Euripedes Guilherme de Oliveira Nobrega

Graduated in Electronic Engineering from Instituto Tecnológico de Aeronáutica (1973), Master in Mechanical Engineering from Universidade Federal da Paraíba (1979) and PhD in Electrical Engineering from Universidade Estadual de Campinas (1992). Developed a postdoctoral project at the University of Houston (2001). He is currently Full Professor MS-6 (retired) at the State University of Campinas, Department of Computational Mechanics at the Faculty of Mechanical Engineering, where he teaches since 1988. He was one of the creators of the Control and Automation Engineering course, which started its activities in 1998, having been the course coordinator for 6 years, starting in August 2005. He was a visiting professor at ENSIM - Le Mans Université, INPG - Université Grenoble Alpes and Arts et Métiers Paristech - Paris. Having multidisciplinary training and experience, he develops research in several areas, with an emphasis in recent years on the use of Artificial Intelligence (AI) and Machine Learning methods applied to signal processing in the context of Monitoring, Detection and Fault Diagnosis Systems, the which also extends to Fault Tolerant Control of machinery and mechanical structures subject to damage. Having also worked continuously in the design and programming of intelligent digital instrumentation, for the various monitored systems, he has designed intelligent embedded equipment, with the use of microcontrollers and processors of various types. In the last decade, monitoring and diagnostic systems have made use of FPGA, associated with microcontrollers, aiming at encapsulating AI methods hardware in distributed embedded systems, using modern languages, in particular Python and OpenCL, in addition to C ++, aiming at parallel programming, and thus achieve high performance by parallelizing algorithms in real time. Such methods are applied to time series processing, with signals and data from systems in various areas of engineering, having worked on the diagnosis of rotating machines, detection of arrhythmias on electrocardiograms, detection and location of damage in flexible mechanical structures, in addition to control damage-tolerant asset. Intelligent signal processing involves programming new method algorithms such as spectral analysis, time-frequency analysis, perceptron neural networks, convolution networks, recurrent networks, and statistical machine learning methods, both supervised and unsupervised, which have been used in several applications and publications, using conventional computers, microcontrollers, digital signal processors (DSP's), multi-core graphics processors (GPU's) and FPGA's, either alone or in distributed networks. (Source: Lattes Curriculum)