Neural networks offer a flexible, general purpose approach to building the complex highly nonlinear models that are required for a complex system such as a nuclear power plant. Recent developments in learning algorithms for neural networks provide alternatives to the traditional pattern recognition techniques for analysis of diagnostics. This research planning grant involves a comprehensive bench mark test for feasibility study for nuclear reactor signal analysis via a neural network approach. The test consists of a comparison between equipment degradation monitoring by traditional time series analysis versus a neural network approach.
