Fault diagnosis of an industrial once-through benson boiler by utilizing adaptive neuro-fuzzy inference system

Nowadays, the increasing request to have more electric power and the growing complexity of advanced thermal power systems, make it ever more important to improve the performance and reliability of the systems. Hence, an attention is concentrated on fault diagnosis systems to compensate the adverse effects automatically, under conditions of noisy measurement. In order to improve the proficiency of process monitoring and increase accuracy of fault diagnosis (FD) for the once-through Benson type boiler, this article proposed a data driven method based on the configuration of six adaptive neuro fuzzy inference systems (ANFIS). In the proposed structure, due to strong interaction between measurements each ANFIS classifier has been developed to diagnose one particular fault. Finally to evaluate the effectiveness and performance of the proposed FD system against 6 major faults of once-through Benson type boiler under conditions of noisy measurement, different set of test scenarios have been performed.