in sensors, actuators or interruptions in signal communications. In this paper we compare the classical H∞ output feedback controller with two different types of Markov Jump Linear Systems (MJLS) controllers. We show that, on a network with 25% packet loss rate, the proposed MJLS controllers outperform the classical H∞ controllers by a factor of 10.