HiDimFD

Generation of Interpretable Residuals for Fault Diagnosis based on Projection Techniques: Leveraging Variable Redundancy

Abel Alberto Cuadrado Vega, Ignacio Díaz Blanco, José María Enguita González, Diego García Pérez and Ana González Muñiz

Paper submitted for publication in IEEE Transactions on Control Systems Technology

---

Requirements

Statistics and Machine Learning MATLAB Toolbox

SR3 MATLAB Toolbox

Dataset

Dataicann: vibration and current data of an induction motor

Download the dataicann.zip archive and place the contained dataicann.mat file in the directory with the scripts (only needed for sporadic_and_random_faults).

Usage

• Simulations in Section III:

Execute:

	eftsloop

• Experiments in Section IV:

First select in the script sporadic_and_random_faults the type of fault (sporadic or random) by commenting out the line not corresponding from the following two, which can be found in said script:

	fault_type = SPORADIC_FAULT;
	fault_type = RANDOM_FAULT;

Then execute the script:

	sporadic_and_random_faults