| Title |
An Experimental Study on Multi-Fault Detection and Diagnosis Analysis of HVAC System |
| Keywords |
Fault detection and diagnosis ; 고장검출 및 진단 ; Variable air volume ; 가변풍량 ; Neural network ; 신경망 ; HVAC system ; 공조시스템 ; Multi-fault ; 중복고장 |
| Abstract |
The objective of this study is to detect the multi-fault of HVAC system using a new pattern classification technique. To classify the effect of single-fault in determining the pattern, supply air temperature, OA-damper, supply fan, and air flowrate were chosen as experimental parameters. The combination of supply temperature, flow rate, supply fan and OA- damper were chosen as multi-fault conditions. Three kinds of patterns were introduced in the analysis of multi-fault problem. To solve multi-fault problem, the new pattern classification technique using residual ratio analysis was introduced to detect the multi-fault as well as single- fault. The residual ratio could diagnose single-fault or multi-fault into several patterns. |