Fault detection of gas turbine engines using dynamic neural networks

The main objective of this paper is to develop a neural network-based scheme for fault detection of an aircraft engine. Towards this end, a set of dynamic neural networks (DNN) are developed to learn the dynamics of the jet engine. The DNN is constructed based on a dynamic multilayer perceptron network which uses IRR filters to generate dynamics between the input and the output of the system. Our proposed DNN does not require a delayed sample of the output and is developed based on a single-input single-output (SISO) network. Consequently, the structure of the network would be the same in the training and the recall phases. The dynamic neural network that is described in this paper is developed to detect component faults that may occur in a dual spool turbo fan engine. Various simulations are carried out to demonstrate the performance of our proposed fault detection scheme.