Fault diagnosis in split-system air conditioner by analyzing the chaos in the electrical current signal

Computational technological advances for data processing have increasingly contributed to the evolution of predictive techniques for fault detection and diagnosis (FDD) of motors and machines. As a novelty, this article presents the application of the Signal Analysis based on Chaos using Density of Maxima (SAC-DM) technique in split-system air conditioner equipment to diagnose the occurrence of faults. To this end, a short-term acquisition of the electrical signal at the equipment input was carried out in a non-invasive way, subsequently, the proposed technique is used for diagnosing compressor capacitor degradation and incrustation in the condenser and evaporator units at different blocking levels. The results prove a better accuracy compared to the Fast Fourier transform (FFT), making it possible to detect and diagnose air conditioner faults, with an accuracy of 100 % for FDD of single faults, 96.55 % for detection and 82.76 % for diagnosis of dual faults. In addition, the technique demands only short-term data acquisition and a single experiment to detect and diagnose the faults.