Thrust Estimation of Small-scale Turbojet Engine based on K-Nearest Neighbor Bat Algorithm and Gaussian Process Regression

In order to achieve direct thrust control of small-scale turbojet engines, the model of K-nearest neighbor bat algorithm-Gaussian process regression (KNBA-GPR) is proposed in this paper. First, the population is initialized with Beta function and the information of K-neighborhood is regarded in the velocity update. To jump out of the local optimum, adaptive variation is also considered. Furthermore, the hyperparameters of Gaussian process regression are optimized by KNBA with the objective of minimizing the sum of squares due to error(SSE). Finally, the effectiveness of KNBA-GPR is verified by numerical simulations.