Thrust estimator design based on K-means clustering and reduced least squares support vector regression

In this paper,a design scheme for thrust estimator was proposed based on K-means clustering and reduced least squares support vector regression. Firstly,the K-means clustering algorithm was utilized to cluster the data in the full flight envelope,and the,a sub-estimator was designed in each cluster using the recursive reduced least squares support vector regression (RRLSSVR). In the process of designing the sub-estimator with RRLSSVR,the Cholesky factorization was utilized to replace the original iteration method for achieving higher numerical stability. Finally,the simulation experiments show that the thrust estimator can satisfy the requirement of direct thrust control for aeroengines,and as compared with other scheme for thrust estimator,the proposed scheme in this paper is also superior.