تشخيص عيب در توربين گاز به كمك يك سيستم گسسته پيشامد

Fault Diagnosis of Gas Turbine Using Discrete Event System

اين مقاله به تشخيص عيب در توربين گاز به كمك يك سيستم گسسته پيشامد ( Discrete Event System) مي پردازد. توربين مورد بحث توربين 94.2 است كه تعداد زيادي از اين توربين ها در ايران نصب شده اند. سيستم گسسته پيشامد مورد استفاده يك اتومات زماني است كه كاربرد خوبي در سيستمهاي كنترل نظارتي و تشخيص عيب دارد. در اين مقاله با استفاده از زمان رخداد پيشامدها، يك سيستم عيب ياب جديد طراحي مي شود. براي استخراج اتومات زماني نيز، با افراز فضاي پيوسته خروجي، زمان رخداد پيشامدهاي گسسته محاسبه و سپس مدل اتومات زماني سيستم به دست مي آيد. نتايج حاصل از به كارگيري اين روش براي توربين V94.2 نشان دهنده سرعت خوب در تشخيص عيوب و عدم تاثير اختلال هايي مانند تغييرات بار مصرفي است.

This paper deals with fault diagnosis of V94.2 Gas Turbines (GTs) which have been widely used in Iran power plants in the recent years (more than 150 units). Theses turbines are expensive and must work continuously to have an admissible performance. So, precise fault detection of these plants is an important and serious issue and it has several technical and economic advantages for users of these units. As a result, researches have been increasingly developed in this field and some novel techniques and theories have been adopted for fault diagnosis of Gas turbines. In this paper, a model based fault diagnosis system is presented for V94.2 GT. The diagnoser is based on one of the most known Discrete Event Systems (DES), Timed Automaton (TA). Parameters of the TA are obtained by using a GT simulator. The Rowen model is used as a simulator in this research. Faultless and faulty modes of GT are simulated by this simulator, then the needed data are obtained for corresponding TAs. A novel diagnoser is proposed in this research. This diagnoser is a state machine that can detect the faults by observing the sequence of timed events occurred in the GT.