Showing papers by "Ricardo A. Ramirez-Mendoza published in 2011"
TL;DR: In this article, a semi-active control system of an automotive suspension with an Magneto-Rheological (MR ) damper as a key element is considered. And two controllers are proposed from different approaches: LPV control and a Frequency Estimation-Based (FEB ) control.
26 citations
28 Jun 2011
TL;DR: Dynamic Principal Component Analysis (DPCA) and Artificial Neural Networks are compared in the fault diagnosis task to compare the online performance of both approaches in terms of quick detection, isolability capacity and multiple faults identifiability.
Abstract: Dynamic Principal Component Analysis (DPCA) and Artificial Neural Networks (ANN) are compared in the fault diagnosis task. Both approaches are process history based methods, which do not assume any form of model structure, and rely only on process historical data. Faults in sensors and actuators are implemented to compare the online performance of both approaches in terms of quick detection, isolability capacity and multiple faults identifiability. An industrial heat exchanger was the experimental test-bed system. Multiple faults in sensors can be isolated using an individual control chart generated by the principal components; the error of classification was 15.28% while ANN presented 4.34%. For faults in actuators, ANN showed instantaneous detection and 14.7% lower error classification. However, DPCA required a minor computational effort in the training step.
4 citations
25 Jul 2011
TL;DR: In this paper, the authors compared Dynamic Principal Component Analysis (DPCA) and Artificial Neural Networks (ANN) in the fault diagnosis task, which do not assume any form of model structure and rely only on process historical data.
Abstract: Dynamic Principal Component Analysis (DPCA) and Artificial Neural Networks (ANN) are compared in the fault diagnosis task. Both approaches are process history based methods, which do not assume any form of model structure, and rely only on process historical data. Faults in sensors and actuators are implemented to compare the online performance of both approaches in terms of quick detection, isolability capacity and multiple faults identifiability. An industrial heat exchanger was the experimental test-bed system. Multiple faults in sensors can be isolated using an individual control chart generated by the principal components; the error of classification was 15.28% while ANN presented 4.34%. For faults in actuators, ANN showed instantaneous detection and 14.7% lower error classification. However, DPCA required a minor computational effort in the training step.
1 citations
20 Jun 2011
TL;DR: A comparison between different state-of-the-art and a Linear Parameter Varying (LPV) models for Magneto-Rheological (MR) dampers is presented in this paper.
Abstract: An MR damper is a device that exhibits a high nonlinear and complex behavior with a hysteresis phenomenon. A comparison between different state-of-the-art and a Linear Parameter Varying (LPV) models for Magneto-Rheological (MR) dampers is presented. Several experimental datasets validate that Linear Parameter Varying (LPV)-based model outperforms the classical MR damper models for 51 % than any structures considering the Error to Signal Ratio index and 37 % better considering the Squared root of Sum of Squared Errors index.
1 citations
20 Jun 2011
TL;DR: The results showed that the fuzzy-based method was able to correctly introduce the dependency of the model to the electric current and reduce the prediction error by more than 12%.
Abstract: A Magneto-Rheological (MR) damper exhibits a hysteretic and non-linear behavior. This behavior makes it a challenge to develop a model for the system. The present research is centered on proposing and analyzing a fuzzy-based method employed to introduce the electric current dependency into MR damper models, based on experimental data. Among the state of the art, a semi-phenomenological model was selected. The fuzzy-based method was compared against other approaches. The results showed that the fuzzy-based method was able to correctly introduce the dependency of the model to the electric current and reduce the prediction error by more than 12%.