الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 16 |  |  |
| | | from | 16 | | |
Abstract
Diagnosis is a very complex and important task for finding the root cause offaults in nuclear power plants. The objective of this thesis is to investigate the feasibility of using the combination of signed directed graph and artificial neural networks for fault diagnosis in
nuclear power plants.
Signed directed graph is constructed to represent the cause-effect relations among the
dynamic process variables. Signed directed graph consists of nodes representing the process variables and branches representing the qualitative influence of a process variables on the related variable. The main problem in fault diagnosis using the signed directed graph is the unmeasured variables. Therefore, artificial neural networks are used to estimate the values of unmeasured nodes. The cause-effect graph for each fault is constructed trom the signed directed graph. Then in the cause-effect graph we search about the node which does not have an input branch. This node is the fault origin node.
In this work, a new approach for using signed directed graph in nuclear power plant fault diagnosis has been developed. The combination of signed directed graph and artificial neural networks is used for fault diagnosis in the steam generator and primary circuit. The signed directed graphs ofthe primary circuit and steam generator are constructed to represent the cause-effect relations among the variables. The used data are collected trom a basic principle simulator of a pressurized water reactor (925 Mwe).
The result of this work demonstrated that this method, enables us to diagnose a multiple fault and it is not restricted by single fault assumption. Also, it is not restricted by pre-defined faults as expert systems and neural networks. Also by using this method, the fault propagation can be followed. So This method can be used successfully for fault diagnosis in nuclear power plants.