Fault Diagnosis of Regenerative Water Heater Based-On Multi-class Support Vector Machines
- Resource Type
- Conference
- Authors
- Wang, Lei; Zhang, Rui-qing
- Source
- 2009 Fifth International Conference on Natural Computation Natural Computation, 2009. ICNC '09. Fifth International Conference on. 1:489-492 Aug, 2009
- Subject
- Computing and Processing
Water heating
Fault diagnosis
Support vector machines
Artificial neural networks
Cogeneration
Turbines
Hydrogen
Heat engines
Classification algorithms
Constraint optimization
steam turbine
regenerative water heater
fuzzy rules
support vector machines
fault diagnosis
- Language
- ISSN
- 2157-9555
2157-9563
The main idea of multi-class support vector machines (SVMs) is described. a multi-class model for regenerative water heater fault diagnosis is presented combining the fuzzy logic and SVMs. The typical faults set of regenerative water heater is built after thoroughly analyzing the relationships between performance parameters and faults. Finally, the model is inspected and verified by an example in a regenerative water heater of the turbine unit, the result of diagnosis shows that it is simple and practical; it can identify the regenerative water heater faults effectively.