Structural optimization of the neural network model for the gas turbine engine monitoring

The paper presents a technique of the neural network model reduction for the GTE operation mode monitoring during bench tests. The technique is based on the conversion of the neural network training task to the of multi-criteria optimization task.