Evaluating the performance of a fault detection and diagnostic system for vapor compression equipment
This paper presents a detailed evaluation of the performance of a statistical, rule-based fault detection and diagnostic (FDD) technique presented by Rossi and Braun (1997). Steady-state and transient tests were performed on a simple rooftop air conditioner over a range of conditions and fault levels. The steady-state data without faults were used to train models that predict outputs for normal operation. The transient data with faults were used to evaluate FDD performance. The effect of a number of design variables on FDD sensitivity for different faults was evaluated and two prototype systems were specified for more complete evaluation. Good performance was achieved in detecting and diagnosing five faults using only six temperatures (2 input and 4 output) and linear models. The performance improved by about a factor of two when ten measurements (three input and seven output) and higher order models were used. This approach for evaluating and optimizing the performance of the statistical, rule-based FDD technique could be used as a design and evaluation tool when applying this FDD method to other packaged air-conditioning systems. Furthermore, the approach could also be modified to evaluate the performance of other FDD methods.
- Research Organization:
- Purdue Univ., West Lafayette, IN (US)
- OSTI ID:
- 20002323
- Report Number(s):
- CONF-990102--; ISSN 0001-2505
- Country of Publication:
- United States
- Language:
- English
Similar Records
Hybrid Model-Based and Data-Driven Fault Detection and Diagnostics for Commercial Buildings
Hybrid Model-Based and Data-Driven Fault Detection and Diagnostics for Commercial Buildings: Preprint