Smart Sensor for Fault Detection in Induction Motors Based on the Combined Analysis of Stray-Flux and Current Signals: A Flexible, Robust Approach
Israel Zamudio-Ramírez, Roque A. Osornio‐Rios, Jose A. Antonino‐Daviu
Abstract
The most recent trend in the electric motor condition monitoring area relies on combining the information obtained from different machine quantities to reach a more reliable conclusion about the motor’s health. This knowledge is of critical importance today, especially in industrial applications, in which unexpected outages can lead to severe repercussions. This article presents a new intelligent sensor that combines, in a single unit, the information obtained from the analysis of stray fluxes (both axial and radial) and currents by means of a feedforward neural network (FFNN) for classification purposes. Unlike other solutions, the sensor is based on the application of advanced signal processing tools that are adapted to the online analysis of these quantities under transient conditions from a single processing unit (a smart sensor). The combination of these new tools with the classical steady-state analysis of such quantities enables one to obtain a more reliable conclusion on the motor health. The experiments included in the article demonstrate the reliability provided by the sensor, which is being prepared to incorporate a third input based on infrared data.