Thrust modelling of a target drone engine with nonlinear least - squares estimation based on series expansions

Abstract

Purpose The purpose of this study is to build a high accuracy thrust model under various small turbojet engine shaft speeds by using robust, ordinary, linear and nonlinear least squares estimation methods for target drone applications. Design/methodology/approach The dynamic shaft speeds from the test experiment of a target drone engine is conducted. Then, thrust values are calculated. Based on these, the engine thrust is modeled by robust linear and nonlinear equations. The models are benefited from quadratic, power and various series expansion functions with several coefficients to optimize the model parameters. Findings The error terms and accuracy of the model are given using sum of squared errors, root mean square error (RMSE) and R-squared (R-2) error definitions. Based on the multiple analyses, it is seen that the RMSE values are no more than 17.7539 and the best obtained result for robust least squares estimation is 15.0086 for linear at all cases. Furthermore, the R-2 value is found to be 0.9996 as the highest with the nonlinear Fourier series expansion model. Originality/value The motivation behind this paper is to propose robust nonlinear thrust models based on power, Fourier and various series expansion functions for dynamic shaft speeds from the test experiments.