A novel Parsimonious Genetic Programming(PGP) algorithm for aero-engine starting process optimum data-driven nonlinear dynamic modeling was proposed.First

the method used the traditional GP to generate nonlinear input-output models that were represented in a binary tree structure.Then

Orthogonal Least Squares algorithm(OLS) was applied to estimate the contribution of the tree branches to the accuracy of the model.On this basis

the complex redundant subtrees can be eliminated and the GP's convergence speed is enhanced.Finally

the simple

reliable and accurate linear-in-parameters nonlinear dynamic model can be obtained via GP evolution.Applications to real aeroengine starting process modeling validate that the proposed method can generate more robust and interpretable models.