Classifier Fusion in the Accelerometer Sensor Diagnostics

Piotr Bilski

Abstract

The paper presents the application of multiple classifiers to increase the accuracy of the fault detection module. The structure of the module is presented with the applied voting mechanisms. The classifiers implemented to the fault detection operation are briefly described. Issues of the practical implementation are considered. The scheme is used in the diagnostics of the analog part of the MEMS accelerometer. The paper is concluded with the possible prospects for the proposed scheme.

Keywords

analog systems, accelerometer sensor, fault classification, classifier fusion

Download

IMEKO-TC10-2014-037.pdf

DOI

-

IMEKO TC

TC10 - Measurement for Diagnostics, Optimization and Control

Back to the proceedings