DEVELOPMENT OF FAULT TOLERANT CONTROL SYSTEM FOR CONDENSATION POWER TURBINE

An application of fault tolerant control (FTC) system for condensation turbine has been presented in the paper. The tolerance of instrument faults as well as the main actuator fault has been achieved by the use of functional redundancy which consists in performance of appropriate changes in the operation manner of the faulty system. The method of fuzzy neural networks (FNN) has been used for fault detection and isolation (FDI) purposes. The FNN models of turbine power, steam pressure and steam mass flow rate were created and verified. Also “2 from 3” hardware redundancy has been applied for the most important measurement, which is the turbine rotary speed. Appropriate control algorithms for different faulty situations have been prepared. The application has been developed in the Institute of Heat Engineering on the basis of ISaGRAF software and VME-DEVSYS VM30 industrial computer (Contron, former PEP Modular Computers).