2014

Performance Analysis of Extreme Learning Machine for Automatic Diagnosis of Electrical Submersible Pump Conditions.

INDIN 2014 - 12th IEEE International Conference on Industrial Informatics, 2014, Porto Alegre, RS. Proc. of 12th IEEE International Conference on Industrial Informatics, 2014.

Pesquisadores: Francisco de Assis Boldt, Thomas W. Rauber, Flávio Miguel Varejão, Marcos Pellegrini Ribeiro


2013

Automatic diagnosis of submersible motor pump conditions in offshore oil exploration

IECON 2013 - The 39th Annual Conference of the IEEE Industrial Electronics Society, 2013, Vienna. Proc. of the 39th Annual Conference of the IEEE Industrial Electronics Society, 2013.

Pesquisadores: Alexandre Rodrigues Loureiros, Fábio Fabris, Flávio Miguel Varejão, Thomas W. Rauber, Marcos Pellegrini Ribeiro


2013

Computational Intelligence for Automatic Diagnosis of Submersible Motor Pump Conditions in Offshore Oil Exploration

IEEE International Conference on Electronics, Circuits, and Systems, 2013, Abu Dhabi. Proc. of IEEE International Conference on Electronics, Circuits, and Systems, 2013.

Pesquisadores: Thomas W. Rauber, Francisco de Assis Boldt, Flávio Miguel Varejão


2013

Feature Extraction and Selection for Automatic Fault Diagnosis of Rotating Machinery

ENIAC 2013 - Encontro Nacional de Inteligência Artificial e Computacional, 2013, Fortaleza, CE. Proc. of Encontro Nacional de Inteligência Artificial e Computacional, 2013.

Pesquisadores: Francisco de Assis Boldt, Thomas W. Rauber, Flávio Miguel Varejão


2013

Motor pump fault diagnosis with feature selection and Levenberg-Marquardt trained feedforward neural network

15th International Conference on Computer Analysis of Images and Patterns (CAIP 2013), 2013, York. Proc. of the 15th International Conference on Computer Analysis of Images and Patterns, 2013.

Pesquisadores: Thomas W. Rauber, Flávio Miguel Varejão

Palavras-chave: Fault diagnosis, feature selection, feedforward neural network, Levenberg-Marquardt

We present a system for automatic model-free fault detection based on a feature set from vibrational patterns. The complexity of the feature model is reduced by feature selection. We use a wrapper approach for the selection criteria, incorporating the training of an artificial neural network into the selection process. For fast convergence we train with the Levenberg-Marquardt algorithm. Experiments are presented for eight different fault classes.


2013

Using GA for the stratified sampling of electricity consumers

IEEE Congress on Evolutionary Computation (CEC), 2013, Cancun. 2013 IEEE Congress on Evolutionary Computation, 2013. p. 261-268

Pesquisadores: Estevão Costa, Fábio Fabris, Alexandre Rodrigues Loureiros, Hannu Ahonen, Flávio Miguel Varejão, Rodrigo Marin Ferro

Palavras-chave: Genetic Algorithm, simulated annealing, energy loss and sampling

Non-technical energy losses mostly arise from illegal use of energy and force energy distribution companies to inspect large batches of clients in order to make decisions on actions for reducing these losses. Since an exhaustive inspection is impractical due to the high inspection cost and the very large number of clients, a carefully designed sampling procedure is needed. A useful strategy is offered by stratified sampling based on a division of the clients into homogeneous subgroups (strata). In this work we formulate the stratification task as a non-linear restricted optimization problem, in which the variance of overall energy loss due to the fraudulent activities is minimized. Solving this problem analytically is difficult and an exhaustive algorithm is intractable even for small problem instances. Therefore, we propose a Genetic Algorithm for finding practical solutions for the problem. Numerical experiments and a comparison with Simulated Annealing algorithm and a proportional allocation scheme are presented.


2012

A Clustering-Based Approach for Estimating Energy Loss from Non-Metered Installations

4th Asia-Pacific Power and Energy Engineering Conference, 2012, Xangai. Proceedings of the 4th Asia-Pacific Power and Energy Engineering Conference, 2012

Pesquisadores: Gustavo Corteletti Venturini, Fábio Fabris, Alexandre Rodrigues Loureiros, Flávio Miguel Varejão, Rodrigo Marin Ferro


2012

Kernel Enhanced Multilayer Perceptron for Industrial Process Diagnosis

IJCNN International Joint Conference on Neural Networks, 2012, Brisbane. Proc. of the 2012 IJCNN International Joint Conference on Neural Networks, 2012

Pesquisadores: Lucas Henrique Sousa Mello, Flávio Miguel Varejão, Thomas W. Rauber

We perform an empirical performance analysis of the Multilayer Perceptron applied to the fault diagnosis of motor pumps installed on oil rigs. The conventional Multilayer Perceptron architecture is compared to a recently developed enhancement of this general purpose regression/classification paradigm, using an intermediate opaque layer which maps the original patterns to a reproducing kernel Hilbert space prior to learning the usual functional mapping of the network. State of the art statistical tools are used to corroborate our hypotheses that the kernel enhanced version improves the classification performance.


2012

Otimização Amostral para Obtenção da Matriz de Perdas Não-Técnicas

XX SENDI - Seminário Nacional de Distribuição de Energia Elétrica, 2012, Rio de Janeiro. Anais do XX SENDI - Seminário Nacional de Distribuição de Energia Elétrica, 2012

Pesquisadores: Gustavo Corteletti Venturini, Fábio Fabris, Flávio Miguel Varejão, Rodrigo Marin Ferro, Alexandre Rodrigues Loureiros


2011

Condition Monitoring based on Kernel Classifier Ensembles

IEEE 9th International Conference on Industrial Informatics, 2011, Lisboa. Proceedings of IEEE 9th International Conference on Industrial Informatics, 2011.

Pesquisadores: Flávio Miguel Varejão, Thomas W. Rauber, Eduardo Mendel do Nascimento, Rodrigo J. Batista

The objective of this work is the model-free diagnosis of faults of motor pumps installed on oil rigs by sophisticated kernel classifier ensembles. Signal processing of vibrational patterns delivers the features. Different kernel-based classifiers are combined in ensembles to optimize accuracy and increase robustness. A comparative study of various classification paradigms, all performing implicit nonlinear pattern mapping by kernels is done. We employ support vector machines, kernel nearest neighbor, Bayesian Quadratic Gaussian classifiers with kernels, and linear machines with kernels.


Pages