Aero-engine nonlinear control system modeling method

Abstract

The invention discloses an aero-engine nonlinear control system modeling method. The method comprises the following steps: setting an initial value of a nonlinear adaptive filter for identifying inputand output of a control system; recording current and historical moment inputs of the control system; performing inner product operation on an input vector obtained on the basis of the input and theweight of the FIR filter; obtaining FIR filtered output, obtaining a local interval index and a local normalized coordinate of the local interval index in the spline interpolator; generating spline interpolator output; output of the control system at the current moment is collected through a sensor. The output of the spline interpolator is compared with the output of the spline interpolator; obtaining filtering errors, respectively calculating the weight of the FIR filter and the gradient of the spline interpolation point at the current moment according to the filtering error; respectively calculating the gradient momentum of the weight of the FIR filter and the gradient momentum of the spline interpolation point according to the gradient, respectively updating the weight vector of the FIRfilter and the coordinates of the spline interpolation point, and continuing to carry out the adaptive process of the next moment from the second step.

Description

technical field [0001] The invention belongs to the technical field of aero-engine control, in particular to a modeling method for a non-linear control system of an aero-engine. Background technique [0002] Aeroengine is a highly complex and precise equipment, which is the product of deep integration of mechanical, thermodynamic, electrical, control, material and other disciplines. The control system is the "nerve center" of an aero-engine, and the reliability and safety of the control system are crucial to the safe operation of an aero-engine. The model is the basis of control. Due to the harsh working environment and changing working conditions of the aeroengine, the controlled object is time-varying. Therefore, a fast controlled object modeling method is more needed. [0003] Traditional modeling methods mainly focus on the study of component mechanism models. On the basis of analyzing the composition of the controlled system, the mathematical relationship between the ...

AI Technical Summary

Problems solved by technology

On the other hand, the current nonlinear system identification methods mainly focus on the research of neural network, support vector machine, Hammerstein-Wiener model, etc., which perform poorly in the rapidity and robustness of system identification, and are difficult to identify online.

Images