STOCHASTIC CONFIGURATION NETWORK-BASED HEALTH PARAMETER ESTIMATION METHOD FOR TURBOFAN ENGINE

摘要

The present invention belongs to the technical field of aeroengine fault diagnosis, and provides a stochastic configuration network-based health parameter estimation method for a turbofan engine. According to the stochastic configuration network-based health parameter estimation method for a turbofan engine designed in the present invention, a model-based Kalman filtering algorithm is combined with a data-driven-based stochastic configuration network, i.e. the output of the stochastic configuration network being used as compensation for the Kalman filtering algorithm, thereby comprehensively considering the estimation result of the Kalman filter and the estimation result of the stochastic configuration network, and improving the estimation accuracy of the original Kalman filtering algorithm when the measurable parameters of the turbofan engine are less than the health parameters to be estimated. In addition, the present invention effectively reduces the accuracy loss caused by a poor structure of a neural network by means of the stochastic configuration network, and improves the generalization capability of the network. In addition, a firefly algorithm is used to optimize parameters in a stochastic configuration network-based Kalman filter structure, increasing the estimation accuracy of the algorithm.