Markovian theory applied for the development of control strategies in rehabilitation robotics

Author(s)

Mitschka, Christoph M. ; Terra, Marco H. ; Siqueira, Adriano A. G. ; IEEE

• Document type: Journal article
• Source: 2017 AMERICAN CONTROL CONFERENCE (ACC); v. N/A, p. 6-pg., 2017-01-01.

Abstract

Markovian theory found its application in plenty of areas where control strategies are taken advantage of. However in rehabilitation robotics, which are safety critical applications, such strategies are not a common practice. This paper shows the potential of Markovian theory applied in rehabilitation robotics and proposes two different approaches for a development of such type of control strategies. This work proposes a variable control strategy driven by a Markovian decision process of which one contains off-line calculated gains using linear matrix Inequalities techniques for computation and a recursive on-line controller based on robust linear quadratic regulators. (AU)

• FAPESP's process: 12/14074-3 - Multisensor-fusion and regulation of robotic rehabilitation systems via robust filtering and Markovian control
• Grantee: Christoph Michael Mitschka
• Support Opportunities: Scholarships in Brazil - Doctorate (Direct)