Combination of ANN and RSM to Evaluate Performance of a UAV Turbojet Engine

The aim of this research is to evaluate the performance parameters of a small-scale turbojet engine under the conditions of mixing kerosene-based JP-8 fuel, which is widely used in military aviation, with biodiesel fuel at different ratios. Artificial neural networks (ANN) and response surface methodology (RSM) were used for this evaluation. In the study, the jet engine was tested under various operating conditions such as load percentage and fuel mixture ratio. Energy and exergy analyses were performed to calculate the relevant parameters and determine the overall energy and exergy efficiencies of the engine. ANN and RSM models were created to estimate the response values. The ANN and RSM models exhibited remarkable predictive accuracy with high coefficients of determination (R2). Notably, the ANN models outperformed the RSMs, providing even higher R2 values. For example, the ANN model provided R2 results exceeding 0.98, while the RSM models achieved values around 0.97. In addition, the mean relative error (MRE) and root mean square error (RMSE) values for all response variables remained quite low.