A FDD model of VAV systems based on neural-networks and residual statistics

Author

Han Qi ; Wei Dong

Author_Institution

Beijing Univ. of Civil Eng. & Archit., Beijing, China

fYear

2014

fDate

May 31 2014-June 2 2014

Firstpage

1555

Lastpage

1559

Abstract

Components and sensors in VAV (Variable Air Volume) air distribution systems often suffer from failure easily, which result in energy waste, performance degradation or totally out of control. However, there is no applicable automatic commissioning tool for the VAV systems by now. Fault detection and diagnosis (FDD) models for VAV terminal units based on heat-mass balance of air conditioning areas are proposed in this study. Two BP neural network prediction models are built up for predicting the required air flow volume and the demand values of VAV damper opening, Fault detection can be implemented by means of statistics of the residuals between the measured values and the model predictions.

Keywords

air conditioning; backpropagation; control engineering computing; fault diagnosis; mechanical engineering computing; neural nets; statistical analysis; temperature control; BP neural network prediction models; FDD model; VAV air distribution systems; VAV damper opening; air conditioning areas; air flow volume; automatic commissioning tool; backpropagation; fault detection and diagnosis; heat-mass balance; residual statistics; variable air volume; Air conditioning; Atmospheric modeling; Buildings; Predictive models; Shock absorbers; Temperature sensors; BP neural network; Fault detection and diagnosis; Residual statistical; VAV air conditioning;

fLanguage

English

Publisher

ieee

Conference_Titel

Control and Decision Conference (2014 CCDC), The 26th Chinese

Conference_Location

Changsha

Print_ISBN

978-1-4799-3707-3

Type

conf

DOI

10.1109/CCDC.2014.6852414

Filename

6852414