The difference of single component characteristics which is caused by manufacture and installation can make the performance discrepant.A new entropy criterion particle swarm optimization (PSO) has been presented to revise the engine model based on the trial run data.The new algorithm adjusted the speed of inertia weight and migrated the particles of part poor fitness at the same time based on entropy discrimination.The presented algorithm overcame the defect of the original algorithm.The simulation results indicate that the single engine model correction based on entropy criterion PSO is better than the correction based on influence coefficient matrix (ICM).It is verified that the maximum error of the performance parameter is under 1.5%,which means the single engine model and real engine match better.