Aero-engine steady-state model modeling method based on simplex spline functions

Abstract

The invention discloses an aero-engine steady-state model modeling method based on simplex spline functions. Based on flight parameters and control parameters of a given engine, real-time estimation is conducted on the steady state of an aero-engine, an areo-engine steady-state model is built, the method comprises the steps of obtaining training data and testing data of the areo-engine steady-state model; conducting triangle partition on engine working areas, and calculating a barycentric coordinate system of a simplex where a working point is located; calculating basic functions of a simplex spline; solving coefficients of the functions of the simplex spline; utilizing the testing data to verify the generalization ability of the model, if the precision is poor, returning to the second step, re-conducting the triangle partition, and adjusting orders of the basic functions of the simplex spline; if the precision is good, establishing the steady-state model of the areo-engine. The modeling method is low in algorithm complexity, small in stored data size, good in real-time property and excellent in fitting effect, and the defect that large sample data cannot be fitted by support vector regression machine is effectively avoided.

Description

technical field [0001] The invention belongs to the technical field of system control and simulation in aerospace propulsion theory and engineering, and relates to a method for building a steady-state model of an aeroengine, in particular to a method for modeling a steady-state model of an aeroengine based on a simplex spline function. Background technique [0002] Aeroengine is a multivariable, strongly nonlinear and time-varying complex system, and its stable and safe operation places high demands on the engine control system. The nonlinear mathematical model of aero-engine is the basis of engine numerical simulation and the basis of establishing an adaptive real-time model of the engine. To fully tap the engine performance potential or perform fault diagnosis, it is first necessary to establish a high-precision, real-time airborne engine mathematical model that reflects the full flight envelope engine state, so as to ensure that the calculated engine parameters will not d...

AI Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is to overcome the deficiencies of the prior art, provide a kind of modeling method based on the simplex spline function of the aeroengine steady state model, can avoid the support vector machine method of the prior art such as multiple input multiple output Reduced iterative minimum quadratic support vector machine (MRR-LSSVR) has poor real-time model training in the case of large sample data, effectively reducing algorithm complexity and improving real-time training

Images