Monitoring system for an experimental facility using GMDH methodology

Abstract

This work presents a Monitoring System developed based on the GMDH - Group Method of Data Handling methodology to be used in an Experimental Test Facility. GMDH is a combinatorial multi-layer algorithm in which a network of layers and nodes is generated using a number of inputs from the data stream being evaluated. The GMDH network topology has been traditionally determined using a layer by layer pruning process based on a pre-selected criterion of what constitutes the best nodes at each level. The traditional GMDH method is based on an underlying assumption that the data can be modeled by using an approximation of the Volterra Series or Kolmorgorov-Gabor polynomial. The Fault Test Experimental Facility was designed to simulate a PWR nuclear power plant and is composed by elements that correspond to the pressure vessel, steam generator, pumps of the primary and secondary reactor loops. The nuclear reactor core is represented by an electrical heater with different values of power. The experimental plant will be fully instrumented with sensors and actuators, and the data acquisition system will be constructed in order to enable the details of the temporal analysis of process variables. The Fault Test Experimental Facility can be operated to generate normal and fault data. These failures can be added initially with small magnitude, and their magnitude being increasing gradually in a controlled way. The database will interface with the plant supervisory system SCADA (Supervisory Control and Data Acquisition) that provides the data through standard interface.