Gas turbine engine prices vary widely. Any organisation planning to invest in a project involving the use of gas turbine engines, as prime mover, must perform a robust economic analysis to guide the organisations investment decisions. One major element that could greatly influence the outcome of an economic analysis, and eventual organisational decisions and planning, is gas turbine engine acquisition price.
This study applies artificial neural networks to estimate gas turbine engine price. A supervised network learning strategy has been adopted to train the network from a dataset of historical records of gas turbine engine performance parameters and engine price. Numerical gradient checking has been performed to validate the computed cost function with quantified similarity obtained in the order of 10−9. The challenge of neural network overfitting has been minimized by applying a regularization technique. As such, the developed network closes reflects real world observations. To validate the network predictions, the developed neural network has been used to estimate the price of known gas turbine engine units with 95% to 99.9% accuracy.