Fault Detection for Air Conditioner using PCANet

Studies on fault detection are particularly important to keep air conditioner in good situation. This paper uses a principal component analysis network(PCANet) to achieve fault detection. The PCANet can carry out feature extraction and model training simultaneously. This model relies on cascaded pri...

Full description

Saved in:
Bibliographic Details
Main Authors: Yang, Hailong, Yang, Zuyuan, Yang, Hao, Xie, Yuhui
Format: Conference Proceeding
Language:English
Subjects:
Air Conditioner
Atmospheric modeling
Fault detection
Feature extraction
Histograms
Home appliances
PCANet
Principal component analysis
Training
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Studies on fault detection are particularly important to keep air conditioner in good situation. This paper uses a principal component analysis network(PCANet) to achieve fault detection. The PCANet can carry out feature extraction and model training simultaneously. This model relies on cascaded principal component analysis which is employed to learn multistage filter banks and followed by simple binary hashing and block histograms for indexing and pooling. In this study, a series of experiments have been carried out against air conditioner noise datasets. Experimental results indicate that the PCANet model's outperforms are satisfactory in terms of accuracy and efficiency.
ISSN:2161-2927
DOI:10.23919/ChiCC.2019.8866100